U.S. patent application number 15/694750 was filed with the patent office on 2018-07-19 for meteorological sensing systems and methods.
The applicant listed for this patent is Physical Optics Corporation. Invention is credited to Ihor Berezhnyy, Tomasz Jannson, Gabriel Kaplan, Andrew Kostrzewski, David Miller, Koyiro Minakata, Edward Patton, Gregory Peng, Sookwang Ro, Dmitry Starodubov, Rodion Tikhoplav, Christopher Ulmer.
Application Number | 20180203158 15/694750 |
Document ID | / |
Family ID | 59982004 |
Filed Date | 2018-07-19 |
United States Patent
Application |
20180203158 |
Kind Code |
A1 |
Ulmer; Christopher ; et
al. |
July 19, 2018 |
METEOROLOGICAL SENSING SYSTEMS AND METHODS
Abstract
A portable weather station, including an lower body portion; an
upper body portion disposed on the lower body portion in a spaced
apart relationship thereby forming an open channel between the
upper body portion and the lower body portion; and a plurality of
weather condition sensors wherein a first set of one or more of the
plurality of weather condition sensors is mounted on the upper body
portion of the portable weather station and a second set of one or
more of the plurality of weather condition sensors is mounted on
the lower body portion of the portable weather station.
Inventors: |
Ulmer; Christopher; (San
Pedro, CA) ; Starodubov; Dmitry; (Reseda, CA)
; Peng; Gregory; (Redondo Beach, CA) ; Tikhoplav;
Rodion; (Santa Monica, CA) ; Miller; David;
(San Pedro, CA) ; Kostrzewski; Andrew; (Garden
Grove, CA) ; Minakata; Koyiro; (Irvine, CA) ;
Kaplan; Gabriel; (Calabasas, CA) ; Jannson;
Tomasz; (Torrance, CA) ; Patton; Edward;
(Torrance, CA) ; Ro; Sookwang; (Glendale, CA)
; Berezhnyy; Ihor; (La Jolla, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Physical Optics Corporation |
Torrance |
CA |
US |
|
|
Family ID: |
59982004 |
Appl. No.: |
15/694750 |
Filed: |
September 1, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14457511 |
Aug 12, 2014 |
9784887 |
|
|
15694750 |
|
|
|
|
61865069 |
Aug 12, 2013 |
|
|
|
61923457 |
Jan 3, 2014 |
|
|
|
61947886 |
Mar 4, 2014 |
|
|
|
61953603 |
Mar 14, 2014 |
|
|
|
61989660 |
May 7, 2014 |
|
|
|
62005840 |
May 30, 2014 |
|
|
|
62017745 |
Jun 26, 2014 |
|
|
|
62020574 |
Jul 3, 2014 |
|
|
|
62026549 |
Jul 18, 2014 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01W 1/02 20130101; G08B
17/005 20130101; G08B 21/18 20130101; G01S 7/4813 20130101; G01S
7/003 20130101; Y02A 90/19 20180101; G01S 17/95 20130101; G01W 1/08
20130101; Y02A 90/10 20180101; G08B 21/10 20130101; B64D 1/08
20130101; G01S 17/86 20200101 |
International
Class: |
G01W 1/02 20060101
G01W001/02; G08B 21/18 20060101 G08B021/18; B64D 1/08 20060101
B64D001/08; G01S 17/95 20060101 G01S017/95 |
Goverment Interests
STATEMENT OF RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED
RESEARCH
[0002] This invention was at least partially supported by the
Government under Contract # H92222-11-C-0034 awarded by U.S.
Special Operations Command.
Claims
1. A portable weather station, comprising a lower body portion; an
upper body portion disposed on the lower body portion in a spaced
apart relationship thereby forming an open channel between the
upper body portion and the lower body portion; a plurality of
weather condition sensors wherein a first set of one or more of the
plurality of weather condition sensors is mounted on the upper body
portion of the portable weather station and a second set of one or
more of the plurality of weather condition sensors is mounted on
the lower body portion of the portable weather station.
2. A wind sensing apparatus, comprising: a thermal generator
coupled to a power source; a plurality of temperature sensors
arranged in a predetermined pattern with respect to the thermal
generator; a detection module configured to determine wind speed or
wind direction based on temperatures measured by the temperature
sensors.
3. The wind sensing apparatus according to claim 2, wherein the
detection module is configured to determine wind direction based on
differences in temperatures sensed by one or more of the plurality
of temperature sensors.
4. The wind sensing apparatus according to claim 3, wherein the
wind direction is determined based on an increase in temperature
sensed at one or more of the plurality of temperature sensors.
5. The wind sensing apparatus according to claim 3, wherein the
wind direction is determined based on a temperature differential
measured at one or more of the plurality of temperature sensors
relative to the other of the plurality of temperature sensors.
6. The wind sensing apparatus according to claim 2, wherein the
detection module is configured to determine wind speed based on a
temperature measured by one or more of the sensors relative to a
predicted temperature for that sensor.
7. The wind sensing apparatus according to claim 6, wherein the
predicted temperature for said sensor is determined based on one or
more of an amount of energy applied to the thermal generator, a
distance between the thermal generator and said sensor and the
ambient temperature.
8. A cloud-ceiling sensing apparatus, comprising: an optical light
source disposed to transmit light toward the cloud ceiling; a
photodetector disposed in an orientation to receive light from the
light source that has been reflected from the cloud ceiling; a
shroud at least partially surrounding the optical light source at a
predetermined height, wherein the predetermined height is selected
based upon a height needed to prevent light from the optical light
source from directly impinging on the photodetector.
9. The cloud-ceiling sensing apparatus according to claim 8,
further comprising a ceiling height calculation module configured
to determine a time of flight for the light to travel from the
light source to the cloud ceiling and to the photodetector and to
compute a distance from the cloud-ceiling sensing apparatus to the
cloud ceiling based on the determined time of flight.
10. A method for detecting weather anomalous events using weather
sensor fusion, comprising: receiving at a cybersensor weather data
samples from first weather sensing equipment; the cybersensor
evaluating the weather data samples from the first weather sensing
equipment against weather data in a database, determining whether
or not the weather anomalous event exists based on the evaluation,
generating the first alarm indicating the presence of a weather
anomalous event when the determination is positive, and not
generating the first alarm indicating the presence of a weather
anomalous event when the determination is negative; and receiving
at a second cybersensor a subset of the weather data samples,
evaluating the weather data samples against weather data in a
database, determining whether or not the weather anomalous event
exists based on the evaluation, generating a second alarm
indicating the presence of a weather anomalous event when the
determination is positive, and not generating the second alarm
indicating the presence of a weather anomalous event when the
determination is negative; wherein the first cybersensor has a
higher likelihood of a false positive determination than the second
cybersensor; generating a final alarm indicating the presence of a
weather anomalous event when both the first and second cybersensors
determined that the weather anomalous event exists, and not
generating the final alarm indicating the presence of a weather
anomalous unless both the first and second cybersensors determined
that the weather anomalous event exists.
11. The method of claim 10, wherein the subset of weather data
samples received at the second cyber sensor comprises only those
weather data samples for which the first cybersensor positively
determined the presence of a weather anomalous event.
12. The method of claim 10, further comprising receiving at the
second cybersensor one or more additional weather data samples
corresponding to weather anomalous events positively determined by
one or more additional cybersensors.
13. The method of claim 10, further comprising receiving at one or
more successive cybersensors, a corresponding subset of the weather
data samples, each corresponding subset of the weather data samples
comprising weather data samples for which an immediately prior
cybersensor determined a weather anomalous event exists, and
wherein the final alarm indicating the presence of a weather
anomalous event when is generated when all of the cybersensors have
determined that the weather anomalous event exists, and not
generated unless all of the cybersensors have determined that the
weather anomalous event exists.
14. The method of claim 10, wherein the final alarm is the second
alarm generated by the second cybersensor.
15. A system for detecting weather anomalous events, comprising: a
database storing weather events and corresponding weather data for
the weather events; a plurality of cybersensors arranged in series
relative to one another, each cybersensor having an input coupled
to receive weather data samples generated by weather sensing
equipment, and each cybersensor configured to evaluate the weather
data samples against weather data in a database, determine whether
or not a weather anomalous event exists based on the evaluation,
generate a signal indicating the presence of a weather anomalous
event when the determination is positive, and not generating the
signal indicating the presence of a weather anomalous event when
the determination is negative; wherein each successive cybersensor
is configured as having a progressively lower likelihood of a false
positive determination than its preceding cybersensor.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a Divisional of U.S. patent application
Ser. No. 14/457,511, filed on Aug. 12, 2014, which claims the
benefit under 35 U.S.C 119(e) of U.S. Provisional Application Nos.
62,026,549, filed Jul. 18, 2014, 62/020,574, filed Jul. 3, 2014,
62/017,745, filed Jun. 26, 2014, 62/005,840, filed May 30, 2014,
61/989,660, filed May 7, 2014, 61/953,603, filed Mar. 14, 2014,
61/947,886, filed Mar. 4, 2014, 61/923,457, filed Jan. 3, 2014,
61/865,069, filed Aug. 12, 2013, each of which are hereby
incorporated herein by reference in their entirety.
TECHNICAL FIELD
[0003] The disclosed technology relates generally to meteorological
sensing, and more particularly, some embodiments relate to
portable, remote weather sensing devices.
DESCRIPTION OF THE RELATED ART
[0004] Meteorology, including fields of weather measurement,
weather forecasting, climatology, atmospheric chemistry, and
atmospheric physics has long been an important field of observation
and study. Indeed, there is evidence that as early as 400 BC (and
perhaps earlier), advanced societies attempted to predict weather
and climate patterns. Since at least as early as the 15th century,
there have been efforts to provide equipment to measure atmospheric
variables with a fairly high degree of accuracy. Rain, wind,
barometric pressure, temperature and humidity are examples of
weather-related variables typically measured by meteorological
sensing equipment. Early meteorological sensing equipment included
items such as the rain gauge, the anemometer, and the hygrometer.
Sometime later, equipment such as the barometer and the Galileo
thermometer were developed. The 20th century brought with it
developments in remote sensing devices such as weather radar,
weather satellites, and other technologically advanced
weather-sensing equipment.
[0005] In many cases, meteorological sensing equipment can be
configured for remote operation, allowing sensing instruments to
collect data regarding weather events at a location remote from the
user's base of operations. Such remotely collected data can be
transmitted back to the base of operations for collection, study
and record keeping. The base of operations may include, for
example, instrumentation and equipment to receive and analyze
collected data. The analyzed data can be provided to personnel for
operational purposes, or it can be observed and studied for
purposes such as, for example, weather forecasting, climate study
and so on.
[0006] Weather observation plays an essential role in human life.
Precise detection and recording of key weather parameters such as
wind, pressure, temperature, visibility, and cloud layer height are
essential, for example, for the safety of airplane flights. As
further examples, continuous weather recording is important for
wild fire prevention and disaster response. Currently available
weather observation systems such as the TMQ-53 Tactical
Meteorological Observing System have sufficient accuracy to record
weather parameters for aviation, however the system requires
stationary power connectivity (external power source connection
such as the grid) and larger than desirable, difficult to assemble,
and too expensive for wide deployment. Alternatively, there are
smaller handheld weather observation units, such as the Kestrel
4000 Pocket Weather Meter, which can record wind, temperature,
pressure, and humidity. Such units cannot, however, measure
visibility or cloud layer height and cannot operate remotely and
independently from the operator. These deficiencies hinder the
ability of current devices to function autonomously and or for use
in automated weather recording.
[0007] Currently, there is no weather observation device that would
be small and light enough for easy deployment (size, weight and
power constrains), accurate enough for aircraft operation, and
capable of providing sustainable weather recording and transmission
from a remote location.
[0008] As noted above, the current weather sensor solutions are
much heavier than desired for remote applications and are too large
for easy transport and implementation. In addition, the typical
configuration used in existing weather stations requires isolation
of each component, which requires the use of separate arms mounted
onto the base unit and adds to the bulk and size of the system.
Various embodiments provide a Portable Imaging Weather Observation
System (PIWOS) and features thereof, that can be configured
addresses the need.
[0009] A key weather parameter that is desired to be measured in
most full-capability weather monitoring stations is precipitation
type and amount. Precipitation type refers to the identification of
the falling moisture into categories such as drizzle, rain, hail,
small hail, snow, etc. Precipitation amount is a volumetric
assessment of the amount of moisture falling per unit of time,
typically measured in inches per hour (or other like units). These
parameters are often measured by existing sensors that use
mechanical and optical means to assess precipitation type and
amount, for the purposes of sensing microclimates.
BRIEF SUMMARY OF EMBODIMENTS
[0010] According to various embodiments disclosed herein, various
novel configurations for a portable, field-operable weather station
can be provided. Various embodiments also provide unique
technologies for sensor configurations, data communication,
information security and data analysis.
[0011] In various embodiments, a weather station device can be
provided that utilizes thermal components for windspeed and wind
direction measurement. For example, in various embodiments, a
thermal anemometer can be configured to measure wind speed and wind
direction with few or no moving parts. Heat generator(s) and
thermal sensor(s) can be configured to take advantage of the
cooling effects that generally accompany wind, and the
understanding that the cooling effects are generally greater at
higher wind speeds. In some embodiments, a novel apparatus and
method can be included to simultaneously determine the wind
direction as well as the velocity of the wind.
[0012] A thermal generator or other heat source can be included
along with a plurality of thermal sensors (e.g., thermistors or
thermocouples) disposed proximal to the heat source. In such
embodiments, the system can be configured to determine the
direction of the wind based on the temperature differentials among
the sensors or the temperature changes detected by the sensors.
This can be used to determine the direction in which heat from the
heat source is most predominantly lost to the air. In some
embodiments, this heat source could be a resistor or other
resistive element heated by an electrical current. Other heat
sources can be used as well.
[0013] Embodiments can be configured to use a plurality of thermal
sensors arranged in a pattern (e.g., in a ring) around the heat
source to determine the direction in which the heat from the heat
source is blown. This design measures wind direction by allowing
the heat emitted by the heat source to be detected by the thermal
sensors around the heat source. Those thermal sensors that are
downwind from the heat source will detect the heat that is carried
by the wind from the heat source to the sensor.
[0014] Accordingly, in various embodiments, a new thermal sensor
design can be implemented to enable the determination of wind speed
and direction using purely thermal means allowing for extreme
reduction in size of the device and the complete elimination of
moving parts. Embodiments can be implemented to: Reduce size versus
mechanical and acoustic approaches; eliminate binding due to dust
contamination that is suffered by mechanical approaches; eliminate
freezing due to ice accumulation environments that is suffered by
mechanical approaches.
[0015] In some embodiments, a new micro visibility sensor
configuration can be included to measure visibility in an extremely
compact device with limited power.
[0016] Various embodiments can be configured to use a novel
Co-Axial Configuration for the weather sensor that organizes the
sensors into specific groups mounted co-axially so that the remote
weather system can be combined as a single unit instead of multiple
components connected through cables. Embodiments can be implemented
to: Eliminate multiple setup steps required to assemble components
in a conventional system; eliminate large, bulky towers; reduce
overall system profile against wind loading; reduce cost by
eliminating multiple enclosures and interconnections; efficiently
enable protection of exposure-sensitive sensors by stacking them
axially underneath sensors and components that require full
exposure; enable potential air-droppable capability; improve
ruggedness and durability.
[0017] A new precipitation sensor design can be implemented that
uses an accelerometer to detect the contact of precipitation with
the enclosure of the MWS enables detection of precipitation over a
large area using a micro-size component. Embodiments can be
implemented to: Analyze the impact signature to allow rain, hail,
small hail, and drizzle to be distinguished; use electrodes
embedded in hydrophobic plastic to enable detection of active
moisture in the air rather than standing water; use the entire
system body for detection to allow for a radical size reduction
compared to conventional approaches.
[0018] A new camera system can be included. In some embodiments,
the new camera system uses state-of-the art micro-optics allowing
for panoramic images to be taken, digitally compressed, and
transmitted via satellite from any location. Embodiments can be
implemented to include: Mounting of micro cameras on a flexible
printed circuit board enables positioning of cameras for panoramic
operation; use of optical lenses and prisms allows for a plurality
of camera configurations; compression of images followed by breakup
into packets enables delivery of images via Iridium satellite
connection.
[0019] A new power system design that coordinates the collection of
solar power, recharge and discharge of batteries, temperature
control, and sleep-cycle control enables the system to operate
autonomously. Embodiments can be implemented to include: Use of
hybrid system of batteries and electric double-layer capacitors
(EDLC) enables high-power boost transmissions using a dramatically
smaller battery source than can be achieved otherwise.
[0020] In various embodiments of the disclosed technology, systems
and methods for deploying weather station equipment or other
equipment or instrumentation may also be provided. For example,
apparatus can be provided to deploy instrumentation from an
aircraft or other airborne vehicle or platform, including, for
example, fixed-wing or rotor aircraft. In another example,
apparatus can be provided to integrate the instrumentation both
physically and electronically with a fixed or moving platform such
as a vehicle.
[0021] In various embodiments, aspects of the deployment systems
disclosed herein can be configured to provide one or more of the
following features for integration with fixed or moving platforms.
[0022] Allowance for a rigid mounting to a platform via common
mounting features such as a single threaded insert or four standard
screw holes in the weather station device [0023] Allowance for the
weather station device to be powered by external sources, such as a
vehicle battery [0024] Allowance for the weather station device to
transmit data via cabling to an external screen such as a laptop or
vehicle-mounted display [0025] Allowance for the weather station
equipment to transmit data via radio frequency (HF, VHF, UHF,
Bluetooth, etc) or existing external satellite communications.
[0026] In various embodiments, aspects of the deployment systems
disclosed herein can be configured to provide one or more of the
following features for the deployment of equipment (in some
embodiments, sensitive equipment). [0027] Protects the equipment
from being damaged upon impact [0028] Anchors the equipment to the
ground upon impact [0029] Creates a stable platform in
non-penetrable soils [0030] Can be used on uneven or sloped terrain
[0031] Ensures that the equipment will be oriented vertically
[0032] Raises the equipment to a desired height
[0033] Various aspects of the deployment systems can be configured
to include several features that may work together to form a
comprehensive device, but individual aspects of the technology can
be used alone or in subsets to achieve desired results or
objectives. Examples of such aspects of the design can include:
[0034] Aerodynamic Shape--the shape of the deployment device can be
configured to ensure verticality (or near verticality) of the
deployed system. The shape can also be configured to provide a
consistent or somewhat consistent impact speed regardless of drop
height. In other words, the shape can be configured to yield a
terminal velocity at a desired speed or within a desired range of
velocities. [0035] Anchoring Mechanism--in various embodiments, the
device can include an anchor mechanism to anchor the system into
the ground in penetrable soils. For example, a weighted spike can
be included at the bottom of the system (e.g., at the bottom of the
mast) to penetrate the soil where the device lands, anchoring the
device to the ground or with pneumatic cushioning, friction-based
energy absorbers, or other means of absorbing impact energy. [0036]
Recoiling mast--the mass can be configured to absorb some or all of
the shock of the impact to help protect the equipment upon impact.
For example, a multi-segment mast can be provided with
spring-loaded segment or segments to help absorb the shock of
impact of the system with the ground. [0037] Ratcheting Fins--fins
can be included to provide single- or dual-functionality. For
example, aerodynamic fins can be included to provide orientational
stability to the system during travel from the deployment vehicle
(e.g., the aircraft) to the deployment location (e.g., ground).
Fins can also be included to provide a stable base for the system
at the deployment location. Further, translatable fins can be
provided to serve as stability fins during flight at the aft end of
the system body, and to move along the body to the base upon impact
with the deployment surface (e.g., the ground). A ratcheting or
spring-like mechanism can be used to allow the fins to move along
the body (e.g. the mast) of the device from the aft end to the
base. For example, a ratcheting mechanism can be included that
requires sufficient force to prevent the fins from moving from the
aft end to the fore end during flight; and this force can be set at
such a level that the momentum of the fins upon impact with the
deployment surface allows the fins to overcome this force and
travel along the body to the base of the device at the deployment
surface. Additionally, the ratcheting mechanism can be configured
to prevent the fins from moving back up the mast after deployment
on the deployment surface to provide a stable base for the system.
These fins can be used to provide stability to the system in both
non-penetrable penetrable deployment surfaces. [0038] Flash
Parachute--a parachute can be included to arrest horizontal motion
to facilitate more accurate placement for deployment from
low-flying high-speed aircraft.
[0039] As noted above, the various aspects of the technology
described herein can be used individually or in various
combinations as may be desired or appropriate for a given
application or situation. For example, it may be desirable to use
fins without the parachute, the spike without fins, fixed rather
than ratcheting fins, and so on.
[0040] Accordingly, in various embodiments the technology disclosed
herein can provide systems and methods for deploying sensors,
weather equipment, and other electronic equipment by aircraft
without requiring hand installation by soldiers on the ground (or
other ground personnel). After reading this description, it will
become apparent to those of ordinary skill in the art how the
systems and methods described herein can be used for deployment of
other apparatuses as well. Other examples include, for example,
seismic, chemical, radiological, reconnaissance, indications, image
capture, weaponry, or other equipment or sensors. As will also
become apparent to those of ordinary skill in the art after reading
this description, the size and shape of the system can be varied
from that depicted in the figures herein without departing from the
spirit and scope of this technology.
[0041] It is desirable that weather system devices can be
configured in various embodiments to include a design that is
smaller and more compact than presently available options, while
also weighing less (e.g., less than a pound). The weather sensor
would ideally also have higher power efficiency than current
solutions by two orders of magnitude. In addition, the device would
ideally also be capable of automatic collection of weather data and
transmission of the data via satellite communications, with the
data capable of being reported hourly in a format suitable for
conversion to the METAR weather format. The device would ideally
also be operational in varying weather conditions, including in
snow accumulation environments. Computation of dew-point, station
pressure, and altimeter setting are desired features. The device
would ideally be capable of operation over the entire range of
terrestrial temperatures. Measuring capabilities should include
temperature, pressure, humidity, wind velocity, wind direction,
gust velocity, gust direction, and lightning. To prevent
unauthorized access and data corruption, the device would ideally
be capable of detecting tilt and tampering. For visualization, the
device should be capable of collecting panoramic images of the
surrounding area and be capable of transmitting those images over a
satellite link.
[0042] In various embodiments, an apparatus for air-drop deployment
of a payload comprising instrumentation or equipment is provided.
The deployment apparatus includes, in various embodiments an
elongate body member having a first end and a second end; a
mounting ring disposed on and at least partially surrounding a
portion of the body member at the first end; a fin assembly
comprising a plurality of fins attached to the mounting ring; a
weighted tip at the second end of the body member; and a connector
at the first end configured to engage the payload.
[0043] The mounting ring may include, for example, a release
mechanism and is slidably mounted to the elongate body, wherein the
release mechanism is configured to maintain the mounting ring at
the first end during flight and to release the mounting ring upon
impact of the apparatus with a deployment surface, allowing the
mounting ring to move from the first end toward the second end upon
impact. A ratcheting mechanism may also be included to connect the
fin assembly to the mounting ring, wherein the ratcheting mechanism
allows the fin assembly to pivot from an in-flight position to a
deployment position.
[0044] In various configurations, the elongate body may include a
plurality of coaxially arranged sections slidably disposed in an
end-to-end arrangement, a locking mechanism configured to retain
the plurality of coaxially arranged sections in a retracted
position; and a spring mechanism applying pressure against the
coaxially arranged sections.
[0045] In other embodiments, a portable weather station includes a
lower body portion; an upper body portion disposed on the lower
body portion in a spaced apart relationship thereby forming an open
channel between the upper body portion and the lower body portion;
a plurality of weather condition sensors wherein a first set of one
or more of the plurality of weather condition sensors is mounted on
the upper body portion of the portable weather station and a second
set of one or more of the plurality of weather condition sensors is
mounted on the lower body portion of the portable weather
station.
[0046] A wind sensing apparatus can include a thermal generator
coupled to a power source; a plurality of temperature sensors
arranged in a predetermined pattern with SMRH:485900265.1 respect
to the thermal generator; a detection module configured to
determine or estimate wind speed or wind direction based on
temperatures measured by the temperature sensors. The detection
module may be configured to determine wind direction based on
differences in temperatures sensed by one or more of the plurality
of temperature sensors.
[0047] The wind direction may be determined, in some embodiments,
based on an increase in temperature sensed at one or more of the
plurality of temperature sensors. The wind direction may be
determined based on a temperature differential measured at one or
more of the plurality of temperature sensors relative to the other
of the plurality of temperature sensors.
[0048] In various embodiments, the detection module may be
configured to determine wind speed based on a temperature measured
by one or more of the sensors relative to a predicted temperature
for that sensor.
[0049] The wind sensing apparatus according to claim 6, wherein
predicted temperature for said sensor is determined based on one or
more of an amount of energy applied to the thermal generator, a
distance between the thermal generator and said sensor and the
ambient temperature.
[0050] In still further embodiments, a cloud-ceiling sensing
apparatus may be provided an may include: an optical light source
disposed to transmit light toward the cloud ceiling; a
photodetector disposed in an orientation to receive light from the
light source that has been reflected from the cloud ceiling; a
shroud at least partially surrounding the optical light source at a
predetermined height, wherein the predetermined height is selected
based upon a height needed to prevent light from the optical light
source from directly impinging on the photodetector.
[0051] The cloud-ceiling sensing may also include a ceiling height
calculation module configured to determine a time of flight for the
light to travel from the light source to the cloud ceiling and to
the photodetector and to compute a distance from the cloud-ceiling
sensing apparatus to the cloud ceiling based on the determined time
of flight.
[0052] In yet further embodiments, systems and methods for
detecting weather anomalous events using weather sensor fusion may
be provided and may include: receiving at a cybersensor weather
data samples from first weather sensing equipment; the cybersensor
evaluating the weather data samples from the first weather sensing
equipment against weather data in a database, determining whether
or not the weather anomalous event exists based on the evaluation,
generating the first alarm indicating the presence of a weather
anomalous event when the determination is positive, and not
generating the first alarm indicating the presence of a weather
anomalous event when the determination is negative; and receiving
at a second cybersensor a subset of the weather data samples,
evaluating the weather data samples against weather data in a
database, determining whether or not the weather anomalous event
exists based on the evaluation, generating a second alarm
indicating the presence of a weather anomalous event when the
determination is positive, and not generating the second alarm
indicating the presence of a weather anomalous event when the
determination is negative; wherein the first cybersensor has a
higher likelihood of a false positive determination than the second
cybersensor; and generating a final alarm indicating the presence
of a weather anomalous event when both the first and second
cybersensors determined that the weather anomalous event exists,
and not generating the final alarm indicating the presence of a
weather anomalous unless both the first and second cybersensors
determined that the weather anomalous event exists.
[0053] In various such embodiments, the subset of weather data
samples received at the second cyber sensor comprises only those
weather data samples for which the first cybersensor positively
determined the presence of a weather anomalous event.
[0054] The systems and methods may further include receiving at the
second cybersensor one or more additional weather data samples
corresponding to weather anomalous events positively determined by
one or more additional cybersensors. They may also include
receiving at one or more successive cybersensors, a corresponding
subset of the weather data samples, each corresponding subset of
the weather data samples comprising weather data samples for which
an immediately prior cybersensor determined a weather anomalous
event exists, and wherein the final alarm indicating the presence
of a weather anomalous event when is generated when all of the
cybersensors have determined that the weather anomalous event
exists, and not generated unless all of the cybersensors have
determined that the weather anomalous event exists. In some
configurations, the final alarm is the second alarm generated by
the second cybersensor.
[0055] A system for detecting weather anomalous events may include
a database storing weather events and corresponding weather data
for the weather events; a plurality of cybersensors arranged in
series relative to one another, each cybersensor having an input
coupled to receive weather data samples generated by weather
sensing equipment, and each cybersensor configured to evaluate the
weather data samples against weather data in a database, determine
whether or not a weather anomalous event exists based on the
evaluation, generate a signal indicating the presence of a weather
anomalous event when the determination is positive, and not
generating the signal indicating the presence of a weather
anomalous event when the determination is negative; wherein each
successive cybersensor is configured as having a progressively
lower likelihood of a false positive determination than its
preceding cybersensor.
[0056] Other features and aspects of the disclosed technology will
become apparent from the following detailed description, taken in
conjunction with the accompanying drawings, which illustrate, by
way of example, the features in accordance with embodiments of the
disclosed technology. The summary is not intended to limit the
scope of any inventions described herein, which are defined solely
by the claims attached hereto.
BRIEF DESCRIPTION OF THE DRAWINGS
[0057] The technology disclosed herein, in accordance with one or
more various embodiments, is described in detail with reference to
the following figures. The drawings are provided for purposes of
illustration only and merely depict typical or example embodiments
of the disclosed technology. These drawings are provided to
facilitate the reader's understanding of the disclosed technology
and shall not be considered limiting of the breadth, scope, or
applicability thereof. It should be noted that for clarity and ease
of illustration these drawings are not necessarily made to
scale.
[0058] Some of the figures included herein illustrate various
embodiments of the disclosed technology from different viewing
angles. Although the accompanying descriptive text may refer to
such views as "top," "bottom" or "side" views, such references are
merely descriptive and do not imply or require that the disclosed
technology be implemented or used in a particular spatial
orientation unless explicitly stated otherwise.
[0059] FIG. 1 is a diagram illustrating an example of stacked
sensor sets in accordance with one embodiment of the technology
disclosed herein.
[0060] FIG. 2 is a functional block diagram illustrating an example
weather station device in accordance with one embodiment of the
technology described herein.
[0061] FIG. 3 is a diagram illustrating an example embodiment of an
extension module for the weather station device in accordance with
one embodiment of the technology described herein.
[0062] FIG. 4A is a diagram illustrating a top view of an example
mechanical layout for exemplary thermal sensors in accordance with
one embodiment of the technology disclosed herein.
[0063] FIG. 4B is a diagram illustrating a side view of an example
mechanical layout for exemplary thermal sensors in accordance with
one embodiment of the technology disclosed herein.
[0064] FIG. 5A is a diagram illustrating a front view of an example
of an opto-mechanical arrangement in accordance with one embodiment
of the technology described herein.
[0065] FIG. 5B is a diagram illustrating a side view of an example
of an opto-mechanical arrangement in accordance with one embodiment
of the technology described herein.
[0066] FIG. 6 is a diagram illustrating a close-up view of a
portion of a visibility measurement system such as that shown
above.
[0067] FIG. 7 is a diagram illustrating an example of a diffraction
problem in accordance with various embodiments.
[0068] FIG. 8 is a diagram illustrating an example of an expanded
version of FIG. 7, including a source geometry, to illustrate an
example of the diffraction problem
[0069] FIG. 9 is a diagram illustrating an example of this relation
between the inclination angle, .alpha., and diffraction angle,
.delta..
[0070] FIG. 10 is a diagram illustrating a Generalized Lambertian
Source Model in polar coordinates r=r(.alpha.).
[0071] FIG. 11A is a diagram illustrating an example of diffraction
edge profiling in accordance with various embodiments of the
technology disclosed herein.
[0072] FIG. 11B is a diagram illustrating an example of diffraction
edge profiling in accordance with various embodiments of the
technology disclosed herein.
[0073] FIG. 12 is a diagram illustrating an Exemplary Look-Up Table
for Dual-Control-Variable Calibration Parameter, K(.delta..sub.i,
r.sub.j)=K.sub.ij.
[0074] FIG. 13A is a diagram illustrating diffraction
efficiency.
[0075] FIG. 13B is a diagram illustrating photodiode quantum
efficiency.
[0076] FIG. 13C is a diagram illustrating light source power
density.
[0077] FIG. 14 presents a 2D cross-section of a 3D geometry with
tube axial symmetry 681 in accordance with various embodiments of
the technology disclosed herein.
[0078] FIG. 15 is a diagram illustrating an example precipitation
characterizer subsystem in accordance with various embodiments of
the technology disclosed herein.
[0079] FIG. 16 is a diagram illustrating an example precipitation
quantifier subsystem in accordance with one embodiment of the
technology disclosed herein.
[0080] FIG. 17 is a diagram illustrating a droplet former
quantifier in accordance with one embodiment of the technology
disclosed herein.
[0081] FIG. 18 is a diagram illustrating a side view of a
deployment assembly in accordance with one embodiment of the
technology described herein.
[0082] FIG. 19 is a diagram illustrating an example of a
multi-segment riser pole before, during, and after impact in
accordance with one embodiment of the technology disclosed
herein.
[0083] FIG. 20 is a diagram illustrating an example of this
displacement and ratcheting in accordance with one embodiment of
the systems and methods described herein.
[0084] FIG. 21 is a diagram illustrating an example of deployment
on the slope using ratcheting fins in accordance with various
embodiments of the technology disclosed herein.
[0085] FIG. 22A illustrates an example of horizontal travel as a
result of speed of the deployment aircraft.
[0086] FIG. 22B illustrates an example of horizontal travel as a
result of speed of the deployment aircraft.
[0087] FIG. 23 is a diagram illustrating the operational release of
a flash parachute in accordance with one embodiment of the
technology described herein.
[0088] FIG. 24 is a diagram illustrating an example Truthing-based
Anomalous Event Software Engine (TAESE) that can be used to
implement a WAES in accordance with one embodiment of the
technology described herein.
[0089] FIG. 25 is a diagram illustrating an example of WAEVENT
Sensor Fusion (WSF) for two cascaded sensors in accordance with one
embodiment of the systems and methods described herein.
[0090] FIG. 26 is a diagram illustrating an example of a WAEVENT
Sensor Fusion (WSF) Software Engine using two (2) cybersensors in
cascade in accordance with one embodiment of the systems and
methods described herein.
[0091] FIG. 27 is a diagram illustrating an example of a WAEVENT
Sensor Fusion (WSF) engine for 4 cascaded cyber sensors in
accordance with one embodiment of the systems and methods described
herein.
[0092] FIG. 28 is a diagram illustrating an example of a Weather
Data Event Format (WDEF) in accordance with various embodiments of
the systems and methods disclosed herein.
[0093] FIG. 29 is a diagram illustrating this relationship.
Particularly, FIG. 29 provides an illustration of optimum
performance for .PHI.=.PHI..sub.0 for y(x)-dependence, where x=(CR)
and y=.epsilon..sub.FEC.
[0094] FIG. 30 is a diagram illustrating the relationship presented
by Eq. (51).
[0095] FIG. 31A is a diagram illustrating a data transmission (Tx)
transfer sequence, which includes examples of characteristic
parameters representing each step. A similar sequence, but in the
inverse, may occur for the data receiving (Rx) transfer
sequence.
[0096] FIG. 31B is a diagram illustrating the data receiving (Rx)
transfer sequence, in which the characteristic operations are
defined rather than representative parameters.
[0097] FIG. 31C is a diagram illustrating the data Tx transfer
sequence, equivalent to the data Rx transfer sequence of FIG.
31B.
[0098] FIG. 32 is a diagram illustrating an example of
(PSNR)-dependence as a function of internal (BER).sub.1-control
variable, defined by Eq. (84) in accordance with one embodiment of
the systems and methods described herein.
[0099] FIG. 33 is a diagram illustrating an example of U-dependence
as a Function of (CR) for various (BER).sub.1-values in accordance
with one embodiment of the systems and methods described
herein.
[0100] FIG. 34 is a diagram illustrating an example of z Function
dependence as a Function of y, with the x variable as a parameter,
in which the ex-variable as values x.sub.1, x.sub.2, x.sub.3.
[0101] FIG. 35 is a diagram illustrating an example of a z Function
as a Function of y, with the variable x as a parameter in
accordance with one embodiment of the systems and methods described
herein.
[0102] FIG. 36 is a diagram illustrating an exemplary z surface in
(x, y, z) space, including contour lines in accordance with one
embodiment of the systems and methods described herein.
[0103] FIG. 37 is a diagram illustrating an example of planes
perpendicular to the x axis in accordance with one embodiment of
the systems and methods described herein.
[0104] FIG. 38 is a diagram illustrating an example of x
cross-sections of an exemplary z surface in accordance with one
embodiment of the systems and methods described herein.
[0105] FIG. 39A is a diagram illustrating an exemplary relation
between state and control variables due to causation principle.
[0106] FIG. 39B is a diagram illustrating an exemplary relation
between state and control variables due to causation principle.
[0107] FIG. 40 is a diagram illustrating an example of a reflection
between subspaces and parameters.
[0108] FIG. 41 is a diagram illustrating a familiar contour mapping
with contour lines at z: 100 m, 110 m, 120 m--elevations.
[0109] FIG. 42 is a diagram illustrating an example of non-linear
contour mapping in accordance with various embodiments of the
technology disclosed herein.
[0110] FIG. 43 is a diagram illustrating an example of a z manifold
in (x, y, z) space in accordance with various embodiments of the
technology disclosed herein.
[0111] FIG. 44 is a diagram illustrating an example of non-linear
contour lines in accordance with various embodiments of the
technology disclosed herein.
[0112] FIG. 45 is a diagram illustrating an example of hysteresis
in the case of the desert rain phenomenon in accordance with
various embodiments of the technology disclosed herein.
[0113] FIG. 46 is a diagram illustrating an example of a typical
relation (44) for normal dispersion in accordance with various
embodiments of the technology disclosed herein.
[0114] FIG. 47 is a diagram illustrating an example of Linear
R(c)-dependence in accordance with various embodiments of the
technology disclosed herein.
[0115] FIG. 48A illustrates an example of non-linear oscillator
catastrophes in accordance with various embodiments of the
technology disclosed herein.
[0116] FIG. 48B illustrates an example of non-linear oscillator
catastrophes in accordance with various embodiments of the
technology disclosed herein.
[0117] FIG. 48C illustrates an example of non-linear oscillator
catastrophes in accordance with various embodiments of the
technology disclosed herein.
[0118] FIG. 49 is a diagram illustrating an example of a CONOPS
2100 for a weather station such as a C2 Weather Sensor System
(C2WS2) in accordance with one embodiment of the technology
described herein.
[0119] FIG. 50 is a diagram illustrating a cross-domain DT2
structure in accordance with various embodiments of the technology
disclosed herein.
[0120] FIG. 51 is a diagram illustrating an example of an RF TOE
key implemented as an RF proximity key in accordance with one
embodiment of the technology disclosed herein.
[0121] FIG. 52 is a diagram illustrating an example of a weather
station cartridge in accordance with one embodiment of the
technology disclosed herein.
[0122] FIG. 53A is a diagram illustrating an example of a
return-to-maximum procedure, including a decreasing U-value.
[0123] FIG. 53B is a diagram illustrating an example of a
return-to-maximum procedure, including an increasing U value.
[0124] FIG. 54 is a block diagram illustrating an example system
including a transmit/receive physical layer and a wireless (or
wired) Sensor Star Communication Interface, which may be configured
to perform compression and decompression as well as OVH-operations
such as, for example: IA, FEC (Forward Error Correction), cipher,
and others, within Data Transfer System (DTS).
[0125] FIG. 55 is a diagram illustrating an example of a necessary
condition of probability of a no-error per data stream.
[0126] FIG. 56 is a diagram illustrating an example of the "two
nines" criterion is illustrated.
[0127] FIG. 57 is a diagram illustrating an example application of
an unmanned operation using two IA keys with time synchronization
in accordance with one embodiment of the technology disclosed
herein.
[0128] FIG. 58 is a diagram illustrating an example of a typical
relation for normal dispersion.
[0129] FIG. 59 is a diagram illustrating an example of dotless time
synchronization in accordance with one embodiment of the technology
disclosed herein.
[0130] FIG. 60 illustrates an example computing module that may be
used in implementing various features of embodiments of the
disclosed technology.
[0131] The figures are not intended to be exhaustive or to limit
the invention to the precise form disclosed. It should be
understood that the invention can be practiced with modification
and alteration, and that the disclosed technology be limited only
by the claims and the equivalents thereof.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0132] Embodiments of the technology disclosed herein is directed
toward a devices and methods for providing weather condition
sensing and reporting, including from remote locations. Further
embodiments of technology disclosed herein include embodiments of a
deployment mechanism for instrumentation, including weather sensing
instrumentation. Still further embodiments of the technology
disclosed herein include weather event detection systems and
methods as well as data privacy and communication technology.
[0133] As previously discussed, one of the major limitations to
current weather station designs is the use of separate mounting
arms to allow isolation of each component implemented in the
system. While embodiments can be implemented that use separate
mounting arms, other embodiments can be configured to avoid such
structures. For example, various embodiments of the technology
disclosed herein can include a novel axial configuration that may
be used to stack different sensor sets along a single vertical axis
or otherwise vertically stack the components.
[0134] FIG. 1 is a diagram illustrating an example of stacked
sensor sets in accordance with one embodiment of the technology
disclosed herein. In this example embodiment, the station includes
an upper part 100 and lower part 105 separated by an air gap 110.
The air gap 110 between the upper and lower portions of the device
provides a shaded area for measurement of ambient temperature and
humidity, and an air channel to allow for measurement of wind speed
and direction. The upper part 100 and lower part 105 in this
example are connected together by supports 112. The supports 112
could be made as an element of the upper part 100 or the lower part
105, or they could be separate components. Although 2 supports 112
can be seen in the example of FIG. 1, any of a number of supports
can be used to separate upper part 100 and lower part 105 by a
desired predetermined distance. The distance of separation can be
chosen keeping in mind the goals of air gap 110, which can include
providing adequate shade for ambient temperature and humidity
measurement, and providing adequate space between upper part 100
and lower part 105 to allow sufficient airflow for measurement of
wind speed and wind direction. The quantity, size and shape of
supports 112 can also affect airflow in air gap 110, and can thus
be chosen with this in mind.
[0135] Upper part 100 and lower part 105 can be fabricated using
any of a number of different fabrication techniques including, for
example, injection molding process. In other embodiments, the upper
part 100 and lower part 105 could be machined or printed using 3D
printing methods, or other well-known manufacturing techniques
could be used. The lower part 105 may have a mounting means 113 for
accepting or attaching desired mounting elements. For example, in
some embodiments, mounting means 113 could be implemented as a hole
or cavity at the bottom of lower part 105. In some embodiments, the
cavity could include threads such a mounting element (e.g., a
tripod) could be screwed into the cavity to attach the device to
the mounting element. As a further example, the cavity could
include a 1/4'' threaded hole that would allow mounting the device
on commercially available tripods. This example illustrates, the
cavity can have a coupling that is complementary to a coupling
element of the mounting element such that the mounting device can
be fixedly or removably attached to the meteorological device. For
example, friction fit, screw fit, snap fit, socket/pin fit and
other mounting and fitting configurations can be used.
[0136] In the illustrated embodiment, the mounting means 113 is a
cavity that is used to accept a pin at the top of a spike 115. In
such a configuration, Spike 115 can be secured to bottom part 105
such that the device can be secured by driving Spike 115 into the
ground, into a tree, or in some other mounting location. In still
further embodiments, the mounting means could be used to mount the
weather device to a permanent structure (e.g., a building, a
bridge, and so on), a moving vehicle, a manned or unmanned
aircraft. In other embodiments, the mounting means could allow the
weather device to be secured to additional larger auxiliary sensors
or allow smaller auxiliary sensors to be mounted to the larger
weather device. In yet other embodiments, the weather device could
be mounted in a deployment mechanism such as, for example, that
described below with reference to FIGS. 18-23.
[0137] The upper part 100 may have an upper surface element 118
attached to the top of upper part 100. This upper surface element
118 could include, for example, a printed circuit board (PCB). The
upper surface element 118 could be used to mount sensor elements,
including skyward facing sensor elements (125, 130, 135, and 140 as
shown in FIG. 1) for purposes such as, for example, meteorological
condition sensing. These weather condition sensors can include, for
example, illumination measuring sensors to measure sunlight,
moonlight, or lighting in general; precipitation measuring sensors;
impact measuring sensors; cloud height sensors; visibility sensors;
lightning measurement sensors; and other sensors that can be used
to identify and measure weather conditions as they occur.
[0138] The device may also include solar panels 127 or other like
photovoltaic cells to receive solar energy and convert it into
electricity for operation of the weather station device. In the
illustrated example, solar panels 127 are mounted on upper surface
element 118 to provide exposure to the sun or other sources of
optical energy. In other embodiments, other power sources can be
used in place of or in addition to solar panels 127. For example,
battery cells can be provided to power the device as can other
renewable sources of energy such as, for example, wind energy.
[0139] In some embodiments, the meteorological measurement device
can utilize a novel method to measure precipitation type and amount
using a solid-state pressure or force sensor to sense the force of
precipitation such as, for example, falling rain or hail. In other
embodiments, an accelerometer can be used to sense the force of
precipitation. For example, a piezo or other like sensor can be
used to generate electrical current in response to impact made by
precipitation impinging on the sensor. The electrical current can
be proportional to an amount of displacement of the piezo electric
material, which can be used to provide information in addition to
merely the presence or absence of precipitation.
[0140] In addition, a moisture sensor can be provided and can
include metal electrodes protruding from a hydrophobic material.
The electrodes can be used to detect moisture by detecting
increased electrical conductivity between the electrodes. The
following scenarios illustrate examples of how a force sensor and
conductivity sensor can be used to detect precipitation. When the
solid-state force sensor detects impacts by the precipitation and
the moisture sensor detects moisture, the device reports the
presence of rain. When the solid-state force sensor detects impacts
having the singature of precipitation and the moisture sensor
detects no moisture, the device reports the presence of hail. When
the solid-state force sensor detects no impact but the moisture
sensor detects moisture, the device reports the presence of
drizzle. Embodiments can be implemented to: Analyze the impact
signature to allow rain, hail, small hail, and drizzle to be
distinguished; use electrodes embedded in hydrophobic plastic to
enable detection of active moisture in the air rather than standing
water; use the entire system body for detection to allow for a
radical size reduction compared to conventional approaches.
[0141] To estimate the amount of precipitation, the device measures
the quantity and size of impacts and multiplies the collective
total impact amount by a coefficient that depends on whether the
type of precipitation is rain or hail. To estimate the size of hail
stones, the device assesses the maximum impact strength recorded.
Other embodiments can calculate the average or medium impact. Other
computations can be used as well.
[0142] In another embodiment, the method for measuring the quantity
and force of impacts by falling rain or hail can be implemented
using one or more accelerometers mounted to one or more surfaces
that are exposed to such precipitation. For example, mounting
accelerometers to the enclosure of the MWS enables detection of
precipitation over a large area using a micro-size component.
[0143] In various embodiments, one or more accelerometers can be
mounted beneath upper surface element 118. For example, an
accelerometer can be mounted on either side of the PCB without
having a substantial adverse effect on measuring performance. For
example, it could be mounted on the bottom side of upper surface
element 118, so that it does not otherwise take up space they can
be used for top side electronics and components such as sensors
that rely on external exposure to sense whether parameters. The
accelerometer setting may be sensitive enough to detect the
vibration caused by rain or hail impacting any surface of the top
exposed PCB or impacting any exposed component solidly attached to
the PCB.
[0144] The quantity and strength of impacts are both detectable by
the accelerometer and measurable with the signals output from the
accelerometer. Analyzing the output signal to look for particular
impact characteristics, such as vibration frequency and amplitude,
can be done to filter out and disregard output from the
accelerometer that is caused by non-precipitation sources, such as,
for example, wind turbulence and physical movement. In some
embodiments, the output can be digitized and analyzed by a DSP or
other processing device to determine whether the impacts are due to
precipitation and to identify or classify the precipitation.
[0145] Because the entire upper surface element 118 (including the
components mounted thereon) can be used to detect precipitation,
this technique offers a potentially larger surface area to provide
detection of precipitation than does a dedicated Piezo (or other)
element that would have to share space with other sensors mounted
on the printed circuit board. Accordingly, this can result in more
accurate precipitation estimates. In yet another embodiment,
accelerometers can be mounted on support elements 112 to likewise
detect precipitation impacting upper part 100.
[0146] Yet another alternative for measuring the quantity and force
of impacts from precipitation uses a microphone or other like
element to sense the impacts on the external surfaces of the
device. A sound analysis module can be used to analyze the
microphone output to detect and discern the sound of precipitation
impacting upper surface element 118 and components mounted thereon
from other sounds. For example, the microphone output can be
digitized and a DSP or other processing device used to analyze the
output to identify certain defined characteristics that indicate
that detected sounds are from the impact of rain drops or hail
stones on the device, and to classify the type of precipitation.
Accordingly, the system can filter out and ignore detected sounds
that are caused by non-precipitation sources such as wind and
movement.
[0147] Whether by analog or digital (e.g., DSP or other processing
device) means, a module can be included to quantify the quantity
and force of impacts and evaluate the impact signature to
distinguish between classify precipitation as rain, hail, small
hail, drizzle, etc., and to filter out other impacts as
non-precipitation impacts.
[0148] These and other various approaches for detecting and
quantifying the presence of precipitation can be used as a sole
sensor solution or in combination with other impact measurement
solutions for precipitation sensing. The combination of several
precipitation sensing solutions can be used to improve the
reliability of detection and hence proper weather reporting.
[0149] Although any of a number of different moisture sensors can
be used, as noted above one embodiment employs electrodes embedded
in a hydrophobic material. The hydrophobic material encourages the
transport of liquid away from the electrodes. Accordingly, the
electrodes can remain dry during dry conditions, yet when any
moisture is dropped on the sensor such as from precipitation, the
sensor can respond to the moisture and produce a detectable
electrical output signal indicating the presence of ongoing
precipitation.
[0150] In various embodiments, the precipitation sensing system can
be configured to use the following signals: [0151] Disdrometer
Average--Integrates both the quantity and strength of impacts of
precipitation elements. [0152] Disdrometer Peak--Measures the
impact strength independent of quantity. [0153] Moisture
Sensor--Measures falling moisture of either rain or snow. [0154]
Temperature Sensor--Suggests whether falling moisture is rain or
snow.
[0155] With one or more sensors, the following precipitation
conditions can be distinguished: [0156] No Precipitation: No signal
is present on the disdrometer or moisture sensor. [0157] Rain: Both
the disdrometer and the moisture sensor detect a clear signal. The
average disdrometer signal measures the quantity. [0158] Hail: The
disdrometer registers a signal (impacts) but the moisture sensor
remains untriggered. The peak signal distinguishes it from ice
pellets or snow pellets. [0159] Ice Pellets or Snow Pellets: The
disdrometer registers a signal (impacts) but the moisture sensor
remains untriggered. The peak signal distinguishes it from hail.
[0160] Rain: The disdrometer registers no signal (no impact) but
the moisture sensor detects falling moisture and the temperature is
above 3.degree. C. [0161] Snow, Snow Grains, or Ice Crystals: The
disdrometer registers no signal (no impact) but the moisture sensor
detects falling moisture and the temperature is below -3.degree.
C.
[0162] With the hybrid sensor system in some applications, snow and
drizzle may be difficult to distinguish from one another in the
zone between -3.degree. C. and 3.degree. C., where snow can turn to
drizzle and vice versa with slight changes in elevation and
temperature. A second challenge is that the structural character of
snow forms (snow, snow grains, and ice crystals) and small hail
forms (ice pellets and snow pellets) may be difficult to
distinguish without including optical detection and analysis
modules. Nonetheless, even without optical detection and sensing,
the system can be implemented to provide reliable and distinct
determinations of six precipitation categories, including the "no
precipitation" state.
[0163] Embodiments of the meteorological sensing device can utilize
forward-scattering modules to estimate visibility, as is a common
technique understood in the art. However, in some embodiments, the
device uses a novel system and method to perform this measurement
using a low-power, optical light emitting diode (LED) transmitter
and a large-area photo-diode optical detector. The optical
transmitter may be modulated on and off under microprocessor or
other module control. The detector may be digitally sampled. With
such apparatus, forward scattering can be measured by subtracting
the sum of the detector output when the transmitter is off from the
sum of the detector output when the transmitter is on. In some
embodiments, a single LED transmitter can be used to limit the
amount of power consumed by the sensor while the detector uses an
array of photodiodes for increased sensitivity.
[0164] In various embodiments, the weather sensing device includes
a module configured to calculate the cloud layer height. In some
embodiments, the measurement is made using a laser based range
finder approach. A pulsed laser source can be used to transmit a
light pulse up toward the cloud layer. The light pulse reflects off
the cloud layer and is returned to the weather station, which can
also include a receiver to detect the reflected light pulse. A
computing module measures the time of flight of the laser light
pulse from the laser transmitter to the receiver and uses this
time, based on the speed of light, to calculate the distance to the
cloud layer. Other similar optical or sonic ranging techniques can
be used. In various embodiments, the laser system can be
implemented using a microchip laser with a passive Q switch that
provides the high peak power of the laser pulse but requires very
low electrical power. Typically, peak power will be in kilowatts.
In some embodiments, the system can use a miniature pulsed laser of
higher energy, i.e. hertz, that can operate in low frequency
operating mode and can reach even higher peak powers. Typically,
the even higher peak power can be in megawatts. The laser emitting
aperture can be different from the detecting aperture to improve
signal to noise ratio.
[0165] The receiver used to detect the reflected light pulse can be
a high-speed detector with high sensitivity. In a preferred
embodiment, the detector is an avalanche photodiode. In other
embodiments, the detector can be a photomultiplier tube or
photodiode. Optical filters can be used (e.g. bandpass filters at
the bandwidth of the laser light) to filter out unwanted noise
(e.g. ambient light) to prevent noise signals from impinging on the
detector. This can improve the signal-to-noise ratio of the system
and hence, the detectability of the ranging pulses.
[0166] The solar panels 127 can be arranged across the upper
surface element 118 in a pattern that minimizes shadowing of the
panels by the various sensor components mounted on and extending
above upper surface element 118. Circuitry can be included with the
photovoltaic elements 127 to allow electric power from illuminated
elements to be collected while blocking electricity drainage by
photovoltaic elements that may be in a shadow. The upper surface
element 118 preferably should have four or more solar panels 127.
In another embodiment, the solar panels 127 can be arranged in a
vertical or semi-vertical orientation in order to increase
efficiency of solar energy capture at higher latitudes. For
example, solar panels can be configured vertically on lower part
105.
[0167] An antenna 140 can also be mounted on upper surface element
118. This antenna 140 could be used for receiving transmit signals
between the weather device and external devices. Examples of
received signals can include operating commands for the weather
station from a control center or base of operations. For example,
an operator (e.g. a human operator or computer control device) can
send commands to the weather station to power on or off certain
functions, to change operational parameters, or to otherwise affect
the function or operation of the system. Such signals can be sent,
for example, through a satellite network such as, for example, the
Iridium satellite network and via GPS satellite signals. Examples
of signals transmitted by the weather station device can include,
for example, weather sensor readings, camera images, and GPS
coordinates for weather station.
[0168] Imaging systems can be included as well to enhance the
performance and capabilities of the weather station. The use of
imaging for weather condition analysis can substantially improve
awareness of the weather observation personnel and can also be used
to assist in troubleshooting or verifying the performance of
complex sensor sets. Imaging systems can also supplement the
information received by the sensors and may be used to verify the
information received by the sensors. Imaging capabilities
implemented in the weather station device can therefore, in some
embodiments, be implemented to improve the reliability of the
remotely acquired data and can help to prevent or reduce false
readings, tampering and damage related sensor misreading.
[0169] Accordingly, imaging systems can be implemented, in some
cases with multiple cameras, to allow images to be taken in
multiple directions, digitally compressed, and transmitted (e.g.,
via satellite) from the weather station device at its remote
location. In the embodiment illustrated in FIG. 1 an example camera
module 120 is included. This example camera module 120 includes
four cameras 125 (three can be seen in the figure) facing in four
different directions. Particularly, in this example, the cameras
are arranged to be facing at 90.degree. intervals from one another.
The use of four cameras is not a requirement; the device may employ
any quantity of cameras, but this quantity may be limited in some
cases by size restrictions of the weather station device or data
bandwidth constraints of the communications link.
[0170] In some embodiments, additional memory or data storage can
be provided to allow images to be taken in excess of available
bandwidth and stored for later transmission. In this manner,
additional images of weather events can be captured and stored for
transmission such a system is not purely bandwidth constrained. In
various embodiments, regardless of whether such additional memory
is included, sufficient buffering is provided to allow capture,
processing (if any) and transfer of images without unduly burdening
the system by processing and bandwidth constraints.
[0171] In some embodiments, the cameras may be oriented such that
their viewing angle is slightly above horizontal. With such an
orientation, images obtained from the cameras can include both the
surrounding terrain and the sky. Typically, with weather station
applications, it is desirable to be able to view the sky to enable
the capture of current weather events and the possible prediction
of future weather events.
[0172] In various embodiments, the one or more cameras included
with the weather station can be mounted on an adjustable mount so
that their viewing direction and angle can be changed. For example,
in the illustrated embodiment, camera module 120 can be configured
to rotate about its axis so that the cameras can scan beyond their
normal horizontal field of view. Additionally, the cameras
themselves can be adjustable in azimuth, elevation, and zoom so
that the scene detected by the cameras can be adjusted and chosen
for a given situation. Optical zoom, electronic zoom, or a
combination of both optical and electronic zoom can be used to
change the focal length of the camera or cameras.
[0173] In the illustrated embodiment, the four cameras 125 are
arranged to be facing 90.degree. apart about the plane of the
panorama. Imaging of the sky allows for verification of the sensor
readings and enables an operator to visually verify weather
conditions. For example, imaging the clouds may be used to identify
cloud coverage and determine the presence, direction and the type
of an incoming storm. Typical up tilt angles for the cameras could
be from 5 to 45 degrees with optimal angle around 10 to 20 degrees.
Other angles above and even below the horizon can be used.
[0174] With the proper field of view (horizontal), the cameras can
be configured to cover 360 degrees of view around the weather
station device. In some embodiments, the cameras may have gaps in
the field of view, which may be acceptable for various
applications. For example, in one embodiment four cameras with
60-degree field of view can be provided, covering 240.degree. of
the total 360.degree. degrees of surround with 30-degree gaps
between the images of each camera. Other angles and gap spacing may
be used in different embodiments and overlapping fields of view can
also be provided.
[0175] The cameras may be mounted in different ways. In some
embodiments, X-Y or azimuth-elevation mounts can be used to mount
one or more cameras on the weather station to provide control of
the directionality of the cameras. In other embodiments, the
cameras, including those configured as illustrated in FIG. 1, may
be mounted on a single rigid-flex or completely flexible PCB to
allow flexibility in mounting. For example, the flexible printed
circuit board can be folded or bent into the desired shape to
provide proper camera orientation. This can also, in some
embodiments, facilitate a compact shape for the module. A
rigid-flex or completely flexible PCB can also be implemented to
avoid requiring wire interconnects or board-to-board connections
between multiple separate PCBs (e.g. one for each camera).
Accordingly, this can provide ease of manufacturability and a
compact size and shape. This mounting also allows for the
manufacture of a single panorama camera system PCB, saving
materials, cost, and time as compared to embodiments using multiple
printed circuit boards.
[0176] In some embodiments, the cameras are oriented with inclines
above the horizontal plane to achieve an image with more view of
the sky than the ground. In the embodiment illustrated in the
example of FIG. 1, the PCB board is folded into the shape of a
four-sided truncated pyramid, with one camera on each face of the
pyramid. In this example, each camera is positioned at the same
corresponding location on each respective face of the pyramid.
Other shapes may be used in different embodiments, including a
conical shape with a circular base. In addition, the shape of the
board may be modified to allow for the use of a different number of
cameras and different orientations.
[0177] In some embodiments, supporting electronic components and
circuitry are mounted on the flexible or rigid-flexible PCB, such
as on the pyramids faces. In other embodiments, the supporting
electronic components and circuitry are mounted on a "tail" section
of the PCB board, where the tail section extends off the bottom of
one or more of the pyramids faces. This tail may be foldable out of
the way of the pyramid, either by folding the tail under the
pyramid or to the side of a bottom edge. A shared common bus can be
included to electrically connect all the camera sensors, thereby
reducing the number of signal traces that must be routed. The
shared bus extends from the camera on the "end" pyramid face,
through each of the adjacent pyramid faces, and to the supporting
electronics on the tail section. Circuitry can be provided to allow
signals to be multiplexed onto this shared bus for data transfer.
Likewise, separate buses can be used for parallel data
transfer.
[0178] When the flexible or rigid-flexible PCB is used, the board
may be maintained in the folded truncated pyramid shape by being
mounted on a solid internal support structure. The support
structure can be made of any of a number of different types of
solid materials. Some example solid materials can include plastics,
metals, or wood. The support structure may be sized and shaped to
fit inside the folded PCB shape to contact and hold each pyramid
face securely in its corresponding location and orientations. The
structure can further include a lip that extends around the
underside and front of the bottom edge of each pyramid face to hold
that face in place against the internal structure.
[0179] The top edge of each pyramid face may be held in place in
some embodiments using a removable cap secured by a screw or other
fastener. The removable cap can be included to provide a
non-permanent mounting for the rigid-flex or flex PCB into the
support structure. In other embodiments, the flexible or
rigid-flexible PCB can be secured to a simple internal support
structure without a bottom lip and removable cap. This embodiment
may be secured by means, such as, for example, glue or epoxy,
screws, and other fasteners. As these examples serve to illustrate,
any of a number of different support structures and mounting
brackets can be used to secure the printed circuit board in place
in the desired shape or configuration.
[0180] The rigid-flex or flex PCB can be designed to orient the
cameras at other inclines or declines by adjusting the dimensions
and shape of the PCB so that it folds into a one of the variety of
potential shapes with the desired incline or decline tilt. The
height of the PCB board can be sized to provide space required for
the components and circuitry. As an illustration, a 0.degree.
incline would result in a folded shape with vertical faces and may
include a vertically uniform footprint or horizontal plane shape
that is constant from the bottom through to the top. Using a fully
flexible circuit with a 0.degree. incline of mounted components can
also be implemented using a PCB formed into a cylindrical shape
with a circular base. For incline angles above horizontal, a PCB
formed into a conical shape with a circular base can also be used.
As these examples serve to illustrate, there are number of
different PCB configurations that can be provided to allow mounting
of cameras at desired positions and orientations.
[0181] The use of cameras, including image sensors and optics (e.g.
lenses), mounted on rigid-flex or flex PCBs is not limited to
applications for capturing panoramic images or images about
360.degree.. The rigid-flex or flex PCB can be designed to be bent
or folded to orient any number of cameras in any position and any
direction. This maintains the benefits of manufacturing and
assembling a single or multi-camera system (with specific placement
and orientation requirements) all as one PCB unit.
[0182] The use of a rigid-flex or flexible PCB is not limited to
applications for mounting imaging devices. Indeed, in various
embodiments, other sensors or other weather station components can
be mounted on flexible printed circuit boards or rigid-flex printed
circuit boards to accomplish the same or similar features as
described above with respect to the cameras. Accordingly, the use
of rigid-flex or flex PCBs for sensor mounting can allow mounting
of sensors and other components in desired positions and
orientations as described above. In some embodiments, other sensor
components therefore can be mounted on a rigid-flex or flex PCB
with the truncated pyramid, cone, or vertically uniform shape as
described above.
[0183] The upper part 100 may also include a lower surface element
143, which in the example of FIG. 1, is shown as being mounted on
the lower side of the upper part 100. This lower surface element
143 could be, for example, a PCB to which components of the weather
station may be mounted. Because lower surface element 143 is below
the "canopy" of upper part 100, or surface element 140 may in some
embodiments be used to mount elements or components of the weather
station that can be employed without requiring an upward facing
orientation (e.g. they do not require direct sun or precipitation
exposure has available on the top surface). Such elements could
include, for example, sensor elements such as humidity sensor 145
and wind sensor 150, for example, which do not require
precipitation or direct sun exposure that is available to
components on the upper surface element 118. Other examples of such
sensor elements include a temperature sensor, a pressure sensor, a
thermal wind speed and thermal wind direction sensor (including,
for example, wind sensors employing hot element or "hot wire"
functionality as further described below).
[0184] The air gap between the upper and lower portions of the
device provides a shaded area for measurement of ambient
temperature and humidity, and an air channel for measurement of
wind speed and direction. Although not illustrated in FIG. 1,
sensors such as, for example, micro sensors for pressure, magnetic
orientation, tilt, position (e.g. GPS), and a microprocessor can be
included in internal portions of upper part 100.
[0185] The open area between the upper part 100 and the lower part
105 can be configured to provide an area for wind sensing by
thermal, mechanical, sonic, or other methods, and which can be
heated to avoid accumulation of freezing rain, snow, or other
precipitation. In some embodiments, the wind sensor 150 includes a
thermal element 154 surrounded by temperature sensors 152 arranged
in a pattern (only two shown). Preferably, temperature sensors 152
are arranged in a ring pattern evenly spaced equidistant from
thermal element 154, but other patterns and arrangements can be
used. The preferred quantity of temperature sensors is four or
greater, but other quantities of temperature sensors 152 may be
employed in various embodiments. In some embodiments, a constant
energy is applied to thermal element 154 to raise its temperature,
and reductions in temperature of the thermal element 154 due to
removal of the heat by wind is used to calculate wind speed. The
temperature of the air around thermal element 154 is measured by
temperature sensors 152, which are arranged in a pattern about
thermal element 154. Analysis of the temperature readings from the
various temperature sensors 152 can be performed to determine a
pattern of air flowing from the thermal element 154 to the various
temperature sensors 152. Heat generated by thermal element 154 is
carried by the wind to one or more temperature sensors 152 that are
downwind from thermal element 154. Accordingly, changes in absolute
temperature readings for temperature sensors 152 or relative
temperature readings among temperature sensors 152 can be used to
identify wind direction. For example, if the northernmost
temperature sensor 152 is reading at a higher temperature relative
to the remaining temperature sensors 152, this indicates a
southerly wind carrying the heat from thermal element 154 to the
northernmost temperature sensor 152.
[0186] Accordingly, the temperature at each of the temperature
sensors 152 is measured to determine the direction of greatest
temperature increase, or to determine the temperature differential
among sensors. The sensor measuring the highest temperature is in
the downwind direction of the wind. In other embodiments, absolute
changes in temperature for each sensor are measured as the heat
source is cycled from the on state to the off state (or vice versa)
and the sensor with the greatest change in absolute temperature is
in the downwind direction of the wind. Accordingly, in various
embodiments, the system is configured to measure only the
difference in temperature between cycles, or the difference in
temperature among sensors, and may therefore be insensitive to
absolute temperature.
[0187] Thermal element 154 can include a resistive heating element
such as, for example, a resistor, resistive wire or resistive
heating element. Current can be supplied to thermal element 154 to
increase its temperature above ambient temperature. To conserve
power, algorithms can be employed to determine the amount of
current needed to raise the temperature of thermal element 154
sufficiently above ambient temperature to allow operation of the
thermal sensor.
[0188] The thermal sensors, for example, can comprise small
negative temperature coefficient (NTC) thermistors mounted on
0.3-inch long pins. Thermal element 154 and temperature sensors 152
can, in some embodiments, be mounted such that they are at or near
the center of the air channel provided by opening 110 such that
they receive relatively unimpeded airflow. Additionally, in various
embodiments, the sensors are preferably smaller in diameter than
the separation distance between the sensors. Utilizing small
sensors reduces interference with air flow around the sensors.
Additionally, using smaller sensors typically yields a smaller
thermal mass, allowing the temperature of the temperature sensors
152 to rise and fall more quickly in response to heat generated by
thermal element 154. As seen in the example of FIG. 1, temperature
sensors 152 and thermal element 154 are mounted toward the center
of opening 110 to maximize the amount of shade from sunlight
provided by upper part 100. Shading from the sun can be desired so
that the operation of the sensors is not affected by movement of
the sun, direct impact by the sun's rays, or the effects of
changing cloud cover. For similar reasons, surfaces surrounding
wind sensor 150 can be made from, covered with or painted with
non-reflective materials.
[0189] As these examples serve to illustrate, with some
embodiments, thermal wind sensor designs can be implemented to
enable the determination of wind speed and direction using thermal
means that allow for reduction in size of the weather station
device and the elimination of moving parts as compared to
conventional mechanical and acoustic approaches. This can also
provide the benefit of eliminating binding that affects mechanical
devices due to dust and contamination or due to freezing or ice
accumulation.
[0190] In various embodiments, wind sensor 150 can include a module
employing software or other algorithms to compute wind speed and
direction. In some embodiments, a algorithm can be implemented to
first determine the temperature sensor 152 experiencing the largest
amount of temperature rise as compared to other sensors in the ring
of thermal sensors 152. From this, the approximate wind direction
can be determined. The algorithm can further be configured to
calculate a more precise wind direction by using a weighted average
of temperature changes among the thermal sensors 152. The algorithm
uses the measure of actual heat rise for each temperature sensor
152 over the average rise in temperature among the set of
temperature sensors 152. The method can be configured to also
employ an algorithm to compute the average wind speed and wind
direction by averaging the measurements over multiple measurement
periods. In some embodiments, the algorithm can be configured to
compute an average wind direction over multiple time periods only
when the multiple direction readings are within 90.degree. of one
another. The system can further be configured to compute parameters
for wind gusts by determining the highest wind speed recorded over
a given period of time and the wind direction at the time the
highest wind speed was recorded.
[0191] Below the wind sensing area in the example of FIG. 1, is an
ambient sensing area used for sampling ambient air for purposes of
determining for example humidity and temperature. The ambient
sensors, however, are not limited to being positioned below the
wind sensing area and can be placed in other locations about
chamber 100 including above or at the same level as the wind sensor
components. Preferably, the ambient sensors are not mounted in such
a way so as to interfere with the flow of wind about the wind
sensor components.
[0192] The weather station device can utilize high-accuracy,
solid-state thermal sensors to measure ambient air temperature. In
various embodiments, conventional off-the-shelf temperature sensors
can be used for this purpose. In many embodiments, the weather
sensing device can be configured to use a novel method to measure
the amount of solar heating present at the weather sensing device,
and to compensate for that solar heating to more accurately assess
ambient air temperature. In accordance with one embodiment, the
method relies on multiple temperature sensors. An external
temperature sensor exposed to the air, is provided preferably in a
shaded area. An internal temperature sensor (e.g., internal to the
body of upper part 100) to measure the temperature of the shaded
surface is also provided. The method determines the amount of solar
heating that is affecting the ambient temperature sensor and
subtracts this from the temperature measurement to arrive at a more
accurate or better approximation of the temperature measurement.
The amount of solar heating in one embodiment can be measured as
the internal temperature sensor reading minus the external sensor
reading. This provides an estimate of the amount of heat
contributed to the device by solar effects. This temperature
differential can be subtracted from the temperature measured by the
ambient temperature sensor to arrive at a more accurate estimation
of the ambient temperature.
[0193] From the analysis of temperature data, consistent behavior
was noted to assess the proper correction factor, which is found to
be very consistent for wind speeds greater than a knot. With the
technology disclosed herein, true air temperature can be calculated
by subtracting a fraction (e.g., as determined by device-specific
calibration) of the difference between the external and internal
readings from the external reading. The only deviation from the
consistent correction factor occurs for air that is very still
(less than 1 knot) for a prolonged amount of time. This is an
uncommon condition in most real-world applications.
[0194] The weather station device can also utilize capacitive
humidity sensors to measure ambient humidity. Capacitive humidity
sensors are readily available as off-the-shelf products. In various
embodiments, the device can be configured to use a novel method to
measure the amount of solar heating and to compensate for errors
introduced by solar heating in the humidity measurement. In various
embodiments, the method utilized one external temperature sensor
placed in a position to be exposed to the air, but in a shaded area
of the device. The method also uses an internal temperature sensor
that measures the temperature of the shading surface. The amount of
solar heating may be assessed as the internal sensor reading minus
the external sensor reading, multiplied by a correction coefficient
and added to the humidity reading.
[0195] The use of multiple sensors across the weather station
device, for example, multiple temperature, humidity and pressure
sensors on upper surface element 118 and lower surface element 143
can be configured to give the system the capability to reliably and
intelligently deduce or better estimate the true weather parameters
from multiple sensor readings in a small form factor. For example,
high intensity sun illumination would result in a heating
differential between the upper mounting surface element 118 and the
lesser heating of lower surface element 143. The difference in
temperature for the temperature sensors on upper mounting surface
element 118 and lesser heating of lower surface element 143 allows
deducing the heat flow across the weather station device and
calculating the true temperature reading without adding additional
sun shading for the weather station device.
[0196] The use of multiple pressure sensors allows reducing random
calibration variations for more accurate pressure readings and
provides redundant pressure reporting in the event one or more of
the sensors fails. Both interfaces of the upper mounting surface
element 118 and lower surface element 143 could be sealed.
Alternatively, upper mounting surface element 118 could be
hermetically sealed to upper part 100, for example, and the lower
mounting surface element 143 mounting can include gaps to prevent
condensation and moisture accumulation in the inner volume of part
100. The electrical connections of components of the upper mounting
surface element 118 and the lower mounting surface element 143 are
preferably sealed from environment using potting materials, such
as, for example, epoxy potting.
[0197] The weather observation unit can include additional
capabilities in various embodiments to expand the range of missions
and applications for the unit. These capabilities can be
implemented inside or integrated physically with the portable
weather station unit. In other embodiments, these capabilities can
be supplied as external add-ons connected to the device by a cable,
connector, or through a wireless link. The capabilities may
include, for example, the detection of radiation or nuclear
material with the inclusion of an ionizing radiation detector
module. Chemical detection capabilities can be implemented by
adding a chemical presence detection and identification unit.
Biological agent presence detection could be implemented by using a
specific biological agent detection and identification unit. In a
preferred embodiment, nuclear, chemical, and biological agent
presence detection may be done by sampling the environmental medium
such as air or water through the detector unit for accumulation and
analysis for improved sensitivity of detection.
[0198] The lower part 105 as well as upper part 100 can be powered
by solar panels 127 mounted on the upper part 100. Lower part 105
may additionally or alternatively include a battery and capacitors
mounted in the inner volume thereof. These power sources can be
used to power either or both lower part 105 and upper part 100.
Where batteries or capacitors are included with solar panels 127,
solar panels 127 can be used to recharge the capacitors and
batteries to allow operation during periods of little or no light.
The electrical connection from lower part 105 to the upper part 100
could be done through the mounting means 112.
[0199] Lower part 105 can be designed such that it is a separable
stand-alone power module in any of the power configurations noted
above. This power module can be designed with universal mounting
features on top, for example, to allow attachment of upper part 100
and create the weather station configuration defined in FIG. 1.
Alternatively, other sensors or sensor systems, such as the
extension module depicted in FIG. 3, can be mounted on top of lower
part 105 and be powered by its power system. Such integration of
the power module to other sensors is not limited to weather
sensors, but can include other sensors such as chemical,
radiological, biological, imaging, motion, etc. Such a
configuration enables a universal power module capable of
integrating with and powering any plurality of sensors.
Alternatively, the power module can be wired to the full weather
station, as depicted in FIG. 1, and provide additional backup power
to the main power system of the weather station.
[0200] Although the batteries or capacitors can be placed in either
or both the upper part 100 and lower part 105, placement in lower
part 105 lowers the center of mass for the weather station device.
Having a lower center of mass improves the deployability of the
unit for air-drop deployments, such as those in which the weather
station device is dropped from an airborne platform to a desired
location. A lower center of mass for the weather station device
would allow the system to stabilize its direction and orientation
during the descent phase of air-drop deployment for proper
orientation and placement in the ground.
[0201] A power switch 155 can be included to turn the weather
station device on and off. Although power switch 155 is illustrated
as being positioned on lower part 105, power switch 155 can be
located elsewhere on the weather station device. Also, the weather
station device may include a local connection port 160 and an
extension port 165. Although illustrated as being located on lower
part 105, such ports can be placed on either upper part 100 or
lower part 105.
[0202] The local connection port 160 can be used for data transfer
and to supply electric power to the weather station device from an
external battery, solar cell or other power supply. This external
power could be used to charge the weather data station device's
batteries or to power the device during operation. Because the port
can be used for data transfer, it can also be used to verify the
device's performance after activation. Additionally, this port can
be used to receive sensor data from the weather station device
using a hardwired connection.
[0203] The extension port 165 can be used to connect additional
modules such as, for example, sensor modules or other devices to
the weather station device to augment its functionality. One
example of such a module is an enhanced LIDAR system that can be
used for visibility and cloud height measurement. Additional
functionality can include, for example, modules to provide the
capability to detect wind speed and wind direction above the
weather station device using sound or light, such as sonic
anemometer or laser anemometer. Another example is the use of a
module to detect the presence of hazards such as radiological
hazards, nuclear material hazards, chemical hazards or biologic
hazards. In various embodiments, external modules connected through
extension port 165 can include their own power sources such as, for
example, batteries or solar panels, and can also be configured to
exchange electric power with (from or to) the weather station
device.
[0204] Using battery power to the exclusion of solar panels 127
frees up the use of the entire top surface of the device for
placement of different sky-facing sensors--such as LIDAR,
precipitation sensors, and photo sensors, and so on--without the
added size and weight of additional mounting arms as in
conventional solutions. Accordingly, various embodiments can be
implemented considering trade-offs between unit size, real estate
available for multiple sensors, and the long-term availability of
power through the use of solar panels.
[0205] Existing weather sensors are much heavier than may be
desired for remote deployability and are much too large to be
easily carried. The typical configuration used in existing weather
station consists of individual components mounted using separate
arms suspending the individual components at some distance from one
another. This is done so that the sensors do not interfere with one
another and their various effects can be isolated. In contrast,
with the current weather station system, an axial configuration
(e.g., such as that shown in FIG. 1) can be utilized in various
embodiments to provide different sensor mounted in a vertical
configuration or otherwise vertically orient the component
arrangement.
[0206] In the example illustrated in FIG. 1, the sensors are
organized and arranged into specific groups mounted on upper and
lower portions of the device so that the remote weather system can
be provided as a single unit instead of multiple components
connected through cables or mounted on arms and extensions. With an
integrated package such as the example shown in FIG. 1, embodiments
can be implemented to: eliminate multiple setup steps otherwise
required to assemble components in a conventional system; eliminate
large, bulky towers; reduce the overall system profile to reduce
wind loading; reduce cost by eliminating multiple enclosures and
interconnections; enable potential air-droppable capability;
improve ruggedness and durability.
[0207] This configuration of sensors as shown in FIG. 1 allows for
sky-facing sensors such as LIDAR, precipitation sensors, and photo
sensors to be placed at the top of the device and to use all (if
needed) available area on the top surface for sensors and
photovoltaic cells. Below the top surface section is the open area
used for wind sensing by either thermal, mechanical, sonic, or
other wind sensing systems and which can be heated to avoid
accumulation of ice or freezing rain. Below the wind sensing area
is an ambient sensing area that can be used to mount sensors for
ambient air sampling such as humidity and temperature sensors. This
overall arrangement allows sky-facing sensors, heated sensors, and
ambient sensors to all exist as an integrated unit within a single
compact device rather having such components separated on a
sprawling device with multiple arms and extensions. The arrangement
also allows for variations such as combining the wind sensing and
ambient sensing areas as a single area if the wind sensor requires
no heating or if the area is large enough such that when sensor
heating does not impact the ambient sensors.
[0208] The weather sensing device can utilize this stacked, axial
configuration with a combined area for wind and ambient sensing to
meet stringent size and weight requirements. The device also
utilizes most of the upper surface for solar cells to recharge
batteries contained in the lower portion of the device; the
remainder of the upper surface may be used for the satellite
communications antenna, the lightning detection antenna, optical
ambient light sensors, a moisture sensor, a precipitation force
sensor, and a panoramic camera system. The air gap between the
upper and lower portions of the device provides a shaded area for
measurement of ambient temperature and humidity, and an air channel
for measurement of wind speed and direction. Sensors for pressure,
magnetic orientation, tilt, GPS, and a small microprocessor can be
included internal to the device such as, for example, in the upper
or lower portion of the device.
[0209] In various embodiments of the weather sensing device, the
power system can be configured to utilize constant current charging
from the solar panels to charge lithium-ion cells or other energy
storage devices in parallel. A low-power linear regulator power
supply can be included to maintain continuous operation of the
microprocessor system even in the sleep mode, while separate
switched mode supplies can be included and activated under
microprocessor control. Supercapacitors in parallel with the solar
cells can be included and used to deliver surge currents drawn from
the switched-mode supplies that would otherwise exceed the
current-delivering capability of the small battery pack.
[0210] In embodiments using supercapacitors, additional modules can
be used to manage the supercapacitors. One module can include a
charge dissipation circuit that automatically discharges the
supercapacitors and turns off the device when the power switch is
placed in the off position. Another module can be included and
implemented to prevent high in-rush currents from occurring when
the device is switched on and the batteries are connected to the
supercapacitors. In one embodiment, the circuit can be implemented
as a soft-start circuit that uses a high-power MOSFET (Metal Oxide
Field Effect Transistor) driven by an RC (resistor-capacitor)
network connected to the discharge load. This circuit can be
implemented to limit current as the supercapacitors charge to the
level of the batteries and for the connection to be of minimal
electrical resistance when the device is fully on. This circuit can
also be configured to reduce spark wear on the mechanical switch as
well as the surge current stresses that would otherwise occur.
[0211] A number of additional features may also be included to
protect the electronics of the system from environmental conditions
for outdoor use, and, in some embodiments, particularly from rain
and other forms of precipitation or condensation. Although sealed
enclosures are typically used to achieve this type of protection,
the weather station system can, in some embodiments, include one or
more various features to achieve the same or a similar level of
protection without the bulk of an enclosure. To protect the device
without requiring excessive weight, the weather station device can
be implemented to use un-enclosed printed circuit boards on which
exposed sensors, and solar cell elements can be directly mounted.
Antennas can also be directly mounted on an exposed print circuit
board and epoxy encapsulation can be provided to shield the
antennas from environmental exposure. The absence of gaskets and
seals prevents moisture from becoming trapped within the enclosure
and can also avoid problems associated with pressure variations
such as may arise due to changes in elevation. Additionally,
drainage holes can be included to prevent the build-up of moisture,
which can be detrimental if the weather station device later
encounters freezing conditions, which can cause water expansion due
to the phase change.
[0212] In various embodiments, the device can include a completely
self-contained power system. Power demands in a device with a small
form factor are therefore ideally reduced or minimized. In many
embodiments, to achieve reduced power consumption, the system is
configured to power down all components. In further embodiments,
all components except for the microprocessor are powered down, and
the microprocessor is placed in a low-power sleep mode. The
microprocessor can be configured to keep track of time in sleep
mode and wake-up at predetermined intervals for scheduled
measurements. At the time for scheduled measurements, the
microprocessor awakens of sleep mode and powers on the sensors and
other components of the device used for the scheduled measurement.
In some embodiments, the microprocessor can be configured to power
on only certain sensors that are used for particular measurements
that are scheduled for a given time interval. In other embodiments,
the microprocessor can be configured to power on the entire weather
sensing device for measurements.
[0213] Accordingly, in some embodiments, the system can be
configured to awaken at periodic intervals (e.g., every twenty
minutes, every hour, every three hours, or other predefined
intervals, regular or otherwise) to conduct scheduled measurements.
In various embodiments, the intervals can be programmed via the
communication interface (e.g., by messages received through an
Iridium satellite transmission). In other embodiments, the system
can be configured to wake-up at scheduled times to conduct
measurements scheduled for those particular times. In some
embodiments, the system can be programmed to awaken at given times,
take particular measurements at the various awake times, and
transmit the results to a base of operations (e.g., via satellite
relay). In some embodiments, the data can be transmitted as it is
being measured or at the end of each awake time, while in other
embodiments, the data can be transmitted in batch form after a
series of measurements are taken over a plurality of awake
times.
[0214] Accordingly, in some embodiments, the system can be
configured to wake multiple times (e.g., at regular intervals or as
scheduled) leading up to a designated weather report. For example,
in one embodiment, the system is configured to wake four times, at
two-minute-and-thirty-second intervals during a ten-minute window
immediately preceding a weather report. In some embodiments, for
example, the timing is structured so that the weather messages are
sent at roughly 55 min after the hour in hourly mode or every 15,
35, and 55 min after the hour in continuous mode. In three-hour
mode, the unit can be configured to follow the hourly timing but
only during hours 0, 3, 6, 9, 12, 15, 18, and 21. In other
embodiments, other timing factors can be used.
[0215] Firmware in the device can be included to control the
wake-up and sleep behavior as well as the operational modes:
continuous, scheduled and periodic. When the weather station device
is in sleep mode, a crystal oscillator or other timing source can
be used to keep track of the sleep and wake times. For example, in
some embodiments the device operates on a 32 kHz crystal oscillator
to consume less than 100 uA of current. The microprocessor
maintains state during sleep mode and is able to maintain time
throughout sleep mode. In other embodiments, a low power receiver
can remain powered on and the wake instructions sent from a remote
device.
[0216] When the device is in active mode, the microprocessor in one
example implementation operates on a 12 MHz oscillator so that
sampling and computation can occur rapidly. When in this active
mode, a sensor sampling state is executed every 10 mS during which
a different action is performed each time. Since an interrupt is
triggered every 10 mSec to handle timing during the active state,
the firmware is designed so that all state executions take less
than 10 mSec to execute. Computations that take extra time are
broken into smaller subtasks. Static variables within the state
handling routines maintain the intermediate calculation values
between function calls so that long computations can be picked up
wherever they are left off. Other clock rates and interrupt cycles
can be used. The 12 MHz oscillator frequency and 10 mS trigger
intervals are provided by way of example only.
[0217] Image compression is a well-known technique used in data
communications. Compression helps to reduce the amount of redundant
information in data, such as still images, in order to achieve
efficient storage and transmission of the data over a
communications link. In some embodiments, the weather station
device compresses the still images captured by camera sensors on
the device for easier transmission over the communication link. The
camera sensor in the weather station device outputs image frames at
a determined number of frames-per-second (FPS). In some
embodiments, the sensor is capable of 30 FPS, but may be configured
to output at a lower frame rate, for example 15 FPS. When the
weather station device captures an image with a camera, image can
be saved for compression and transmission.
[0218] A main control processor (e.g., a microcontroller in the
weather station device) asserts the power-down control signals for
each camera to cause all cameras to be transition to a disabled,
powered-down state by default to save power when the cameras are
not actively being used. When an image is to be captured from a
camera, the microcontroller de-asserts the power-down control
signal to the camera to enable/power-up the cameras. Cameras can
all share a common bus for their output signals to reduce the
number of components used and to reduce the number of traces needed
for signaling, thereby simplifying the system design. Accordingly,
in some embodiments, the cameras can be enabled one at a time to
prevent bus contention.
[0219] Once a camera is powered up, the microcontroller can, in
some embodiments, use a shared/common I2C communication bus to
configure the camera's various settings (resolution, zoom, pixel
output format, picture settings, etc.). When the camera has been
fully configured and given a few seconds to adjust to the exposure
and other picture conditions for the environment that it is
sensing, the microcontroller sends a control signal to a CPLD,
FPGA, or other programmable integrated circuit or module designed
to capture and image and store the image data to external
memory.
[0220] In some embodiments, the CPLD, FPGA, or other programmable
integrated circuit or module interfaces with the camera sensor
clock, vertical sync, horizontal sync, and data output signals.
When an image capture occurs, the module detects the assertion
(rising edge) of the vertical sync signal to identify the beginning
of data output for a new image frame. At the same time, a
horizontal sync signal is received that identifies the beginning
(with its rising edge) and end (with its falling edge) of data
output for a new row of pixels for the image frame. The module uses
each clock cycle of the camera sensor's clock output to read in the
pixel data for storing. The clocking in of pixel data continues for
the expected number of data bytes for one row of the image frame at
the configured resolution and pixel data output format. When the
expected number of data bytes for one row has been clocked in, the
module waits for the next horizontal sync signal to identify the
beginning of the output of the next row of pixel data. This row
pixel data is read by the module in the same way as before. This
process repeats for each row of pixel data of the image frame,
until all rows of pixel data for the configured image resolution
have been read.
[0221] When the module reads pixel data of an image frame from a
camera sensor, it immediately writes the pixel data to an external
memory IC (e.g., SRAM), using sequential memory addresses to write
pixels in the same order that they are read from the camera
sensor.
[0222] When all pixel data bytes of an image frame have been read
from the camera sensor and stored to memory, the module outputs a
status signal to the microcontroller to indicate that an image
frame has been captured to memory and is ready for
reading/processing. Upon receipt of this status signal, the
microcontroller initiates a clock/data/enable interface to read the
pixel data out of the external memory IC, routed through the
module. The microcontroller provides the clock and enable signals
to the module since it is operating at a lower frequency than the
module. The enable signal is asserted to trigger the module to
enter a memory reading state. The enable signal remains asserted
throughout the entire pixel data reading process. De-assertion of
the enable signal causes the module to return to an idle state if
it is in the middle of the memory reading state. The module also
automatically returns to an idle state after all pixel data bytes
have been read out of memory (determined by a count of the expected
number of bytes based on image frame resolution and pixel output
format).
[0223] The microcontroller asserts the data interface clock signal
to the module to trigger the module to read data from the next SRAM
address. The module immediately reads and outputs the pixel data
byte to the microcontroller for the SRAM address. After a delay to
guarantee that the new SRAM data is stable on the module to
microcontroller interface, the microcontroller reads the pixel data
byte into an internal FIFO buffer. The microcontroller then
de-asserts and re-asserts the data interface's clock signal to the
module to request the data for the next SRAM address for reading
and buffering. This process repeats until all pixel data bytes have
been read from the SRAM.
[0224] During the process of reading pixel data from the SRAM,
routed through the module, and output to the microcontroller, the
pixel data is accessed from the SRAM addresses in blocks of eight
horizontal pixel data sets by eight vertical pixel data sets. A
pixel data set includes all of the pixel data bytes that represent
one image frame pixel. These blocks of eight-by-eight pixel sets
are read in rows from left to right and top to bottom of the image
frame, in order to support the JPEG compression that is performed
by the microcontroller.
[0225] During the process of reading pixel data from the SRAM,
since the microcontroller is controlling the read process with the
assertion and de-assertion of the data interface's clock signal,
the microcontroller can pause the reading of pixel data in order to
perform the JPEG compression on the eight-by-eight pixel sets block
and transmit the result out through a different interface (e.g. I2C
or UART) to free up the FIFO buffer before reading in the next set
of pixel data for JPEG compression of the next pixel data
block.
[0226] Image compression, e.g., via the JPEG standard, can be
implemented in the microcontroller (from the Microchip dsPIC33F
family), taking advantage of its built-in computational functions
for various steps of the JPEG compression algorithm. Currently,
there is one fixed set of quantization tables for luminance and
chrominance to achieve the desired level of compression to balance
the sufficient image detail with minimal image data size. However,
multiple tables can be stored in memory and selected by the user
via external commands to change the effective compression levels.
In addition, other imaging standards may be used based on
requirements of the system.
[0227] This basic architecture of a CPLD/FPGA (or other module) to
receive and store data from a sensor that outputs at a high
frequency for future processing can be applied for any number of
sensors. In some embodiments, the architecture can be used to
record the output data of an analog-to-digital converter rated in
the megasamples-per-second (MSPS) range and allowing a slower
processor to analyze and process the samples later. This allows for
greater flexibility in the type of processor used in the device and
adding more flexibility in meeting stringent size and weight
requirements.
[0228] In some embodiments, the weather station uses the Iridium
short burst data modem (SBD) satellite communications network for
data communications. Commanding a weather station, however, does
not have to be limited to the Iridium SBD network. The Iridium SBD
communications can be replaced or even augmented by any of a
variety of other communications interfaces to allow for other
commanding and data reporting options. Such options include, for
example, UART, ZigBee wireless, IEEE 802.11 (Wi-Fi) wireless, other
proprietary or custom standards or protocols, or even custom
infrared or laser-based optical data transmission solutions. For
ease of illustration only, the transmission of compressed still
images will be described in terms of the Iridium SBD network. After
reading this example, one of ordinary skill in the art will
understand how this can be implemented using other communication
protocols and devices.
[0229] In embodiments using the Iridium system, the Iridium SBD
(short burst data) service is limited to data burst transmissions
of up to a few kilobytes at a time. The Iridium 9602 modem, for
example, is limited to transmissions of 340 bytes, which is not
sufficient for transmitting a complete image of a reasonable
resolution and detail. In some embodiments, the camera sensor
output resolution can be chosen to be 320.times.240 pixels and use
JPEG compression performed by the imaging microcontroller. These
characteristics can significantly reduce an image to a data size
that can be broken up and transmitted via several Iridium SBD
bursts over a time that fits within the normal operating mode of
the weather station device (hourly sensor readings and
transmissions). In some embodiments, data from other weather
systems external to the system can be fed in to the communication
sub-system and treated as if the sensor readings were taken by the
transmitting device itself.
[0230] Image data packets transmitted over the Iridium SBD contain
a header that identifies each packet as image data. The header of
each packet contains an incrementing sequence number to ensure that
image data can be recombined in the correct order and to mark the
total number of packets transmitted in the last packet. The header
also contains a status byte to indicate whether the packet is the
last data packet of an image. The header of the first packet
contains a timestamp that corresponds to the timestamp of the
weather sensor readings taken just prior to the image capture.
[0231] The image data portions of each image data packet include
the JPEG compressed image data streamed from the imaging
microcontroller. Image data is requested from the imaging
microcontroller by the processor (a different microcontroller) that
is assembling and transmitting the Iridium SBD packets, so only the
amount of image data needed to fill the current packet is
requested. The image data fills the remainder of each 340-byte
Iridium SBD packet. In the last packet, if image data does not
completely fill to 340 bytes, benign JPEG data fill bytes of
hexadecimal 0xFF are used. These fill bytes do not affect the
reassembly and display of a completed JPEG image.
[0232] The Iridium SBD packets are received via e-mail attachments
by a remote computer. Each packet is received as a file attachment
in an e-mail message, one e-mail message and attachment per
transmitted Iridium SBD packet. Once the attachments are
downloaded, the JPEG image can be reassembled into a file that is
viewable on a computer by appending the image data portions of each
packet, in the correct sequence, to a predefined JPEG file header
containing the proper attributes. The standard JPEG end of image
marker bytes of 0xFFD9 is then appended to the end of the image
data bytes to complete the file.
[0233] This method of transmitting data for a complete image can be
applied to sending any kind of large data format over any other
interface and transmission protocol with a limited transmission
packet size. It can also be used as a method to limit the bandwidth
and effective transmission rate of one particular set of data, in
order to make room for other data sets to share bandwidth over the
interface for time sensitive transmission events.
[0234] Control of the weather station device is also capable using
the satellite communications link. Commands are sent to the weather
station device via the satellite communications link, such as the
Iridium SBD network in certain embodiments. The remote end user
creates an e-mail message targeted at the unique IMEI
(International Mobile Equipment Identity) number of the Iridium
9602 modem on the weather station device. The user attaches a text
file containing a single command or a list of commands to be
processed and executed by the weather station device. The contents
of this text file are queued for the weather station device to
download via the Iridium SBD network. The next time the weather
station device transmits data (any data, weather or image) it will
download and read the message queues from the Iridium SBD network
and any valid commands are processed at the completion of the
original data transmission (after all weather or image data has
been sent), to not interrupt the original process.
[0235] If multiple instances of the same command type are received
during a single data transmission state (transmission of one group
of weather packets or one group of image packets), depending on the
command type, either the latest/most recent command received
overrides the previously received command of that type (e.g.
operating mode) or the latest/most recent command can be appended
to the previously received command of that type (e.g. image request
from certain cameras).
[0236] Various functions of the weather station system may be
controlled by sending commands remotely over the satellite link.
Commands may be sent to change the operating mode of the weather
station device between continuous, hourly, every three hours, or
some other operating mode employed in the design. By changing the
operating mode, it is possible to change the frequency with which
the weather sensing components collect weather data remotely as
well as the frequency of transmission of that data, both of which
are controlled by the operating mode. Commands may also be sent to
request images to be captured and transmitted from the weather
station device, independent on the operating mode selected. When
such commands are sent, the weather station device may be manually
forced to collect and transmit data without waiting for the next
scheduled iteration of data collection. In some embodiments, this
command may also reset the timing of the operating mode. For
example, if a command is sent to tell the weather station device to
collect data from weather sensors in between the programmed
sampling interval, the weather station device will sample the data
and the sampling interval will reset at that time. In other
embodiments, the weather station device will not reset the sampling
interval. In this case, the weather station device will sample the
data as requested and then again at the scheduled interval time.
Commands may also be sent to request an update of the GPS location
of the weather station device. Requesting GPS information can be
used to ensure that the GPS sensor is functioning properly by
comparing the known location with the transmitted result. In
addition, the GPS data can be used to ensure that the device has
not been compromised by checking its known location against the
received GPS data.
[0237] The commanding of a weather station device does not have to
be limited to the Iridium SBD network. The Iridium SBD
communications can be replaced or even augmented by another
communications interface in the design to allow for other
commanding and data reporting options. Such options include UART
(already implemented in the PIWOS), ZigBee wireless, IEEE 802.11
(Wi-Fi) wireless, or even custom infrared or laser-based optical
data transmission solutions.
[0238] FIG. 2 is a functional block diagram illustrating an example
weather station device in accordance with one embodiment of the
technology described herein. The use of a three (3) axis
accelerometer and three (3) axis compass provides a means for
correcting the offset of weather-station-device orientation during
placement and operation in remote location. The accelerometer also
serves as a powerful tool for identifying attempts at tampering
with the weather station device or changes of the orientation due
to the environment or other factors. The use of the compass allows
for the addition of magnetic orientation so that the photographic
direction of the camera images from the cameras of the imaging
module can be determined, and therefore the direction for incoming
weather conditions determined. The collection of sensor data from
multiple weather station devices at the central server can be
configured to provide an efficient, sustainable, reliable, and
redundant way of remote weather monitoring for the needs of
multiple users.
[0239] Required weather parameters may be recorded using modules
such as a lightning detector, ambient light sensors, cameras,
LIDAR, multiple moisture sensors, multiple temperature sensors,
pressure and humidity sensors, and disdrometer sensors. The battery
system, in combination with the capacitor bank and solar panels,
can be used to provide power to the system. System firmware, in
combination with the main processor, collect the readings from the
sensors, adds the required processing, corrections and encryption,
if needed, to recorded data and then transmit the resulting data to
the central server through the communications interface, such as
the Iridium satellite network.
[0240] FIG. 3 is a diagram illustrating an example embodiment of an
extension module for the weather station device in accordance with
one embodiment of the technology described herein. The module
includes a body 305 and a cover 310. The module may further include
a connection port 315 that allows connecting the extension module
to weather station device. The module may also have an electric
switch 320 to turn the module on and off. The cover 310 may also
include solar panels 325 or other power sources for electric power
generation or supply. The section 330 of the cover 310 may
configured to tilt to allow water or precipitation to run off of
the cover or to orient the solar panels 325 toward the equatorial
latitudes for better sun exposure. The cover section 330 may have
optical windows 335 and 340 for laser output and laser return
collection correspondingly. These optical windows can be used to
protect the laser range finder inside the extension module. The
advanced module can also include an access port 350. This port can
be used to take samples of air or precipitation inside the module
for contamination analysis. The contamination analysis can be, for
example, for chemical, biological, and radiation contamination.
[0241] As noted above, in some embodiments, new micro weather
sensing devices can be included to detect atmospheric
characteristics such as, for example, atmospheric visibility, wind
speed, and wind direction.
[0242] As introduced above, another sensor that can be included is
a new thermal sensor design capable of measuring wind speed and
wind direction using no moving parts. Embodiments can be
implemented that have no moving parts and that are smaller than
conventional wind sensors. FIGS. 4A and 4B are diagrams
illustrating top and side views, respectively, of an example
mechanical layout for exemplary thermal sensors in accordance with
embodiments of the technology disclosed herein.
[0243] Referring now to FIG. 4A, in this example sensor 150
includes a plurality of thermal sensors 152 mounted on a printed
circuit board. In various embodiments, these thermal sensors 152
are mounted so that the temperature sensing mechanism of thermal
sensors 152 (e.g., thermocouple, thermistor, etc.) are placed at a
predetermined spacing above the surface of the print circuit board.
Preferably, the temperature sensing mechanisms of thermal sensors
152 are positioned so as to be located within the channel formed
between the upper part and the lower part of the weather sensing
device. In this manner, these temperature sensors can be positioned
to sense the temperature of air moving through this channel. In
various embodiments, these thermal sensors 152 are passive thermal
sensors. Although 8 thermal sensors 152 are shown as being arranged
in a circular ring about a central element 340 (with central
element 340 being at approximately the center of the ring), other
quantities of thermal sensors 152 can be used, and they can be
disposed in alternative arrangements or geometries.
[0244] In this example, central element 340 is provided and
includes a thermal sensor and a thermal generator or heater (not
separately illustrated). In operation, the thermal generator of
central element 340 is powered to generate heat. In various
embodiments, a controlled amount of power can be provided to the
central thermal generator to heat the central thermal generator
above ambient temperature. The amount of power can be determined
based on the ambient temperature. As wind blows through the channel
between the upper and lower parts of the weather station (e.g.,
channel 110 of the example in FIG. 1), this airflow causes heat to
be transferred from the central thermal generator to one or more of
the thermal sensors 152. Particularly, the airflow, or wind, causes
the heat from the central thermal generator to be transferred to
the one or more thermal sensors 152 that are downwind from the
central thermal generator.
[0245] An example of this is illustrated at FIG. 4A. In this
example, the wind is blowing in the direction from right to left
across the page as indicated by arrow 341. This causes heat
produced by the central thermal generator to travel in the
direction of arrow 341 toward the ring of thermal sensors 152. This
is illustrated by the elliptical shaded area 343. As can be seen in
this example, this higher temperature air flows toward and around
the leftmost thermal sensor 152 (i.e., the one at the "9 o'clock"
position on the ring). As a result, leftmost thermal sensor 152
reads a higher temperature than do the other thermal sensors 152.
From this, it can be determined which direction the wind is blowing
relative to the weather station device itself. If the orientation
of the weather station device as deployed (e.g., as mounted in the
ground) is known, the actual wind direction can be determined.
[0246] The faster the wind flows over this central element, the
less the temperature rise it will experience due to more heating
being lost to the moving air in what can be referred to as a heat
bloom. Accordingly, wind speed can be determined by measuring the
ambient temperature (e.g., from a temperature sensor removed from
the thermal array) measuring the temperature at the thermal
generator (e.g., via a temperature sensor of central element 340)
and measure the temperature increase that the affected thermal
sensor 152.
[0247] In some embodiments a micro-visibility sensor can be
included and implementations of it can be configured to: allow for
measurement of visibility in an extremely compact device with
limited power; use common, individually focused photodiodes with
integrated spectral filtering but can also be used with more
sophisticated optical elements; and use common, individually
focused light (e.g., 850 nm LED (light-emitting diode)) but can
also be used with more sophisticated optical elements.
[0248] FIGS. 5A and 5B are diagrams illustrating a front view and a
side view, respectively, of an example of an opto-mechanical
arrangement in accordance with one embodiment of the technology
described herein. Referring to FIG. 5A, the arrangement of this new
sensor 430 includes a transmitter tower including a light source
(e.g., one LED internal to the confining element 432) and a
receiver tower 433 containing a photoreceptor 431 (e.g., an array
of photodiodes). The light source is surrounded by a light director
or confining element 432. In this example, confining element 432 is
a tube (e.g., substantially cylindrical in shape), but other
configurations can be chosen. The top of the tube is formed or cut
at an angle so that the optical receiver array 431 is below the
shadow line 435 from the transmitted beam 434, more clearly seen in
FIG. 5B. This arrangement prevents not only direct light from the
transmitter from entering the receiver but also prevents
reflections from within the transmitter tube from entering the
receiver. The transmitter tube can also be internally coated with a
non-reflective material to further avoid reflections from the
transmitter from reaching the receiver. Therefore, the only light
from the transmitter that can enter the receiver is from
reflections of the light above the shadow line, in particular from
atmospheric scattering. The reason for this is that there is no
straight-line path from any point inside the hollow transmitter
tube 432 (FIG. 5A) to any point on the optical receiver array 431
(FIG. 5A) including any secondary point sources caused by optical
edge interference. Therefore, all secondary reflections 436 (FIG.
5B) from inside the transmitter tube occur only above the shadow
line 435 (FIG. 5B). Because of the soft edge and low-coherence
light source, illumination below the shadow line by secondary
Fraunhofer diffraction is also reduced or avoided. Therefore, the
only signal from the transmitter that is received by the
photo-detectors is the light reflected from the atmosphere.
[0249] Electronically, the sensor operates by driving the LED with
a square wave at a specific frequency of 32.768 kHz, but those
skilled in the art will note that the approach can also be applied
to other waveforms and to higher or lower frequencies. During
operation, the photodiode array detects light that is scattered by
the atmosphere into the array. To detect this small amount of light
(as low as one-millionth of the transmitted power), the photodiode
signal is first AC-coupled and amplified to a sufficient level to
allow detection due to scattering.
[0250] Measurement of the scattered light occurs by separately
averaging the received light level when the LED is on and when it
is off. The difference between these two levels is then amplified
again and digitized to provide a clear measure of the amount of
scattered light. Accordingly, a 50% duty cycle square wave is
preferred, but other waveforms and duty cycles can be used.
[0251] As this example illustrates, devices for visibility
measurement may be based on return optical signals, which can be
very weak, especially in the case of high atmospheric signal
attenuation. Therefore, reducing Intra-System Optical Noise (ISON)
may be important for the high sensitivity typically desired for
such devices.
[0252] In accordance with embodiments of the technology disclosed
herein, and as discussed immediately above, a hollow optical
transmitter tube, Tx, may be cut to shield the optical receiver,
Rx, (e.g., photoreceptor 431 described above in FIG. 5A) from
intra-source optical noise (ISON), such as that from the light
source of the visibility measurement system itself. The typical
optical source in various embodiments is an LED, and the typical
receiver Rx may be photodiode. FIG. 6 is a diagram illustrating a
close-up view of a portion of a visibility measurement system such
as that shown above in FIGS. 5A and 5B. As can be seen from FIG. 6,
the cut angle, .alpha., is smaller than .alpha.=90.degree.. O
[0253] In the example of FIG. 6, the optical Tx/Rx-sub-system
including optical tube 507, which can be implemented using the same
optical tube 432 as illustrated in FIGS. 4A and 4B. Optical tube
507 includes a light source such as, for example, an LED or other
light source (not shown in FIG. 6). As described above, an optical
receiver 502 (e.g., photodetector array 431 in FIGS. 5A and 5B) is
used to gather light scattered by moisture in the air. FIG. 6 also
illustrates exemplary characteristic optical rays 503 as well as a
Ray diffracted from edge 504 of tube 507.
[0254] The cutoff angle, .alpha., denoted as 505, is smaller than
.alpha.=90.degree.. The auxiliary angle, .delta., denoted as 506,
can be used for further analysis. This diffraction angle, 506, can
be very small as shown in FIG. 5B, but it is shown large in FIG. 6
to illustrate the theory of operation. The diffraction boundary,
507, is critical for diffraction phenomenon. As this example
further illustrates, light from the light source does not have a
direct path to the receiver 502. Accordingly, receiver 502 is
shielded from the light source thereby reducing ISON.
[0255] FIG. 7 is a diagram illustrating an example of a diffraction
problem in accordance with various embodiments. The illustrated
diffraction problem is illustrated using a diffraction edge 521
(e.g., analogous to edge 504 of FIG. 6). A diffraction (local)
half-plane 522 (e.g., analogous to diffraction boundary 507 (i.e.
the edge of the tube) in FIG. 6) is shown with an exaggerated
thickness for illustration purposes. However, Sommerfeld's
canonical diffraction theory is provided for an infinitely thin
half-plane.
[0256] Angle, .phi., denoted as 523, depicts a diffraction ray 524
including other diffraction rays 525 resulting from incident
optical ray 526. With a growing angle, .phi., diffraction rays 525
become weaker and weaker, as symbolically shown by the arrows'
length. The incident ray, or rather electromagnetic ray with
electrical (polarization) vector, {right arrow over (E)}, 528 is
perpendicular to the direction of propagation 529 and can be
arbitrarily-polarized 528. However, the diffraction ray 524 may be
considered as a superposition of two elementary polarizations: TE
(transversal-electric, or {right arrow over (E)}.sub..perp.), and
TM (transversal-magnetic, or {right arrow over
(Eu)}.sub..parallel.), denoted as 530, and 531, respectively. The
electrical vector complex amplitude, U, (Sommerfeld's canonical
diffraction solution) for a perfectly conducting diffraction
half-plane, (r is the distance from edge 521):
U = 1 + i 4 .pi. kr e ikr ( 1 cos ( .PHI. - .alpha. 2 ) .-+. 1 cos
( .PHI. + .alpha. 2 ) ) ( 1 ) ##EQU00001##
where k is wavenumber (k=2.pi./.lamda., where .lamda.-optical
wavelength), i= {square root over (-1)}, and e is natural
logarithmic base, while .phi.-angle is 523, and .alpha.-angle, is
532, in FIG. 7. The upper minus (-) is for TE-polarization, and
lower plus (+) is for TM-polarization. As this diagram illustrates,
the sign (-) shows that for diffraction ray 524 tangential for
half-plane, U=0, as expected, because, in such a case, the
electrical vector tangential component, E.sub.t, should be equal to
zero:
E.sub.t=0 (2)
[0257] However, this condition is only satisfied for a
perfectly-conducting half-plane. In various embodiments, the
half-plane maybe a dielectric (e.g., glass, or plastic). Thus, the
polarization-sensitive terms can be omitted, and Eq. (1) reduced
to:
U = 1 + i 4 4 kr e ikr 1 cos ( .PHI. - .alpha. 2 ) ( 3 )
##EQU00002##
which describes the cylindrical wave with propagation waveform, in
the form:
e ikr r ( 4 ) ##EQU00003##
[0258] while the wave intensity will be proportional to |U|.sup.2,
where |U| is module of the complex amplitude, U.
[0259] FIG. 8 is a diagram illustrating an example of an expanded
version of FIG. 7, including a source geometry, to illustrate an
example of the diffraction problem. This example includes an
optical source 601 and hollow tube walls 602, 603. An incident Ray
604 represents source intensity, I.sup.(s), at angle .alpha., which
hits the edge 605 at angle, .beta., where
.beta.=90.degree.-.alpha.. Ray 606 is reflected at surface 605
representing Fresnel reflection. This ray 606 is incident at edge
607, creating diffracted ray 608, which is diffracted at
diffraction angle, .delta., which denotes diffraction at
.delta.-angle from shadow boundary 609. At locations farther from
shadow boundary 609 the diffraction beam becomes weaker. In order
to see this quantitatively, the following angular relation is
noted:
.phi.=.alpha.+.pi..delta. (5)
thus: .phi.-.alpha.=.pi.+.delta., and the diffraction beam
intensity, |U|.sup.2, becomes,
U 2 = 1 8 .pi. kr 1 sin 2 .delta. 2 ( 6 ) ##EQU00004##
[0260] From this relationship, it can be seen that |U|.sup.2
decreases with increasing angle, .delta.. For exemplary angle,
.delta..sub.0:
U 0 2 = 1 8 .pi. k r 1 sin 2 .delta. 0 2 ( 7 ) ##EQU00005##
assuming .delta..sub.0=3.degree., it can be further assumed that
.delta..gtoreq..delta..sub.0. In this case, the relative intensity
ratio, is
.eta. = U U 0 2 = sin 2 .delta. 0 2 sin 2 .delta. 2 ; .delta.
.gtoreq. .delta. 0 ( 8 ) ##EQU00006##
[0261] This relation is tabulated in Table 1. For example, for
.delta.=30.degree., and .delta..sub.0=3.degree., we obtain:
.eta.=0.01.
TABLE-US-00001 TABLE 1 .eta.-RATIO vs. .delta.-ANGLE, for
.delta..sub.0 = 3.degree. .delta. 3.degree. 5.degree. 10.degree.
20.degree. 30.degree. 40.degree. 50.degree. 60.degree. .eta. 1 0.35
0.09 0.023 0.01 0.00 0.004 0.003
[0262] According to Eq. (3), for .phi.=constant, we obtain
.DELTA..delta.=-.DELTA..phi. (9)
i.e., increasing of .phi.-value causes decreasing of .delta.-value,
and vice versa. FIG. 9 is a diagram illustrating an example of this
relation between the inclination angle, .alpha., and diffraction
angle, .delta..
[0263] According to FIG. 9, it can be seen that for constant .phi.
values:
U 1 U 2 2 = sin 2 .delta. 2 2 sin 2 .delta. 1 2 ( 10 )
##EQU00007##
i.e., for larger .alpha.-values, the .delta.-value is smaller,
which results in larger |U|.sup.2 values. Thus, a more inclined the
`cut` at the end of the tube (smaller .alpha.-value) results in a
smaller |U|.sup.2-value.
[0264] In order to address the global diffraction problem, it is
useful to also consider the source angular spectrum and Fresnel
reflection from the tube walls. FIG. 10 is a diagram illustrating
an example of a generalized Lambertian source model in
cross-section for the 2D case.
[0265] Referring again to FIG. 8, the source intensity spectrum,
is
I.sup.(s)(.alpha.)=I.sup.0 cos.sup.n.alpha. (11)
where the Generalized Lambertian Source model has been assumed, as
shown in polar coordinates in FIG. 10, where n=1 for a Lambertian
source and n>1 for a non-Lambertian source.
[0266] In FIG. 10, a Generalized Lambertian Source Model in polar
coordinates r=r (.alpha.) is presented. This example is shown for
scenarios including a Lambertian source (n=1), and a non-Lambertian
source (n>1). For a Lambertian source, there is a right angle
triangle ABC, where:
r=1 cos .alpha..sub.0=cos .alpha..sub.0 (12)
which becomes Eq. (11) for n=1. Half-Max-Half-Angle (HMHA) values
.alpha..sub.1 and .alpha..sub.2 are obtained for the Lambertian and
non-Lambertian case, respectively.
[0267] In FIG. 10, a generalized Lambertian source model 630
includes a Lambertian case 631, a non-Lambertian case 632, and a
half-max case 633, in which the maximum intensity value has been
normalized by one (as denoted by 634).
[0268] For the Lambertian case, the half-max angle value,
.alpha..sub.1, may be obtained by crossing sphere 631 with
hemisphere 633. For the non-Lambertian case, .alpha..sub.2 may be
obtained by crossing rotational ellipsoid 632 with hemisphere 633.
FIG. 10 presents the 2D cross-section of the 3D case. According to
Eq. (11), n-value can be found from the following equation:
0.5 = cos n .alpha. 1 / 2 ( 13 ) or , n = log ( 0.5 ) log cos
.alpha. 1 / 2 ( 14 ) ##EQU00008##
[0269] For example, for .alpha..sub.1/2=60.degree., n=1 (Lambertian
case), while for .alpha..sub.1/2=30.degree., n=4.8
(non-Lambertian). This results in .alpha..sub.1=60.degree., and
.alpha..sub.2<60.degree., as in FIG. 10.
[0270] According to Eq. (11), the source intensity, I.sup.(s)
(.alpha.) is a monotonically decreasing .alpha.-function. Also,
according to FIG. 8, the Fresnel reflection intensity coefficient,
R, is a monotonically increasing function of .beta.. For example,
for .beta.=0.degree., R.apprxeq.8%; and for .beta.=90.degree.,
R=100%=1. Inversely, the R-value is a monotonically decreasing
function of .alpha.. For .alpha.=90.degree., R.about.8%, and for
.alpha.=0.degree., R.about.1. However, the diffraction intensity,
|U|.sup.2, is a monotonically increasing function of .alpha..
Therefore, there are two contradictory tendencies, summarized in
the form:
I(.alpha.)=|U(.alpha.)|.sup.2I.sup.(s)(.alpha.)R(.alpha.) (15)
[0271] In order to minimize the I(.alpha.)-value representing the
ISON factor, it is useful to reduce .alpha.-value to some value
<90.degree.. However, then, both I.sup.(.alpha.)(.alpha.) and
R(.alpha.) would increase. Therefore, this leaves the challenge of
determining how to minimize ISON factor, the solution to which is
not obvious.
[0272] Furthermore, it may be desirable to minimize diffraction
effects by smoothing the tube edge profile. FIGS. 11A and 11B are
diagrams illustrating examples of diffraction edge profiling in
accordance with various embodiments of the technology disclosed
herein. In the example illustrated in FIG. 11A, the diffraction
edge profiling 660 illustrates the hollow tube wall without
profiling 661. In the example illustrated in FIG. 11B, the
diffraction edge profiling illustrates the hollow tube wall with
profiling 662. Referring to FIG. 11A, a large diffraction beam
results as symbolized by relatively large ray arrow 663. In
contrast, referring to FIG. 11B, a smaller diffraction beam results
as symbolized by smaller ray arrow 664.
[0273] As noted above, in important goal for the system is to
minimize ISON. According to Eq. (15), three factors should be
minimized by mitigating the contradiction tendency between the
1.sup.st factor and the other two factors, leading to satisfying
the following global condition:
I(.alpha.)<I.sub.T(.alpha.) (16)
where I.sub.T(.alpha.) is some threshold value in [W/m.sup.2], say:
1 pW/(50 .mu.m).sup.2=410.sup.-4 W/m.sup.2, for example.
[0274] All 3 factors can be found numerically or experimentally and
calculated for a specific .alpha.. Situations may be common in
which the |U|.sup.2-factor will be very small, but factors
I.sup.(s)(.alpha.) or R(.alpha.), or both, are still very large.
Because the Fresnel R(.alpha.)-factor is derivative of the source
factor, I.sup.(s)(.alpha.), only the latter can be manipulated.
This can be accomplished by either changing the geometry, or by
changing a type of source (by increasing n-power factor).
[0275] It is noted that typically all three factors are spectrally
dependent, either on wavelength, .lamda.; frequency, f; or angular
frequency, .omega.. Any of these parameters may be chosen because
for small variations:
.DELTA. .lamda. .lamda. = .DELTA. f f = .DELTA. .omega. .omega. 1 (
17 ) ##EQU00009##
[0276] For example, according to Eq. (6), the diffraction beam
intensity, is
U ( .delta. , r ; .lamda. ) 2 = .lamda. 16 .pi. 2 r 1 sin 2 .delta.
2 ( 18 ) ##EQU00010##
[0277] However, the remaining two factors are weakly
spectrally-dependent, while the diffraction factor, according to
Eq. (18), is proportional to wavelength, .lamda.. Thus, only the
diffraction factor is strongly spectrally-dependent.
[0278] However, diffraction .lamda.-dependence is a simple
proportionality. Therefore, by integration, instead of Eq. (18),
the following relation can be obtained:
.intg. .lamda. 1 .lamda. 2 U ( .delta. , r ; .lamda. ) 2 d .lamda.
= U ( .delta. , r ; .lamda. _ ) 2 .DELTA. .lamda. ( 19 ) where :
.lamda. _ = .lamda. 1 + .lamda. 2 2 , and .DELTA. .lamda. = .lamda.
2 - .lamda. 1 . ##EQU00011##
[0279] Therefore, the primary relation (17) is preserved while
having adding extra bandwidth factor, .DELTA..lamda., while it is
assumed that the source spectrum is within
(.lamda..sub.1,.lamda..sub.2)-range.
[0280] The approximate formula for a diffraction beam has the
following form:
I ( .delta. , .alpha. , r ) = K .lamda. _ .DELTA..lamda. 16 .pi. 2
r 1 sin 2 .delta. 2 I ( s ) ( .alpha. ) R ( .alpha. ) ( 20 )
##EQU00012##
where the double angle (.delta.,.alpha.) emphasizes the fact that
the narrow incident beam (defined by its angular distribution,
I.sup.(s)(.alpha.) and its intensity Fresnel reflection,
R(.alpha.)) is spread into a set of diffraction "rays", as shown by
525 in FIG. 7. Thus, only the diffraction factor seems to have
double angle dependence. However, by introducing the diffraction
angle, .delta., this double dependence is reduced into a single
angle dependence, on .delta., as in Eq. (20). The proportionality
constant, K, is introduced for calibration purposes. Thus Eq. (20)
presents factorized dependence in the form:
I(.delta.,.alpha.,r)=K|U(.delta.,r)|.sup.2I.sup.(s)(.alpha.)R(.alpha.)
(21)
where r-distance from diffraction edge.
[0281] It is very difficult, if possible at all, to analytically
derive the optical detection factor. There are many reasons for
this, including the semiconductor process of photonic detection,
complex incidence angle dependence, and so on. Therefore, various
embodiments utilize a calibration constant, K. Typically, for
high-sensitive CCD photo detectors, the final optical intensity of
the beam, I, will be about:
I ~ 1 pW ( 10 m ) = 10 - 12 W 10 - 10 m 2 = 0.01 W / m 2 ( 22 )
##EQU00013##
[0282] After I-value measurement (at fixed distance, r), we can
approximately predict I-intensity behavior for various source/tube
configurations. This is because the 1.sup.st diffraction factor is
known while two remaining source and Fresnel factors can be
obtained from general optical propagation considerations. Thus, by
knowing the K-factor, I-value distribution for various source/tube
configurations can be predicted in order to minimize I-value, thus
satisfying condition (16).
[0283] More detailed consideration shows that the calibration
constant, K, is, in fact, a calibration parameter with two control
variables: .delta., and r. Thus, this parameter may be referred to
herein as a Dual-Control-Variable Calibration Parameter, in the
form:
K=K(.delta.,r) (23)
[0284] This parameter is without subscript, which is in contrast to
I.sup.(s)(.alpha.) and R(.alpha.) which can include a subscript.
This is, because, the latter functions can be different for various
types of sources for or different average wavelengths, .lamda.. In
contrast, Eq. (23) depends only on the receiver geometry (unless
the receiver type is changed).
[0285] FIG. 12 is a diagram illustrating an Exemplary Look-Up Table
for Dual-Control-Variable Calibration Parameter, K(.delta..sub.i,
r.sub.j)=K.sub.ij. In FIG. 12, exemplary mapping table is shown,
as
K.sub.ij=K(.delta..sub.i,r.sub.j) (24)
[0286] This relation has been obtained after quantization, or
digitization, for specific average wavelength, .lamda..
[0287] In the example of FIG. 12, exemplary look-up table for
dual-control-variable calibration parameter, K.sub.ij, includes two
indices i, j. These indices may be numbered by integers: 1, 2, 3,
4, . . . . The index, i, can be used to denote discrete values of
the diffraction angle, .delta., in the form: .delta..sub.i. The
index, j, can be used to denote discrete values of distance radius,
r, in the form: r.sub.j. This calibration parameter can be used to
characterize photo detector behavior as a function of diffraction
geometry. In general, a monotonically-decreasing dependence on
r.sub.i and .delta..sub.l can be expected as shown in FIG. 12. The
question marks show the values that were not measured
experimentally. The r.sub.j are some normalized values that depend
on methodology.
[0288] One important yet nonobvious point is that knowing the
K-value for specific .lamda., and .DELTA..lamda.-bandwidths,
I-values for different light sources can be predicted. FIGS. 13A,
13B, and 13C are diagrams illustrating a comparison of diffraction
efficiency (FIG. 13A), photodiode quantum efficiency (FIG. 13B),
and light source power density (FIG. 13C). A further remarkable
coincidence is that photodiode quantum efficiency, .eta..sub.ph, is
also proportional to wavelength, .lamda., for entirely different
reasons. This is shown in FIGS. 13A, 13B, and 13C, where comparison
of diffraction efficiency (FIG. 13A), photodiode quantum efficiency
(FIG. 13B), and light source wavelength power density (FIG. 13C),
is shown. It may be assumed that the light source (LED, for
example) linewidth, .DELTA..lamda., is within the (.lamda..sub.1,
.lamda..sub.2)-range, according to Eq. (19). Assuming that, also,
.lamda..sub.2<.lamda..sub.3, which is not needed, in general,
the source relation similar to that in Eq. (19) exists, namely:
G(.lamda.).DELTA..lamda..
[0289] FIG. 13B pairs further explanation. For absolute zero (T=0
in K.degree.) temperature, the curve has a sharp triangular form.
However, for T>0 (again, T in K.degree.), the curve is smoothing
within range (.lamda..sub.3, .lamda..sub.4), which is defined by
the .DELTA.E-parameter. It is remarkable that two fundamental
quantum physics relations defining .lamda..sub.g and .DELTA.E
contain so many (3) fundamental physical constants, in the
form:
.lamda. g = hc E g , .DELTA. E = kT ( 25 ab ) ##EQU00014##
where E.sub.g-energy of semiconductor gap (few electron volts, eV)
and .DELTA.E-energy of thermal fluctuations, while h, c, k-are
three (3) fundamental constants, namely, Planck's constant (h), the
speed of light in vacuum (c), and Boltzmann's constant (k). In
particular for T.about.300.degree. K (room temperature),
.DELTA.E.about.0.02 eV.
[0290] In various embodiments, the Approximate Prediction Procedure
(AP2) may begin with selecting other light sources that may be
useful for prediction. Such a source may be characterized by
manufacturer power angular dependence, I.sup.(s)(.alpha.), which is
given. Then, the Fresnel formula, R(.alpha.), is used, which
(presumably) has already been tabulated. An
.alpha.=.alpha..sub.0-value is selected for a given incidence ray,
.alpha., following source and tube geometry as in FIG. 8. For these
.alpha..sub.0-values, G(.alpha..sub.0), and R(.lamda..sub.0)-values
may be obtained. Then, the |U|.sup.2 value is given according to
Eqs. (6), (18), or (20), according to the following formula:
I 0 ( .delta. 0 , .alpha. 0 , r 0 ) = K ( .delta. 0 , r 0 ) .lamda.
_ .DELTA..lamda. 16 .pi. r 0 1 sin 2 .delta. 0 2 I ( s ) ( .alpha.
0 ) R ( .alpha. 0 ) ( 26 ) ##EQU00015##
[0291] It can be seen that Eq. (26) is identical to Eq. (20), or
Eq. (21), except general values, (.delta., .alpha., r), have been
replaced by specific (.delta..sub.0, .alpha..sub.0,
r.sub.0)-values, according to the following reduction
operation:
(.delta.,.alpha.,r)(.delta..sub.0,.alpha..sub.0,r.sub.0) (27)
[0292] An example of this situation is illustrated in FIG. 14, for
the specific example: source, tube, and photodetector geometries.
Particularly, FIG. 14 illustrates an example 2D Geometry for
Approximate Prediction of the I.sub.0-value.
[0293] This geometry leads to the Approximate Prediction Procedure
(AP2). FIG. 14 presents a 2D cross-section of a 3D geometry with
tube axial symmetry 681. The tube wall cross-sections 682, 683 can
be made using glass, plastic or other tube materials (the glass is
preferably hardened, however). The specific light source geometry
684 defines a spherical angle, .alpha..sub.0, which also has axial
symmetry. Therefore, incidence ray 685 is also defined with an
inclination angle, .alpha..sub.0.
[0294] Because the photodetector geometry 686 is also defined, the
diffraction angle, .alpha..sub.0, and cylindrical distance,
r.sub.0, are also given. This includes an incidence angle,
.gamma..sub.0. According to FIG. 14, the following trigonometric
relation exists:
.alpha..sub.0+180.degree.+.delta..sub.0+.gamma..sub.0=360.degree.
(28)
thus,
.gamma..sub.0=180.degree.-.alpha..sub.0-.delta..sub.0 (29)
[0295] Therefore, the photodetector incidence angle, .gamma..sub.0,
may be determined by incidence angle, .alpha..sub.0, and
diffraction angle, .delta..sub.0. This is because the shadow
boundary 687 is also defined. The intensity, I.sub.0, at
photodetector surface is related to power, P.sub.0, by the relation
P.sub.0=I.sub.0A, where A is the photodetector area. This is
because the intensity is normal projection of Poynting vector,
according to fundamental rules of radiometry.
[0296] Applying AP2-specific indexing, Eq. (26) may be written in
the form:
I.sub.I(.alpha..sub.1,.delta..sub.1r.sub.1)=K(.delta..sub.1,r.sub.1)|U|.-
sup.2(.delta..sub.1,r.sub.1)I.sub.1.sup.(s)(.alpha..sub.1)R(.alpha..sub.1)
(30a)
I.sub.2(.alpha..sub.2,.delta..sub.2,r.sub.2)=K(.delta..sub.2,r.sub.2)|U|-
.sup.2(.delta..sub.2,r.sub.2)I.sub.2.sup.(s)(.alpha..sub.2)R(.alpha..sub.2-
) (30b)
[0297] These are basic AP2 formulas that may be used for
approximate prediction. The subscripts (1) and (2) denote different
types of a light sources. For example, in Eqs. (30ab), they are
applied for two sources: 1 and 2. These intensity subscripts: "1",
and "2" are only on the left side of Eqs. (30ab), and for denoting
I.sup.(s) values; i.e., for both intensity values in W/m.sup.2.
[0298] The remaining factors do not have intensity subscripts and
they are dimensionless. This is because they are universal for all
types of light sources with the same .lamda.-value. Therefore,
measuring a K-factor for a given source, e.g., 1-source, also
provides valid K-values for other sources, e.g., 2-source. The
other not-indexed factors, namely, |U|.sup.2 and R, are also
universal, and obtained analytically, namely the
|U|.sup.2-diffraction factor, by Sommerfeld, and R-Fresnel factor,
by Fresnel. Thus, knowing formula indexed by "1", for example,
allows to predict other formula, indexed by "2", for example.
[0299] An Approximate Prediction Procedure Example (AP2) is now
described. In the first step, a universal diffraction factor is
provided. This can be accomplished using the classic Sommerfeld
diffraction formula (26) or other analytic available formulas
useful for the AP2 problem. The universal diffraction factor,
|U|.sup.2, may be provided as a function of distance, r.sub.1, and
diffraction angle, .delta..sub.1. In one embodiment, the
diffraction formula can be used in the form of a look up table.
[0300] In a second step, a universal Fresnel factor is provided.
For example, the universal Fresnel factor, R, can be provided as a
function of incidence angle, .alpha..sub.1, in a similar manner as
accomplished in the first step.
[0301] In a third step an averaged wavelength is selected. So far,
both universal factors |U|.sup.2 and R, are valid for any averaged
wavelength, .lamda.. However, in this step a .lamda.-value is
selected. This determines the class of light sources with the same
.lamda.-value. In the case of the next .lamda.-value, the procedure
may be repeated for this next value. This selected .lamda.-value is
preferably chosen using constraints explained in the description of
FIGS. 13A, 13B, and 13C. This .lamda.-value has been denoted in
FIG. 13C.
[0302] In a fourth step, a look up table for K is developed. This
can be accomplished, for example, as illustrated in FIG. 12. These
factor K-values will hold for entire class of light sources with
the same averaged wavelength .lamda.-value.
[0303] In a fifth step, an AP2 formula is applied. Various
embodiments apply a universal AP2 formula with an arbitrary
k-index, where k=1, 2, . . . , N, in a form that is a
generalization of Eq. (30):
I.sub.k(.alpha..sub.k,.delta..sub.k,r.sub.k)=K(.delta..sub.k,r.sub.k)|U|-
.sup.2(.delta..sub.k,r.sub.k)I.sub.k.sup.(s)(.alpha..sub.k)R(.alpha..sub.k-
) (31)
where N is number of light sources of interest, belonging to the
same class, defined by specific .lamda.-value.
[0304] Because K-values are valid for whole .lamda.-class of light
sources (i.e., light sources with the same .lamda.-value), and
|U|.sup.2, R-factors are universal for all .lamda.-classes, only
angular source characteristics I.sub.k.sup.(s) (.alpha..sub.k), for
a given light source need to be known.
[0305] In a sixth step, the angular source characteristics are
determined. These can be determined from manufacturer data. From
this data the operation can apply angular source characteristics,
I.sub.k.sup.(s) (.alpha..sub.k), for k-source. Using the edge-cut
tube geometry, as in FIG. 14, for example, the operation can
determine incidence angle, .alpha..sub.k, denoted as,
.alpha..sub.0, in FIG. 14. From this, the R(.lamda.)-value is also
given.
[0306] In a seventh step, diffraction beam coordinates are
determined. For example, this can be accomplished using the
geometry of FIG. 14. This step determines diffraction beam
coordinates: r.sub.k, .delta..sub.k, denoted as r.sub.0,
.delta..sub.0, as shown in FIG. 14. The operation may also use the
K-look-up table, to find K(.delta..sub.k, r.sub.k)-value from the
look-up table, using the format of FIG. 12.
[0307] In an eighth step, the final intensity value is determined.
Because all factors, on the right side of Eq. (31), are given, the
I.sub.k-intensity value can be calculated, thus predicting this
value for any light source of interest with a given .lamda.-class
where the photodetector area, A, is known, the system can also
predict the power of photodetector, P.sub.k, in Watts:
P.sub.k=I.sub.kA (32)
[0308] It can be seen that the area, A, is not indexed, because it
is not defined for only the .lamda.-class, but, for all sources
available. The P.sub.k-value is intra-system optical noise power,
describing the ISON factor.
[0309] The example eight-step AP2 procedure described above can be
used to predict intra-system optical noise power, P.sub.k, as
defined by Eq. (32), and Eq. (31). This is based on the assumption
that angular source characteristics I.sub.k.sup.(s) do not depend
strongly on wavelength, and therefore, it is defined by averaged
wavelength, .lamda.. Otherwise, the AP2 process should be repeated
for a spectrum of wavelengths, .lamda., and averaged over a source
power density, G.sup.(s)(.lamda.). However, the basic principle of
this procedure does not change. Also, it should be noted that the
intra-system optical noise power is optical, and not electrical
noise power, the latter one being well-known as
Noise-Equivalent-Power, (NEP).
[0310] In the case of electrical noise power, (NEP), however, the
non-linear rule, which is not commonly known, should be observed.
However, this is a result of the basic photodetection law. In this
sense, it is, in general, also valid here. The general
photodetection law is such that electrical current of a
photodetector is proportional to the optical power signal. On the
other hand, Signal-to-Noise-Ratio (SNR) is usually defined as a
ratio of signal and noise of electrical power rather than the ratio
of optical power. Therefore, in order to obtain (SNR)=n, it can be
important to have (optical) signal power only {square root over
(n)}-times larger than (NEP)-value. For example, in order to obtain
(SNR)=10=10 dB, it is ideal to have optical signal power, only
{square root over (10)}-times larger ( {square root over
(10)}=3.16).
[0311] By applying a procedure such as the one described above for
AP2 in general, and a basic AP2 formula (see; Eq. (31)) in
particular, a set of I.sub.k-values, for k=1, 2, 3, . . .
representing various hypothetical light sources can be obtained.
Thus, the optimization path leading to the ISON minimization can be
predicted, including a variety of practical factors, such as, for
example:
[0312] a) Type of photodetector and its geometry
[0313] b) Particular diffraction case
[0314] c) Specific Fresnel reflection effects
[0315] d) Various types of light sources
as summarized in Table 2, below.
[0316] The ISON values are represented by I.sub.k-values. The
practical factors: 1, 2, 3, 4, 5, 6, as summarized in Table 2, can
be varied, leading to a diversification of ISON-values (represented
by I.sub.k-values, as in Eq. (31), leading in turn to a broad
variety of possible scenarios that can be exercised for ISON
optimization purposes.
TABLE-US-00002 TABLE 2 Summary of Practical Factors for ISON
Minimization *) No. Factor Description Critical Parameter Symbol 1
Type of Photodetector Calibration Parameter K 2 Geometry of
Photodetector Diffraction Coordinates r.sub.k, .delta..sub.k 3
Cutoff Geometry Incidence Angle .alpha..sub.k 4 Fresnel Reflection
Fresnel Intensity Reflection R Coefficient 5 Particular Diffraction
Complex Amplitude |U|.sup.2 Case Modu!e Square 6 Light Source
Intensity Source Intensity I.sup.(s) Distribution *) For specific
averaged wavelength .lamda.-value.
[0317] In various embodiments of the weather system devices
disclosed herein, the precipitation sensing and measurement systems
may be included. In some embodiments, such systems can comprise
three primary subsystems. These can include a precipitation
characterizer, a precipitation quantifier, and a precipitation
classifier. The precipitation characterizer in some embodiments
includes a subsystem configured to determine a precipitation type.
The precipitation quantifier may in some embodiments include a
subsystem that primarily measures precipitation amounts. The
precipitation classifier may, in some embodiments, include a
subsystem that makes a final determination of a type of detected
precipitation. Each of these subsystems are now described.
[0318] FIG. 15 is a diagram illustrating an example precipitation
characterizer subsystem. The precipitation characterizer subsystem
701 in this example includes a precipitation receiver 702, which in
the illustrated embodiment is a flat plate upon which precipitation
falls, and an accelerometer 703 that detects the impact of falling
precipitation on that plate by the detection of shock waves
travelling through the plate and manifesting as accelerations or
micro-accelerations at the accelerometer. By monitoring and
evaluating the signal from the accelerometer, the difference
between hail 704, small hail or rain 705, or drizzle 706 can be
largely distinguished according to the classification of peak
impact 707 resulting from the impact of each droplet or hailstone.
(The precipitation classifier, which is described in detail below
is the subsystem that evaluates the signal and makes the final
determination.)
[0319] Although the precipitation receiver is depicted here as a
simple flat plate, this surface can be of any shape, but preferably
a shape that will not allow water to pool on the top surface, which
would otherwise mute the impacts of successive precipitation.
Another embodiment of the disclosed technology utilizes a
mechanical case of the system itself as the precipitation receiver
without requiring the addition of a separate plate. An example of
this uses the plastic or metal (or other material) enclosure of a
weather sensor as a precipitation sensor by the addition of a
simple accelerometer.
[0320] Although the arrangement of such a simple system as a means
of determining may seem obvious, this is not so without the use of
hindsight. Arrival at this solution occurred out of an evolution of
ideas that first began with the use of an electromagnetic voice
coil (e.g., a conventional acoustic speaker) to detect impacts on
the speaker cone directly. This method was later refined to use a
Piezo element that detected impacts on the Piezo element directly.
It was only after extensive experimentation that the discovery was
made that impacts on the board to which the Piezo element was
mounted produced a muted, yet measurable signal. This, after
further work, led to the discovery that the Piezo element acted as
an accelerometer in this mode of operation, thereby enabling this
solution.
[0321] Although a precipitation amount (as typically noted in
inches per hour) can be measured by an extension of the technique
used to measure precipitation type, this method is not highly
accurate. The wide variation in impact strength of different
droplet sizes a makes the counting of impacts inaccurate for the
assessment of volume, and heavy mist or drizzle is not detected by
this means. To accurately assess the precipitation amount, a means
of assessing true volume in various embodiments utilizes a
different subsystem with a completely different structure and
methodology.
[0322] FIG. 16 is a diagram illustrating an example precipitation
quantifier subsystem in accordance with one embodiment of the
technology disclosed herein. Referring now to FIG. 16, in this
example, the precipitation quantifier subsystem 721 includes a
funnel 722, a drop of former 723, a droplet detector 724 and a
droplet counter 725. In this example apparatus, funnel 722
accumulates moisture from a number of different forms of
precipitation including, for example, drizzle, rain, snow, etc.
over the collection area of the funnel 722. The collected
precipitation is directed to the droplet former 723. Droplet former
723 in this example utilizes a bi-conical shape 726 that draws
water in from the funnel 722 while forming droplets at its output
727. Preferably, is configured such that the droplets are of a
uniform physical size that drop onto the detector below.
[0323] In this example, the droplet detector includes two
electrodes spaced apart from one another (e.g., 0.2 inches apart)
that are disposed on a hydrophobic substrate. Because of the
hydrophobic nature of the substrate, droplets landing on the
detector are in effect repelled by the substrate and cause to move
quickly across the electrodes. Because precipitation droplets have
a much higher conductivity (typically ranging from 200.OMEGA.cm to
200,000.OMEGA.cm) compared to air, the conductivity between the
electronics increases (e.g. spikes) when a droplet passes across
the electrodes. This increase or spike can be detected by applying
voltage to the electrodes and measuring the electric current
through the droplet. An electronic pulse counter 725 can be used to
count the drops from the drop former, and inaccurate measure of the
true volume can be assessed across a number of different types of
precipitation.
[0324] In various embodiments, the accuracy of the sensor may be
governed by the degree to which the droplet former creates
consistently-sized droplets, since the technology for accurate
counting (e.g. a pulse counter) is well understood.
[0325] FIG. 17 is a diagram illustrating a droplet former
quantifier in accordance with one embodiment of the technology
disclosed herein. As can be seen from this example, droplet former
741 includes an internal taper 742, which is configured with a
geometry so as to allow a specific amount of mass to pass through
it until it overcomes the surface tension that a droplet encounters
at the outlet orifice 743. The internal taper may be defined by an
optimized ratio of the inlet 744 to the outlet orifice diameters as
well as its overall length 745. If the internal taper 742 is too
restrictive, then the droplets may take a considerable amount of
time to form. On the other hand, if the internal taper 742 is too
large, then the droplets either combine to form a single stream, or
have no uniformity as they exit the outlet orifice.
[0326] With regard to detection of the drops, those versed in the
art will also note that water droplets can be detected by means
other than conducting electrodes. For example, the presence of
droplets can be detected by capacitive, mechanical, and optical
means.
[0327] It is noted that the use of a funnel as a means of
concentrating precipitation for the purposes of measurement is a
conventional technique. However, conventional techniques have not
used a funnel-shaped apparatus for gathering moisture for
converting to a homogenized (all of equal size) droplet stream and
exploiting the fact that counting of equal-sized droplet yields a
measure of total volume or volumetric rate.
[0328] The two above-describe subsystems are nearly sufficient for
the measurement of precipitation type and amount. However, as
introduced above, the precipitation characterizer cannot fully
distinguish all types of precipitation. Specifically, embodiments
of the characterizer cannot detect drizzle or snow, and cannot
distinguish the impacts of small hail from heavy rain.
[0329] Accordingly, a precipitation classifier can be included to
perform such classification. The precipitation classifier is the
third subsystem of the sensors set, and includes logic that
combines inputs from both the precipitation classifier and the
precipitation quantifier into a final determination of
precipitation type. Table 3, below, shows an example of a logical
truth table that the classifier can implements to determine the
final precipitation type.
TABLE-US-00003 TABLE 3 Logical Truth Table for Classifier Impact
Level Moisture Measured Temperature Type High No n/a Hail Medium No
n/a Small Hail Low/High/Medium Yes n/a Rain None Yes Above Freezing
Drizzle None Yes Below Freezing Snow
[0330] Any combination that doesn't map to one of the input values
is labeled "unknown precipitation" until the classifier is able to
map the precipitation to one of the known types. Types which remain
"unknown" can be assumed to be one of several more rare types
including ice grains, snow grains, blowing sand, or volcanic
ash.
[0331] Instrumentation in various forms, including instrumentation
as described above can be delivered via a number of different
mechanisms, including delivery and installation by personnel. In
some applications or deployments, it may be desirable to deliver
the instrumentation via air deployment. For example, it may be
desirable to deliver the instrumentation by deploying it from an
airborne platform such as a fixed- or rotor-winged aircraft.
Accordingly, delivery mechanisms can be provided that can house the
instrumentation and be dropped, released or otherwise deployed from
an aircraft to carry the instrumentation to the ground with or
without power.
[0332] In various embodiments, the deployment can be configured
with an aerodynamic shape (e.g., somewhat similar to that of a
dart) to facilitate flight from the airborne platform to the ground
(or other deployment location). The mechanism can include a
weighted tip (or a forward-end weight bias), an elongated body
section, and a set of stabilizing fins at the aft end. This shape
facilitates a vertical orientation during free-fall flight from the
airborne platform to the deployment site. With sufficient weight at
the fore-end, or tip, a vertical or near-vertical fall trajectory
can be achieved even in high winds and at high initial deployment
velocities. The weight at the fore-end can even assist causing
penetration of the tip into the soil or other deployment surface.
Accordingly, the weight can be chosen based on anticipated weather
conditions at the drop location and soil hardness or density at the
deployment site.
[0333] FIG. 18 is a diagram illustrating a side view of a
deployment assembly in accordance with one embodiment of the
technology described herein. The example illustrated in FIG. 18
also illustrates the deployed equipment carried by the assembly.
Referring now to FIG. 18, the example assembly includes a mast 750,
a fin assembly 752, a stopper 755, and an anchor spike 754. Also
illustrated in the example of FIG. 18 is an example payload 757,
which, in this example, is a portable weather station. Payload 757
can be any of a number of different payloads, including portable
weather stations as described in this document. Although not
illustrated, a connection mechanism or other connector can be
included to allow the payload to be mounted to the deployment
assembly. For example, a threaded mount can be provided to allow
the payload to be screwed onto the deployment assembly. Likewise, a
snap fit assembly, bayonet mount, friction fit assembly or other
like connector can be provided to engage the payload (i.e., to
engage a complementary connector or fitting of the payload).
[0334] In the illustrated example, fin assembly 752 comprises four
fins arranged circumferentially about the body of the assembly,
although other quantities of fins can be provided. One purpose of
the fins is to provide stability during flight to allow the
assembly to maintain its orientation and intended flight path. This
can be used to help prevent the assembly from tumbling during
return to the ground. Fins can accomplish this by moving the center
of pressure of the assembly aft of the center of gravity. A
weighted tip also helps to accomplish this objective by moving the
center of gravity forward. A different number of fins can be used
as appropriate for the given application or environment. For
example, 2, 3, 4, 5, 6 or more fins can be used as appropriate,
although fewer than three fins is preferably avoided. Three to four
fins are adequate for most applications. Likewise, fin area, fin
shape (i.e., planform), thickness, and so on, can vary from that
shown.
[0335] Although not illustrated, the fins can also include fairings
at the joint between the fin and the body. The fairings can be used
to help reduce interference drag if desired. However, it is noted,
that fairings can also affect (e.g., potentially increase) the
terminal velocity as well. Also not illustrated are features that
may be included such as radial tapers on the leading and trailing
edges, which can be used to reduce drag and provide a more
efficient shape; and airfoils, which can further reduce drag. As
noted above, however, reduction of drag should be considered as a
factor in determining the desired terminal velocity of the
assembly. It may not be desirable to have too high a terminal
velocity, as a higher terminal velocity tends to create a greater
shock to the equipment on impact.
[0336] In the illustrated example, the body of the assembly is
formed by mast 750. In this example, mast 750 is a single piece,
cylindrical in shape, and has a relatively uniform diameter from
the fore to the aft end. In other embodiments, the body can be of
other shapes, geometries and sizes, and can be made from multiple
pieces and have a tapered or otherwise varying diameter along its
length. As described in more detail below, the body can comprise a
multi-segment mast or rod that can serve functions such as
cushioning the shock of landing, and extending the height of the
assembly upon landing. Additional shock absorbency or cushioning
can be provided at the point at which the instrumentation 757 is
mounted to the assembly.
[0337] Anchor tip or spike 754 can be tapered as illustrated to
allow easier penetration into the soil or other deployment surface.
Stopper 755 can be included to help the penetration of the assembly
into the deployment surface at a predetermined depth. Either or
both of anchor spike 754 and stopper 755 can be weighted to bias
the weight of the unit toward the nose. Likewise, body section 750
can be weighted toward the nose. Weighting toward the nose of the
assembly assists in orienting the assembly in a nose-forward
position (e.g., nose-down) during its return to Earth.
[0338] As noted above, the terminal velocity is affected by factors
such as the total mass of the assembly and its drag coefficient.
The terminal velocity, V.sub.t can be determined mathematically as
follows:
V t = 2 mg .rho. A C d ( 33 ) ##EQU00016##
where m is the mass of the assembly, g is the acceleration due to
gravity, .rho. Is the density of the fluid (e.g. the air) through
which the assembly is traveling in its return to Earth, A is the
projected area of the assembly and C.sub.d is the drag coefficient
of the assembly.
[0339] Table 4, below, provides example design parameters for
example assembly having a terminal velocity of 100 mph. in this
example, the unit has a total mass of 3,150 grams, and a drag
coefficient of 0.75.
[0340] As noted above, aerodynamic analysis of this design
estimates a terminal velocity of roughly 100 mph. Based on the time
it takes for the assembly to reach its terminal velocity, it is
estimated for this design that a speed of 35 mph (sufficient to
fully penetrate most soils) will be reached after only a 40 ft
drop. However, those of ordinary skill in the art reading this
disclosure and its teachings will see that the design parameters
can be easily changed to achieve a desirable terminal velocity or
impact shock, or other related objective, for the deployment
assembly. In addition to providing verticality and relative
insusceptibility to winds, this approach utilizes the downward
force of the fall to anchor the system into the ground. Even if
small rocks or other surface irregularities are encountered, the
assembly can be configured such that its moment of inertia is large
enough to overcome any forces that would bias it from vertical. In
addition, because the terminal velocity is limited at 100 mph (or
other velocity as designed), the system can be dropped from any
height. A cushioning mechanism in the riser pole (described below)
spreads the deceleration due to impact over time, thereby limiting
the shock force that the equipment experiences.
[0341] Because the tip of the falling assembly is expected to hit
the ground with sufficient force to penetrate into the earth for
anchoring, it is desirable to reduce the level of G-shock
encountered by the equipment being deployed. To reduce that shock
load, the mast above the spike may be designed to both re-coil
(e.g., be compressed) and then expand upon impact. This mechanism
can be configured to not only reduce the total shock level but also
to increase the overall height of the deployed equipment without
requiring the pole to be at its full deployed length when dropped
from the aircraft. Accordingly, any of a number of design
mechanisms can be included with the assembly to accomplish these
objectives. For example, mast 750 can be spring-loaded (e.g. at its
junction with, or somewhere above anchor tip 754) to absorb the
shock of impact. As another example, mast 750 can be segmented to
provide shock absorbency and the ability to telescope to a fully
deployed height. In various embodiments, a segmented mast 750 can
be implemented as two or more interlocking coaxial tubular members
that can be configured to be compressed for storage in flight, and
to telescope to an expanded height upon impact with the ground.
[0342] FIG. 19 is a diagram illustrating an example of a
multi-segment riser pole before, during, and after impact in
accordance with one embodiment of the technology disclosed herein.
In this example, two segments are shown, however, after reading
this description, one of ordinary skill will appreciate that more
than two segments can be used depending on the final height
desired. For clarity of description and to better facilitate
understanding by the reader, the assembly is depicted without fins.
As seen in FIG. 19, the length of the riser pole can be configured
to reduce in length in the moments after impact to absorb the
G-shock that the equipment experiences and then to gradually
lengthen to full length after impact to fully elevate the equipment
to the desired height.
[0343] As this example illustrates, at 762, moments before impact,
the assembly is at a reduced height selected for the unit's
return-to-ground flight. The upper segment of the mast is
partially, but not necessarily completely pushed into the lower
segment. In the next four images 764, during and immediately after
impact with the ground surface, it can be seen that the mast
compresses to absorb the shock of impact with the ground surface.
In various embodiments, the upper (inner) mast segment is
spring-loaded to absorb the shock of impact. Upon impact, the upper
mast segment is translated by the force of gravity to move further
into the lower (outer) mast segment thereby increasing the stopping
distance of the instrumentation, and cushioning or absorbing the
shock on the instrumentation mounted at or near the top of the
mast.
[0344] In the next six images 766, the mast telescopes to desired
height. In various embodiments, the telescoping operation can be
triggered by the impact with the ground surface. For example, a
spring-loaded locking or catch mechanism can be used to hold the
mast in its in-flight position (762) during its fall to the earth.
Depending on the design, the shock upon impact can cause the catch
mechanism to release, allowing the spring force applied to the
upper mast segment to cause the upper mast segment to expand as
shown at 766. As another example, a locking or catch mechanism can
be provided that is released when the lower end of the inner
segment reaches a certain point within the outer segment.
[0345] As will be appreciated by one of ordinary skill in the art
reading this description, the number and length of the segments can
be chosen to achieve the desired in-flight and deployed lengths.
For example, in one embodiment, when the assembly is released from
an aircraft, it is roughly 2 ft in length to make it more
manageable prior to deployment. Although it initially shrinks to
absorb impact, it eventually lengthens to 3 ft after it is fully
anchored. This ideal combination can be readily achieved by the use
of mechanical springs, pneumatic cushioning, or other shock
absorbency techniques to absorb and dissipate impact energy, a
simple spring that pushes the pole to expand at final deployment,
and a quick-release latch that keeps the pole in a shortened state
prior to drop. Those versed in the art can see that this mechanism
can be scaled up or down to larger and smaller sizes.
[0346] Although the stabilizing fins are desirable to facilitate or
ensure vertical deployment of the equipment, they can be withdrawn
from the area around the equipment in order to avoid or minimize
interference with operation of the equipment. In various
embodiments, a mechanism can be provided that not only achieves the
displacement of these fins, but also utilizes the fins for added
stability of anchoring. In further embodiments, the fins can be
designed such that they can even stabilize the assembly on a solid
or impenetrable surface such as, for example, rock, frozen ground,
a building rooftop, and so on. The fins can also provide additional
stability on extremely soft surfaces such as granular sand, peat,
mud, and so on.
[0347] In various embodiments, the mechanism for causing
displacement of the fins from an in-flight position to a deployed
position can be configured to work by allowing the shock-force of
impact to carry the fins down to the ground upon impact and to
pivot the fins down to the point that they meet the ground.
Furthermore, in various embodiments, a mechanical ratcheting
mechanism can be used to lock each fin in place once the fins reach
the downward, deployed position.
[0348] FIG. 20 is a diagram illustrating an example of this
displacement and ratcheting in accordance with one embodiment of
the systems and methods described herein. Referring now to FIG. 20,
image 772 depicts the assembly moments before impact. As seen by
this example illustration, the fins are in their in-flight position
toward the aft end of the assembly to provide a center of pressure
aft of the center of gravity, thereby making a more stable
flight.
[0349] As discussed above, a mechanism can be provided (e.g. a
spring-locking or other mechanism) that locks the fins into their
in-flight position during flight, but releases upon the shock of
the impact. Immediately after impact, as shown by images 774, after
the catch mechanism is released, the momentum of the fins and the
force of gravity cause the fins to move along the body in a
downward direction toward the deployment surface. The continued
momentum of the outer ends (tips) of the fins causes them to
continue to move in a downward direction once the body upon which
the inner end of the fins is mounted ceases its downward movement.
At 776, the fins are fully deployed, providing a more stable base
to the unit. As also shown at 776, the mast is forced by its spring
mechanism to rise to its final deployed height.
[0350] As also seen in this illustration, the fins can be mounted
on a sliding ferrule or other tubular or ring-like member
configured to be able to slide from the aft end to the tip for
deployment. A catch or spring-loaded mechanism can be provided to
maintain the fins at the aft end during flight and to allow the
fins to fall to the deployed position as a result of the shock of
impact. A catch mechanism can also be provided at the tip end to
lock the fins in place upon deployment. As FIG. 20 illustrates, a
pivot mechanism can also be provided at the ferrule or ring to
allow the fins to pivot from their upright in-flight position to
the deployed position contacting the ground. For example, a pin,
hinge or other like mechanism can be provided about which the fins
may pivot.
[0351] Rubber or rubberlike feet can be provided at the tips of the
fins to provide a more stable base for deployment. The contribution
of rubberlike feet to the drag of the assembly should be considered
for designs that include such a feature. In the embodiment
illustrated in images 776 of FIG. 20, the fins are shown to include
a pointed protrusion at the tips of the fins, which can be used to
penetrate the deployment surface and provide increased ability.
[0352] Although ideal for conditions with flat and penetrable
ground, in various embodiments the mechanism can also be configured
for deployment on steep slopes. FIG. 21 is a diagram illustrating
an example of deployment on the slope using ratcheting fins. As can
be seen by the embodiment of FIG. 21, the fins can be configured to
ratchet beyond 90.degree. from the mast to allow stable footing on
a sloped or irregular surface. Additionally, each fin can be
positioned independently of the others to allow stable footing on
such sloped or irregular surfaces. As seen at image 784, in this
example, the assembly includes the same in-flight configuration for
the fins. At 786, the fins are released and travel downward along
the mast until each fin reaches the deployment surface. As seen in
these images, once each fin reaches its grounded position, its
movement ceases. A stepped ratchet mechanism can be used to lock
the fins in place at their downward-most position. Continuing to
images 7808, it is seen in this example that the remaining two fins
continued until they contact the deployment surface, ceasing their
downward motion and locking into place by the ratcheting mechanism.
As this illustrates, the mast can remain in a vertical or near
vertical position and the fins can be deployed to conform to the
contour of the ground or other deployment surface to provide
stability.
[0353] In cases where the penetrating ground spike hits a rock or
other impenetrable surface and is unable to penetrate, or where it
hits a muddy or sandy surface with poor stability, the fins provide
a backup that enables the platform to remain stable and vertical.
Using the combination of both a penetrating ground spike and a
lateral base formed by the fins gives the assembly stability and
effectiveness on a wide variety of terrain.
[0354] It is noted that the ground spike can be configured to
provide anchoring in penetrable surfaces such as soil, sod, gravel,
clay, firm mud, firm sand, and combinations of these. Although not
illustrated, the ground spike can include a relatively sharp tip to
allow penetration and a shoulder (like a broadhead arrow rotated in
360.degree.) to provide a more firm hold into the ground.
[0355] Also, as the above illustrates, the fins can be configured
to provide a stable platform on impenetrable surfaces (rock) and
provide additional stability on soft and extremely soft surfaces
including sugar sand and soft mud. Both the spike and fins can be
configured to be capable of working just as effectively on uneven
terrain and on slopes of 45.degree. or more without affecting their
performance.
[0356] The combination of the re-coiling pole and ratcheting fins
as described above can, in most embodiments, be configured to
provide air emplacement from low-speed rotorcraft or from
high-speed fixed wing aircraft flying at high altitudes. However,
when dropped from a low altitude at a high speed, enough horizontal
speed from the drop is carried through on the flight to the ground,
which can lead to a non-vertical impact angle. FIGS. 22A and 22B
illustrate examples of horizontal travel as a result of speed of
the deployment aircraft 789.
[0357] To address these deployment scenarios where equipment must
be dropped from a low altitude and a high speed simultaneously, the
assembly can be configured to include a "flash parachute" or other
like mechanism that can be attached to the air-drop fin system
prior to release. The parachute in such embodiments can be referred
to as a "flash" parachute because it is designed to be open for a
short period of time (e.g., only 2-3 or 3-5 seconds, or other time
interval as appropriate) to stop the lateral motion of the system.
After deployment, the parachute is configured to be released to
drift away while the assembly and its equipment drops vertically to
the ground. Using a parachute that remains deployed longer than
this short period of time could cause the assembly to be carried
off course due to wind conditions in the area and could slow the
descent to such a rate that penetration of the anchor spike is
hindered. In other words, in various embodiments, the parachute is
not used to deliver the equipment to the ground or to limit the
downward fall speed--it is used to arrest its horizontal
movement.
[0358] FIG. 23 is a diagram illustrating the operational release of
a flash parachute in accordance with one embodiment of the
technology described herein. As seen in this example, upon initial
release from the aircraft (image 792), the assembly begins to fall
but will continue to travel at or near the airspeed of the delivery
aircraft, but is slowed by the atmosphere. The flash shoot begins
to deploy, slowing the unit's horizontal airspeed. At images 794
and 796, it can be seen that the parachute opens further as it
resists the flow of air, further slowing the horizontal airspeed.
As also shown at images 794 and 796, the weight bias toward the tip
of the assembly continues to cause the assembly to move to a
vertical orientation. Accordingly, to avoid having the parachute
unduly impact the vertical airspeed of the assembly, the parachute
is released from the assembly as shown in image 798. Accordingly,
the timed-release mechanism holds the parachute in place to arrest
horizontal movement and releases to allow the unit to fall to the
ground and preferably reach its terminal velocity. In some
embodiments, the timed-release mechanism can be implemented using a
timer that triggers the release mechanism after a certain amount of
time has elapsed from release from the aircraft. In other
embodiments, motion or acceleration sensors can be used to trigger
the release of the parachute from the assembly at the moment the
sensors detect that sufficient horizontal movement has been
arrested. In yet another embodiment, the parachute release
mechanism can be configured to release the parachute once
sufficient pull is provided by the parachute. For example, force
sensors can be used to detect the amount of pull and trigger the
release mechanism. As another example, the parachute release
mechanism can be a friction coupling that releases with sufficient
force from the parachute packs on the coupling.
[0359] In various embodiments, the weather station system/device
can include other advanced capabilities such as weather event
predictive analysis and information protection tools, the latter
ones related to so-called Information Assurance (IA). Such tools
and technology can also be used at a central weather station or at
any other location or facility used to perform predictive analysis.
It is not known, nor would it be obvious to use these capabilities
as applied or tuned to weather station specificity or to weather
event prediction. Accordingly, a detailed discussion follows. Also,
in various embodiments, these capabilities are achieved by
including software engines and algorithms, or other like modules,
allowing them to address the SWaP2 constraints, where P2 refers to
both the power supply and processing power.
[0360] One aspect of non-obviousness in the weather event
predictive analysis context arises in microclimate weather
prediction, which requires prediction of so-called Weather
Anomalous Events, or WAEVENTS, based on a heuristic software engine
and so-called cybersensing, based on Bayesian inference.
[0361] One non-obvious aspect of information protection tools that
can be included in embodiments herein is in IA-key management and
wireless/RF transmission, including through harsh weather
communication channels. In some embodiments, the IA-keys include
encryption keys and injection keys, as described below. This is,
because, the allowance of even a single uncorrected error can
inhibit the ability to use the keys as intended, while a high
assurance of error correction by error-correcting-codes can be
bandwidth overhead (OVH) cost prohibitive. Therefore, in some
embodiments, specific solutions to this challenge are
incorporated.
[0362] In some embodiments, the detection and identification (ID)
of Weather Anomalous Events, or WAEVENTS, detrimental to weather
predictive analysis, especially in microclimate conditions, relies
on detecting and identifying Digital Topologic Singularities (DTS)
as part of Digital Singular Mapping (DSM). This can be the case for
both linear and non-linear DTS, the latter including catastrophes
as defined by mathematical theory of catastrophes. While the theory
of catastrophes is known, their heuristic detection and
identification is unobvious and unknown in prior art, and may be
incorporated in various embodiments of the technology disclosed
herein. Various embodiments related to weather predictive analysis
and information protection tools, may be software-based and may
thus, avoid the use of hardware constraints of the weather station
system/device, especially in cases in which the RF networking and
satellite communication are useful, or even required.
[0363] Various embodiments of the systems and methods described
herein can be configured to provide an optimum or more ideal
solution for SWaP (Size, Weight and Power) constraints.
Additionally, embodiments can be implemented to provide a systemic
solution to command-and-control (C2) issues, including high
bandwidth communication issues and satellite communication issues.
In further embodiments, modules can be implemented (e.g., using a
software engine) for: detection, identification and recognition of
weather anomalous events. The system can also be configured to
provide novel weather sensing schemes, including optical visibility
measurement.
[0364] Various embodiments can be configured to provide a module
for weather anomalous event detection, identification, and
recognition. The Weather Anomalous Event System (WAES), may be
based on binary cybersensing, including a double-alarm digital
decision generation (DDG) scheme. It may be based on a cybersensor
fusion making decision: determining whether a weather event is
anomalous (alarm), or not (no alarm); and Bayesian Figure of Merit
(FoM) (e.g., a Positive Predictive Value (PPV), similar to that as
used previously extensively in medicine, in general, and in X-ray
Mammography in particular).
[0365] Embodiments can also be configured to apply a Bayesian
Inference, in a novel way, by applying the Bayesian statistics for
weather anomalous events, which can be referred to herein as
WAEVENTS. An exemplary way in which WAEVENT may be configured as a
target to be detected by the WAES is due to a cybersensor set:
C.sub.1, C.sub.2, C.sub.3, . . . C.sub.n, where n is the total
number of cybersensors. These sensors may be connected with two
buses: a sensor bus, or C-bus; and, a microprocessor (.mu.P)-bus,
or P-bus. The cybersensors are fed by the computer cloud or
database, upgraded from PC-interface. The output WAEVENT, or red
alarm, is outputted to PC, while yellow alarms are stored in .mu.P
for further consideration.
[0366] FIG. 24 is a diagram illustrating an example Truthing-based
Anomalous Event Software Engine (TAESE) that can be used to
implement a WAES in accordance with one embodiment of the
technology described herein. In the example illustrated in FIG. 24,
the WAES schematic 1100 is shown as including two data buses 1101
and 1102, as well as database 1103, microprocessor (.mu.P) 1104,
cybersensors 1105, 1106 and 1107, and Binary Decision Generator
1108 as basic WAES sub-systems or components. Weather data 1109 is
used as input to the WAES system 1100. In various embodiments, this
can be a formatted weather event, or Structured Weather Event
(SWE). In various embodiments, the weather event is input to the
system as a statistical quant or sample 1109 and is investigated by
the cybersensors in the system (e.g., 1105, 1106, 1107 in the
illustrated example). In various embodiments, a greater or lesser
quantity of cyber sensors can be utilized. The cyber sensors can be
used to produce an anomaly ranking. The anomaly ranking can be
summarized at Voting Logic Gate (VLG) 1110 to produce a yellow
alarm or no-alarm. Also memory (such as flash memory, for example)
1111 can be added.
[0367] In various embodiments, the received weather data 1109 may
be pre-structured, or formatted in the form of a statistical quant
of information. This may be referred to herein as a Weather
Information Statistical Quant (WISQ). The WISQ may include metadata
such as, for example, geospatial coordinates: (x, y, z), temporal
coordinate, t, as well as weather data parameters, such as, for
example: [0368] Temperature [0369] Atmospheric pressure [0370] Wind
speed and direction [0371] Rain volume [0372] Relative humidity
[0373] Insolation (solar radiation) [0374] Camera cloud high [0375]
Visibility [0376] Others
[0377] Cyber sensors 1105, 1106, 1107 can be used to verify the
weather data parameters to determine whether or not to produce a
weather anomalous event (WAEVENT). Therefore, in various
embodiments, the weather event space is binary, producing either a
WAEVENT (or signal, S); or noise, N (no-WAEVENT). Based on sensor
readouts made available to voting gate logic (VLG) 1110, the binary
decision generator (BDG) 1108 produces either an alarm, S'; or
no-alarm, N'. In various embodiments, the basic paradigm of
Bayesian Inference may be based on two absolute event
probabilities: p(S), p(N) which may be exclusive events. Therefore,
the sum of these two exclusive events results in certainty:
p(S)+p(N)=1. In terms of a sensor response, the BDG causality
relation results in two exclusive readout probabilities:
p(S')+p(N')=1.
[0378] Also, there may be four (4) direct conditional
probabilities: p(S'|S), p(N'|N), p(S'|N), p(N'|S). These are,
respectively: probability of detection, probability of rejection,
probability of false positives and probability of false negatives.
Those conditional probabilities satisfy two conservation relations:
p(S'|S)+p(N'|S)=1, and p(S'|N)+p(N'|N)=1. Based on Bayes Theorem,
embodiments can be configured to also introduce four (4) inverse
(Bayesian) conditional probabilities: p(S|S'), p(N|N'), p(S|N'),
and p(N|S'). The probability, p(S|S') is referred to as Positive
Predictive Value (PPV). By using the Bayesian Truthing Theorem
(BTT), it can be determined that: (PPV)=p(S|S') is equal to the
ratio of true alarms, a.sub.1, to total number of alarms, a, which
is the basic Key Performance Parameter (KPP) of so-called Bayesian
Truthing, introduced by analogy to RADAR truthing in 1960s.
Accordingly, Bayesian Truthing (BT) may be used in various
embodiments to introduce absolute measurable quantities in the
analysis, rather than relying solely on relative parameters such as
statistical probabilities. This, in various embodiments, can result
in simplifying the WAEVENT structure to be more useful for
experimental measurement and experimental verification. However,
there is equivalence between Bayesian statistical formulas and
truthing formulas.
[0379] For example, the probability of false positives (PFP),
is
( PFP ) = p ( S ' | N ) = lim n -> .infin. ( a 2 n ) ( 34 )
##EQU00017##
where a.sub.2 is the number of false alarms, n is the number of
no-targets, and asymptotic limit (n.fwdarw..infin.) provides the
equivalence between Bayesian Statistics and Bayesian Truthing, the
latter one based (for binary case) on a number of parameters. In
various embodiments, there may be nine (9) parameters: m, s, n, a,
a.sub.1, a.sub.2, b, b.sub.1, b.sub.2, which are the number of:
statistical quants, targets, no-targets, alarms, true alarms, false
alarms, no-alarms, true no-alarms and false no-alarms,
respectively. In order to have statistics valid, various
embodiments make the m-value a large number (e.g., assume:
m.gtoreq.10.sup.9). The challenge with anomalous events, such as
WAEVENTS, is such, that, as anomalous events, the targets are rare
(e.g., anomalous weather, by definition, is a rare event in
comparison with normal weather); i.e., s-number is small
quantity:
S<<m (35)
[0380] However, for targets as rare events, the PPV-value is
usually small. This is because, in good approximation, the PPV
value is:
( PPV ) = 1 1 + P ( S ' | N ) p ( S ) = 1 1 + ( PFP ) p ( S ) ( 36
) ##EQU00018##
[0381] In order to illustrate this case, assume the probability of
false positives, PFP=10.sup.-6, which is a rather low number, and
calculate positive predictive value (PPV) as a function of target
population, p(S). An example of this is shown in Table 5.
TABLE-US-00004 TABLE 5 (PPV) vs. p(S), for (PFP) = 10.sup.-6 p(S)
10.sup.-8 10.sup.-7 10.sup.-6 10.sup.-5 10.sup.-4 10.sup.-3
10.sup.-2 (PPV) 0.0099 0.09 0.5 0.91 0.99 0.999 0.9999
[0382] According to Table 5, for a very low target population such
as s=10 (then, p(S)=10.sup.-8, for m=10.sup.9), the result is a
very low PPV-value of 0.0099. Only for s.gtoreq.1000
(p(S)=10.sup.-6), the PPV-values are large. The threshold is:
p(S)=(PFP)(PPV)=0.5 (37)
i.e., when (PPV)-value is 50%.
[0383] Therefore, there is a fundamental problem with recognizing
anomalous events, such as WAEVENTS, which embodiments of the
systems and methods disclosed herein are configured to solve in an
unique way. In various embodiments, the solution is implemented
using a multi-step process (in a similar way as in X-ray
mammography in the case of breast cancer diagnosis) by applying a
sequence of sensors with very low target misses, i.e., for
p(N'|S)=(PFN)<<1, when false negatives are very low (or,
probability of false negatives, PFN, is very low).
[0384] In order to solve this rare target problem, embodiments can
be configured to apply a WAEVENT Sensor Fusion (WSF), (discussed in
detail below). The WSF is based on Bayesian Truthing rather than on
Bayesian statistics. Therefore, a number of conservation relations
may be applied for in the nine truthing parameters mentioned above.
For example, this can be in the form of:
m=s+n (38)
m=a+b (39)
a=a.sub.1+a.sub.2 (40)
b=b.sub.1+b.sub.2 (41)
s=a.sub.1+b.sub.2 (42)
n=b.sub.1+a.sub.2 (43)
[0385] Among these six (6) equations, only five (5) of them are
independent. Therefore, among nine (9) truthing parameters, four
(4) of them are free.
[0386] For the sake of clarity, consider a 2-step process, or
WAEVENT Sensor Fusion (WSF) with two cybersensors, or, simply, two
sensors: SENSOR 1 and SENSOR 2. Theoretically, the single ideal
sensor is possible with zero false positives (PFP=0) and zero false
negatives (PFN=0). In practice, however, using a medical analogy, a
more practical approach that can be implemented with various
embodiments is to apply the set of sequential sensors, all of them
with very low false negatives (target misses), yet, with
monotonically decreasing false positives. The number of sensors
cascaded is preferably two, or larger than two for cases in which
there is a problem with (PPV)-value increasing. Here, for
simplicity, we consider only two (2) sensors in cascade. For
purposes of providing an understanding, this discussion describes
the WSF quantitative analysis as an example.
[0387] For SENSOR 1, consider Input (Free) Parameters: m=10.sup.9,
s=10, b.sub.2=1, (PPV)=10.sup.-3. From Eq. (42), we obtain:
a.sub.1=s-b.sub.2=10-1=9. Since: (PPV)=a.sub.1/a, then
a=910.sup.3=9000. Also, from Eq. (40):
a.sub.2=a-a.sub.1=9000-9=8991; and, from Eq. (39):
b=m-a=10.sup.9-9000=999991000. However, according to Eq. (41) we
have: b.sub.1=b-b.sub.2=999990999, and from Eq. (43), we obtain:
n=b.sub.1+a.sub.2=9999999990. For checking, we verify that indeed
we have: n+s=999999990+10=10.sup.9=m, according to Eq. (38).
[0388] The output parameters of SENSOR 1 may become the input
parameters of SENSOR 2. Therefore, for SENSOR 2, we have Input
Parameters: m=9000, s=10, b.sub.2=1, (PPV)=0.99. Due to the
1.sup.st sensor, this 2.sup.nd sensor can afford very low both
false positives (high PPV) and false negatives (low b.sub.2-value).
Because of small s and b.sub.2 values, there is some uncertainty
with keeping all values as integer numbers. For example, it is
unclear whether the s-number should be 10, or 9 (because of
b.sub.2=1, in the 1.sup.st sensor case). Also, we should observe
that due to cascade values, the number of alarms from the previous
sensor (a=9000) becomes the number of statistical samples (quants),
in the case of the 2.sup.nd sensor:
m.sup.(2)=a.sup.(1) (44)
where the upper index is for sensor numbering.
[0389] Using the same approach as in the case of the 1.sup.st
sensor, we obtain: a.sub.1=s-b.sub.2=9, and:
a=9/0.99=9.09.apprxeq.10, where we approximate to higher integer.
Also, a.sub.2=a-a.sub.1=10-9=1, and: b=m-a=9000-10=8990. Then,
b.sub.1=b-b.sub.2=8990-1=8989, and: n=8989+1=8990. For checking:
n+s=8990+10=9000=m.
[0390] The results of both sensors in this example are summarized
in Table 6. We apply diagonal line: "/", as "or", due to the
difficulties associated with integer approximation. This is because
the number of statistical quants should always be an integer number
rather than a fractional number.
TABLE-US-00005 TABLE 6 Example Values of 9 Truthing Parameters for
Two Sensors' Cascade, Including: (2a)-SENSOR 1, (2b)-SENSOR 2 m s
b.sub.2 n b b.sub.1 a.sub.1 a a.sub.2 2a. SENSOR 1 10.sup.9 10 1
999999990 999991000 999990999 9 9000 8991 2b. SENSOR 2 9000
10/9*.sup.) 1 8990 8990 8989 9/10*.sup.) 10 1/0*.sup.) *.sup.)Due
to integer accuracy. In any case, both b.sub.2 and a.sub.2 values
are very low (b.sub.2 = a.sub.2 = 1); leading to the perfect
performance.
[0391] FIG. 25 is a diagram illustrating an example of WAEVENT
Sensor Fusion (WSF) 1200 for two cascaded sensors in accordance
with one embodiment of the systems and methods described herein. In
the example illustrated in FIG. 25, the cascade sensor function
includes SENSOR 1 1201, and SENSOR 2 1202. Pre-structured input
data 1203 are received and provided to cyber sensor 1201, resulting
in intermediate output data 1204. Intermediate output data 1204 can
be in the form of yellow alarms a.sup.(1), which are shown with an
example number of 9000, according to Table 6. These yellow alarms
become the inputs to SENSOR 2; thus, satisfying Eq. (44). This
results in the final output (in this example a red alarm) 1205.
This example illustrates that the first sensor 1201 has very low
false negatives, but rather high false positives. Indeed, its
(PPV)-value, (PPV).sub.1=10.sup.-3 (i.e. high false positives),
while its b.sub.2.sup.(1)=1 (i.e., low false negatives). The reason
for that is with a low target population 1206
(s.sup.(1)=s.sup.(2)=10), it is difficult, if possible at all, to
produce both low false positives and low false negative for the
1.sup.st sensor. Indeed, for (PPV).sub.1=10.sup.-3, 1207, and s=10,
a high number of false alarms, a.sub.2=8991, results.
[0392] FIG. 26 is a diagram illustrating an example of a WAEVENT
Sensor Fusion (WSF) Software Engine using two (2) cybersensors in
cascade. In the example illustrated in FIG. 26, WAEVENT Sensor
Fusion Software Engine 1300 is illustrated with two cyber sensors
1301, 1302 in cascade. However, in contrast to the example
illustrated in FIG. 25, the example of FIG. 26 also considers
outputs from other "first-type" sensors. The "first-type" sensors
may be 1.sup.st sensors in cascade, with low false negatives, yet
high false positives 1303--i.e., sensors with a large number of
false alarms. Additional first-type sensors, shown in this example
as as a-sensor 1304 and b-sensor 1305, are also contributing to the
input of the second-type sensor 1302. Strictly speaking, in
embodiments using a number of sensors greater than two (2), the
first-type sensor may be referred to as an introductory sensor, and
the second-type sensor may be referred to as a final sensor. This
is, because, in addition to these two sensors, there may be a
number of intermediate sensors. For example for a total number of
four (4) sensors, the number of intermediate sensors is two
(2).
[0393] In the example shown in FIG. 26, sensors 1304, 1305 need not
be considered intermediate sensors in the sense discussed above.
Therefore, in the example of FIG. 26, there are no intermediate
sensors. Contributions from other introductory sensors 1304, 1305
may come from the soft decision concept of the WAES, assuming that
some yellow alarms produced by these sensors may have been left for
further consideration. In such a case, the input sample space for
sensor 1302, m.sup.(2), may be the sum of yellow alarms a.sup.(a)
1306 and a.sup.(b) 1307 as shown by summation formula 1308.
[0394] The second sensor 1302 as the final sensor in this example
has both low false negatives and low false positives 1309. In other
words, the number of output alarms 1310 producing a red alarm 1311
includes only a low number of false alarms, thus realizing the
ultimate goal of a high PPV for the WSF software engine.
[0395] FIG. 27 is a diagram illustrating an example of a WAEVENT
Sensor Fusion (WSF) engine 1400 for four cascaded cyber sensors.
This WAEVENT Sensor Fusion (WSF) includes 4 cyber sensors 1401,
1402, 1403 and 1404. Among these sensors, sensor 1401 can be
referred to as the introductory sensor and sensor 1404 may be
referred to as the final sensor. Using this terminology convention,
sensors 1402 and 1403 are the intermediate sensors. These four
sensors have input sampling spaces, m.sup.(1), m.sup.(2),
m.sup.(3), and m.sup.(4), respectively. These four sensors have
output alarms: a.sup.(1), a.sup.(2), a.sup.(3), and a.sup.(4),
respectively, with a.sup.(4) being a red alarm. In all these
elements of cascade, the input space can be larger than number of
output alarms, in the form:
m.sup.(i+1).gtoreq.a.sup.(i) (45)
for i=1, 2, 3, as in FIG. 27, used as an example. The wings 1405,
1406, 1407, 1408, 1409, 1410, and possibly others, come from an
extra yellow alarm contribution due to the soft decision structure
of the system, resulting in an output figure with both low false
alarms, and false no-alarms.
[0396] Because formal structurization of weather data is either
very difficult, or even impossible, embodiments of the systems and
methods disclosed herein utilize the pre-structurization of the
weather data. In particular, embodiments utilize Weather Anomalous
Event (WAEVENT) pre-structurization. For example, embodiments may
be configured to create the Weather Data Event Format, including
its temporal and geospatial coordinates, in either 3D time-space
(x, y; t), or in 4D time-space (x, y, z; t), where z is the
altitude of the Region of Interest (Rol), while (x, y)-are terrain
coordinates, and t-time is a coordinate. Then, all Parameters of
Interest (PoIs), such as temperature, humidity, wind directions,
etc., (denoted by PoI1, PoI2, PoI3, etc.) may be introduced. A
Weather Anomalous Event Ranking, or WAER, may be provided to one or
more of the totality of PoIs. Voting logic and a Digital Decision
Support Engine (DDSE) may be used to produce a digital decision,
which can be a soft decision or a hard decision, in the form of a
dual-alarm structure, including yellow and red alarms. With such
embodiments, the system can be configured to categorize
Catastrophic Anomalous Events, or other Weather Anomalous Events,
as summarized in FIG. 27.
[0397] FIG. 28 is a diagram illustrating an example of a Weather
Data Event Format (WDEF) 1500 in accordance with various
embodiments of the systems and methods disclosed herein. In this
example, WDEF 1500 includes PoI name 5101 PoI value 1509 Weather
Anomalous Event Ranking (WAER) for single event 1502 and the WAER
for group of PoIs 1503 and PoI average 1510. The WAER-value is
provided as a percentage, as for exemplary WAERs; 1504, 1505, 1506,
1507, and 1508. For example, in this particular example, the
anomaly for geospatial coordinates (x.sub.1, y.sub.i) is only 10%,
denoted by 1504, while the anomaly for the group of parameters (#1
an #2) 1505 is 15%, and the same value, 15%, for #2, as denoted by
1506. Similarly, this example shows a single WAER 1507 and the same
value, 5%, for 1508. The number of parameters, n, is denoted by
1511, and the WAER ranking threshold 1512 is 65%. This also shows
the same value for 1513, 1514, and 1515; thus, there is no alarm
1516. However, if the value of 1513 is 72%, for example, these
would be an "alarm" for 1516.
[0398] The number of samples, m, is preferably be sufficiently
large to satisfy the Bayesian Statistics. For each sample, or
statistical quant, the WDEF document such as that illustrated in
FIG. 28 can be used as an element of pre-structurization.
[0399] Weather C2 Sensor (WC2S) in various embodiments provide a
systemic solution to weather sensing, and can also be configured to
include Command and Control (C2) capability for the Weather
Station. The 1.sup.st element of the WC2S, is a software engine,
which, in various embodiments, is a software engine such as that
introduced above. This document now describes two additional
aspects of the WC2S. These relate to satellite communication and
Information Assurance (IA).
[0400] Satellite communication and related bandwidth control are
important factors for weather C2 sensing. This is because the
limited bandwidth, B, can be transmitted through satellite
channels, for example. Of course, due to signal compression, the
original (raw) bandwidth, B.sub.0, is larger than the bandwidth, B,
typically transmitted. However, for the quantitative purposes,
there may be a significant bandwidth load due to overhead (OVH),
B.sub.OVH, which is some fraction, .epsilon., of the transmission
bandwidth, B:
B.sub.OVH=.epsilon.B (46)
where: .epsilon.-coefficient is the sum of various overhead
components:
.epsilon.=.epsilon..sub.Crypto+.epsilon..sub.FEC+.epsilon..sub.Net
(47)
where .epsilon..sub.Crypto, .epsilon..sub.FEC and .epsilon..sub.Net
are due to: crypto (IA), forward error correction, and network
load, respectively. On the other hand, from the link budget point
of view, the RF bandwidth, B, is inversely proportional to distance
square, in the form:
B = CONSTANT R 2 = C R 2 ( 48 ) ##EQU00019##
where C is constant and R-distance between Transmitter (Tx) and
Receiver (Rx).
[0401] The overall bandwidth equation, is
B = C R 2 = B 0 ( CR ) + OVH = B 0 ( CR ) + B ( 49 )
##EQU00020##
where (CR) is compression ratio. Usually, this equation may be
applied to the highest bandwidth data; i.e., video data, while
audio and textual data, or rather numerical data (in the case of
weather C2 sensor), are less bandwidth-intensive. Using Eq. (49),
we can write,
B ( 1 - ) = B 0 ( CR ) ( 50 ) ##EQU00021##
or, using all components of Eq. (49), we obtain,
B = B 0 ( CR ) ( 1 - ) = C R 2 ( 51 ) ##EQU00022##
Example 1
[0402] Assuming R=50 km, .epsilon.=50%, and (CR)=100:1, what is the
raw bandwidth, B.sub.0? In order to find the solution to this
problem, the value of the constant, C, should be known. Typically,
for various applications it can be assumed that:
C=100 Mbps0.1 km.sup.2 (52)
i.e., the RF bandwidth for R=1 km, including a typical network load
and other loads, is about 100 Mbps. Then, this leads to:
B 0 = C ( CR ) ( 1 - ) R 2 = 100 Mbps 1 km 2 ( 100 ) ( 0.5 ) ( 50
km ) 2 == ( 100 ) ( 50 ) ( 50 ) 2 Mbps = 5000 2500 2 Mbps ( 53 )
##EQU00023##
Example 2
[0403] Using data from EXAMPLE 1, a display resolution, or pixel
number, PN=740.times.480 ((VGA) standard), and the RGB
(red-green-blue) mode of 24 bpp (bits per pixel), find the maximum
video frame rate, FR.
[0404] The original, uncompressed video bandwidth, B.sub.0, is
B.sub.0=(PN)(BPP)(FR) (54)
[0405] Assuming B.sub.0=2 Mbps, the maximum frame rate, is
( FR ) = B 0 ( PN ) ( BPP ) = 2 Mbps ( 355 , 000 ) ( 24 ) = 0.23
fps ( 55 ) ##EQU00024##
i.e., about one video frame per 5 sec. This is quite satisfactory
for typical meteorological camera measurements, such as cloud
height, for example.
Example 3
[0406] Assume a crypto load of 10%, an error correction load of
5%,
[0407] and a network bandwidth load of 50%. Further assume:
(CR)=100:1, C=100 Mbps1 km.sup.2, pixel resolution: 740.times.480,
RGB-mode, (FR)=0.2. In this scenario, what is the maximum distance,
R?
[0408] The overhead coefficient, is: .epsilon.=10%+5%+50%=0.65, and
the raw bandwidth, B.sub.0, is
B.sub.0=(355,000)(24)(0.2)=1.704.10.sup.6 bps (56)
and, the RF bandwidth, B, is
B = B 0 ( CR ) ( 1 - ) = 1.704 10 6 ( 100 ) ( 1 - 0.65 ) = 4.87 10
4 = 48.7 Kbps ( 57 ) ##EQU00025##
Using Eq. (51), we obtain,
R = C B = 100 Mbps 1 km 2 48.7 Kbps = 45 km ( 58 ) ##EQU00026##
Which is a relatively short distance for satellite
communication.
Example 4
[0409] Assuming C=100 Mbps1 km.sup.2, R=100 km, (CR)=100:1, and
.epsilon.=0.5, what is the maximum raw video bandwidth,
B.sub.0?
[0410] By applying Eq. (51), we have:
B = C R 2 = 100 Mps lkm 2 10 4 km 2 = 10 Kbps ( 59 )
##EQU00027##
also, we have:
B.sub.0=B(CR)(1-.epsilon.)=(10 Kbps)(100)(0.5)=500 Kbps (60)
Example 5
[0411] Assuming B.sub.0=500 Kbps, black-white 8 bpp, and (FR)=0.2,
what is the maximum pixel resolution, PN-value?
( PN ) = B 0 ( FR ) ( BPP ) = 500 Kbps ( 0.2 ) ( 8 ) = 312.5 10 3
pixels ( 61 ) ##EQU00028##
Using the same ratio: 740: 480=1.54, we obtain the shorter side of
450 pixels, and longer side of 693 pixels; i.e., almost VGA format
for gray color; i.e., VGA gray screen can be applicable for
satellite communication, assuming R=100 km.
[0412] Information Assurance (IA) typically increases the bandwidth
requirements. That is, where more information assurance (IA) that
is desired, more bandwidth is typically required to provide such
assurance. Depending on the type of communication medium
(data/numerical, audio, video), there may be different
relationships between the available bandwidth, B.sub.A, and the
original or raw bandwidth, B.sub.0.
[0413] In the case of a data/numerical medium, lossless compression
may be applied with the network provision that, after Forward Error
Correction (FEC), the resulting (BER), or Bit-Error-Rate, is almost
zero. In the case of audio and video, however, the bandwidth cost
may in all practicality prohibit such an ideal situation. This is
why in various embodiments of systems and methods may apply a lossy
compression, in order to reduce the B.sub.A-level significantly
below the B.sub.0-level. This may be especially true in the case of
video compression where the relationship between compression ratio
(CR) and the overhead (OVH) components is critical. For the purpose
of discussing the IA Bandwidth Cost, consider three (3)
OVH-components: .epsilon..sub.FEC (error correction),
.epsilon..sub.NET (network), and .epsilon..sub.CRYPTO (cyber cost),
the latter one addressing the IA in a narrow sense.
[0414] In order to analyze quantitatively the IA Bandwidth Cost,
Eq. (50) may be rewritten in the form:
B A = B 0 ( CR ) ( 1 - ) ( 62 ) ##EQU00029##
[0415] where B is the original (raw) bandwidth, B.sub.A is the
available bandwidth, (CR) is the compression ratio, and E is the
overall OVH coefficient, which is the sum of the above three OVH
coefficients (FEC, crypto, network), according to Eq. (47). For
further estimation purposes, it may be observed that:
(CR).gtoreq.1 (63a)
0.ltoreq..epsilon..ltoreq.1 (63b)
Then, for low (CR)-values, the following relationship should
hold:
B.sub.A>B.sub.0, for low CR-values (64)
[0416] However, for the video/audio case, circumstances are usually
such that:
B.sub.A<B.sub.0 (65)
[0417] Accordingly, they can be derived from Eq. (47) that:
.epsilon.=.epsilon..sub.FEC+a;
a=.epsilon..sub.CRYPTO+.epsilon..sub.NET (66)
and, Eq. (62) becomes,
B A = B 0 ( CR ) ( 1 - a - FEC ) = B 0 ( CR ) ( A - FEC ) ( 67 )
##EQU00030##
where
A=1-a (68)
[0418] For parameterization purposes, both B.sub.A and B.sub.0 can
be set as constant values:
B.sub.A=constant,B.sub.0=constant (69ab)
[0419] While B.sub.A is generally a constant due to communication
channel limitations (e.g., in the case of satellite communication),
the B.sub.0 value does not need to be constant. However, for
practical purposes it can be assumed that also Eq. (69b) holds. By
putting Cartesian coordinates: x, y, in the form:
x=(CR),y=.epsilon..sub.FEC (70ab)
Eq. (67) becomes
B A = B 0 x ( A - y ) or , ( 71 ) x ( A - y ) = B 0 B .LAMBDA. ; or
, ( 72 ) A - y = 1 x B 0 B A ( 73 ) ##EQU00031##
Which is equivalent to:
A = y + 1 x B 0 B A or , ( 74 ) 1 = y A + 1 x ( B 0 AB A ) or , (
75 ) 1 = y A + D x ; D = B 0 AB A ( 76 ) ##EQU00032##
[0420] This equation can be parameterized using the well-known
trigonometrical identity: sin.sup.2.PHI.+cos.sup.2.PHI.=1, in the
form:
sin 2 .phi. = FEC A = y A sin .phi. = y A or , ( 77 ) .phi. = arc
sin ( y A ) ( 78 ) ##EQU00033##
where .PHI.-parameter, in radians; thus,
sin 2 .phi. = D ( CR ) = D x cos .phi. = D x ( 79 )
##EQU00034##
[0421] According to Eq. (78), .PHI.-grows with increases in y,
where: x=(CR) and y=.epsilon..sub.FEC. Also, from Eq. (76), the
y-value can be maximized, when the x-value is also maximized, and
vice versa. According to Eq. (30):
x.gtoreq.1 (80a)
0.ltoreq.y.ltoreq.1 (80b)
[0422] Therefore, starting with x=1; then, y=y.sub.MIN, and
.PHI.=.PHI..sub.MIN. Then, for x.fwdarw..infin., y.fwdarw.A, and
.PHI..fwdarw..pi./2. The latter conclusion follows since, for
infinitely large CR values, the raw bandwidth, B.sub.0, is reduced
to zero; thus, all available bandwidth, B.sub.A, can be used for
the OVH. Thus: y=.epsilon..sub.FEC=A=1-a; and therefore,
.epsilon.=(1-a)+a=1, as it should be.
[0423] Both limited cases: x=1, and x=.infin., are, of course,
extreme, and never met in practice. Thus, the optimum performance
is somewhere between .PHI.=.PHI..sub.MIN and .PHI.=.pi./2, for
.PHI.=.PHI..sub.0, where:
.PHI..sub.MIN<.PHI..sub.0<.pi./2 (81)
[0424] FIG. 29 is a diagram illustrating this relationship.
Particularly, FIG. 29 provides an illustration of optimum
performance for .PHI.=.PHI..sub.0 for y(x)-dependence, where x=(CR)
and y=.epsilon..sub.FEC. The parameter, which characterizes the
system performance, may be referred to as the
Peak-Signal-to-Noise-Ratio (PSNR), denoted as U, in the form:
U=(PSNR) (82)
[0425] The PSNR may be defined as bit-by-bit average difference
between the original (un-compressed) and compressed image, in
decibels. (For no-difference (PSNR)=.infin.; for good image
quality, (PSNR).gtoreq.30 dB). In the view of the above comments, U
is a function of the .PHI.-parameter, with a maximum somewhere
between .PHI..sub.MIN and .pi./2, in the form:
U=f(.phi.) (83)
[0426] FIG. 30 is a diagram illustrating the relationship presented
by Eq. (83). Particularly, FIG. 30 illustrates the behavior of the
U-Function as a function of .PHI.-parameter, with maximum out
.PHI.=.PHI..sub.0, where U=(PSNR). In this figure, U=(PSNR), and
the U function has maximum 1450 at .PHI.=.PHI..sub.0. This function
is cut at 1451 and 1452, because their behavior outside maximum
must generally be determined by specific experiment, and the cases
well outside the maximum vicinity are non-practical cases.
[0427] In typical embodiments, the system is configured such that
input data (e.g. such as video data) are firstly compressed; then,
encrypted; then error corrected; then, networked. This data
transmission (Tx) transfer sequence is shown in FIG. 31, which
includes examples of characteristic parameters representing each
step. A similar sequence, but in the inverse, may occur for the
data receiving (Rx) transfer sequence. An example of this is
illustrated in FIG. 31B in which the characteristic operations are
defined rather than representative parameters. In FIG. 31C, the Tx
data sequence, equivalent to Rx data sequence, as in FIG. 31B is
shown.
[0428] In the example illustrated in FIG. 31A, the Tx Video Data
Transfer Sequence 1600 is shown. In this example transfer sequence
1600 input data 1601 is received for transmission. At operation
1602, data compression is applied by a data compression module 1602
to achieve a desired compression ratio CR. Data is then encrypted
by a data encryption module 1603 to provide a measure of data
security. This is represented by .epsilon..sub.CRYPTO-coefficient.
A forward error correction (FEC) module 1604 can be included to
perform a Forward-Error-Correction (FEC) operation, represented by
.epsilon..sub.FEC-coefficient. The resulting data can be configured
for networking a block 1605 and output as output data 1606. The
data can be modulated onto an RF carrier for transmission as an
RF-transmitted wavefront 1607.
[0429] In the example of FIG. 31B, the Receiving (Rx) Video Data
Transfer Sequence 1630 receives video data 1631. Where the data was
modulated onto an RF carrier, it can be demodulated before
providing it to the network receive operation 1632. A forward error
correction module 1633 can be included to perform error correction
on the received data based on the forward error correction
operation supplied by the transmitter. The data can be decrypted
using decryption module 1634, and decompressed using decompression
module 1635. The resultant received data 1636 can e output for use
by the system.
[0430] In the example of FIG. 31C, the sequence of FIG. 31A is
shown as being repeated, except, in this example only the
characteristic operations have been shown, in direct equivalence to
FIG. 31B. This example Tx Video Data Transfer Sequence including
Characteristic Operations 1660 includes Tx-input data 1661 received
by the system. The sequence further includes data compression
module 1662, data encryption module 1663, error correction module
1664, and networking operation 1665. The resulting data is
modulated and transmitted as a Tx-wavefront 1666.
[0431] In various embodiments, these modules and operations can be
implemented as they are commonly implemented in various medications
systems for communications of data, video, and other content. This
description illustrates, however, that each operation has its
bandwidth cost, which should be estimated in order to analyze any
type of C2 (Command-Control), especially those that operate with
all three types of media: data/numerical information, audio, and
video.
[0432] For all these types of data, FIG. 31, which comprises FIGS.
31A, 31B and 31 C applies, except there are typically going to be
quantitative differences in value ranges of parameters: CR,
ECRYPTO, EFEC, and ENET, which may be referred to herein as control
variables. All these control variables may include intra-system
variables in such a sense that they represent system responses to
factors such as environmental, terrain, and latitude changes.
[0433] In order to complete the control variable set, it may be
useful to consider the Bit-Error Ratio (BER).sub.0, which is an
external control variable in the sense that it comes from the
external environment. However, after the error correction, this
external (BER).sub.0-value may be transformed to an internal
(BER).sub.1-value, in the form (other operations such as Frequency
Hopping can also be included):
(BER).sub.0.fwdarw.(BER).sub.1; (BER).sub.1<(BER).sub.0 (84)
[0434] The (BER)-figure is usually defined as a so-called ensemble
average figure; [0435] i.e., averaged over a statistical ensemble
of specific ensemble realizations. However, in practice, systems
and methods may be configured to operate with ensemble
realizations. Also, it may be noted that there may be two basic
types of data stream errors: bit-by-bit errors, or b.sub.3-errors;
and burst-errors, or b.sub.1-errors, the latter of which are
typically more severe than the former. Usually, the b.sub.1-errors
are mitigated (not fully eliminated, however) by reshuffling of
bits before sending, which, of course, does cost additional
latency.
[0436] In the context of control variables, the (PSNR)-figure may
also be introduced as a state variable. Both control variables and
state variables create so-called phase space variables, the
nomenclature used in the mathematical theory of catastrophes which
will be applied here as an embodiment of the invention, as an
unobvious generalization of Singular Mapping (SM) concept.
[0437] The Singular Mapping (SM) approach may be configured to
extract and identify topological singularities, both linear and
non-linear, coming from SM-visualization, or SMV. The SMV may be
applied as a general concept, which can be applied Singular Mapping
(SM) to C2-communication parameters: .epsilon.-parameters, CR, and
(PSNR), but also to any other parameters describing a situation in
a weather station, for example. Then, the control variables are
weather variables such as: temperature, relative humidity, air
pressure, etc.; while the single state variable or several state
variables represent some resulting parameters such as ranking
number of anomalous event, for example.
[0438] The SMV can be configured to create discrete topological
singularities (DTS), which can be either linear or non-linear ones.
The linear DTS may include maxima, minima, inflection points, etc.;
while the non-linear DTS may be referred to as catastrophes, which
are sudden drops, or jumps of state variable value. The 3.sup.rd
category of DTS may include threshold DTS, which characterize
off-the-expectation (OTE) incidents, or anomalous events, which are
rather rare targets. Thus, the OTEs include non-linear DTS, or
N/DTS, and threshold DTS, the latter of which may be characterized
by some exceeding-threshold values. These may be referred to herein
as T/DTS--for brevity.
[0439] This rather comprehensive description of Discrete
Topological Mapping (DTS) is provided to help explain the rather
difficult DTS-concept by using the C2-communication example,
discussed below. This example has two goals: first, to explain the
WAES as C2-sensor; and second, to explain the general concept of
Discrete Topological Mapping.
[0440] FIG. 32 is a diagram illustrating an example of
(PSNR)-dependence as a function of internal (BER).sub.1-control
variable, defined by Eq. (84). Higher (BER).sub.1-value results in
a lower (PSNR)-value, or U-value. In the extreme case when (CR)=1,
and (BER).sub.1=0, the result it that U=.infin.. However, for
(CR)>1, and (BER).sub.1=0, U=.infin., typically only in
situations in which the (CR) operation is lossless. However, for
video/audio data, the (CR) operation is typically lossy. In this
case, U<.infin., even for (BER).sub.1=0. This means that for
scenarios with lossy compression, the reconstructed image according
to FIG. 31B will be, in general, different from original video
image.
[0441] As noted, FIG. 32 illustrates an example of U-Dependence
(i.e., PSNR-dependence) as a Function of (BER).sub.1, for various
(CR)-values. In this example, curves 1700 are presented including
three typical (CR)-values: 1701, 1702, and 1703, where:
(CR).sub.3>(CR).sub.2>(CR).sub.1 (85)
where (CR).sub.1=1. Therefore, for (BER).sub.1=0 (no errors, after
correction), we obtain,
U=(PSNR)=.infin., for (CR)=1 (86)
characterized in FIG. 32 by point 1704. Eq. (86) can also be
satisfied for (CR)>1, if the compression is lossless. In FIG.
32, however, the (CR).sub.2 and (CR).sub.3 values are for lossy
compression. Therefore, U<.infin., for points 1705 and 1706.
[0442] It can also be observed that all three curves 1701, 1702,
1703, are monotonically decreasing, as shown by decreasing tendency
of curve tails 1707, 1708, 1709. This monotonic feature generally
holds unless some hidden parameters exist, especially for
b.sub.1-errors, which also should be classified as anomalous
events.
[0443] FIG. 33 is a diagram illustrating an example of U-dependence
as a Function of (CR) for various (BER).sub.1-values. The example
chart 1750 includes three curves 1751, 1752, and 1753, for various
(BER).sub.1-parameters: (BER).sub.1.sup.(1), (BER).sub.1.sup.(2),
and (BER).sub.1.sup.(3), respectively. Since,
(BER).sub.1.sup.(1)=0, U=.infin. for (CR)=1, (illustrated by
reference character 1754). The curves are also monotonic 1755,
1756, 1757. However, the monotonicity does not need to be
satisfied, and the (CR)-value can be lower than unity. This more
general case is discussed in detail below.
[0444] In order to emphasize the non-linear singular sets, this
document refers to digital mapping more precisely as Digital
Singular Mapping (DSM).
[0445] In order to provide a more general case, or the
DTM-generalization, the Cartesian variables (other coordinate
variables can also be used) can be placed in the form:
x=(CR),y=(BER).sub.1,z=U=(PSNR) (87abc)
where control variables: (CR), and (BER).sub.1, are denoted by
"horizontal" (x, y)-coordinates, while state U variable is denoted
by vertical z-coordinate.
[0446] In the context of the theory of catastrophes', the
single-state variable case may be categorized as a co-rank-1
catastrophe, while multiple state-variable cases may be categorized
as co-rank 2, 3, etc. catastrophes, respectively. For clarity of
discussion, this section of the document focuses on co-rank 1
catastrophes.
[0447] Firstly, it may be observed that although typically the
function set is such that such functions such as functions 1707,
1708, 1709 do not cross (e.g., as in FIGS. 32 and 33); this is not
always the case. In fact, such functions can cross each other, as
shown in the example of FIG. 34. FIG. 34 is a diagram illustrating
an example of z Function dependence as a Function of y, with the x
variable as a parameter, in which the ex-variable as values
x.sub.1, x.sub.2, x.sub.3. The function set 1800 with three member
functions 1801, 1802, and 1803 is presented. The example of FIG. 34
also illustrates function 1803 as a non-monotonous function 1803,
which is crossing function 1802 at two points: A and B. FIG. 34 is
a generalized version of FIGS. 32 and 33. This generalization is
provided to illustrate the general principle the z-function may be
considered in the form:
z=z(y;x.sub.n) (88)
[0448] In which z is state variable and y is control variable, with
x.sub.n as a parameter. The x, y variables can also be reshuffled
in the form:
z=z(x;y.sub.n) (89)
where y.sub.n is a parameter. In general, the z-function of two
variables, x, y, can be written in the form:
z=z(x,y) (90)
[0449] As noted above, this general form may always alternatively
be presented in the form of either Eq. (88), or in the
complementary form of Eq. (89). The only material difference
between Eq. (90) and Eqs. (88) and (89), is that in Eq. (88), two
variables (z, y) are continuous and the x.sub.r, variable is
discrete ("n" is integer), while in Eq. (89), two other variables
(z, x), are continuous and the y.sub.n variable is discrete. In
contrast, in Eq. (90), all three variables x, y, z are continuous.
It can be determined that the function set in the example of FIG.
34 is sampling a version of the 2D-function (90), which is
continuous in all three variables.
[0450] Eq. (90) may further be generalized into the following
convoluted form:
F(x,y,z)=0 (91)
[0451] Similarly, Eqs. (88) and (89), can be generalized into the
convoluted forms:
F(y,z;x.sub.n)=0; F(x,z;y.sub.n)=0 (92ab)
in which Eq. (92a) is the convoluted form of Eq. (88), and Eq.
(92b) is the convoluted form of Eq. (89).
[0452] It should be noted that not all functions (91) are capable
of being presented in an un-convoluted form (90). For example, the
following function can be presented only with sign uncertainty:
x.sup.2+y.sup.2=z.sup.2z=.+-. {square root over (x.sup.2+y.sup.2)}
(93)
[0453] In the context of the theory of catastrophes this can be in
important distinction because function (90) does not contain
catastrophes, while function (91) sometimes includes
catastrophes.
[0454] Examples of co-rank-1 catastrophes as anomalous events are
described in: T. Jannson, et al., "Catastrophic Extraction of
Anomalous Events," SPIE Proc. Vol. 8359-19, 2012, which may be
useful to the reader as a reference for background purposes. For
the purpose of the exemplary DSM operations, this document
discusses the 3D space (x, y, z) which can be easily visualized.
However, as would be known to one of ordinary skill in the art the
procedure is valid for the 4D-space and higher spaces, as well as
for co-rank-2, co-rank-3, and higher co-rank catastrophes. However,
heuristication of the Digital Singular Mapping (DSM) procedure, is
believed to present a novel and non-obvious approach.
Heuristication of the DSM procedure includes performance of the
procedure automatically, such as by computer or other processing
system.
[0455] For ease of discussion and to facilitate understanding,
consider only 3-D space (x, y, z), in which z is the state
variable, and x, y are control variables. In 3D space, Eq. (91)
describes a 2D surface in 3D (which is sometimes referred to herein
in shorthand as a 2-surface). Similarly, in 2D space a 1D surface,
or curve, can be referred to herein as a 1-surface. However, in
n-dimensional space, only (n-1) continuum can be referred to as a
"surface."
[0456] In order to obtain the simplest so-called fold catastrophe,
it may only be necessary to slightly deform FIG. 33 into the form
as in FIG. 34.
[0457] FIG. 35 is a diagram illustrating an example of a z Function
as a Function of y, with the variable x as a parameter. In this
example, the curves 1830 include function set 1831, 1832 and 1833.
As seen in this example, function 1833 crosses function 1832 at the
points designated as A and B. However, in contrast to the example
of FIG. 34, function 1833 is more "folded" then is function 1803 in
FIG. 34. Indeed, in the example of FIG. 35, function 1833 is folded
to a level such that a new topological quality occurs.
Particularly, the normal to function 1833 at point 1834 is
perpendicular to a line (illustrated by the broken line) parallel
to the z-axis. In this example, the normal is represented by arrow
1836 and perpendicularity is shown by the right-angle symbol:
[0458] A similar perpendicularity feature does occur at point 1835,
as illustrated and defined by the normal, shown as arrow 1837.
Accordingly, in FIG. 35, two fold catastrophes do occur at points
1834 and 1835.
[0459] Prior to this disclosure, there may have existed difficulty
with realizing that functions 1832 and 1833 can cross. This may be
because it can be difficult to provide inductive thinking in
generalizing 2D-views into a 3-D perspective. This can be seen in
FIGS. 36, 37 and 38 which illustrate, in steps, a generalization
from 2D to 3D, or, more, generally, from (n-1)-space to
n-space.
[0460] FIG. 36 is a diagram illustrating an exemplary z surface in
(x, y, z) space, including contour lines. FIG. 37 is a diagram
illustrating an example of planes perpendicular to the x axis. FIG.
38 is a diagram illustrating an example of x cross-sections of an
exemplary z surface.
[0461] In the example of FIG. 36, exemplary z surface in (x, y, z)
space is illustrated with contour lines 1851, 1852 and 1853. These
contour lines illustrate cross-sections of a z surface in planes
normal to the z-axis in the form:
z=constant (95)
[0462] However, in FIG. 37, the surface cross-sections 1881, 1882,
1883 are normal (perpendicular) to x-axis, in the form:
x=constant (96)
[0463] As these examples illustrate, the cross-sections of the z
surface described by Eq. (95) are perpendicular to the z axis,
while the cross-sections of the same z surface, described by Eq.
(96), are perpendicular to the x axis; i.e., the cross-sections are
perpendicular to each other. However, z cross-sections, defined by
Eq. (95) can be considered as having a special status because they
are normal to the state variable, while x cross-sections and y
cross-sections are only perpendicular to the control variables.
Nevertheless, any of those cross-sections, if sufficiently dense,
are sufficient to reconstruct z surface, such as in FIG. 36. In
particular, such cross sections as 1880 or as represented by planes
1811, 1882, and 1883, if they are sufficiently dense to be
sufficient to reconstruct, the z surface. How dense they should be
may be defined by the Sampling Theorem.
[0464] Accordingly, it can be concluded that a kind of deductive
thinking for going from a z surface to its contours 1851, 1852,
1853, or to x cross-sections 1891, 1892 and 1893, is relatively
straightforward. In contrast, inductive thinking for going from
x-cross-sections 1891, 1892 and 1893, into the z surface can create
issues if turned into some anomalous case, as in FIG. 35, or even
as in FIG. 34. This is, because, in transformation from FIG. 38
into FIG. 36, it appears as if an extra dimension is added, namely
the x variable; while in fact, what occurred was a transformation
of a discrete variable, x.sub.n, (which may also be referred to as
a parameter) into a continuous variable, x.
[0465] After this explanation, it is appropriate to introduce a
process for handling the fold catastrophe in an automatic way
(i.e., fully heuristically), defining its location on the
convoluted z surface:
F(x,y,z)=0; z-state variable (97)
satisfying the following equation for its z-dependent partial
1.sup.st differential:
.differential. F .differential. z = 0 ( 98 ) ##EQU00035##
[0466] According to Eq. (98), in the fold catastrophe location (x,
y, z) and on z surface (Eq. (97)), the normal to this surface has
zero z-component; i.e., indeed it is perpendicular to the z axis,
as in points 1834 and 1835 in FIG. 35. This is, because, the normal
vector to the z surface in Cartesian coordinates: (x, y, z) is
proportional to a gradient to this surface, in the form:
n = A gradF = ( .differential. F .differential. x , .differential.
F .differential. y , .differential. F .differential. z ) ( 99 )
##EQU00036##
[0467] where {right arrow over (n)} represents a normal vector to
the F surface, and is a proportionality constant. Therefore,
indeed, Eq. (98) is equivalent to the following formula:
n.sub.z=0 (100)
[0468] With this foundation in mind, the Digital Singular Mapping
(DSM) automatic procedure is now described according to various
embodiments. For the sake of clarity, this procedure is given for a
3D-space, with Cartesian variables (x, y, z), for z-state variable
and (x, y)-control variables. However, this procedure is
straightforward for all co-rank-1 catastrophes as: fold, cusp,
swallowtail, butterfly, wigwam and higher, as well as for co-rank-2
catastrophes: elliptic umbilic, hyperbolic umbilic, and parabolic
umbilic, as well as for higher co-rank catastrophes, and other than
Cartesian coordinates, based on prior art mathematics of the Theory
of Catastrophes.
[0469] A Digital Singular Mapping (DSM) procedure may be
implemented heuristically. "Heuristicity" as used herein refers, in
some embodiments, to the process feature eliminated from
non-heuristic elements, which assumes some involvement of
intelligence or consciousness as explained by Erwin Schrodinger in
his famous book: What is Life?: With Mind and Matter and
Autobiographical Sketches, Cambridge Univ. Press, 1945. This is
because in various embodiments, the DSM Procedure can be
implemented as an automatic (or, autonomous) process that could be
handled by the computer system.
[0470] In the theory of catastrophes, the phase space may be
constituted by state variables and control variables. This approach
may be continued in various embodiments, but it should be noted
that the DSM contains more general Digital Topologic Singularities
(DTS), which include both linear and non-linear DTS. The linear DTS
can include such singular curves and points as: maxima, minima,
inflection points, and other singular areas of standard function
analysis. In contrast, the non-linear DTS can include various types
of mathematical catastrophes including those with a single state
variable (co-rank 1), two state variables (co-rank 2), etc. The
number of control variables can be arbitrary.
[0471] By applying a causality principle, state variables may be
separated from control variables in such a way that the control
variables are input or cause variables, while state variables are
output, effect, or result variables. However, by treating this
problem heuristically, it may be impossible to separate them in a
unique way. FIGS. 39A and 39B are diagrams illustrating an
exemplary relation between state and control variables due to
causation principle. FIG. 39A includes an example for linear
digital topological singularities (catastrophes) 1922. FIG. 39B
includes an example for non-linear digital topological
singularities (catastrophes) 1924.
[0472] As the examples illustrate, the curves for both cases are
identical. Nevertheless, the state and control variables have been
replaced. In particular, referring to FIG. 39A, for singularities
and 1922, the points C, D, E, determine linear DTS locations of
maxima (C, D) and minima (E). On the other hand, referring to FIG.
39B, for singularities 1924, the identically located points C', D',
E', determine non-linear DTS fold catastrophe locations in 2D phase
space, for simplicity. Accordingly, as the examples illustrate, a
simple reshuffling of state and control variables changes the DTS
meaning in a material way.
[0473] Therefore, in the case of the example DSM procedure
discussed below, the 1.sup.st step is partially non-heuristic,
while the next steps are rather heuristic in a sense of the
Schrodinger definition of consciousness. These heuristic
distinctions can be used in various embodiments to provide an
automated computer system process, which in some cases can be a
fully automated process that does not require human intervention.
Such a feature can be useful for a C2 weather station operating in
the field to allow it to autonomously distinct between anomalous
and normal meteorological events.
[0474] For clarity of discussion, this document describes this
process in terms of Cartesian coordinate systems, defining
nD-space, where n is the number of dimensions. For n=3 we obtain
3D-space. For any nD-space, or shortly n-space, embodiments can be
implemented using (n-m) sub-spaces, where m is a number of excess
variables that can be treated as parameters. This is shown in FIG.
40, which is a diagram illustrating an example of subspaces and
parameters.
[0475] In the example shown in FIG. 40, 3D-space (x, y, z) has been
discretized in such a sense that the x-dimension is treated as a
parameter. The discretized 2-surface 1950 is presented in the form
of discrete set, of 1-surfaces 1951, 1952 and 1953, which are
cross-sections of the 2-surface, with the 3.sup.rd.times. variable
treated as a constant. (These constants x.sub.1, x.sub.2, x.sub.3
may be referred to herein parameters.) This paradigm can be
generalized in such a sense that the number of discrete dimensions
(parameters) can be larger than one, as well as a number of
dimensions can be larger than two. For example, we can consider a
2-surface with z parameters in the form:
F(x,y,z; u,v)=0 (101)
[0476] This 2-surface (x, y, z), with two parameters (u, v) is a
cross-section of a 4-surface F(x, y, z, u, v)=0. In particular, the
1-surface in FIG. 40 can be presented in the form (where the
semicolon separates continuous variables from discrete
parameters):
F(y,z; x)=0 (102)
which is a cross-section of the 2-surface F(x, y, z)=0. The
essential variables can be defined as such variables that contain
catastrophes. They should contain at least one state variable. Such
essential variables define a sub-surface, or cross-section, that
contains a catastrophe, or catastrophes.
[0477] The understanding of relations between topologic surfaces,
cross-sections and projections is important for heuristic
development of the DSM. In order to provide such simple (heuristic)
development, it is convenient to apply Cartesian systems of
coordinates such as (x, y, z) in 3D space, or (x, y, z, u, v) in 5D
space, for example. Therefore, generalized, or curvilinear
coordinates need not be provided unless specifically required.
[0478] In 3D space, a 2-surface satisfies the following convoluted
relation:
F(x,y,z)=0 (103)
[0479] In non-convoluted form, however, Eq. (103) simplifies into
the form:
z=z(x,y) (104)
where z is a state variable, and (x, y) are control variables.
Comparing Eq. (104) with Eq. (98) it can be seen that in such a
case (x, y, z) variables are not essential variables. This is
because, by presenting Eq. (104) in convoluted form (103),
yields:
F(x,y,z)=z-(x,y)=0 (105)
Thus,
[0480] .differential. F .differential. z = 1 .noteq. 0 ( 106 )
##EQU00037##
[0481] Thus, indeed, there are no catastrophes in this case. In
fact, Eq. (104) presents familiar geophysical contour mapping,
which, usually does not include catastrophes. This specific example
is crucial to understand the meaning of essential coordinates. It
will be discussed as EXAMPLE 6 in the next section.
[0482] According to Eq. (103), surface cuts or, cross-sections, can
be made in three (3) possible ways; i.e., as x-cross-sections,
y-cross-sections, and z-cross-sections. In the case when z
coordinate is a state variable, while (x, y) are control variables;
then, z cross-sections are also called contour lines. These contour
lines may be in the form:
F(x,y; z.sub.n)=0 (107)
where z.sub.1, z.sub.2, . . . , z.sub.n are locations of
z-coordinate. When, these locations are uniformly distributed, then
there are familiar mapping contours, as shown in FIG. 41. FIG. 41
is a diagram illustrating a familiar contour mapping with contour
lines at z: 100 m, 110 m, 120 m-elevations. FIG. 42 is a diagram
illustrating an example of non-linear contour mapping.
[0483] While the contour mapping as in FIG. 41 is rather familiar,
the non-linear contour mapping as in FIG. 42 is not familiar. This
is because the contour lines of FIG. 42 cross each other, leading
to some possible catastrophe locations. In non-linear contour
mapping as in FIG. 42, the contour lines at z=100 m and z=110 m do
cross each other at points A and B. The example of x-cross-sections
is shown in FIG. 37. In general, the cross-sections such as x, y, z
cross-sections can lead to full surface reconstructions, assuming
that the distance between cross-sections, .DELTA.z, tends to zero
value:
{ lim F ( x , y , x i ) = 0 } = { F ( x , y , z ) = 0 } .DELTA. z
-> 0 i = 1 , 2 , 3 n -> .infin. ( 108 ) ##EQU00038##
i.e., the continuum of surface cross-sections leads to continuous
surface, for number of cross sections, n, tending to infinity.
[0484] Surfaces vs. Projections.
[0485] Combining Eq. (103) with Eq. (98), leads to two (2) surface
equations that may be satisfied, simultaneously, in the form:
F ( x , y , z ) = 0 ( 109 a ) .differential. F .differential. z = G
( x , y , z ) = 0 ( 109 b ) ##EQU00039##
[0486] The 1.sup.st surface Eq. (109a) can be used to determine the
possible manifold surface, while the 2.sup.nd surface equation
(109b) can be used to determine the geometrical locii of fold
catastrophes. In order to obtain their projections, the z variable
may be eliminated from Eq. (109), resulting in a z projection
referred to as a bifurcation set, in the form:
K(x,y)=0 (110)
[0487] DSM and Manifolds.
[0488] FIG. 43 is a diagram illustrating an example of a z manifold
in (x, y, z) space. In FIG. 43, the z-manifold is shown in (x, y,
z) space, with a z state variable and (x, y) control variables. As
this example illustrates, a vertical, straight line, satisfying
equation: x=x.sub.1, y=y.sub.1, crosses the z-manifold in three
points, with z-coordinates z.sub.2, z.sub.3. If the number of
crossing points is larger than one, this may be referred to as a
manifold; otherwise it can be deemed that there is no manifold.
[0489] The DSM procedure according to various embodiments may be
carried out in process steps that may be "heuristicized" as much as
possible or practical. Table 7 is a table illustrating example
steps for the DSM procedure in accordance with one embodiment of
the systems and methods disclosed herein.
[0490] In the first step (step number 1), the process identifies
state and control variables. For simplicity of description,
consider 3D-space, (x, y, z), with z state variable and (x, y)
control variables. The input data, obtained either from experiment
or by estimation, may be in the form of a set of points: (x.sub.i,
y.sub.i, z.sub.i); i=1, 2, 3, . . . , n; where n is the number of
points. By applying standard sampling procedures, the sampling
points (x.sub.i, y.sub.i, z.sub.i), are usually formatted in the
form of curves, with the 3.sup.rd coordinate (x, or, y) as a
parameter. Only control variables can be parameters. Thus, such
sampling curves as in FIG. 40, for example can be obtained.
[0491] In the second process operation (step number 2), the process
provides x.sub.i-parameters, for example, denser and denser (using
rules of standard sampling theorems), until, a continuum DSM
surface is obtained. See, Eq. (76), for example. In the third
process operation (step number 3) the process identifies whether
this continuum surface is a z manifold. For these purposes, the
process can be configured to apply a bundle of vertical lines, as
in FIG. 43, and find, whether their cross-sections with a given
surface produce multi-value solutions. If the answer is yes, then
it means that a given surface is a manifold. On the other hand, if
the answer is no, a given surface is not a manifold. Equivalently,
in the 1.sup.st case the (x, y, z) variables may be deemed to be
essential; otherwise, they are not essential. This is, because, the
presence of the manifold means at least the existence of fold
catastrophe/catastrophes.
[0492] In the fourth process operation (step number 4) the process
identifies a location of any fold* catastrophes that exist (i.e.,
non-linear DTS, in the manifold case). Otherwise, the process
identifies possible locations or regions of linear DTS. In hybrid
situations, it is possible that both linear and non-linear DTS
exist. In the fifth process operation (step number 5), the process
provides DTS synthesis by finding bifurcation sets, or (x,
y)-projections of fold catastrophes, in the case of non-linear DTS.
Otherwise, the process locates linear singular sets, only. In the
sixth process operation (step number 6) methodology summarizes the
results in a proper format, which can be defined, for example, by
Pre-Structuring. In the seventh process operation (step number 7)
system provides the necessary generalizations, including providing
more dimensions, etc.
TABLE-US-00006 TABLE 7 Example Summary of Digital Singular Mapping
(DSM) Step Number Step Name 1 Identify state and control variables
2 Develop DSM continuous surface: F(x, y, z) = 0 3 Identify if the
DSM surface is manifold. Find essential variables. 4 Identify
location of fold*.sup.) catastrophes and linear Digital Topologic
Singularities (DTS) 5 Provide DTS synthesis 6 Summarize the results
in pre-structuring format 7 Provide necessary
generalizations**.sup.) *.sup.)And this procedure can be
generalized to higher order catastrophes such as cusp and others
**.sup.)Mostly, apply larger number of dimensions, if needed.
[0493] Three examples of fold catastrophes presented in heuristic
way are presented including geophysical, meteorological, and
physical ones.
Example 6. Geophysical Contour Mapping
[0494] In the case of well-known geophysical contour mapping, the
contour lines are z-cross-sections, where z is the elevation
coordinate, and (x, y) are the geophysical coordinates. In this
case, the DTS step 1, as in Table 7, is automatically provided with
a z coordinate as a state variable and (x, y) as control variables.
The familiar contour mapping lines are z-cross-sections. In FIG.
41, exemplary linear contour lines are presented; while in FIG. 42,
the exemplary non-linear contour lines are presented. By comparing
FIG. 41 with FIG. 42, it can be seen that, rarely, we see such
contour lines as in FIG. 42, except, perhaps, when the detailed
mapping of mountain caves, or coves is provided. This is, because,
typically, the geophysical mapping does satisfy the un-convoluted
2-surface Eq. (104) condition, which does not result in fold, or
higher co-rank-1 catastrophes (such as: cusp, swallowtail,
butterfly, wigwam, etc.). In the case of such unusual rocky
mountains, such as the "finger" mountains in Arizona, for example,
some unusual shapes can be found including those as shown in FIG.
44, which resembles a human body, for example. FIG. 44 is a diagram
illustrating an example of non-linear contour lines.
[0495] In the example illustrated in FIG. 44, Non-Linear Contour
Lines are illustrated such as those that can be described by using
Eq. (107) in convoluted form. Accordingly, these do not lead to Eq.
(105). Therefore, at least, fold catastrophes can exist, in such a
case.
[0496] In fact, in the example of FIG. 44, the convoluted 2-surface
2000, characterized by the following convoluted 2-surface equation,
is presented in the form:
F(x,y,z)=0.revreaction.CONVOLUTED 2-SURFACE (111)
with four (4) exemplary z-cross-sections 2001, 2002, 2003, and
2004, representing four (4) elevations: z=250 m, 2005; z=200 m,
2006; z=150 m, 2007, and z=100 m, 2008, respectively. Their
projections are 2009, 2010, 2011, and 2012, respectively. As this
example illustrates, these projections do cross each other in such
a sense that projections 2011 and 2012 coincide. This is, because,
contours 2003 and 2004, at different elevations 2007 and 2008, are
identical as a peculiar specific non-linear case. Of course, these
contour lines do not need to be axially symmetrical as in in the
example illustrated in FIG. 44.
Example 7. Meteorological Fold Catastrophe
[0497] Consider a peculiar phenomenon of creating a rain in a
desert by putting fire on a cactus forest. This unusual anomalous
effect does occur when relative humidity in the air is very high
(e.g., so high that there is no rain because there are not
sufficient condensation centers in the air). By creating the fire,
however, a smog results, producing the required condensation
centers; thus, resulting in unexpected rain, which is a kind of
non-linear DTS, or mathematical catastrophe. The term
"mathematical" catastrophe, may be used herein to refer to the
effect of sudden drop or jump of state variable. While, the drop is
considered as normal (regular) catastrophe, the jump is usually not
considered as normal catastrophe, in a familiar sense.
[0498] In order to "heuristicize" this phenomenon, the rain rate,
R, may be introduced as state variable, and temperature, T; time,
t; and (x, y), may be used as control variables. Accordingly,
embodiments can be implemented having a 5D space: (T, t, x, y, R),
and 4-surface, which is so-called an equilibrium surface in the
theory of catastrophes, in the form:
F(T,t,x,y,R)=0 (112)
[0499] To visualize this so-called hyper-surface (i.e., higher than
2-surface in 3D), embodiments can be implemented to "discretize"
two state variables as parameters (x.sub.n, y.sub.n), resulting in
an equilibrium 2-surface:
F(T,t,R; x.sub.n,y.sub.n)=0 (113)
[0500] Such a surface has, indeed, a "catastrophic" jump of an
R-variable for a given (T, t)-values.
[0501] In the context of this example, the "paramaterizing" process
may be considered, in a way, as indicative of the presence
(existence) of other additional variables (e.g., by presenting them
as parameters). For example, in Eq. (113), two other (spatial)
coordinates (x.sub.n, y.sub.n) are indicated. However, there may be
some other hidden variables that are essential but omitted in the
process. In some embodiments, the process may fail to indicate such
variables in equations such as Eq. (113), which can result in some
non-linear DTS mapping missed.
[0502] Consider again the fire phenomenon as discussed in Example
7, by analyzing two essential variables R and T. FIG. 45 is a
diagram illustrating an example of hysteresis in the case of the
desert rain phenomenon discussed above with reference to example 7.
With reference to FIG. 45, in this example, the evolution path,
ABC, is applicable only in the case of arrows as indicated in FIG.
45. Otherwise, when the evolution path goes from right to left
(e.g. as in DEF), this path is different from the previous one. The
graphical difference between paths ABC and DEF is denoted by the
crosshatched area under the curve. If there were no difference
between the two paths, the crosshatched area would be zero.
Therefore, the hysteresis is proportional to the crosshatched
area.
[0503] The example of FIG. 45 includes two fold catastrophes: a
jump catastrophe AB; and a drop catastrophe EF. This example
illustrates that time variable, t, is not shown even as a
parameter. Therefore, the t variable is a hidden variable in this
case. In fact, air temperature, T, is direct function of t, as
shown in the example of FIG. 46, which plots example of temperature
versus time dependence. This also illustrates that FIG. 46 does not
contain non-linear singularities, only linear ones (a maximum).
[0504] FIG. 45 demonstrates a kind of heuristication effect. This
is because, as shown in FIG. 45, two fold catastrophes can be found
automatically, by measuring the R(T) dependence (i.e., these
non-linear singularities were found without Schrodinger's
consciousness). In fact, the catastrophic phenomenon in FIG. 45 is
a result of choosing non-primary coordinate, T, while using the
primary coordinate such as condensation center (aerosol)
concentration, c, as in FIG. 47 (which illustrates a Linear
R(c)-dependence), would not show any non-linear singularity. It is
noted that the definition of an essential variable (as variable
involved in the catastrophic effect) is heuristic rather than
non-heuristic, which can be considered as deficiency. However, in
the context of automatic extraction of singularities (both linear
and non-linear singularities), this can be an advantage because it
allows the process in various embodiments to maximize the
identification (ID) sensitivity of catastrophe detection.
Example 9
[0505] This example considers a well-known example. However, the
heuristicity analysis in accordance with various embodiments is
new, leading to a better understanding of the relation between
heuristic and non-heuristic examples of the catastrophe ID.
[0506] Consider the standard linear oscillator equation for the ID
case, where x is the oscillation coordinate; {dot over (x)}=dx/dt,
is its first time-differential; and, {umlaut over
(x)}=dx.sup.2/dt.sup.2, is its second time-differential, with added
non-linear force term, F.sub.x(x), in the form:
{umlaut over (x)}+.omega..sub.0.sup.2x+k{dot over (x)}=F cos
.gamma.t+F.sub.x(x) (114)
where .omega..sub.0 is linear resonance angular frequency, k is
viscosity coefficient (in frequency units), F is amplitude of
stimulating force, y is stimulating force frequency, and F.sub.x(x)
has the form:
F.sub.x(x)=-ax.sup.3 (115)
[0507] This 3.sup.rd-order non-linearity may lead to both the
3.sup.rd harmonic and to the contribution to the linear term, the
latter of which may be essential for the catastrophes, due to the
following solution of Eq. (114), in the form:
A = F ( .omega. 0 2 - .gamma. 2 + 3 aA 2 4 ) 2 + .gamma. 2 k 2 (
116 ) ##EQU00040##
where A is resulting amplitude of oscillations in the form:
x=A cos(.gamma.t-.PHI.) (117)
where .PHI. is the phase. We see that for a=0, Eq. (116) becomes
the standard solution of the linear oscillator. In the vicinity of
the resonance (.gamma..apprxeq..omega..sub.0), the stimulating
angular frequency, .gamma., can be presented in the form:
.gamma.=.omega..sub.0(1+.epsilon.); .epsilon. (118)
[0508] By substituting Eq. (118) into Eq. (115), and assuming k
.omega..sub.0, the following equilibrium surface equation
(z=A.sup.2) is obtained:
F ( , a , z ) = z ( 3 az 4 - 2 ) 2 + k 2 z - F 2 = 0 ( 119 )
##EQU00041##
[0509] The square of amplitude, z, is the state variable, while
(.epsilon., a) are control variables. By differentiating function,
F, twice in respect to z, the locations of the fold and cusp
catastrophes can be obtained. In particular, the location of two
cusp catastrophes, is given by
(a,.epsilon.)=.+-.(32k.sup.3 {square root over (3)}/27F.sup.2,k
{square root over (3)}/2) (120)
[0510] FIGS. 48A-C illustrates example of non-linear oscillator
catastrophes. FIG. 48A is an example of a bifurcation set for
non-linear oscillator catastrophes. FIG. 48B is an example of
linear singularities for a non-linear oscillator for a=a.sub.1.
FIG. 48C is an example of non-linear singularities for a non-linear
oscillator for a=a.sub.0.
[0511] In FIG. 48A, bifurcation set 2050 for non-linear oscillators
is presented including cusp catastrophes 2051 and 2052 and fold
catastrophes. Each cusp catastrophe in various embodiments can be
configured to generate two branches of fold catastrophes for the
so-called hard oscillator 2053 when a>0 and soft oscillator 2054
when a<0. The hard non-linear oscillator 2053 is discussed in
detail.
[0512] In FIG. 48B, the linear singularities for non-linear
oscillator 2055 are discussed outside of the bifurcation region,
which is inside two fold catastrophes branches 2056 and 2057, for
a=a.sub.1 following evaluation path 2058 in FIG. 48A. In FIG. 48B
the resonance curve 2059 is in the form:
z=z(.epsilon.) (121)
which is shown, in a normalized form. As this example illustrates,
the resonance width 2060 may be equal to k, which is in agreement
with standard linear oscillator theory. According to Eq. (120), the
location of cusp catastrophe for hard case, is
= 2 = k 3 2 = 0.87 k > 0.5 k ( 122 ) ##EQU00042##
i.e., it is outside the resonance curve width, as in FIG. 48B. Of
course, the resonance is located at .epsilon.=0, as shown by the
curve 2059. Therefore, FIG. 48B shows the linear singularity (a
maximum).
[0513] In contrast in FIG. 48C, the non-linear singularities
(catastrophes) 2061 are shown including a characteristic hysteresis
effect 2062. This is because the evolution path, FEDBA, does not
coincide with the evolution path, ABCEF, in the other direction. As
described above, this creates the hysteresis effect (in FIG. 48C)
the resonance at .epsilon.=0 has been omitted for sake of
simplicity). The hysteresis also creates bi-stability (i.e. a
situation where the system is bifurcating between two states: upper
(FED) and lower (CBA)). For DB-catastrophe 2063, a drop is
indicated, while for CE-catastrophe 2064, a jump is indicated.
Because FIGS. 48B and C present z-cross-sections, the catastrophes
2063 and 2064 can be reduced to points 2065 and 2066 respectively
in FIG. 48A. The full evolution path ABCEFEDBA is also shown in
FIG. 48A as 2067. This illustrates that the region of catastrophes
or bifurcations 2053 is symmetrical to region 2054 for a soft
oscillator. Therefore, both hard and soft oscillators behave
symmetrically in respect to catastrophes.
[0514] Heuristicity Analysis.
[0515] Comparisons of Examples 7 and 8 provide heuristicity
comparisons. This is, because, EXAMPLE 7 is for an extremely
heuristic case, while EXAMPLE 8 is extremely non-heuristic. Indeed,
in the case of EXAMPLE 8, it is important to know oscillator theory
very well in order to find catastrophes, while in the case of
EXAMPLE 7, is somewhat easier to select measurement variables. It
can be seen that this difference is drastic only for non-linear
singularities, while for linear singularities, the process, in
general, can be very heuristic, assuming a causality principle. In
contrast, in the case of non-linear singularities we have a full
spectrum of estimations, some of them very heuristic, other very
non-heuristic. In this context, we see that non-heuristic solutions
are rather narrow in a sense of application, while heuristic ones
are generally broader. However, referring to EXAMPLE 8, the
solution is not so narrow because the resonance phenomenon is
rather broad, with applications in mechanics, electronics,
acoustics, biophysics, etc.
[0516] Table 8 provides a summary comparison between example linear
and non-linear singularities. Smooth, regular continuous functions
and manifolds, lead to discrete singularities, the prototype of
anomalous events.
TABLE-US-00007 TABLE 8 Summary Comparison of Linear and Non-Linear
Singularities Non- No. Feature Linear Linear 1 Typical Examples
Maximum, Catastrophes Minimum, (Jumps) Inflection Points *) 2
Topological Type Continuum Continuum 3 Analythic Geometry can be
Yes Yes Applied **) 4 Heuristicity High Lower 5 Causality Principle
Yes Yes 6 Phase-Space is Valid Yes Yes 7 State/Control Variables
***) Yes Yes 8 Discretization of Continuity ****) Yes Yes 9 ID of
Anomalous Events Yes Yes 10 Regression can be Applied Yes Yes 11
Sampling Theorem can be Applied Yes Yes *) And higher inflection
points (higher differentials than of the 2.sup.nd order). **) This
feature is a consequence of continuum topology. ***) Feature (7) is
equivalent to (6). ****) The essential point.
[0517] Various embodiments of the technology disclosed herein
include a C2 Weather Sensor System (C2WS2), which is the particular
case of the C2 Sensor System. The C2WS2 contains Command-Control
(C2) system structure applicable to Weather Station. In the Weather
Station case, the anomalous event is a Weather Anomalous Event, or,
shortly, WAEVENT based on the Software Engine, or, more
specifically, on Truthing-based Anomalous Event Software Engine
(TAESE), as described in FIG. 24. The WAEVENT Pre-Structuring, as
discussed above, can be important to define the sampling space used
for Bayesian Truthing of the TAESE, including Weather Data Event
Format (WDEF) as a sample, shown in FIG. 28.
[0518] The optimum selection and identification (ID) of a WAEVENT
can take advantage of the WAEVENT Sensor Fusion (WSF) software
engine, an example of which is discussed above in reference to
FIGS. 26 and 27. The selection and identification of WAEVENTS can
be done either in the digital domain (e.g., as in FIG. 24) or in
the topologic (analog) domain, leading to a yellow/red, dual-alarm
Autonomous Decision Generation Process (ADGP). Any of the decision
paths, whether digital, or topologic (analog), can result in a
yellow alarm, while two yellow alarms from both paths and at the
same time create the red alarm, within the dual-alarm structure,
introduced for Information Quality (IQ) purposes.
[0519] Four (4) or more information structures may be included in
various embodiments. These include the well-known Information
Assurance (IA), Information Security (IS), Information Hardening
(IH), and Information Quality (IQ).
[0520] The Bayesian Inference, based on the PPV figure of Merit
(FoM), is introduced only for Performance Metrics purposes rather
than for decision process purposes. This is an important
distinction because the latter option suffers from the autonomous
system's low actionability. As used herein PPV can be taken to
refer to the Positive Predictive Value.
[0521] The cost of bandwidth is also an important factor as
discussed above. This is because both RF-power and processing power
are approximately proportional to bandwidth, within
SWaP2-constraints (in which "P2" refers to both RF and processing
powers).
[0522] While IA may be thought of in various applications as a
familiar cipher term relating to the encryption and decryption
process, within .epsilon..sub.CRYPTO-coefficient, and Information
Quality (IQ) has been discussed above, the IS and IH require
further explanation. In particular, the IS is related to the
protection of a location of source information, while the IH may be
considered a more general term introduced by analogy to device
(hardware) hardening against harsh environmental conditions. This
is, because, even in the case of ideal crypto system, the lossy
(video) compression and environmental noise introduces additional
errors, as discussed above.
[0523] FIG. 49 is a diagram illustrating an example of a CONOPS
2100 for a weather station such as a C2 Weather Sensor System
(C2WS2) in accordance with one embodiment of the technology
described herein.
[0524] This exemplary systemic Weather Station with C2 capability
has three (3) sources of weather data: (1) weather data from its
own sensors 2101, (2) weather data from other weather stations, and
(3) weather data from the Command Control Center (CCC) 2102. In one
embodiment, weather data such as weather data 2102 can be received
through one or more wireless (RF) communication channels, and its
own database 2103. The double arrow 2104 is provided to illustrate
the fact that in various embodiments the database 2103 can support
various summaries. Examples of these summaries can include: tables,
look-up tables, and other lists, which can work for both TAESE
(Truthing-based Anomalous Event Software Engine) 2106 and DT2 (Data
Topologic Transfer) 2107 (the 1.sup.st one working in the digital
domain), while the 2.sup.nd one may be working in the analog
domain. The phrase "analog domain" can be used to refer to digital
experimental and estimation data that are transferred into a
topologic continuous domain, including linear and non-linear DTS
(Digital Topologic Singularities). The example DTS structure shown
in FIG. 49 shows an example in which digital data is transferred to
an analog (topological) domain, and, by heuristic, or
semi-heuristic processes is transferred to a digital DTS. In
various embodiments, a linear DTS may include: maxima, minima,
inflection points, higher inflection points (for higher order
differentials), and above-threshold points, while non-linear DTS
may include various types of catastrophes.
[0525] The TAESE 2106 can be configured to generate a yellow alarm
or rather a T-alarm (or no alarm) 2108; while the DT2 2107 can also
generate a yellow alarm or rather a D-alarm 2109 (or no alarm).
Both yellow alarms may be synthesized within an Autonomous Decision
Generation (ADG) sub-system 2110. If two yellow alarms are produced
for the same sample, the ADG may be configured to produce a red
alarm. If only one yellow alarm is produced by either of 2108 or
2109, the T/D yellow alarm may be produced 2112. If neither yellow
alarm is produced no alarm is generated, thus defining the
soft-decision process. In parallel, the database 2103 can be
configured to produce messages 2114 as a kind of relay with, for
example, minimum data micro-processing (.mu.P) or only
micro-controlling (.mu.C). These alarm outputs 2111, 2112, 2113 and
messages 2114 may be transmitted to output data interface 2115,
which can further re-transmit these messages to devices such as,
for example, a PC cartridge or other device 2116 whether via a
wired link or wirelessly 2117.
[0526] FIG. 50 is a diagram illustrating a cross-domain DT2
structure in accordance with various embodiments of the technology
disclosed herein. In the example shown in FIG. 50, an example of
the detailed cross-domain DT2 structure 2109 is presented. This
example includes Including digital (input) measurement and
estimation data 2150, topological (analog) domain 2151, and, again,
digital domain, 2152. As described above, this "cross-domain"
operation is useful to obtain discrete events 2152 from continuous
2151 which was synthesized from digital input data 2150.
[0527] Referring again to FIG. 49, the output data 2115 may be
constrained by Information Assurance (IA) 2118, Information
Security (IS) 2119, Information Quality (IQ) 2120 and Information
Hardening (IH) 2121 as explained briefly above in accordance with
the example embodiments. These are also discussed in greater detail
below.
[0528] Further embodiments relating to specific solutions for
Information Constraints IA, IS, IQ, and IH are now described.
[0529] Information Assurance.
[0530] The Information Assurance (IA) may be related to data cipher
operations such as, for example, data encryption and decryption
(i.e., transformation of data from plain text or clear text to
cipher text and vice versa). The cipher key in various embodiments
can include two parts or two half-keys: encryption keys and
decryption keys. The cipher key can be a symmetrical key or an
asymmetrical key. In the case of asymmetrical keys, the encryption
half-key is usually a public key while the decryption half-key is
usually a secret key. Therefore, if the public key is transmitted
to a transmitter party, this party can only encrypt the data and it
is not able to decrypt encrypted data.
[0531] Dealing with situations involving a cross-domain of red and
black data they present challenge. However, various embodiments of
the disclosed technology are not concerned with the crypto key but
are instead concerned with the other half-key. In various
embodiments, the other key can be a key used to enable
communications. Accordingly, by way of nomenclature, this key can
be referred to from time to time as a "turn-on-engine" key or a TOE
key for short. Conventional keys for turning on or turning off a
process are typically mechanical or electronic keys. In contrast,
in various embodiments the TOE key may be implemented, for example,
as an RF key. The "turn-on-engine," or TOE-operation, can be
analogized analogous to that function for car key. In other words,
without a "turn-on-engine" key applied to a weather station, the
weather station will not operate because it cannot be turned on.
Likewise, after the TOE-key is removed, the weather station will in
various embodiments stopped communicating.
[0532] While various of these features may be known, embodiments of
the disclosed technology include an additional feature of the
TOE-key. For example, in some embodiments the TOE-key is an RF-key.
More specifically, the TOE-key may be implemented in various
embodiments to be wirelessly connected with the key owner in such a
way that he/she keeps in his/her pocket the additional sub-key,
which is RF-connectable with the TOE-key. This can be implemented,
for example, as an RF proximity connection with a maximum
connection distance (e.g., up to 50 m i.e., for >50 m--this
connection is broken). Therefore, even if RF proximity TOE-key is
lost or stolen, it cannot be used to turn-on the weather station
(all functions can be locked or just communications, for example)
unless the thief is nearby.
[0533] FIG. 51 is a diagram illustrating an example of an RF TOE
key implemented as an RF proximity key in accordance with one
embodiment of the technology disclosed herein. As illustrated in
the example of FIG. 51, the RF TOE key 2202 comprises 2 parts 2203,
2204. These two parts 2203, 2204 can be communicatively coupled via
wireless connection 2205. In various embodiments, wireless
connection 2205 can be implemented as an RF connection, however
alternative wireless connections can be implemented. In various
embodiments, the communication between parts 2203 and 2204 can be
two-way communication to allow information to flow between the two
parts 2203, 2204 in both directions.
[0534] As noted above, in various embodiments, the wireless
connection 2205 can be distance limited to provide a measure of
security by requiring the 2 keys to be within a certain distance of
one another for operation. Accordingly, in the example illustrated
in FIG. 51, the wireless connection 2205 is only operable when the
2 parts are within the maximum connection distance, d, 2206. In
some embodiments there can be user specific requirements for the
use of key 2203. However, in other embodiments, key 2203 can be
used by anybody, but the holder 2207 of subkey 2204 (e.g. in his or
her pocket, or in his or her possession) must remain within the
predetermined distance, d, 2206 such that:
d.ltoreq.d.sub.0 (123)
where d.sub.0 is some threshold distance, defined by the
specification of the wireless electronics. In various embodiments,
the maximum distance, d, is in the range of d.about.20-50 m,
however other distances can be used.
[0535] Accordingly, in order to operate the weather station 2201,
the key owner or holder 2207 must either put key 2203 into its key
slot 2208 him or herself, or must be within the maximum distance
when another person put key 2203 into its key slot 2208. Although
illustrated and described as a conventionally shaped key with a
corresponding key slot, other shapes sizes and configurations of
keys can be used for key 2203. Key 2203 can be implemented in any
of a number of forms of mechanical, electronic, or
electromechanical keys they can be used to "unlock" weather station
2201.
[0536] Because in such embodiments one person can be designated as
the keeper of key 2203, and another different person can be
designated as keeper of the subkey 2204, this further increases the
security of the connection (i.e., IA) and information security
(IS). Where the condition as set forth in equation (123) is broken,
weather station 2201 cannot be turned on for operation. In some
embodiments, the key arrangement requires that the keeper of key
2204 be in proximity to key 2203 at all times during operation or
operation of the weather station 2201 is shut down. This can result
in a high security operation for the weather station 2201, assuming
that the battery system of subkey 2204 is sufficiently
hardened.
[0537] In addition to the information security and information
assurance aspects provided by security solution described above
with reference to FIG. 51, additional or alternative Information
Security (IS) solutions can be considered as protection of the
location of the RF source related to Weather Station. This is
because weather stations operating (especially those operating in
remote locations) may be or designed to communicate through
satellite channels that can be spotted. Accordingly, for purposes
of information security, it may be preferable to not disclose the
location of some critical person or unit.
[0538] Therefore, some embodiments use a cartridge type solution
2300. FIG. 52 is a diagram illustrating an example of a weather
station cartridge in accordance with one embodiment of the
technology disclosed herein. Referring now to FIG. 52, in this
example a weather station cartridge 2301 is designed to be inserted
into a cavity 2302 (which may be of complementary shape or geometry
to cartridge 2301). As this example illustrates, a cartridge door
2304 can be used to close off cavity 2302. In some embodiments,
cartridge door 2304 can be configured with appropriate seals to
provide weather sealing for cavity 2302 (and for cartridge
2301).
[0539] FIG. 52 also provides an illustration of an expanded view of
an example cartridge 2306 (e.g. weather station cartridge (WSC)
2301). As this example illustrates, cartridge 2306 can include a
housing 2308, memory 2307 (e.g., flash memory) and a GPS Beacon
2309, and vibration protection or dampening substance 2310.
[0540] The weather station cartridge 2301 can be implemented in
such a way as to provide a plurality of functions, examples of
which may include: extraction of high-bandwidth (such as video)
data without troublesome cabling, including an adequate power
source 2311 (e.g., battery power), data ground and power pins 2312
(+5 V, for example) and adequate electronics and mechanics;
concealing a location of the cartridge since it can be brought to a
location that does not use vulnerable RF-communication;
high-quality transport of sensitive data (to avoid vulnerable
RF-communications); etc.
[0541] Aspects of the technology in various embodiments relating to
Information Quality (IQ) is addressed throughout this document.
Various embodiments enhancer and prove information quality through
the use of, for example, Bayesian Truthing (BT) and the sampling
space with anomalous events as rare targets, or signals (S). These
anomalous events are referred to herein as WAEVENTS in the case of
the Weather Station. The WAEVENTS may be detected, selected and
identified (ID) using a Cross Domain DT2 structure, an example of
which is described with reference to in FIG. 49. WAEVENTS may also
be detected, sometimes in parallel, using the TAESE (Truthing-based
Anomalous Event Software Engine), which in some embodiments relies
on an Autonomous Decision Generation (ADG) sub-system 2110.
Particularly, various embodiments can utilize the ADG 2110 as a
digital assistant to facilitate operation of the weather station in
modes beyond that of merely a Data Transfer System (DTS). In
various embodiments, the WAES (Weather Anomalous Event System) can
be implemented using the C2WS2 structure, an example of which is
illustrated in FIG. 49, and may further include the TAESE
(Truthing-based Anomalous Event Software Engine), an example of
which is described in detail with reference to FIG. 24.
[0542] Protection of information in various embodiments may be
referred to herein as Information Hardening (IH). Such reference is
used from time to time by analogy to device hardening, which
typically refers to some level of protection against adverse or
hostile environment (including TEMPEST countermeasures). In
particular, the IH generally refers to hardening of information
against a harsh environment that could otherwise lead to increasing
Bit-Error-Rate (BER). Embodiments of an IH solution are discussed
above as relating to minimization of bandwidth cost. In particular,
increasing bit error rates dictates EFEC-increasing, as in Eq.
(67), in the form:
B A = B 0 x ( A - y ) ; A = 1 - a ; a = CRYPTO + NET ( 124 )
##EQU00043##
where: x=(CR) and y=.epsilon..sub.FEC, as well as:
.phi. = arcsin ( y A ) ( 125 ) ##EQU00044##
where .PHI. is a parameter as explained above with reference to
FIG. 29.
[0543] According to Eq. (126), it can be seen that for B.sub.A,
B.sub.0=constant, the following conservation relation exists:
x(A-y)=CONSTANT=I.sub.0 (126) [0544] (where I.sub.0 is an
invariant). This means that if the y variable increases; then, also
the x variable increases. Thus, from Eq. (77), also .PHI.-parameter
increases, according to the following causation relation:
[0544] (BER).uparw.y.uparw.x.uparw..PHI..uparw. (127)
[0545] Therefore, if the system has already been optimized using
the (PSNR)-criterion, as in FIG. 29; then, with .PHI. increasing as
in Eq. (127), the system moves into: .PHI.>.PHI..sub.0, in the
form:
(.PHI.=.PHI..sub.0)(.PHI.>.PHI..sub.0) (128)
[0546] The optimum solution for various applications may be to
reduce the value B.sub.0 (keeping B.sub.A=constant), thus, reducing
x, y values (and the value of .PHI.), and thus, in turn, returning
to the maximum (PSNR) value (or, to the vicinity of this
value):
(PSNR)=U(.PHI..sub.0)=MAXIMUM (129)
[0547] Therefore, the information hardening (IA), maybe closely
related to minimizing the bandwidth cost, as shown in FIGS. 53A and
53B. Particularly, FIGS. 53A and 53B are diagrams illustrating an
example of a return-to-maximum procedure, including: (FIG. 53A)
Decreasing U-value; (FIG. 53B) Increasing U-value.
[0548] In the example illustrated in FIGS. 53A and 53B, the
procedure of "Return-to-Maximum", or RTM-procedure, is shown,
including a decreasing U-value (FIG. 53A) and an increasing U-value
(FIG. 53B), where U=(PSNR). In FIG. 53A, a decreasing U as a result
of increasing (BER) is shown at 2400. This leads to losing a
maximum U.sub.M value 2401 due to an increasing .PHI. parameter
2402. In order to increase the U value again as in FIG. 53B 2403,
the system can be configured to return the .PHI. value back to
.PHI.=.PHI..sub.0 has seen at 2404. However, this is impossible
unless I.sub.0 value 2405 is reduced, which can only be done by
reducing B.sub.0 value according to Eq. (124).
[0549] An innovation associated with this is in the fact, that, due
to the above procedure, the system can determine a level of
reduction needed in the bandwidth B.sub.0 value in order to return
into the previous (PSNR)-value.
[0550] The reduction of the B.sub.0 value in the case of video
signal transmission can be accomplished by adjusting one or more of
the following factors alone or in combination: [0551] 1) Display
format (resolution) [0552] 2) Pixel dynamic range, in bpp [0553] 3)
Frame Rate [0554] 4) Reducing color into black-white (grey)
[0555] In a similar manner, embodiments can be implemented to
provide the (PSNR) maximization procedure when B.sub.0=constant,
but B.sub.A is not constant. However, the latter case is less
practical than the previous one.
[0556] The designation C2WS2 (C2 Weather Sensor System) as used
herein in various embodiments refers to a Weather Sensor System
with Command-Control (C2) capability. In particular, such a system
should preferably be compliant with Command-Control-Center (CCC),
including in the IA-sense. For example, in the case of highly
sensitive information, the CCC can send a public encryption key in
order to receive weather information data encrypted by an
asymmetric cipher key. In general, the C2WS2 can be configured to
receive data from its own meteorological sensors: S.sub.1, S.sub.2,
. . . , S.sub.n, and from other Weather Station sensors through the
CCC, mostly by satellite communication channel, or by other wired
or preferably wireless communication links. In various embodiments,
these data may be parallelized through Wireless (or wired) Sensor
Star Communication Interface (WSSCI). Then, the weather event
messages may be pre-structured as in FIG. 28, for example, within
Weather Data Event Format (WDEF). When the Weather Anomalous Event
Ranking (WAER) exceeds some threshold value, then such event may be
classified as yellow alarm, due to the TAESE (Truthing-based
Anomalous Event Software Engine). In parallel, the weather data may
be summarized through the DT2 (Data Topologic Transfer), as in FIG.
49.
[0557] FIG. 54 is a block diagram illustrating an example system
2500 including a transmit/receive physical layer 2501 and a
wireless (or wired) Sensor Star Communication Interface 2502, which
may be configured to perform compression and decompression as well
as OVH-operations such as, for example: IA, FEC (Forward Error
Correction), cipher, and others, within Data Transfer System
(DTS).
[0558] The example system illustrated in FIG. 54 also includes
advanced operations and components, including Hardened Flash Memory
(HFM) 2503, a C2WS2 (C2 Weather Sensor System) 2504, and cartridge
2505. Optionally, the information can be transferred by cartridge
2505, by physical means 2506, or downloaded 2507 to a PC (Personal
Computer) 2508 or other work station or computing device. In
parallel, the weather information data can be received by the
weather system's own sensors S.sub.1, S.sub.2, . . . , S.sub.n
2509, and communicated wirelessly 2510 from or to the CCC 2511.
Also, the information can be wirelessly transmitted/received
through satellite communication channel, or other wireless channel
from/into other sources 2512.
[0559] In parallel, the PC 2508 is connected with extra DeCODEC
2513, RAR (Random Access Retrieval) 2514, a Graphical User
Interface (GUI) 2515, display 2516 and other sub-systems 2517.
[0560] The transmission of encryption keys and injection keys
through an RF interface can be a challenge due to errors that can
arise with wireless communications. In many circumstances, any
error (even single error) in communication of that data
representing those keys, sometimes referred to as Information
Assurance (IA) keys, can render an encryption or decryption system
inoperable. This damage often cannot be corrected in sufficient
time relative to the time criticality of the security operation. A
relevant-for-the-encryption operation (REO) time may be defined in
such a way that, during the REO time, breaking an encryption key
would compromise the IA of the system; i.e., if during the
REO-time, t.sub.R, the encryption key (or, injection key) is
broken, then, the IA of the overall system is compromised, leading
to the following relation:
t.gtoreq.t.sub.RIA is not compromised (130a)
t<t.sub.RIA is compromised (130b)
Eqs. (130a) and (130b) define the REO-time, t.sub.R.
[0561] For the purpose of various embodiments, the encryption key
can be defined as comprising two half-keys: an encryption half-key,
EK, and a decryption half-key, DK.
[0562] For asymmetric keys, based on factorization of two large
prime numbers (or, primes), PN1 and PN2; the EK, represented by PN1
is public, while the DK is secret. With asymmetric keys, a receiver
(Rx) of cipher text (i.e., encrypted text, in contrast to
non-encrypted plain text) can be configured to send the public
EK/PN1 to the transmitter (Tx). Then, the transmitter applies this
public half-key to encrypt the text and sends to Rx. In this
moment, the transmitter is not able to decrypt its own cyber text.
After sending the cyber text by Tx into the Rx, the Rx decrypts
this cyber text into plain text using the secret DK, representing
both primes PN1 and PN2.
[0563] For symmetric keys, on the other hand, both half-keys, the
EK and the DK are maintained as secret. In any case (i.e., for both
symmetric and asymmetric keys), the transmission of encryption key
through RF channel is a challenge, related to Eq. (130)
conditioning.
[0564] As noted above, injection keys may be analogized to a
"start-engine" key for an automobile. In various embodiments,
injection keys can be mechanical, electronic, or a combination of
the 2. Therefore, the injection key, IK, is referred to herein in
some embodiments as the "turn-on-key" of the encryption system.
[0565] Embodiments of the technology are related to IA-secure
transmission of both encryption and injection keys through an RF
(wireless) channel, which is typically much more prone to errors
than cable channels, or wired channels. Embodiments may also be
applicable to optical wireless (so-called Free-Space-Optics (FSO))
channels, acoustic channels, and other wireless channels.
[0566] An RF channel accordance with various embodiments of the
technology disclosed herein is now described. Consider an example
binary data stream such as: 1, 0, 0, 1, 1, 1, 0, 1, . . . . , which
should be highly robust, or quasi-robust (QR). Of course, any
binary data stream can be protected by error-correcting codes; such
as, for example, Forward-Error-Correcting (FEC) codes. However,
those codes are not perfect, because they are limited to cases in
which the number of errors, m, for a number of bits, n, is a
relatively small number (e.g., typically m.ltoreq.2, while the case
of m>2 is not protected). This is typically not a problem for
most wired channels when the probability of error per bit, q, is
very small. However, this property may be coming problem when the
communication channel is wireless, whether RF, optical, acoustic,
or, otherwise. This is because, in such cases, the error
probability, q, may not always be sufficiently small because it
depends on weather and other factors that can affect the
communication link.
[0567] In general, due to the statistical nature of bit errors, the
problem does exist, especially for wireless communication channels.
This is, because, applying the FEC codes for large m-numbers would
be costly in terms of bandwidth. In other words, the bandwidth
overhead (OVH) cost of protecting the data stream would be too
high. Embodiments of the technology disclosed herein can be
implemented to address this situation.
[0568] For injection keys, a time delay IA problem may also exist.
This can be validating case in which, in addition to parallelity,
the simultaneousness of the IA keys is issue. This is due to not
only weather conditions, but also because network OVH control
introduces uncontrollable time delays that can be both statistical
and deterministic.
[0569] Weather factors such as, for example, air turbulence, can
introduce multi-path errors and other statistical error problems,
even without obstacles. In particular, air turbulence introduces
dielectric constant, .epsilon., fluctuations, which in turn caused
the refractive index, n, of the communication channel to fluctuate.
One reason for this is that, for non-magnetic media
(.mu..apprxeq..mu..sub.o), the refractive index, n, is n= {square
root over (.epsilon.)}, where .epsilon.-relative dielectric
constant. Higher temperature gradients, .DELTA.T (T in Kelvin),
higher winds, etc., create higher air turbulence, which in turn can
cause higher refractive index, An, fluctuations. High temperature
gradients can exist, for example, in the vicinity of the so-called
marine layer, for example. Typically, the higher the air
transparency, the higher the air turbulence. In addition, high
RF-signal attenuation, due to: fog, mist, etc., may also create
statistical binary errors, (as well as other unwanted effects).
This description sets forth some of the multitude of causes of
statistical errors within RF communication channel.
[0570] For the purpose of this technology this document also
discusses the non-obvious statistical relation between probability
of error, or errors, W.sub.n(m), probability of error per bit, q,
and bit-error-rate (BER).
[0571] Bit-Error-Rate (BER).
[0572] BER testers measure BER values in the following way, based
on a definition of absolute (i.e., not conditional) probability,
p', in the form:
p ' = { lim n -> .infin. ( Number of Errors Total Number of Bits
Per Data Stream ) } Ensemble Average ( 131 ) ##EQU00045##
[0573] The limit shows the number of bits tending to infinity, cc.
For smaller n-numbers, this relation fluctuates, and then it tends,
asymptotically, to define the limit defined by Eq. (131), assuming
that the statistical ensemble is stationary and ergodic, which is
the typical case of an RF-channel if air turbulence is also a
stationary random process. Otherwise, the more complex case of
non-stationary processes must be considered as discussed, for
example, in the book by M. Born, E. Wolf: Principles of Optics,
Cambridge Univ. Press, 7th Edition, 1999; Section 10.2, A Complex
Representation of Real Polychromatic Fields.
[0574] It is noted that the nature of experimentation is that
experimentation is unable to provide a priori information fully
from only experimental data. In fact, only a combination of
experimental and theoretical data allows us to obtain a priori
information (in contrast to a posteriori information which can be
obtained from experiment, only). However, only a priori information
enables a predictive analysis, which is also the subject of this
technology. This is the fundamental epistemologic problem,
discussed by Kant, Mckay, Brillouin, and others, which is relevant
for the disclosed technology. A further complication is the
relation between the theory and experiment as well as a connection
of this relation to entropy, as discussed, for example, by L.
Brillouin in: Science and Information Theory, Academic Press,
1956.
[0575] Various embodiments of the technology disclosed herein can
be configured to formulate practical conditions for obtaining a
quasi-robust (QR) RF communication channel that allows the systems
to send IA-secure encryption and injection keys. In particular, Eq.
(131), defining what is referred to herein as a "smoothed
periodogram," generalized to binary processes, shows that, for
stationary and ergodic statistical processes:
lim n -> .infin. p ' = q = ( BER ) ( 132 ) ##EQU00046##
where q is the probability of error per bit. Then, the probability
of no-error per bit is: p=1-q. Thus, the following conservation
relation can be obtained:
p+q=1 (133)
[0576] In the view of the above discussion, the probability of
m-number of errors within a binary data stream of n-number of bits,
W.sub.n(m), is not equal to q probability:
W.sub.n(1).noteq.q (134)
which is a non-obvious relation.
[0577] It can be shown that, under the above statistical
assumption, the NEP (Number of Errors Probability) equal to:
(NEP)=W.sub.n(m) (135)
[0578] Leads to the Poisson Statistical Distribution for small
m-numbers, and large n-numbers, where W.sub.n(m) is the probability
of m-number of errors and n is the number of bits per given data
stream. It should be noted that the difference between Eq. (131)
and Eq. (135), which is not obvious, leads to the nonobvious
relation (134). This is because Eq. (131) is defined by asymptotic
limit:
n.fwdarw..infin. (136)
while Eq. (135) holds for finite n-numbers.
[0579] The number-of-errors probability, NEP (not to be confused
with noise-equivalent-power for optical signals), is defined by
binomial statistical distribution (Pascal, Newton, Bernoulli), in
the following form:
W n ( m ) = q m p n - m ( n m ) where ( 137 ) ( n m ) = n ! ( n - m
) ! m ! ( 138 ) ##EQU00047##
and n! is a factorial, defined as: n!=n(n-1)(n-2) . . . (1), with
1!=1, and 0!=1, and the NEP-probability, W.sub.n(m), satisfies the
following conservation relation:
m = 0 n W n ( m ) = 1 ( 139 ) ##EQU00048##
[0580] Eq. (139) shows that Eq. (137) is, indeed, valid for a
finite n-number, and, shows that the number of errors, NEP, holds
for any m-number between m=0 (no errors) and n. Thus, it is a
certainty. Also, the following binomial relations hold for the
statistical mean, m, and standard deviation (or, dispersion),
a:
m=nq; .sigma..sup.2=npq; (140ab)
[0581] It should be noted that, for RF-channel: q<<1. In
contrast, the mean value, m, does not need to be small. However,
for a quasi-robust (QR) RF-channel, the m value should be much
smaller than 1:
m<<1 (141)
[0582] Also, although for q<<1, p.apprxeq.1:
q<<1p.apprxeq.1 (142)
the p.sub.n--value can be still small.
[0583] Thus, for q<<1, Eqs. (140ab) become,
m=nq; .sigma.= {square root over (m)}; for q<<1 (143ab)
[0584] For small q-numbers and large n-numbers:
q<<1; n>>1 (144ab)
[0585] This is the case of the IA-keys, because, for typical
RF-channel, q and n can be given as
q=(10.sup.-6-10.sup.-4); and, n=(100-1000) (145ab)
[0586] This is, because, for n=256-key-length, header OVH,
encryption OVH, and FEC-OVH, and further network OVH are added.
[0587] However, for the general RF-channel, the following
NEP-formulas exist for small m-numbers:
W n ( 0 ) = q 0 p n ( n 0 ) = p n ( 146 a ) W n ( 1 ) = qnp n - 1 (
146 b ) W n ( 2 ) = q 2 p n - 2 n ( n - 1 ) 2 ( 146 c ) W n ( 3 ) =
q 3 p n - 3 n ( n - 1 ) ( n - 2 ) 6 ( 146 d ) ##EQU00049##
and, we can continue these exact binomial formulas until m=n.
[0588] For QR-channel, however, we have: m<<1; thus, the
approximate formulas, valid for m<<1, and small m-numbers,
are
W n ( 0 ) = p n ( 147 a ) W n ( 1 ) .apprxeq. m _ p n ( 147 b ) W n
( 2 ) .apprxeq. m _ 2 p n ( 1 2 ) ( 147 c ) W n ( 3 ) .apprxeq. m _
3 p n ( 1 6 ) ( 147 d ) W n ( m ) = W n ( 0 ) ( ( m _ ) m m ! ) (
147 e ) ##EQU00050##
where, Eq. (147e) is valid only for:
m<<n (148)
[0589] According to condition (148) (valid only for small m-numbers
relative to the n-value) Eq. (147a) can be applied in the form:
W n ( m ) = ( 1 - q ) n ( m _ ) m m ! = e n n ( 1 - q ) ( m _ ) m m
! ( 149 ) ##EQU00051##
where n ( . . . ) is natural logarithm, and e is natural logarithm
base. However, for q<<1, we obtain the following linear
Taylor series' term:
n(1-q).apprxeq.-q (150)
thus, Eq. (149) becomes:
W n ( m ) = e - m _ ( ( m _ ) m m ! ) ( 151 ) ##EQU00052##
which is the Poisson distribution, used in the theory of
radioactivity, for example. However, the classic Poisson statistics
are obtained as a limit to infinity (n.fwdarw..infin.) for
q<<1 of binomial distribution, in the form:
lim n -> .infin. W n ( m ) = e - m _ ( m _ ) m m ! ( 152 )
##EQU00053##
which looks identical to Eq. (151), but it is derived in a
different way, as shown, for example, in: H. Margenau and G. M.
Murphy, The Mathematics of Physics and Chemistry, Robert E. Krieger
Publishing Company, 1976. In contrast, Eq. (151) is obtained,
independently, assuming condition (148) is satisfied, which is
exactly equivalent to the QR-wireless communication channel. This
is an argument that Eq. (151) is not obvious for a QR-wireless
communication channel.
[0590] Conditions for the QR-channel (e.g., a communication channel
that allows sending the IA keys (including injection and encryption
key)) satisfying Eq. (130a) as a basic condition of effective
QR-channel is now described. For these purposes, embodiments of the
technology should satisfy the following conditions:
[0591] A. Condition (141) must be satisfied
[0592] B. Condition (148) must be satisfied
[0593] C. The error-correction code protects against:
m.sup.<m.sub.o (153)
[0594] D. Statistical Ensemble (131) must be stationary and
ergodic
[0595] E. Condition (130a) is satisfied
[0596] Then, according to Eq. (153), Eq. (151), for m=m.sub.o,
becomes,
W n ( m o ) = e - m _ ( ( m _ ) m o m o ! ) = W n ( 0 ) ( m _ ) m o
m o ! ( 154 ) ##EQU00054##
[0597] The QR-channel is defined as a wireless communication
channel such that W.sub.n(m.sub.o) is sufficiently small value in
order to satisfy conditions: A, B, C, D, E. It should be observed
that, in contrast to other (theoretical) conditions, Condition C is
an experimental one, which allows avoidance of so-called burst
errors.
[0598] Condition A is based on Eq. (141), in the form:
m=nq<<1 (155)
then, according to Eq. (146a), and approximation (150), the
probability of no-errors, W.sub.n(0), is
W.sub.n(0)=p.sup.n.apprxeq.e.sup.-m=e.sup.-nq (156)
where q is the probability of error per bit, or BER, and n is the
number of bits per data stream, which can include the IA key with
OVH, including a header. However, in various embodiments, for the
purpose of predictive error analysis, only total number of bits, n,
counts.
[0599] In order to explain Condition A, we present number of bits,
n, in decimal base, in the form:
n=10.sup.a (157)
where a is (usually) an integer, for simplicity, such as a=3, for
example. Then, n=10.sup.3=1000. The BER is a small number that can
also be presented in decimal basis:
q=(BER)=10.sup.-b (158)
where b is (usually) integer, for simplicity, such as b=5, for
example; then, q=10.sup.-5.
[0600] For the sake of illustration, consider the constant value of
W.sub.n(0)-probability, leading to the following relation:
nq=constant=10.sup.a-b (159)
thus:
log 10.sup.(a-b)=a-b=constant (160)
[0601] This relation is illustrated in FIG. 55, for: a-b=0; a-b=-1;
and a-b=-2. Particularly, FIG. 55 is a diagram illustrating an
example of a necessary condition of probability of a no-error per
data stream.
[0602] In Table 9, the corresponding values of nq and W.sub.n(0)
are presented. We see, that, for larger negative number of (a-b),
we obtain W.sub.n(0)-values closer and closer to 1, with "number of
nines" equal to |a-b|. For example, for a-b=-4, we obtain: b=a+4,
and, for typical a=3 value, equivalent to n=1000, we obtain b=7, or
BER=10.sup.-7, as shown in Table 10.
TABLE-US-00008 TABLE 9 Corresponding Values of nq and W.sub.n(0)
for (a - b) - Values a - b 0 -1 -2 -3 -4 nq 1 0.1 0.01 0.001 0.0001
W.sub.n(0) 0.37 0.905 0.99 0.999 0.9999
TABLE-US-00009 TABLE 10 Corresponding Values of BER for b = a + 4,
Defining QR-Channel a 2 3 4 5 6 BER 10.sup.-6 10.sup.-7 10.sup.-8
10.sup.-9 10.sup.-10
[0603] According to Table 9, the (a-b)-value defines a QR-channel,
the closer W.sub.n(0) gets to 1, the less space is left for
(probabilities of) errors. The basic issue then is how long of a
data stream can be accommodated for a given BER value (such as
BER=10.sup.-5, for example). Then, b=5, and assuming a "two nines"
criterion, for example, leading to W.sub.n(0)=0.99, the following
can be derived from Table 9:
a-b=-2a=b-2 (161ab)
[0604] In the case of b=5, or BER=10.sup.-5, then a=3, or
n=10.sup.3=1000. In other words this is the maximum data stream
length (1000 bits) that can be accommodated without exceeding the
10.sup.-5 BER value. This is illustrated in Table 10. From this it
can be seen that the "two nines" criterion, equivalent to
W.sub.n(0)=0.99 and BER=10.sup.-5, can be satisfied only for data
streams not longer than 10.sup.3-bits in length. For a=4, for
example W.sub.n(0)=0.9<0.99. Therefore, the "two nines"
criterion is not satisfied. This is explained in Table 11. For
example, for n=10.sup.3, W.sub.n(0)=0.99 and therefore, the
criterion is satisfied. However, for n=10.sup.5,
W.sub.n(0)=0.37<0.99 and the criterion is not satisfied.
TABLE-US-00010 TABLE 11 "Two Nines" Criterion of QR-Channel for BER
= 10.sup.-5 a 3 4 5 6 W.sub.n(0) 0.99 0.9 0.37 0.000045 n 10.sup.3
10.sup.4 10.sup.5 .sup. 10.sup.6 a - b -2.sup. -1.sup. 0 1 nq .sup.
10.sup.-2 .sup. 10.sup.-1 1 10
[0605] In FIG. 56, an example of the "two nines" criterion 2600 is
illustrated. The border line 2601 separates the robust channel area
2602 from non-robust channel area 2603. The robust channel area
2602 is illustrated by shading. The border line illustrates Eq.
(161a), equivalent to W.sub.n(0)-value of 0.99, as in point, A,
denoted by 2604. Moving this point to B is equivalent to a
transition from a-b=-2 into a-b=-3.
[0606] Point B, denoted by 2605, is thus in the robust channel area
2602. In contrast, point C, denoted by 2606, is not in the robust
channel area 2602. This is, because, for such point a-b=-1, which
is equivalent to a W.sub.n(0)-value of 0.9, which is smaller than
0.99.
[0607] Condition B is based on Eq. (148), which leads the set of
Equations (147) into the following form; assuming, Condition A is
satisfied:
W n ( 1 ) .apprxeq. W n ( 0 ) m _ .apprxeq. m _ ( 162 a ) W n ( 2 )
.apprxeq. 1 2 W n ( 0 ) m _ 2 .apprxeq. m _ 2 2 ( 162 b ) W n ( 3 )
.apprxeq. 1 6 W n ( 0 ) m _ 3 .apprxeq. m _ 3 6 ( 162 c )
##EQU00055##
[0608] These equations are essential for the QR-channel. This is
because, by applying the error-correcting code (such as FEC-code),
the number of errors, m, smaller than m.sub.o, according to Eq.
(153) can be corrected. For example, for m.sub.o=3, the data stream
can be protected against one and two (m=1, and m=2) errors. Then,
the W.sub.n(3) probability must be small, according to Eq. (162c).
According to Table 11, for: m=0.01, the "two nines" criterion is
satisfied, while for m=0.001, the "three nines" criterion holds,
according to Table 9. According to Eq. (162c), for m=0.001, the
probability of three errors, W.sub.n(3), is equal to (10.sup.-9/6).
This is close to 10.sup.-10, which is a very small probability.
Table 12 shows this probability for various statistical mean,
m-values.
TABLE-US-00011 TABLE 12 Probability of Three Errors for Various
Statistical Mean Values m 0.1 0.01 0.001 0.0001 W.sub.n(3) 1.67
10.sup.-4 1.67 10.sup.-7 1.67 10.sup.-10 1.67 10.sup.-13
[0609] This shows that, assuming one and two error corrections, the
probability of not-corrected three errors is very low for
m.ltoreq.0.01, which is equivalent to n=10.sup.3, for
(BER)=10.sup.-5. Table 13 shows acceptable BER values for various
channel robustness levels, characterized by different
W.sub.n(3)-values, assuming n=1000.
TABLE-US-00012 TABLE 13 Acceptable (BER)-Values for Different
Channel Robustness Levels, Assuming n = 1000 W.sub.n(3) 1.67
10.sup.-4 1.67 10.sup.-7 1.67 10.sup.-10 m 0.1 0.01 0.001 (BER)
10.sup.-4 10.sup.-5 10.sup.-6
[0610] Condition D relates to the above-described Weather Anomalous
Event System (WAES) and detection by a Weather C2 Sensor of
anomalous events by Truthing-based Anomalous Event Software
Engineer (TAESE). It is also related to two reference papers,
related to C31 (Command, Control, Communication & Intelligence)
systems, presented and published by the Physical Optics
Corporation. These papers are related to a Digital Decision Support
(DDS), presented at SPIE Defense+Security+Sensing (DSS) Symposium;
Baltimore, Md., 6-8 May 2014, and entitled Ref. [1]: T. Jannson, T.
Forrester, A. Kostrzewski, W. Wang. "Bayesian Truthing and
Experimental Validation in Homeland Security and Defense," SPIE
Proc. Vol. 9074-21 (2014); and, Ref. [2] T. Jannson, W. Wang, T.
Forrester, A. Kostrzewski, C. Veeris, and T. Nielsen, "Decision
Generation Tools and Bayesian Inference," SPIE Proc. Vol. 9074-23
(2014).
[0611] Because a Weather Anomalous Event System (WAES) can be
configured to detect and identify weather anomalous events as
described above, and because the Weather Anomalous Event is
particular case of A Bayesian anomalous event thus, Bayesian
interference applies. In particular, for air turbulences and other
weather parameter fluctuations (wind, temperature, pressure,
humidity), leading to non-stationary random processes, a goal of
the IA system in various embodiments is to prevent IA-key data
transmission during non-stationary periods of time. This can result
in an increase in the IA of the system. Furthermore, applying
condition, A, B, C, D may allow the QA-channel criteria to be
preserved.
[0612] A further IA-preserving measure that may be implemented is
to repeat transmission of highly-secure data a number of times,
say, u-times. Then, the W.sub.n(m.sub.o) probability is further
reduced by q.sup.u-factor, which is an extremely small value.
[0613] Condition E is represented by Eq. (130), in which t.sub.R is
a threshold time defining a compromising IA-scenario. Thus, one
solution is to increase the length of the IA-key, n bits, to the
level at which the Eq. (130a) criterion is satisfied; when the IA
system can be broken only in a prohibitive amount of time, t,
larger than the t.sub.A threshold time. Fortunately, in the case of
weather station systems, the t.sub.A value is relatively short
(e.g., 2 days, for example). The solution provides optimization
between a BER value and an n value, defined by conditions A, B, C,
D. Assuming that these two values are (BER)=q, and n value,
equivalent to: m=qn, and fixed, and condition (130a) is not
satisfied (i.e., Eq. (130b) is satisfied); then, the error
correction (OVH) may be increased in order to correct more
errors--in other words, to increase the m.sub.o value to such a
level that the following condition is satisfied:
W n ( m o ) = e - m _ ( ( m _ ) m o m 0 ! ) = W n ( 0 ) ( m _ ) m o
m o ! .ltoreq. T o ( 163 ) ##EQU00056##
where T.sub.o is some threshold value. This is the inverse of the
procedure that can be done numerically by using a non-linear
look-up table. In good approximation, Eq. (163) is reduced to the
following inequality:
( m _ ) m o m o ! .ltoreq. T o ( 164 ) ##EQU00057##
which should be solved for unknown m.sub.o-value, assuming fixed
m-value. For example, for m=0.01, and T.sub.o=10.sup.-11, then
m.sub.o=5. In other words up to four errors per bit stream must be
corrected in order to satisfy inequality (164). This is shown in
Table 14, when various values of W.sub.n(m.sub.o) are tabulated as
a look-up table.
TABLE-US-00013 TABLE 14 Look-up Table for Various m.sub.o-Values,
Assuming m = 0.01 m.sub.o 3 4 5 ( m _ ) m o m o ! ##EQU00058## (
0.01 ) 3 3 ! ##EQU00059## ( 0.01 ) 4 4 ! ##EQU00060## ( 0.01 ) 5 5
! ##EQU00061## W.sub.n(m.sub.o) 1 6 10 - 6 ##EQU00062## 1 24 10 - 8
##EQU00063## 1 120 10 - 10 ##EQU00064##
[0614] According to this look-up table, for m=0.01, and threshold
value of 10.sup.-11, the m.sub.o-value should be, indeed, equal to
5, because, the following relation is satisfied:
1/2410.sup.-8<T.sub.o< 1/12010.sup.-10 (165)
[0615] RF channels are generally prone to statistical errors and
thus can be difficult to control. This can be an important
consideration in the case of IA-critical data streams. In
particular, the IA-key management against RF-errors can be
implemented in various embodiments as an important step for the
purpose of the IA-secure RF-channel, which may be a quasi-robust,
or QR-channel. As described herein, the QR-channel solution may be
based on a Poisson statistical distribution. The Conditions A, B,
C, D, E are necessary and sufficient for definition of the
QR-channel which is quasi-robust against the RF-errors.
[0616] The IA-key management solution is valid for at least the two
major application scenarios described immediately below.
[0617] The first major application scenario is increasing IA
protection. In this scenario, the n-bit length of the IA data
stream should be increased, which creates m mean error value
increasing. In order to mitigate the effect of increasing m value,
the system can be configured to increase m.sub.o number, where
(m.sub.o-1) is the number of errors per data stream to be protected
by an error correcting code. Conditions A, B, C, D, E, provide an
estimate of how much the m.sub.o value should increase. This
provides, of course, extra OVH cost, which can be predicted by this
method, and which is summarized by the following symbolic
relation:
IAnmm.sub.o (166)
where: "" is symbol of increase, and (m.sub.o-1) is number of
errors to be corrected by the FEC code, for example.
[0618] The second major application scenario is a worsening of
meteorological conditions. In such circumstances, the BER value
increases. Thus, the m value increases, which leads to an m.sub.o
value increase to compensate for the BER value increase; thus,
preserving the IA security level. Accordingly, the situation
concludes to the similar outcome as shown in Eq. (166), leading to
the following symbolic relation:
(BER)mm.sub.o (167)
[0619] In Table 15 Conditions A, B, C, D, E are summarized, as
necessary and sufficient conditions of the QR-channel.
[0620] Poission Formula Derivation.
[0621] In order to show that this derivation is different from a
standard one, the steps for for W.sub.n(3) may be repeated in the
form:
W n ( 3 ) = q 3 p n - 3 ( n 3 ) .apprxeq. q 3 p n n ! 3 ! ( n - 3 )
! = q 3 p n n ( n - 1 ) ( n - 2 ) 3 ! .apprxeq. .apprxeq. q 3 p n n
3 ( 1 3 ! ) = ( m _ ) 3 ( 1 - q ) n 1 3 ! = ( m _ ) 3 e - nq 1 3 !
= ( m _ ) 3 e - m _ 1 3 ! ( 168 ) ##EQU00065##
as it should be.
TABLE-US-00014 TABLE 15 Summary of Necessary and Sufficient
Conditions for QR-Channel Relevant No. Math Description Equation A
m << 1 Preliminary condition for statistical (141) error mean
value B m << n Preliminary condition for probabilities (148)
or errors C (m.sub.o - 1) Number of errors to be corrected (153) D
N/A Stationarity*) and ergodicity of (131) RF-channel statistics E
N/A Integrity**) of QR-channel, defined (130ab) by threshold
time***) *)Weather anomalous events do not satisfy this condition.
**)Further protection is by repetition u-times. ***)Time when the
code cannot be broken.
[0622] This document describes various embodiments that can be used
to improve time synchronization of the IA-keys. This novel
solution, which applies elements of the previous sections, can be
especially relevant for injection keys. In such a case, issues may
arise with regard to simultaneously using one, two or more keys for
unmanned or manned operations. In the case of manned operations,
the system can be configured to use one (or more) key to be
synchronized with time of human intervention. However, the
operation can be an unmanned operation as well. FIG. 57 is a
diagram illustrating an example application of an unmanned
operation using 2 IA keys with time synchronization 2700 in
accordance with one embodiment of the technology disclosed herein.
Particularly, in the example illustrated in FIG. 57, two injection
keys are needed to activate the system (which in this case is an
integrative meteorological system (IMS)). In some embodiments,
these keys are injected at the same time and are transmitted (e.g.,
via RF or other wireless communications) from their respective
locations P1, P2. The key locations P1, P2 can be different and
separate with respect to each other.
[0623] In the embodiment illustrated in FIG. 57, both key locations
P1 2703, P2 2704 transmit RF signals 2701, 2702. As noted, these
can be transmitted wirelessly such as, for example, via an RF
communication link. The distances of transmission from their
respective ones of P1 and P2 to IMS 2705 are illustrated as
.DELTA.l.sub.1 and .DELTA.l.sub.2, respectively, which are denoted
by the reference designations 2706 and 2707, respectively.
[0624] One issue that can arise for predictive analysis purposes is
that of identifying material sources of latency which are not
controllable in order to improve time synchronization. There are a
number of candidate sources of latency that can be considered.
[0625] Two candidate sources of latency include turbulence and
speckle. These two sources are typically not serious sources of
latency because, in the 1.sup.st case, they introduce the
non-stationarity of stochastic process; and are thus ruled out by
Condition D, as in Table 15.
[0626] Another issue for consideration is multipath fading.
Multipath fading introduces a reduction in communication bandwidth
rather than added latency. As long as reduced bandwidth is not too
small (i.e., in 10 kbps range), it can be considered as secondary,
rather than primary effect.
[0627] Second order and higher order dispersion can also be
considered as a factor that reduces the effective bandwidth.
Generally, for all these effects, general properties of wave motion
can be applied. Accordingly, in a qualitative sense, all types of
waves (e.g., electromagnetic (including optical) and acoustic) can
be considered.
[0628] The effects of linear dispersion have been discussed by one
of the inventors of the disclosed technology in the paper by T.
Jannson, and J. Jannson, "Temporal Self-Imaging Effect in
Single-Mode Fibers," J. Opt. Soc. Am., 71 no. 11, pp. 1373-1376,
November 1981; and T. Jannson, "Real-Time Fourier Transformation in
Dispersive Optical Fibers," Opt. Lett., 8, No. 4, pp. 232-234,
April 1983.
[0629] The linear dispersion is characterized by the time delay,
At, defined by group velocity, V.sub.g, defined as:
v.sub.g=1/{dot over (.beta.)}.sub.o (169)
where, .beta. is wavenumber, in the form:
.beta. = .omega. c n ( 170 ) ##EQU00066##
and where .omega. is angular frequency (.omega.=2 .pi.f, where
f-frequency in Hz), c is the speed of light in a vacuum, n is the
refractive index equal to {square root over (.epsilon.)}, where
.epsilon. is relative dielectric constant, and .omega..sub.o is the
carrier angular frequency, while {dot over (.beta.)}.sub.o is the
short form of the 1.sup.st differential in respect to .omega..sub.o
in the form:
.beta. . o = d .beta. d .omega. / .omega. = .omega. o ; .beta. o =
.beta. ( .omega. o ) ( 171 ab ) ##EQU00067##
[0630] Using Eq. (170) and notation as in Eq. (171), {dot over
(.beta.)}.sub.o can be written as
.beta. . o = n o c + .omega. c n . o ( 172 ) ##EQU00068##
[0631] Thus, knowing dispersion relation:
n=n(.omega.) (173)
allows computation of {dot over (.beta.)}.sub.o and, then, the
group velocity, v.sub.g.
[0632] FIG. 58 is a diagram illustrating an example of a typical
relation (173) for normal dispersion.
[0633] Using Eqs. (169) and (172):
1 v g = 1 v p h + .omega. c n . o ; v p h = c n o ( 174 ab )
##EQU00069##
where both terms on the right are positive and v.sub.ph is phase
velocity; thus, v.sub.g<v.sub.ph.
[0634] According to Eq. (174) (described above), and Eqs.
(169-174), the time delays for RF-signals, are
.DELTA. t 1 = .DELTA. 1 v g 1 ; .DELTA. t 2 = .DELTA. 2 v g 2 ( 175
ab ) ##EQU00070##
where .DELTA..sub.l and .DELTA..sub.2 can be determined precisely
using GPS system. However, group velocities can vary. Assuming
.DELTA..sub.1 and .DELTA..sub.2 values are fixed, a general time
delay, .DELTA.t, can be described by:
.DELTA. t = .DELTA. v g = Constant v g ( 176 ) ##EQU00071##
then, the relative .DELTA.t-change, .delta.(.DELTA.t), is
.delta. ( .DELTA. t ) .DELTA. t = .delta. ( v g ) v g ( 177 )
##EQU00072##
[0635] Accordingly, if the velocity, v.sub.g, changes by 1%, for
example, then, the time delay also changes by 1%. However, knowing
the dispersion relation (173) allows computation of .DELTA.t for
predictive analysis purposes, which can eliminate this linear
dispersion effect.
[0636] Another element of the example embodiment relates to
injection key management for purposes of time synchronization. With
reference again to FIG. 57, where .DELTA..sub.1 and .DELTA..sub.2
are known and the group velocities, v.sub.g1 and v.sub.g2, are also
known, both time delays, .DELTA.t.sub.1 and .DELTA.t.sub.2 can be
computed. Then, the difference between those delays can be shown
as:
.tau..sub.12=.DELTA.t.sub.2-.DELTA.t.sub.1 (178)
where: .DELTA.t.sub.2>.DELTA.t.sub.1
[0637] For discussion purposes, RF signals 2701, 2702 can be
referred to as RF.sub.1 and RF.sub.2, respectively. With this
nomenclature in mind, and given the difference in time delays, the
system can be configured to send the RF.sub.2 signal earlier than
the RF.sub.1 signal by an amount of time .tau..sub.12. The time
accuracy, .delta.t, may be defined by line width, .DELTA..omega.,
or .DELTA.f, where. .DELTA..omega.=2.pi..DELTA.f. According to the
Heisenberg uncertainty relation, (.DELTA.f) (.delta.t).about.1, and
therefore:
( .delta. t ) = 1 .DELTA. f ( 179 ) ##EQU00073##
[0638] Therefore, both signals RF.sub.1 and RF.sub.2 should be
received by the IMS at the same time or substantially the same
time, with .delta.t-time accuracy. For example, for .DELTA.f=10
kHz, .delta.t=0.1 msec.
[0639] Embodiments of the technology disclosed herein can also be
implemented. This section introduces a new method for improving
time synchronization of IA keys, including the injection keys. In
general, time synchronization of critical signals, such as those
related to the IA keys, can be instrumental to IA key management
and beneficial to improving security.
[0640] FIG. 59 is a diagram illustrating an example of dotless time
synchronization in accordance with one embodiment of the technology
disclosed herein. Particularly, an example geometry of time
synchronization, or time sync, is shown in FIG. 59 for a scenario
in which the integrative metrological system (IMS) is receiving two
synchronization signals, one each from sources P.sub.1 and
P.sub.2.
[0641] Embodiments of the disclosed technology can be implemented
to utilize Sync Time Counting (STC). In other embodiments, Sync
Time Counting can be denoted by (t-t.sub.o) and (t+t.sub.o), where
(t-t.sub.o) means, t-3 sec, t-2 sec, t-1 sec, before the zero time
t-0; and where (t+t.sub.o) means, t+1 sec, t+2 sec, t+3 sec, after
the zero time t-0. The zero time, t-0, may be defined with
.delta.t-accuracy, e.g., as defined by Eq. (179). We assume that
all time delays are controlled within this accuracy including
meteorological effects, network latency and other signal processing
latencies.
[0642] In various embodiments, the system is configured such that
it can identify the zero time, t-0, in a precise way. In one
embodiment, this is performed using what is referred to herein as
the dotless method. The essence of one or more applications of this
embodiment is to apply irrational numbers, such as .pi., e, {square
root over (2)}, etc., without dots. This can be done because, in
various embodiments the irrational numbers are not relied on for
algebraic operations, but are instead used for identification (ID),
or authentication purposes. For example, instead of using an
e-number, which has a natural logarithmic base in the form:
e=2.718281828 . . . , Embodiments can be configured to apply a
dotless e-number in the form e.=2718281828 . . . , where e. is a
dotless e-number. Similarly, the dotless representation of .pi. is
.pi..=3141592654 . . . , and the dotless representation of {square
root over (2)} is: {square root over (2)}.=1414213562 . . . . In
further embodiments, an arbitrary number of digits can be taken
into account (here, this number is 10). As these examples serve to
illustrate, the dotless method can be implemented using dotless
irrational number for IA-purposes. Advantageously, the number of
irrational numbers, or "irrationals," is infinite, and each
irrational is uniquely defined in any modulo-algebra, such as by
modulo-10 algebra, as illustrated in the examples immediately
above. Also, the system can be implemented to avoid floating point
operations after the dot by removing dots and by reorganizing the
irrational number structure in a specific way, depending on number
of digits, n.sub.o, to be taken into account. For these purposes,
the following notation can be adopted:
.pi. n.sub.m (180)
where, ".pi." is the symbol of irrational number, " " is the symbol
of "dotless" operation, and "n.sub.m" is the number of digits to be
taken in modulo-m algebra. Thus, for modulo-10 algebra, modulo-2
(binary) algebra, and modulo-7 algebra, relation (180) becomes,
respectively,
.pi. n.sub.10; .pi. n.sub.2; .pi. n.sub.7 (181abc)
[0643] Assuming, that only four (4) digits are taken, for
simplicity, relations (181abc) become,
.pi. 4.sub.10; .pi. 4.sub.2; .pi. 4.sub.7 (182abc)
[0644] As an example, consider {square root over (2)}, in the
following decimal form:
{square root over
(2)}4.sub.10=1414=110.sup.3+410.sup.2+110.sup.1+410.sup.0 (183)
[0645] This number in binary modulo-2 algebra, is
{square root over
(2.)}4.sub.2=12.sup.10+02.sup.9+12.sup.8+12.sup.7+02.sup.6+02.sup.5++02.s-
up.4+02.sup.3+12.sup.2+12.sup.1+02.sup.0=10110000110 (184)
[0646] For verification, consider
1414=1024+256+128+4+2=1414 (185)
[0647] For (more exotic) modulo-7 base (or, septimal base):
{square root over
(2)}.4.sub.7=4060=47.sup.3+07.sup.2+67.sup.1+07.sup.0 (186)
[0648] For verification:
1414=47.sup.3+07.sup.2+67.sup.1+07.sup.0==(4)(343)+0+(6)(7)+0=1372+42=14-
14 (187)
[0649] According to Eqs. (180-187), the Dotless Method may be
considered as applying irrational numbers, or applying irrationals
in different modulo algebras for cyber identification and/or
authentication.
[0650] For camouflage purposes modulo algebra may be applied even
higher than 10, purposely for creating ambiguity. This can be
deciphered, however, by knowing the symbol .pi. n.sub.m. For
example, it is known that,
1414=111.sup.3+011.sup.2+711.sup.1+611.sup.0==1313+0+77+6=1414
(188)
[0651] Therefore, for embodiments implementing according to these
teachings, the following relation is not ambiguous:
{square root over (2)}.4.sub.11=1076 (189)
[0652] However, for somebody who sees the number "1076" only, Eq.
(189) is highly ambiguous, since, the number of possibilities and
irrationals is infinite.
[0653] Time synchronization, or TimeSync, for short, can be
described as using 2 or more wireless signals from one, two, or
more sources, in order to confirm identification or authentication
of a friendly party. This friendly party can include, for example,
a person or people, a machine, or combination of them.
[0654] A process for verifying the identification or authentication
of a friendly party in accordance with one embodiment of the
technology disclosed herein is now described with reference to FIG.
57. Table 16 illustrates a summary of the dotless timesync
steps.
[0655] In a first step, the IMS 2705 sends the timesync signatures
(TSS) in the form of symbols .pi. n.sub.m, which can be the same or
different for each source via RF signals 2701, 2702 with ciphertext
data modulated or embedded thereon (i.e. Encrypted text) using
symmetric or non-symmetric encryption key to sources 2703,
2704.
[0656] In a second step, based on the TSS, the sources 2703, 2704
send binary data streams with TSS in plaintext (i.e., non-encrypted
text). These data streams can be sent within timesync, in such a
way that the TSS should come at the same zero time. The TSS
zero-time may be defined by the last digit of the .pi.
n.sub.msignature, for example.
[0657] In a third step the external verification of the TimeSync is
done by verifying whether all TSS signatures come in zero time
within a pre-described accuracy.
TABLE-US-00015 TABLE 16 Summary of Dotless Timesync Steps No.
Description of Steps Equation 1 Sending RF signals, with timesync
signatures, by (178) integrative meterological station (Rx) to
sources (Tx), in ciphertext, or plain text 2 Re-sending timesync
signatures by source (Tx) into t - t.sub.0, t - 0 integrative
meterological station (Rx), in plain text, in proper times; t -
t.sub.0, in order to obtain them in zero-time: t - 0 3 Experimental
verification 4 If experimental verification is positive, then the
relevant cyber-operation starts
[0658] In a fourth step, if the experimental verification is
positive then, the relevant operation (such as some IA-operation)
starts; otherwise, it does not. In another embodiment of step 1,
the TSS is known a priori, by sources 2703 and 2704.
[0659] FIG. 59 is a diagram illustrating an example of dotless time
synchronization 2800 in accordance with one embodiment of the
technology disclosed herein. In this example, there are two RF
sources P.sub.1, P.sub.2, denoted as 2801, 2802 and an integrative
metrological station (IMS) 2803. RF sources P.sub.1, P.sub.2, are
configured to send TimeSync signatures (TSS) 2804, 2805 to IMS
2803. These TSS signals 2804, 2805 are received by IMS 2803 at zero
time t-0. The zero time 2806, 2807 is defined by the last digits of
the respective TSS 2804, 2805. The TSS 2804, 2805 symbols may in
some embodiments be different symbols and have different lengths as
illustrated in the example of FIG. 59.
[0660] P.sub.1 2801, P.sub.2 2802 and IMS 2803 have this a priori
information. From the perspective of general IA system knowledge,
the system can determine that the TSS 2804 and 2805 are dotless
irregulars, in the general form of .pi. 7m and .pi. 6m,
respectively, but the .pi.-property (which irregular?) and
m-property (which modulo algebra?) is unknown to an adverse party.
By positively verifying a priori a known zero-time, the system can
confirm this Dotless TimeSync realization 2800. Then, the IMS
IA-system can start. Otherwise, in various embodiments it cannot.
Thus, in order to start the IA-system, in some embodiments two
things happen at once. First, the data streams 2804 and 2805 must
be identified by the IMS 2803 as the TSS, which are a priori known
to their respective sources 2801, 2802 and to the IMS 2803. Second,
their last digits 2806 and 2807 should come at the same zero-time,
denoted as 2808 for 2804 and 2809 for 2805 (the difference in
notations: 2808 and 2809, instead of single one, is, because, there
is uncertainty, .delta.t of the zero-time).
[0661] The question as to whether this realization is true or
false, which in various embodiments can be considered as a
statistical Bayesian question, is addressed below.
[0662] This document now shows that the Dotless TimeSync Operation
can be treated as C2, or rather as a C3I Binary Sensor. The
Bayesian Binary Sensor concept has been discussed by T. Jannson,
et. al., "Bayesian Truthing and Experimental Validation in Homeland
Security and Defense," SPIE Proc. Vol. 9074-21 (2014), a C3I paper,
presented in SPIE Defense+Security+Sensing (DSS) Symp., Baltimore,
Md., 6-8 May 2014, where C3I means: Command, Control, Communication
and Intelligence, while C2 means: Command and Control. In Bayesian
Binary Sensor theory, the Figures of Merit (FoMs) are PPV (Positive
Predictive Value) and NPV (Negative Predictive Value), the latter
FoM defined, as
(NPV)=p(N|N') (190)
where, the NPV is the inverse conditional (Bayesian) probability,
that, under no-alarm, N', the event is not anomalous, N.
[0663] In the description above of a Truthing-based Anomalous Event
Software Engine (TAESE) embodiments provide Bayesian Binary
Cybersensing for detection and identification (ID) of weather
anomalous events, and the weather station is presented as an
exemplary C2 Weather Sensor System (C2WS2). This document now
describes embodiments for the protection of IA-keys, including
encryption keys, injection keys, and others. In particular, the
Dotless TimeSync operation, summarized in Table 16, above,
discusses four (4) operation steps as an example of this operation,
to be sure that zero-time, t-0, is, indeed, the moment in time at
which two or more TimeSync signatures (TSS) represented by dotless
irrationals are received. A question may arise: if the (t-0) time
moment is actually the zero-time moment, or whether this dotless
timesync zero-time event is true or false.
[0664] In order to answer this question, it is important to note
that this directly relates to the existence of a quasi-robust RF
channel or QR-channel, as defined by criteria A, B, C, D, E
described above, and summarized in Table 15. In particular,
according to Condition D, the stationarily and ergodicity of the
QR-channel is the necessary condition of the QR-channel. Not
satisfying this condition creates the possibility of unwanted burst
errors that preclude using RF-channel statistics, based on the
Poission distribution, as defined by Eq. (163), in which
(m.sub.o-1) is the number of errors to be corrected as determined
by Condition C, Table 15 (for sake of clarity, we assume m.sub.o=3;
thus (m.sub.o-1)=2, for example.)
[0665] Therefore, for the purpose of a dotless TimeSync operation,
anomalous event, or, rather, weather anomalous event may be defined
as the existence of burst RF-errors precluding existence of the
QR-channel, thus, making Dotless TimeSync operation and related
zero-time event, false.
[0666] In order to make the Dotless TimeSync operation effective,
it is useful to prove that the probability of a zero-time event is
high, i.e., close to unity. The probability of detection and
identification of a weather anomalous event is defined by the NPV,
as in Eq. (190), while, probability of zero, one, two, . . . ,
(m.sub.o-1) error, is
P ( m o , n ) = m = 0 ( m 0 - 1 ) W n ( m ) ( 191 )
##EQU00074##
[0667] For example, for m.sub.o=3, Eq. (62) becomes
P ( 3 , n ) = m = 0 2 W n ( m ) = W n ( 0 ) + W n ( 1 ) + W n ( 2 )
( 192 ) ##EQU00075##
[0668] Using Eq. (192), for small statistical mean values,
m=nq:
P ( 3 , n ) = W n ( 0 ) + m _ W n ( 0 ) + ( m _ ) 2 2 W n ( 0 ) = W
n ( 0 ) ( 1 + m _ + m _ 2 2 ) = e - m _ ( 1 + m _ + m _ 2 2 )
.apprxeq. .apprxeq. ( 1 - m _ ) ( 1 + m _ + m _ 2 2 ) = 1 - m _ 2 2
- m _ 3 2 .apprxeq. 1 - m _ 2 2 ( 193 ) ##EQU00076##
[0669] In order to make the Dotless TimeSync zero-time event true,
there should be no burst errors, and, at the same time, all
relevant RF errors should be corrected by error-correcting codes,
such as, for example, FEC-codes. The 1.sup.st condition is defined
by the NPV, as in Eq. (190), while the 2.sup.nd one is defined by
the probability P(m.sub.o,n), in general, and by the probability
P(3,n) in particular, for m.sub.o=3. Thus, for m.sub.o=3, the
probability, P.sub.C3I, that the Dotless TimeSync zero-time event,
t-0, is true, is the product of the (NPV) and P.sub.o (3,n), in the
form:
P C 31 = ( NPV ) P ( 3 , n ) .apprxeq. ( NPV ) ( 1 - m _ 2 2 ) (
194 ) ##EQU00077##
[0670] For a Truthing-based Anomalous Event Software Engine, and to
make the Bayesian Binary Cybersensor effective, the Negative
Predictive Value must be close to unity. Accordingly,
(NPV)=1-a; a<<1 (195ab)
and thus, Eq. (194) becomes,
P C 31 = ( 1 - a ) ( 1 - m _ 2 2 ) ( 196 ) ##EQU00078##
[0671] For example, assuming that,
a=m=0.01 (197)
which is a conservative assumption, then
P.sub.C3I=(1-0.01)(1-0.00005).apprxeq.0.99 (198)
[0672] Thus, the 1.sup.st factor dominates, leading to the
probability formula in the form:
P.sub.C3I.apprxeq.(NPV) (199)
[0673] Typical (NPV)-values for Bayesian Inference are extremely
close to unity; much closer, in fact, than the PPV values (PPV:
Positive Predictive Value). A typical value, for example, can be
0.9999, or even closer to 1. In such a case:
P.sub.C3I>0.9999=99.99% (200)
Thus, the probability that the Dotless TimeSync zero-time event is
true is almost a certainty, making the Dotless TimeSync approach
very effective, indeed.
[0674] The term tool can be used to refer to any apparatus
configured to perform a recited function. For example, tools can
include a collection of one or more modules and can also be
comprised of hardware, software or a combination thereof. Thus, for
example, a tool can be a collection of one or more software
modules, hardware modules, software/hardware modules or any
combination or permutation thereof. As another example, a tool can
be a computing device or other appliance on which software runs or
in which hardware is implemented.
[0675] As used herein, the term module might describe a given unit
of functionality that can be performed in accordance with one or
more embodiments of the technology disclosed herein. As used
herein, a module might be implemented utilizing any form of
hardware, software, or a combination thereof. For example, one or
more processors, controllers, ASICs, PLAs, PALs, CPLDs, FPGAs,
logical components, software routines or other mechanisms might be
implemented to make up a module. In implementation, the various
modules described herein might be implemented as discrete modules
or the functions and features described can be shared in part or in
total among one or more modules. In other words, as would be
apparent to one of ordinary skill in the art after reading this
description, the various features and functionality described
herein may be implemented in any given application and can be
implemented in one or more separate or shared modules in various
combinations and permutations. Even though various features or
elements of functionality may be individually described or claimed
as separate modules, one of ordinary skill in the art will
understand that these features and functionality can be shared
among one or more common software and hardware elements, and such
description shall not require or imply that separate hardware or
software components are used to implement such features or
functionality.
[0676] Where components or modules of the technology are
implemented in whole or in part using software, in one embodiment,
these software elements can be implemented to operate with a
computing or processing module capable of carrying out the
functionality described with respect thereto. One such example
computing module is shown in FIG. 60. Various embodiments are
described in terms of this example-computing module 3000. After
reading this description, it will become apparent to a person
skilled in the relevant art how to implement the technology using
other computing modules or architectures.
[0677] Referring now to FIG. 60, computing module 3000 may
represent, for example, computing or processing capabilities found
within desktop, laptop and notebook computers; hand-held computing
devices (PDA's, smart phones, cell phones, palmtops, etc.);
mainframes, supercomputers, workstations or servers; or any other
type of special-purpose or general-purpose computing devices as may
be desirable or appropriate for a given application or environment.
Computing module 3000 might also represent computing capabilities
embedded within or otherwise available to a given device. For
example, a computing module might be found in other electronic
devices such as, for example, digital cameras, navigation systems,
cellular telephones, portable computing devices, modems, routers,
WAPs, terminals and other electronic devices that might include
some form of processing capability.
[0678] Computing module 3000 might include, for example, one or
more processors, controllers, control modules, or other processing
devices, such as a processor 3004. Processor 3004 might be
implemented using a general-purpose or special-purpose processing
engine such as, for example, a microprocessor, controller, or other
control logic. In the illustrated example, processor 3004 is
connected to a bus 3002, although any communication medium can be
used to facilitate interaction with other components of computing
module 3000 or to communicate externally.
[0679] Computing module 3000 might also include one or more memory
modules, simply referred to herein as main memory 3008. For
example, preferably random access memory (RAM) or other dynamic
memory, might be used for storing information and instructions to
be executed by processor 3004. Main memory 3008 might also be used
for storing temporary variables or other intermediate information
during execution of instructions to be executed by processor 3004.
Computing module 3000 might likewise include a read only memory
("ROM") or other static storage device coupled to bus 3002 for
storing static information and instructions for processor 3004.
[0680] The computing module 3000 might also include one or more
various forms of information storage mechanism 3010, which might
include, for example, a media drive 3012 and a storage unit
interface 3020. The media drive 3012 might include a drive or other
mechanism to support fixed or removable storage media 3014. For
example, a hard disk drive, a floppy disk drive, a magnetic tape
drive, an optical disk drive, a CD or DVD drive (R or RW), or other
removable or fixed media drive might be provided. Accordingly,
storage media 3014 might include, for example, a hard disk, a
floppy disk, magnetic tape, cartridge, optical disk, a CD or DVD,
or other fixed or removable medium that is read by, written to or
accessed by media drive 3012. As these examples illustrate, the
storage media 3014 can include a computer usable storage medium
having stored therein computer software or data.
[0681] In alternative embodiments, information storage mechanism
3010 might include other similar instrumentalities for allowing
computer programs or other instructions or data to be loaded into
computing module 3000. Such instrumentalities might include, for
example, a fixed or removable storage unit 3022 and an interface
3020. Examples of such storage units 3022 and interfaces 3020 can
include a program cartridge and cartridge interface, a removable
memory (for example, a flash memory or other removable memory
module) and memory slot, a PCMCIA slot and card, and other fixed or
removable storage units 3022 and interfaces 3020 that allow
software and data to be transferred from the storage unit 3022 to
computing module 3000.
[0682] Computing module 3000 might also include a communications
interface 3024. Communications interface 3024 might be used to
allow software and data to be transferred between computing module
3000 and external devices. Examples of communications interface
3024 might include a modem or softmodem, a network interface (such
as an Ethernet, network interface card, WiMedia, IEEE 802.XX or
other interface), a communications port (such as for example, a USB
port, IR port, RS232 port Bluetooth.RTM. interface, or other port),
or other communications interface. Software and data transferred
via communications interface 3024 might typically be carried on
signals, which can be electronic, electromagnetic (which includes
optical) or other signals capable of being exchanged by a given
communications interface 3024. These signals might be provided to
communications interface 3024 via a channel 3028. This channel 3028
might carry signals and might be implemented using a wired or
wireless communication medium. Some examples of a channel might
include a phone line, a cellular link, an RF link, an optical link,
a network interface, a local or wide area network, and other wired
or wireless communications channels.
[0683] In this document, the terms "computer program medium" and
"computer usable medium" are used to generally refer to media such
as, for example, memory 3008, storage unit 3020, media 3014, and
channel 3028. These and other various forms of computer program
media or computer usable media may be involved in carrying one or
more sequences of one or more instructions to a processing device
for execution. Such instructions embodied on the medium, are
generally referred to as "computer program code" or a "computer
program product" (which may be grouped in the form of computer
programs or other groupings). When executed, such instructions
might enable the computing module 3000 to perform features or
functions of the disclosed technology as discussed herein.
[0684] While various embodiments of the disclosed technology have
been described above, it should be understood that they have been
presented by way of example only, and not of limitation. Likewise,
the various diagrams may depict an example architectural or other
configuration for the disclosed technology, which is done to aid in
understanding the features and functionality that can be included
in the disclosed technology. The disclosed technology is not
restricted to the illustrated example architectures or
configurations, but the desired features can be implemented using a
variety of alternative architectures and configurations. Indeed, it
will be apparent to one of skill in the art how alternative
functional, logical or physical partitioning and configurations can
be implemented to implement the desired features of the technology
disclosed herein. Also, a multitude of different constituent module
names other than those depicted herein can be applied to the
various partitions. Additionally, with regard to flow diagrams,
operational descriptions and method claims, the order in which the
steps are presented herein shall not mandate that various
embodiments be implemented to perform the recited functionality in
the same order unless the context dictates otherwise.
[0685] Although the disclosed technology is described above in
terms of various exemplary embodiments and implementations, it
should be understood that the various features, aspects and
functionality described in one or more of the individual
embodiments are not limited in their applicability to the
particular embodiment with which they are described, but instead
can be applied, alone or in various combinations, to one or more of
the other embodiments of the disclosed technology, whether or not
such embodiments are described and whether or not such features are
presented as being a part of a described embodiment. Thus, the
breadth and scope of the technology disclosed herein should not be
limited by any of the above-described exemplary embodiments.
[0686] Terms and phrases used in this document, and variations
thereof, unless otherwise expressly stated, should be construed as
open ended as opposed to limiting. As examples of the foregoing:
the term "including" should be read as meaning "including, without
limitation" or the like; the term "example" is used to provide
exemplary instances of the item in discussion, not an exhaustive or
limiting list thereof; the terms "a" or "an" should be read as
meaning "at least one," "one or more" or the like; and adjectives
such as "conventional," "traditional," "normal," "standard,"
"known" and terms of similar meaning should not be construed as
limiting the item described to a given time period or to an item
available as of a given time, but instead should be read to
encompass conventional, traditional, normal, or standard
technologies that may be available or known now or at any time in
the future. Likewise, where this document refers to technologies
that would be apparent or known to one of ordinary skill in the
art, such technologies encompass those apparent or known to the
skilled artisan now or at any time in the future.
[0687] The presence of broadening words and phrases such as "one or
more," "at least," "but not limited to" or other like phrases in
some instances shall not be read to mean that the narrower case is
intended or required in instances where such broadening phrases may
be absent. The use of the term "module" does not imply that the
components or functionality described or claimed as part of the
module are all configured in a common package. Indeed, any or all
of the various components of a module, whether control logic or
other components, can be combined in a single package or separately
maintained and can further be distributed in multiple groupings or
packages or across multiple locations.
[0688] Additionally, the various embodiments set forth herein are
described in terms of exemplary block diagrams, flow charts and
other illustrations. As will become apparent to one of ordinary
skill in the art after reading this document, the illustrated
embodiments and their various alternatives can be implemented
without confinement to the illustrated examples. For example, block
diagrams and their accompanying description should not be construed
as mandating a particular architecture or configuration.
* * * * *