U.S. patent application number 13/330198 was filed with the patent office on 2012-04-12 for communicating over a wireless network.
Invention is credited to Sheng Liu, Sokwoo Rhee.
Application Number | 20120087290 13/330198 |
Document ID | / |
Family ID | 37482254 |
Filed Date | 2012-04-12 |
United States Patent
Application |
20120087290 |
Kind Code |
A1 |
Rhee; Sokwoo ; et
al. |
April 12, 2012 |
Communicating over a Wireless Network
Abstract
A first device communicates with a wireless network that
includes nodes that are active for predefined activation times and
that are at least partially dormant when not active. The
communication method includes identifying a start of communication
via a message that exceeds a maximum activation time of nodes on
the wireless network by a factor N, where N is equal to at least a
maximum number of frequencies on the wireless network, and
exchanging information with a second device comprising a node on
the wireless network that is within a transmission range of the
first device, where the information is exchanged following the
message and includes frequency hopping data for the second
device.
Inventors: |
Rhee; Sokwoo; (Lexington,
MA) ; Liu; Sheng; (Cambridge, MA) |
Family ID: |
37482254 |
Appl. No.: |
13/330198 |
Filed: |
December 19, 2011 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12909286 |
Oct 21, 2010 |
|
|
|
13330198 |
|
|
|
|
11443869 |
May 31, 2006 |
7844308 |
|
|
12909286 |
|
|
|
|
60686127 |
Jun 1, 2005 |
|
|
|
60724343 |
Oct 6, 2005 |
|
|
|
Current U.S.
Class: |
370/311 |
Current CPC
Class: |
H04W 52/0216 20130101;
H04W 84/18 20130101; Y02D 30/70 20200801; Y02D 70/22 20180101; H04B
1/7156 20130101; H04W 52/0229 20130101; Y02D 70/40 20180101; Y02D
70/142 20180101; H04W 52/0219 20130101; H04B 1/713 20130101; Y02D
70/144 20180101 |
Class at
Publication: |
370/311 |
International
Class: |
H04W 52/02 20090101
H04W052/02; G08C 17/00 20060101 G08C017/00 |
Claims
1. A method for low power listening on a node, the method
comprising: waking up and listening for a first signal associated
with incoming traffic for a first amount of time; sleeping for a
second amount of time if no signal is sensed in the incoming
traffic; waking up and listening for a second signal associated
with the incoming traffic after the second amount of time;
detecting a preamble in the second signal; capturing a third amount
of time from the preamble, the third time indicative of how much
time before a start of a preamble tail end; resuming normal
operations until an end of the third amount of time; and waking up
and listening for a packet associated with the incoming traffic
after the third amount of time.
2. The method of claim 1, wherein the received packet includes the
preamble tail end.
3. The method of claim 1, wherein the received packet includes a
second packet and the second packet is not the preamble tail
end.
4. The method of claim 1, wherein after waking up and listening for
a packet associated with the incoming traffic after the third
amount of time, the method further comprising: waking up and
listening for a third signal associated with the incoming traffic
for the first amount of time; and sleeping for the second amount of
time if no signal is sensed in the incoming traffic.
5. The method of claim 1, wherein waking up and listening for a
first signal associated with incoming traffic for a first amount of
time comprising waking up and listening for a first signal
associated with incoming traffic for a first amount of time in a
first channel, and wherein waking up and listening for a second
signal associated with the incoming traffic after the second amount
of time further comprising waking up and listening for a second
signal associated with the incoming traffic after the second amount
of time in a second channel.
6. The method of claim 1, wherein waking up and listening for a
first signal associated with incoming traffic for a first amount of
time comprising waking up and listening for a first signal
associated with incoming traffic for a first amount of time in a
first channel, and wherein waking up and listening for a second
signal associated with the incoming traffic after the second amount
of time further comprising waking up and listening for a second
signal associated with the incoming traffic after the second amount
of time in the first channel.
7. The method of claim 1, wherein resuming normal operations until
an end of the third amount of time comprising at least one of:
sleeping for part or all of the third amount of time; sending
packets for part or all of the third amount of time; or listening
for a fourth signal associated with the incoming traffic for part
or all of the third amount of time.
8. The method of claim 1, wherein resuming normal operations until
an end of the third amount of time comprising: determining a
channel number for transmission to a second node, a time for
transmission to the second node, or any combination thereof; and
sending a second packet to the second node based on the channel
number for transmission to the second node, the time for
transmission to the second node, or any combination thereof.
9. The method of claim 1, further comprising: determining a channel
number for transmission to a second node, a time for transmission
to the second node, or any combination thereof based on the packet
associated with the incoming traffic; and sending a second packet
to the second node based on the channel number for transmission to
the second node, the time for transmission to the second node, or
any combination thereof.
10. The method of claim 1, further comprising sending a response to
the received packet.
11. The method of claim 10, further comprising: receiving, by the
second node and from the node, the response to the received packet;
determining, by the second node, a channel number for transmission
to the node, a time for transmission to the node, or any
combination thereof based on the response to the received packet;
and sending, by the second node, a third packet to the node based
on the channel number for transmission to the node, the time for
transmission to the node, or any combination thereof.
12. The method of claim 1, wherein the preamble comprising time
data corresponding to a start of the preamble tail end, a duration
of the preamble, or any combination thereof.
13. The method of claim 1, wherein detecting a preamble in the
second signal comprising: detecting an energy signal associated
with the preamble; and determining if the energy signal is above a
signal level.
14. The method of claim 1, further comprising sending a second
packet after the third amount of time.
15. The method of claim 1, wherein the preamble is transmitted from
a second node and the method further comprising sending a second
packet to the second node after the third amount of time.
16. A node comprising: memory configured to store executable
instructions; and at least one processor configured to execute the
instructions to: wake up and listen for a first signal associated
with incoming traffic for a first amount of time; sleep for a
second amount of time if no signal is sensed in the incoming
traffic; wake up and listen for a second signal associated with the
incoming traffic after the second amount of time; detect a preamble
in the second signal; capture a third amount of time from the
preamble, the third time indicative of how much time before a start
of a preamble tail end; resume normal operations until an end of
the third amount of time; and wake up and listen for a packet
associated with the incoming traffic after the third amount of
time.
17. A method for low power listening on a node, the method
comprising: entering a first low power mode for a first amount of
time, the first low power mode enabling the node to use less power
than needed to receive a packet; entering a first high power mode
after the first amount of time, the first high power mode enabling
the node to receive a first packet and the first packet comprising
a preamble; identifying a second amount of time from the preamble,
the second amount of time is an amount of time until a second
packet; entering a second low power mode until an end of the second
amount of time, the second low power mode enabling the node to use
less power than needed to receive a packet; and entering a second
high power mode after the second amount of time, the second high
power mode enabling the node to receive a packet.
18. The method of claim 17, further comprising determining whether
the first packet comprising the preamble by checking information in
the preamble.
19. The method of claim 17, wherein the first high power mode is in
a first channel and the second high power mode is in a second
channel.
20. The method of claim 17, further comprising: sending, by a
second node, the second packet to the node during the second high
power mode; receiving, by the second node and from the node, a
response to the second packet; determining, by the second node, a
channel number for transmission to the node, a time for
transmission to the node, or any combination thereof based on the
response to the second packet; and sending, by the second node, a
third packet to the node based on the channel number for
transmission to the node, the time for transmission to the node, or
any combination thereof.
21. The method of claim 17, wherein the preamble comprising time
data corresponding to a start of the preamble tail end, a duration
of the preamble, or any combination thereof.
22. The method of claim 17, wherein entering a first high power
mode after the first amount of time comprising: detecting an energy
signal associated with the preamble; and determining if the energy
signal is above a signal level.
23. The method of claim 17, further comprising: receiving the
second packet during the second high power mode; determining a
channel number for transmission to a second node, a time for
transmission to the second node, or any combination thereof based
on the received second packet; and sending a third packet to the
second node based on the channel number for transmission to the
second node, the time for transmission to the second node, or any
combination thereof.
24. A node comprising: memory configured to store executable
instructions; and at least one processor configured to execute the
instructions to: enter a first low power mode for a first amount of
time, the first low power mode enabling the node to use less power
than needed to receive a packet; enter a first high power mode
after the first amount of time, the first high power mode enabling
the node to receive a first packet and the first packet comprising
a preamble; identify a second amount of time from the preamble, the
second amount of time is an amount of time until a second packet;
enter a second low power mode until an end of the second amount of
time, the second low power mode enabling the node to use less power
than needed to receive a packet; and enter a second high power mode
after the second amount of time, the second high power mode
enabling the node to receive a packet.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. Provisional
Application No. 60/686,127, filed Jun. 1, 2005, and of U.S.
Provisional Application No. 60/724,343, filed Oct. 6, 2005. The
contents of U.S. Provisional Application No. 60/686,127 and of U.S.
Provisional Application No. 60/724,343 are incorporated herein by
reference as if set forth in full.
TECHNICAL FIELD
[0002] This application relates generally to communicating over a
wireless network and, more particularly, to communicating over a
wireless network using frequency hopping.
BACKGROUND
[0003] An ad hoc wireless network (a mesh network) is a
self-organizing network in which the network devices themselves
establish communication links with one another. Wireless networks
may be used in different settings. For example, a wireless network
may be established between monitoring and control devices and a
host computer.
[0004] In one example, network devices monitor and control
electrical systems, such as a building lighting system or fire
alarm system. The devices report status information from their
monitored system to the host computer. In response, the host
computer sends control commands, which the devices use to control
the electrical systems.
[0005] There are numerous issues involved in establishing and
maintaining a wireless network, including power consumption and
network congestion. Network devices on a wireless network typically
run off of batteries, which must be checked and changed
periodically. Changing batteries frequently can be inconvenient.
Furthermore, as wireless networks become larger, the amount of data
transmitted across those networks increases, which can result in
data packet collision and, thus, degraded quality of service (QoS).
Nodes (e.g., devices) on a wireless network may communicate using a
process known as frequency hopping. Frequency hopping allows nodes
to communicate using different frequency channels. Nodes on the
network typically cycle through different frequency channels,
during which communications can be sent to, and received from,
other nodes. When not communicating, the nodes typically enter a
low-power state in order to save power. Heretofore, frequency
hopping on a mesh, network was implemented by synchronizing all
devices of the network. That is, each device followed the same
frequency sequence, and entered its communication mode at the same
time and for the same duration. This configuration has proven
difficult to maintain, particularly for large networks.
SUMMARY
[0006] This patent application describes methods and apparatus,
including computer program products, for communicating over a
wireless network.
[0007] In general, in one aspect, this application is directed to a
first device for communicating with a wireless network comprised of
nodes that are active for predefined activation times and that are
at least partially dormant when not active. The communication
method comprises identifying a start of communication via a message
that exceeds a maximum activation time of nodes on the wireless
network by a factor N, where N is equal to at least a maximum
number of frequencies on the wireless network, and exchanging
information with a second device comprising a node on the wireless
network that is within a transmission range of the first device,
where the information is exchanged following the message and
includes frequency hopping data for the second device. This aspect
may also include one or more of the following features. The
information exchanged following the message may include frequency
hopping data for the first device. The information may be exchanged
over multiple frequencies through which the first device hops
following the message. The multiple frequencies may be defined
following the message.
[0008] In general, in another aspect, this application is directed
to a first device for use in conjunction with a wireless network.
The first device outputs a preamble to the wireless to network;
where the preamble contains time data corresponding to a duration
of the preamble, and identifies (e.g., outputs or otherwise
indicates) one or more listening frequencies of the first device
that follow the preamble, where the one or more listening
frequencies comprise one or more frequency bands at which the first
device activates. The first device receives, from a second device
in the wireless network and in the one or more listening
frequencies, sequence data, wake-up data, and duty cycle data,
where the sequence data is usable to obtain a sequence of
frequencies at which the second device activates, the wake-up data
corresponds to times at which the second device activates, and the
duty cycle data is based on durations of time for which the second
device activates. The first device stores (e.g., in memory) the
sequence data, wake-up data, and duty cycle data. This aspect of
the application may also include one or more of the following
features.
[0009] The first device may output a search packet to the second
device. The search packet may be output following the preamble and
before receiving the sequence data, the wake-up data, and the duty
cycle data from the second device. The search packet may identify
the first device and contain sequence data, wake-up data, and duty
cycle data for the first device. The wireless network may include N
(N>2) devices including the second device, where each of the N
devices is configured to activate at a frequency in a sequence of M
(M>1) frequencies. The N devices may activate for time
intervals, one of the N devices may have a maximum activation time
interval of T, and the preamble may have a time duration that is at
least a product of M*T.
[0010] The one or more listening frequencies may comprise a single
listening frequency. The single listening frequency may be a same
frequency at which the preamble is output. The one or more
listening frequencies may comprise multiple listening frequencies.
The search packet may identify sequence data and duty cycle data
for the multiple listening frequencies. The sequence data may
comprise a single number. The first device may generate the
sequence of frequencies at which the second device activates by
processing the single number using a predefined algorithm.
[0011] The first device may, receive, from a third device in the
wireless network, second sequence data, second wake-up data, and
second duty cycle data. The second sequence data may be usable to
obtain a sequence of frequencies at which the third device
activates. The second wake-up data may correspond to times at which
the third device activates, and the second duty cycle data may be
based on durations of time for which the third device activates.
The first sequence data, the first wake-up data, and the first duty
cycle data (which are received by the first device from the second
device) may be received in a same listening frequency following the
preamble, but at different times than, the second sequence data,
the second wake-up data, and the second duty cycle data.
[0012] The first device may receive, from the second device and in
the one or more listening frequencies, sequence data, wake-up data,
and duty cycle data, for a third device in the wireless network.
The third device may not be within a wireless transmission range of
the first device. The first device may send a communication to the
second device using the sequence data, the wake-up data, and the
duty cycle data that was received from the second device. The
sending process may include (i) sending a second preamble, where
the second preamble is sent at a frequency specified in the
sequence data, at about a time specified in the wake-up data, and
for a duration that exceeds a duration specified in the duty cycle
data, (ii) receiving an acknowledgement signal from the second
device in response to the second to preamble, and (iii) sending
information for the communication following the second preamble and
in response to the acknowledgement signal. The information may be
sent in a same frequency as the second preamble. The preamble may
comprise a series of data packets transmitted in sequence, which
may contain the time data from which the duration of the preamble
can be determined.
[0013] In general, in another aspect, the application is directed
to a device that includes memory configured to store instructions
that are executable, and at least one processor configured to
execute the instructions to enter an activation mode to communicate
over a wireless network, and to enter a dormant mode periodically,
where the at least one processor performs fewer tasks in the
dormant mode than in the activation mode. To configure the device
for communication or broadcast over the wireless network, the at
least one processor executes instructions to output a preamble to
the wireless network, where the preamble contains time data
corresponding to a duration of the preamble, and to identify one or
more listening frequencies of the device that follow the preamble.
The one or more listening frequencies comprises one or more
frequency bands at which the device enters the activation mode
following the preamble. The at least one processor also executes
instructions to receive, from a node in the wireless network and in
the one or more listening frequencies, sequence data, wake-up data,
and duty cycle data, where the sequence data is usable to obtain a
sequence of frequencies at which the node activates, the wake-up
data corresponds to times at which the node activates, and the duty
cycle data is based on durations of time for which the node
activates. The at least one processor also executes instructions to
store the sequence data, wake-up data, and duty cycle data.
[0014] In general, in another aspect, the application is directed
to a node in a wireless to network that is configured to receive a
preamble from a device, where the preamble contains time data
corresponding to a duration of the preamble, and to identify one or
more listening frequencies of the device that follow the preamble,
where the one or more listening frequencies comprise one or more
frequency bands at which the device activates. The node is also
configured to send, to the device and in the one or more listening
frequencies, sequence data, wake-up data, and duty cycle data,
where the sequence data is usable to obtain a sequence of
frequencies at which the node activates, the wake-up data
corresponds to times at which the node activates, and the duty
cycle data is based on durations of time for which the node
activates. The sequence data, the wake-up data, and the duty cycle
data may be sent at a time that is obtained to reduce conflict with
other nodes sending data to the device following the preamble.
[0015] In general, in another aspect, the application is directed
to a system comprising a wireless network that includes plural
nodes, at least some which have an activation time during which the
at least some nodes are active. The system includes a first device
that is configured to identify a start of communication via a
message that exceeds a maximum activation time of nodes on the
wireless network by a factor N, where N is equal to at least a
maximum number of frequencies on the wireless network, and a second
device comprising a node on the wireless network. The second device
is configured to respond to the message with frequency hopping
data. The frequency hopping data from the second device identifies
how the second device performs frequency hopping to receive data.
This aspect may also include one or more of the following
features.
[0016] The message may identify one or more frequencies that follow
the message. The second device may be configured to respond to the
message in the one or more frequencies. A response to the message
by the second device may be timed to reduce conflict with possible
messages from other devices on the wireless network. The message
may comprise a preamble containing time data corresponding to a
duration of the preamble. The first device may identify one or more
listening frequencies of the first device that follow the preamble.
The one or more listening frequencies may comprise one or more
frequency bands at which the first device activates. The second
device may be configured to respond to the message in the one or
more listening frequencies. The frequency hopping data may comprise
sequence data, wake-up data, and duty cycle data, where the
sequence data is usable to obtain a sequence of frequencies at
which the second device activates, the wake-up data corresponds to
times at which the second device activates, and the duty cycle data
is based on durations of time for which the second device
activates.
[0017] Each of the foregoing aspects and features can be
implemented via a method, one or more apparatus, one or more
systems, and/or one or more computer program products comprised of
executable instructions stored on one or more machine-readable
media. Furthermore, the foregoing aspects and features may be
combined in any manner.
[0018] The details of one or more examples are set forth in the
accompanying drawings and the description below. Further features,
aspects, and advantages will become apparent from the description,
the drawings, and the claims.
DESCRIPTION OF THE DRAWINGS
[0019] FIG. 1 is a block diagram of a wireless network.
[0020] FIG. 2 is a block diagram of a wireless network that
includes a portion of the network of FIG. 1.
[0021] FIG. 3 is a timing, diagram showing a frequency hopping
sequence and timing for a receiving device on a Wireless network,
such as that of FIGS. 1 and 2.
[0022] FIG. 4 is a flowchart showing a frequency hopping
protocol.
[0023] FIG. 5, comprised of FIGS. 5A and 5B, shows timing diagrams
depicting operation of the frequency hopping protocol, which
includes a sending device's timing and a receiving device's timing
used in the exchange of information between devices.
[0024] FIG. 6 shows timing diagrams depicting individual units of a
preamble used in the frequency hopping protocol in relation to a
normal operational mode of a receiving device.
[0025] FIG. 7 shows timing diagrams depicting how neighboring nodes
respond, at different times, to a search packet transmitted by a
sending node.
[0026] FIG. 8 shows timing diagrams depicting use of a preamble
during the normal course of operation of two network nodes,
particularly where a device's receiving duration is short.
[0027] FIGS. 9A to 9C show examples of devices on the wireless
network of FIG. 1 or 2 that may implement the frequency hopping
protocol.
[0028] Like reference numerals in different figures indicate like
elements.
DETAILED DESCRIPTION
[0029] Described herein is a process for use by devices to
communicate over a wireless network using frequency hopping. FIG. 1
shows an exemplary wireless network 10 on which the process may be
implemented. According to the process, devices on, or entering, the
network may communicate over one or more frequency channels. In
order to enable such communication, frequency hopping data is
exchanged between devices upon entry of a device into the network.
As described below, the frequency hopping data describes how each
device communicates using frequency hopping, thereby enabling the
devices to schedule communications for appropriate frequencies,
times, and durations. Frequency hopping, as used herein, increases
the capacity of the network because it allows network
communications to overlap without substantially increasing
collisions and packet loss.
[0030] Before describing how frequency hopping is used in a
wireless network, we first describe an example of a wireless
network and how it is established.
The Wireless Network
[0031] Wireless network 10 is a heterogeneous network, since
devices on wireless network 10 need not perform the same functions.
Wireless network 10 includes endpoint devices 12 and 14,
intermediary devices 15 to 22, and base stations 24 to 26. Endpoint
devices 12 and 14 and intermediary devices 15 to 22 may communicate
via radio frequency (RF) links. RF links are shown as dotted lines.
Base stations 24 to 26 may communicate to the intermediary devices
via RF links and are wired to a high-speed backbone 29, through
which base stations 24 to 26 communicate with a host computer 30 at
a relatively high speed. High-speed backbone 29 may be any type of
wired or wireless medium, such as Ethernet or Wi-Fi (wireless
fidelity).
[0032] Each of endpoint devices 12 and 14, intermediary devices 15
to 22, and base stations 24 to 26 defines a node of wireless
network 10. Each of these devices includes memory (not shown) that
stores executable instructions and one or more processors (the one
or more processors defining a processing system--not shown) for
executing the instructions to perform functions described herein.
In this implementation, the structure of endpoint devices 12 and 14
may be the same as the structure of intermediary devices 15 to 22,
and the structure each base station 24 to 26 is the same. This may
not be the case in other implementations. Each device is programmed
with appropriate functionality.
[0033] Each node of wireless network 10 may enter a low-power, or
"dormant", mode when not communicating over wireless network 10.
During this low-power mode, a node may maintain some low-level
operations; however, major processing functions are curtailed in
order to conserve power. Following the low-power mode, the node
"wakes-up", i.e., activates and enters its normal operational mode.
During the normal operational mode, the node is again able to
send/receive data over wireless network 10.
[0034] A network node may always enter the low-power mode whenever
it is not communicating over wireless network or the network node
may maintain its normal operational mode during some periods of
non-communication over the wireless network. Reduced power
consumption for network nodes is advantageous, since network nodes
are often powered by low-capacity, small-size batteries, such as
lithium coin cell batteries. Long life for batteries of this type
can generally be achieved when average power consumption of a
remote terminal (e.g., a node of the wireless network) is
relatively low. The frequency hopping protocol described herein
allows network nodes to remain in their low-power modes for
relatively long periods of time, thereby further decreasing the
amount of power consumed by those nodes.
[0035] The base stations, intermediary devices, and endpoint
devices described above may be any type of computing device, such
as a work station, a personal computer, a server, a portable
computing device (e.g., a personal digital assistant or "FDA"), a
cellular telephone, or any other type of intelligent device capable
of executing instructions and connecting to a network. The base
stations, intermediary devices, and endpoint devices can execute
any number of computer programs, including applications that are
configured to generate, receive, and transmit data packets for use
on the network.
[0036] As described in U.S. patent application Ser. No. 10/860,935,
which was filed on Jun. 4, 2004 and which is incorporated herein by
reference, a device entering network 10 may configure itself to
operate as an intermediary device (e.g., a router or repeater) if
necessary based on predefined criteria. Such a device may also have
the capabilities (and configuration) of an endpoint device
described below.
[0037] An endpoint device may be either a source or a destination
of network data. In this implementation, one or more sensing
devices may be connected to an endpoint device; other
implementations may include endpoint devices without sensing
devices. The sensing device(s) may be used to monitor physical
systems, such as a processing plant, and variables, such as
temperature. An endpoint device may acquire analog and/or digital
signals from the sensing device(s) and transmit these signals to a
base station via wireless network 10. An antenna (not shown) may be
included on each endpoint device to effect transmission. Antennas
may also be included on the other wireless devices in the
network.
[0038] One or more actuators may also be connected to an endpoint
device in this implementation. The endpoint device may use analog
or digital command signals to command the actuator(s). These
command signals may originate in the endpoint device or in a host
computer 30. In the latter case, such command signals may be
transmitted from host computer 30, to a base station, and then to
the endpoint device, either directly or through one or more
intermediary devices in wireless network 10.
[0039] An intermediary device is an intermediate node of wireless
network 10 that functions as a router or repeater to forward data
sent by endpoint devices, other intermediary devices, and/or base
stations. Intermediary devices typically send the data in the
format that the data is received and at the same rate as the data
is received. Each endpoint device, described above, may configure
itself to operate as an intermediary device, and vice versa.
Intermediary devices also store routing information, such as a next
hop along a network path to a data packet's intended destination,
and a hop on a return path.
[0040] A base station is a node of the wireless network that may be
connected to high-speed backbone 29. Base stations act as
intermediaries between Wireless network 10 and backbone 29,
performing any necessary data and protocol conversions to permit
data exchange between the two. Base stations that do not connect to
backbone 29 may also be integrated into network 10. Such base
stations (not shown in FIG. 1) may interface to a host device, and
may be connected to other host devices or to the backbone via a
wireless link through, e.g., a Wi-Fi (wireless fidelity)
connection.
[0041] Host computer 30 may also connect to high-speed backbone 29.
Host computer 30 supervises wireless network 10 and performs tasks
that include receiving, processing and storing data generated by
endpoint devices, and issuing command signals to the endpoint
devices. Host computer 30 may also be used to reconfigure an
endpoint device and/or intermediary devices to implement the
wireless network described herein.
[0042] The only requirement in forming wireless network 10 is that
every endpoint device should be within the RF transmission range of
a base station or an intermediary device, and every intermediary
device should be within the RF transmission range of a base station
or another intermediary device. Devices outside of their RF
transmission range are typically not able to communicate directly
with each other over wireless network 10.
[0043] The overall topology of wireless network 10 resembles a
spanning forest, in which the endpoint devices function as leaves,
the intermediary devices function as branches, and the base
stations function as roots. Like in a dense forest where trees can
overlap, communication links among intermediary devices mesh to
form a web-like structure, which enables the endpoint devices
(leaves) and intermediary devices (branches) to communicate with
multiple base stations (roots).
[0044] Multiple networks may occupy the same physical space. Data
packets for such networks are differentiated by a network group
identifier (ID). Thus, the networks remain logically separate even
though they occupy the same physical space. Such networks may even
use the same devices, either in the same or in different roles. For
example, device 24 may act as a router for one network and as a
base station for another network; device 17 may operate as an
endpoint device for one network and as a base station for another
network, etc. Operation of devices in such different networks is as
described herein. Devices may be programmed to operate as base
stations, routers, or endpoint devices in one network by specifying
the group ID of that network and the functionality of the device
for that group ID. A host computer may initiate/control programming
of the various devices or the devices may be programmed directly on
the devices themselves. Network data packets typically contain the
network group ID for their corresponding network.
[0045] In networks, such as network 10, all of the base stations
may be connected to host computer 30. In this case, data packets
may be stored and organized by host computer 30. In networks such
as these, host computer 30 maintains a central database comprised
of packets from one or more endpoint devices. In other networks,
the base stations are not all connected to the same host computer.
In this type of network, different base stations may receive
packets resulting from a single transmission of an endpoint device.
Each base station maintains a separate database of packets that it
receives. The various base stations may synchronize their databases
periodically. Synchronization may occur via wireless connection or
via a wired connection, such as Ethernet (if a wired connection
exists among the various databases). This synchronization results
in each base station containing a complete database of all packets
that reach base stations in the network. Redundant databases such
as these are particularly advantageous because they provide back-up
in the event of failure of a base station on the network.
Establishing the Wireless Network
[0046] As described in U.S. patent application Ser. No. 10/304,528,
which was filed on Nov. 26, 2002 and which is incorporated herein
by reference, there is no connectivity among devices of a wireless
network until at least one endpoint device initiates communication,
e.g., by sending a "a hello message". This process is referred to
here as "terminal-initiated polling". Communication may be
initiated when an endpoint device, such as endpoint device 12, is
first activated. That is, when endpoint device 12 is first
activated, endpoint device 12 broadcasts a hello message to
interrogate its surroundings. The term "broadcast" means to send
(or transmit) to one or more other network devices.
[0047] The hello message is a specialized data packet and is
therefore referred to as a "search packet". The search packet may
contain information, such as the identity of an endpoint device and
a request to enter the wireless network. All intermediary devices
(e.g., routers or repeaters) within the RF transmission range
(typically 30 to 100 feet, but not limited to these values) of the
endpoint device re-broadcast the search packet to seek connections
with base stations or other intermediary devices within their
respective RF transmission ranges. The intermediary devices
re-broadcast the search packet until the search packet reaches all
of the base stations 24 to 26. This technique of propagating the
search packet through the network is referred to as "flooding" the
network.
[0048] When a base station receives a search packet, the base
station responds by generating and broadcasting a confirmation
packet. The confirmation packet is also propagated throughout the
entire wireless network 10 by flooding the network. Eventually, the
confirmation packet reaches the endpoint device that initiated the
hello message. At this point, communication among the network nodes
is possible.
[0049] Along the route that the confirmation packet takes back and
forth through wireless network 10, intermediary devices keep track
of which node sent them the confirmation packet, i.e., an
immediately preceding network node along the route. Each
intermediary device stores a pointer in memory that points to this
node. The pointers enable the intermediary devices to identify
neighboring nodes that can be used in transporting a data packet
closer to a base station. These neighboring nodes are referred to
as master nodes, or simply "masters". A master acts as a primary
recipient of data from its dependent, or "slave", node.
[0050] One advantage of terminal-initiated polling is that an
endpoint device need not wait to join a wireless network. That is,
since the endpoint device initiates entry into the wireless
network, the endpoint device controls when to establish a presence
in the wireless network. The endpoint device is not required to
wait for a periodic beacon signal from a base station before
joining the wireless network.
Joining the Wireless Network
[0051] A heterogeneous network is formed without a rigidly
prescribed hierarchy. In particular, every node in wireless network
10 can assume the role of an endpoint device or an intermediary
device and can change dynamically based on criteria specified by a
user. Essentially, the wireless network takes advantage of the
flexibility of homogeneous networks and the power efficiency of
heterogeneous networks. A process for joining wireless network 10
is described with respect to FIG. 2.
[0052] FIG. 2 shows a wireless network 31, which may be part of
wireless network 10. When a target device, such as device 40 in
FIG. 2, is activated for the first time and wants to join network
31, device 40 attempts to locate a packet forwarding device, such
as an intermediary device or base station, in its neighborhood.
Device 40 does this by broadcasting a search packet to the network,
as described in the preceding and following sections. In one
implementation, the network neighborhood includes all devices to
which device 40 has a direct RF link; although this application is
not limited to this definition of "neighborhood".
[0053] Device 40 receives response(s) to its search packet from its
neighbor(s). Device 40 uses those responses to determine, e.g., if
one or more intermediary devices is within its network neighborhood
(meaning that device 40 can establish an RF link to such a device).
If so, device 40 selects an intermediary device that satisfies
predefined criteria, and designates that device as its primary
master (and others as secondary, tertiary, etc. masters, if
applicable). Device 40 then joins the network as an endpoint device
with a low duty cycle. The same process occurs if device 40
identifies a base station in its neighborhood.
[0054] Criteria for selecting the primary master may include, but
are not limited to, the distance to a base station (or gateway),
the quality of the communication link to the intermediary device,
and the battery capacity of the intermediary device. For example,
device 40 may select a primary master node having less than a
certain number of hops to a base station or endpoint device. In
addition, in this implementation, device 40 may select the primary
master based also (or solely) on other criteria. For example,
device 40 may select a node having a link with low amounts of
noise, and/or having a large battery capacity (thereby ensuring
more reliable operation of the primary master).
[0055] If device 40 cannot find any intermediary device (or base
station) in its network neighborhood, device 40 will start
operating as an intermediary device (e.g., a router). Code
programmed into device 40 initiates its operation as a router,
which includes forwarding data packets, maintaining routing
information, and may enter a low-power mode less often, among other
things. During its operation as a router, device 40 "listens to"
transmissions from neighboring devices, such as node 43. Node 43,
in this case, is an endpoint device, since it is not an
intermediary device or a base station (if it were an intermediary
device or base station, device 40 would have recognized it as such
in response to device 40's initial attempt at entry into the
network). Device 40 "listens for" data transmissions from node 43,
not search packet transmissions. This is because transmission of a
search packet implies that node 43 is looking for connection to
network 10. A data transmission, on the other hand, implies that
node 43 is already connected to network 10.
[0056] After device 40 detects a data transmission from node 43,
device 40 attempts to determine if node 43 is connected to network
10. To do this, device 40 may listen for re-transmission of the
same data packet. Re-transmission of the same data packet implies
that node 43 is not connected to the network. Device 40 may listen
for an acknowledgement (or "ack") packet in response to the
original transmission. The ack packet is sent by another node
(e.g., node 42) that is on the network. This may not always work,
however, because &vice 40 may be unable to receive the ack,
packet due, e.g., to a distance from a node transmitting the ack
packet. Device 40 may send a packet to node 43 to determine its
connectivity. This packet may advise node 43 of the existence of
device 40, and ask node 43 whether node 43 is connected to the
network.
[0057] Assuming that device 40 is able to confirm that node 43 is
connected to the network, device 40 may send a data packet to node
43 asking node 43 (which, as noted above, is an endpoint device) to
reconfigure itself as a router for device 40. Node 43 decides
whether to configure itself as a router based, e.g., on its
capabilities, available bandwidth, network access, and the like.
Assuming that node 43 configures itself as a router; device 40 then
reconfigures itself to be an endpoint device that routes
communications through node 43. Device 40 selects node 43 to be its
own primary master node. If node 43 is not connected to the
network, and no other nodes are available to act as a router for
device 40, after a period of time device 40 may configure itself as
an endpoint device.
[0058] Device 40 may be programmed to reconfigure itself as an
endpoint device after operating as a router for a predetermined
amount of time, e.g., one hour, or more or less than one hour. Any
time period may be used. As noted above, device 40 may enter the
low-power mode periodically. Upon exiting ("awakening") from the
low-power mode, device 40 may again try to establish a connection
to the network as an endpoint device, as described above. If that
is not successful, device 40 may try to establish a connection to
the network as a router in the manner described above, i.e., by
configuring itself as a router, listening for non-search packet
transmissions, etc.
[0059] The amount of time that device 40 remains configured as a
router may be programmed into device 40. Alternatively, this amount
of time may be dictated by network traffic (e.g., search packets)
in the vicinity of device 40. For example, a large number of search
packets detected in the vicinity of device 40 indicates that there
are devices in the neighborhood (which, perhaps, are just not yet
connected to the network). By contrast, few search packets detected
in the vicinity of device 40 indicates that there may be few
devices in the neighborhood (and, perhaps, device 40 is isolated).
Thus, if a large number of data packets are detected in the
vicinity of device 40, device 40 configures itself as a router for
a longer period of time (since more devices implies a higher
probability of achieving a network connection). If few data packets
are detected in the vicinity of device 40, device 40 configures
itself as an endpoint device that can enter low-power mode for a
relatively long period of time. In this case, device 40 "awakens"
from the low-power mode less frequently, since there is less of a
chance of achieving a network connection.
[0060] Device 40 may operate as an intermediary node for one or
more other devices on network 31. Assuming that device 40 has a
link to a master intermediary device, device 40 may decide whether
to remain an intermediary device for neighboring endpoint devices.
From the intermediary device IDs (identifiers) reported by
neighboring endpoint devices, device 40 can identify whether any
one of the neighboring endpoint devices relies on device 40 as its
only intermediary device. For example, device 40 can query other
intermediary devices for routing information. If at least one
device does rely solely on device 40, then device 40 remains an
intermediary device; otherwise device 40 configures itself as an
endpoint device.
[0061] If device 40 does not receive data or an intermediary device
request from its neighboring nodes (endpoint devices or
intermediary devices) for a predefined amount of time, device 40
may configure itself as endpoint device. The predefined time can be
programmed into the device or selected based on one or more
parameters. One example of such a parameter is the frequency with
which a dependent endpoint device communicates with the
intermediary device; i.e., the less frequent the communication from
the endpoint device, the more often device 40 configures itself as
an endpoint device.
[0062] When an intermediary device becomes a master intermediary
device of a high number of nodes, a substantial amount of traffic
can flow into the intermediary device. This can result in data
congestion and, possibly, frequent data packet collision. Excessive
traffic can be diverted for an overloaded intermediary device by
obtaining a new intermediary device from endpoint devices in the
neighborhood. In this situation, an intermediary device can issue
and broadcast a router election request to its neighboring endpoint
devices. Upon receiving this router election request, an endpoint
device responds with a message indicating its operating status,
which may include the ID of intermediary devices that it
communicates with, its data generation rate, its remaining battery
charge, etc.
[0063] After receiving responses from all neighboring endpoint
devices, the intermediary device selects a new intermediary device
based on a certain criteria. For example, an endpoint device that
communicates with the most number of intermediary devices can offer
higher connectivity with the rest of the network; an endpoint
device with a relatively low data generation rate can offer more
bandwidth to route data for other nodes; and an endpoint device
with the most battery charge can afford to operate with a high duty
cycle. The intermediary device instructs the selected node to
configure itself as a router. One purpose of instructing one of the
neighboring endpoint devices to be a router is to divert traffic.
Once a new router is elected, the intermediary device need not
select this new router as its own primary master. The overloaded
intermediary device elects a new router simply to increase the
number of intermediary devices in the neighborhood, so that traffic
congestion through the overloaded intermediary device can be
alleviated.
[0064] Every node in the network may continually search for a
better primary master intermediary device in the manner described
above. In most cases, a node can have one or multiple alternative
masters in addition to its primary master. As such a new primary
master can be selected from the alternative masters. However, there
are situations where a node may find that all its master
intermediary devices are unreliable. In this case, a node can
broadcast a router election request to its neighboring nodes. Upon
receiving the router election request, a neighboring node, either
an endpoint device, an intermediary device or a base station, will
respond with its operating status. After reviewing all responses,
the requesting node selects a new primary master from the
neighboring nodes. If the selected node is already an intermediary
device, it is simply recognized as the new primary master of the
requesting node. If the node is an endpoint device, the requesting
node issues a message to request the endpoint device to become an
intermediary device as well as the primary master of the requesting
node. The endpoint device decides whether to become an intermediary
device based, e.g., on its capabilities and connection to the
network. If it does become an intermediary device, the endpoint
device notifies the requesting node.
[0065] Every node may seek to operate as an endpoint device. A node
will serve as an intermediary device when the node recognizes that
either there is no intermediary device available in its
neighborhood, or all neighboring intermediary devices cannot
provide reliable connectivity due, e.g., to excessive traffic or
radio link issues.
[0066] Ultimately, the process for joining the wireless network may
lead to electing a relatively small number of intermediary device
nodes based on predefined criteria needed to maintain network
connectivity. The criteria can include, but are not limited to, one
or more of the factors mentioned above, such as a number of hops to
a base station, a reliability of the communication link, and
remaining battery life. The probability of an endpoint device
becoming an intermediary device node can be adjusted by tuning
these factors. These factors may be tuned, e.g., via host computer
30 and/or by directly accessing the appropriate network device and
programming the appropriate values.
Frequency Flopping in the Wireless Network
[0067] The frequency hopping protocol described below is described
in the context of a heterogeneous wireless network, such as
wireless network 10. The frequency hopping protocol, however, is
not limited to use with a heterogeneous network, but rather may
also be used with a homogeneous network, e.g., a network in which
all devices have the same structure and/or function. For example, a
wireless network that includes only routers may benefit from the
frequency hopping protocol because the frequency hopping protocol
allows the routers to operate with a reduced duty cycle (as
described below). As a result, batteries that power the routers
will use power less quickly.
[0068] Furthermore, the frequency hopping protocol does not require
a common time base throughout an entire wireless network. That is,
nodes of a wireless network communicate with their neighboring
nodes based on knowledge of the neighboring nodes' wake-up times,
wake-up durations, and/or channel sequences (as described below).
One advantage of this process is that it allows nodes to wake-up
more frequently, each time doing so with a relatively short
duration. Because nodes wake-up more frequently, packets can
propagate through the network relatively quickly. As a result, the
wireless network can operate with increased robustness (proactive
channel switching), relatively high bandwidth, relatively low power
consumption, and relatively low latency.
[0069] Turning now to one example, one or more nodes (e.g., all
nodes) of wireless network 10 (31) may communicate using frequency
hopping. In the context of a wireless network, frequency hopping
may be implemented by providing windows, during which network nodes
can receive communications from other nodes. Referring to FIG. 3, a
receiving node is a node that is to receive communication. The
receiving node wakes, meaning that it activates and operates in its
normal operational mode, at times 45, 46, etc. for predefined
durations. The durations are labeled as MN_RX_time, where "MN"
stands for "Mesh Node" and "RX" for "receive" ("TX" stands for
"transmit"). MN_RX_time may also be changed on the fly depending on
the network status, and each node may have a different MN_RX_time.
The receiving node wakes in predefined frequency channels, which
are labeled Ch 7, Ch 11, etc. At other times 47, 49, etc., the
receiving node may be in a low-power, or dormant, mode, which is
labeled MN_sleep_time.
[0070] We use node 15 (FIG. 1) as an initial example. During its
normal operational mode, the receiving node, here node 15, is able
to receive communications only in specified frequency channels. For
example, at time 46, node 15 wakes-up in channel (Ch) 11 for a
duration of 200 .mu.s (microseconds). If another node on the
wireless network wants to send a communication to node 15, the
other node must do so within the 200 .mu.s duration in Ch 11;
otherwise, it must wait until node 15 wakes-up again (e.g., in
another 10 ms (milliseconds)). When node 15 wakes again, however,
it may wake in a different frequency, channel (Ch 2), although it
will typically (although need not) wake for the same duration. A
node that wants to communicate with node 15 must therefore know the
times at which node 15 wakes, the durations for which node 15 is
awake, and the channel sequence for node 15 (so that the node
wanting to communicate with node 15 can determine the channel in
which to send communications to node 15).
[0071] To complicate matters, in network 10, the nodes need not
have the same wake-up times, wake-up durations, or channel
sequences. For example, node 15 may wake up once every 10 ms for
200 .mu.s, node 16 may wake up once every two seconds for 200
.mu.s, node 17 may wake up once every half second for 50 .mu.s, and
so on. Likewise, node 15 may follow a random channel sequence of 7,
11, 2, 10 . . . 12, node 16 may follow a channel sequence of 16,
15, 14, 13 . . . 1, and node 17 may follow a channel sequence of 1,
2, 3, 4 . . . 16. The channel sequences may be stored in frequency
hopping lists in memory on each node. Although each node may follow
a different channel sequence, each node typically (although need
not) maintains its same sequence. That is, node 15 repeats its
sequence of 7, 11, 2, 10 . . . 12 periodically, node 16 repeats its
sequence 16, 15, 14, 13 . . . 1 periodically, and node 17 repeats
its sequence of 1, 2, 3, 4 . . . 16 periodically. In this
implementation (2.4 GHz RF operation), each node communicates over
sixteen channels; however, in other implementations, some nodes may
communicate over less than, or more than, the sixteen channels
specified herein. For example, in 900 MHz RF operation, there may
be 50 frequency channels.
[0072] A sending, node (i.e., a node that is to send
communication), therefore, should know the frequency sequence,
wake-up time, and duty cycle of a receiving node in order to
determine when to send a communication so that the communication
will be received. A sending node can typically wake and send
communications in any channel irrespective of its own receiving
frequency channel sequence and timing. A receiving node typically
sends an ack packet back to a sending node to acknowledge receipt
of a communication. If the sending node does not receive an
appropriate ack packet in response to a communication, is the
sending node may re-send the communication at a subsequent time and
frequency channel for the receiving node (which may, or may not, be
the time and frequency that immediately follows the current time
and frequency). For example, the sending node may be programmed to
retry the communication at a random subsequent time, thereby
reducing the chances of collisions with other sending nodes. This
process may be repeated until the sending node receives an ack
packet from the receiving node.
[0073] In this implementation, each node stores, and keeps track
of, clocks of its neighboring nodes that are both upstream and
downstream in a communication path. Search response packets include
time stamps and may include the channel sequence of the node that
sends the ack packet. Each time a search response packet is
received from a neighboring node, the receiving device will adjust
the neighbor's clock in its neighborhood database to the received
time. By way of example, for two upstream nodes (primary and
secondary) and 30 downstream nodes, a node should store at least 32
independent clocks (or other next-time-to-wake-up information) in
its neighborhood database. Each clock may include at least two
bytes, which can store 65536 clock ticks. So, with increments of 1
ms (a-clock tick), a clock can cover up to 65 seconds of next
time-to-wake-up information. This may be done for all
communications, not just search response packets.
[0074] One problem for a new node entering a wireless network is
that the new node does not know its neighboring nodes' frequency
sequence, wake-up time, and duty cycle. Accordingly, a protocol
(which is referred to herein as "the frequency hopping protocol")
is provided that enables a device to discover how neighboring
network nodes perform frequency hopping, e.g., to discover their
frequency sequence data, wake-up data, and duty cycle data. In this
context, the frequency sequence data (or simply sequence data) is
usable to obtain a sequence of frequencies at which a neighboring
node activates; the wake-up data corresponds to times at which the
neighboring node activates; and the duty cycle data corresponds to
durations of time for which the neighboring node activates.
[0075] In the context of the processes described above for joining
and establishing the wireless network, the hello message (e.g., the
search packets) is transmitted between a preamble (described below)
used in the frequency hopping protocol and listening frequency(ies)
that follow the preamble. The frequency hopping protocol is also
used to broadcast over the wireless network, as described
below.
[0076] Referring to FIGS. 4 and 5, a sending device (e.g., device
40 of FIG. 2) entering a wireless network outputs (50) a preamble
51. In one implementation, the preamble is a sequence of packets
that are output from the device. The data packets may be
IEEE802.15.4 data packets, and each packet may be, e.g., 1 ms
(.about.25 bytes). The preamble is designed, essentially, to get
the attention of nodes that neighbor the device. In a
non-IEEE802.15.4 context, the preamble may represent a specific
data pattern, e.g., "01010101 . . . ". The preamble identifies a
start of communication between device 40 and one or more of its
neighbors. In this context, a node that neighbors the device is any
node that is part of the wireless network and that is within a
wireless transmission range of the device. In order to ensure that
every neighboring node identifies the preamble, the preamble is
transmitted in one frequency channel (e.g., Ch 8 in FIG. 5), and is
structured so that it exceeds a maximum activation time of all
nodes on the wireless network by a factor N, where N is equal to at
least a maximum number of frequencies on the wireless network. This
ensures that all neighboring nodes of device 40 will wake in the
frequency channel of the preamble and thereby receive the preamble
at least once during its activation cycle.
[0077] By way of example, if the maximum time between activations
for any node on the wireless network is 5 s (where "s" stands for
seconds), and the network recognizes sixteen frequency channels,
the preamble will be at least SOs (16.times.5 s). In another
example, the time between activation (also known as the node's
sleep cycle) is 10.2 ms and there are a total of 50 channels,
resulting in a preamble of 510 ms. In still another example, the
preamble may be about 3.2 s. It is noted that, in some
circumstances (described below), a preamble of greater than 400 ms
may need to be divided among multiple frequency channels in
accordance with Federal Communications Commission (FCC)
regulations.
[0078] Referring to FIG. 6, the duration 52 (MN_RX_time) during
which the receiver listens for communications in a frequency
channel should be at least twice as long as the length of a
preamble packet unit 54. This is done in order to ensure that at
least one complete preamble packet (or "preamble unit") is
recognized by the receiver. If a start of frame delimiter (SFD) is
used to detect the preamble, MN_RX_time need be only slightly
longer than a preamble unit.
[0079] The duration of the preamble may be expressed mathematically
as follows. Each preamble unit contains a time-to-tail end value
(so that any neighboring node that receives the preamble packet
unit knows when to return to the preamble channel to receive a
search packet). Assume that the length of the preamble packet unit
in milliseconds is L_pu. In this case,
MN.sub.--RX_time=2*L.sub.--pu,
MN_cycle_time is the duty cycle of a network node, and is equal to
M_ps*2*L_pu, where M_ps is a predefined power saving multiplier.
MN_sleep_time, which is the amount of time a node is in its
low-power, or dormant, cycle is defined as follows:
MN_sleep_time=MN_cycle_time-MN.sub.--RX_time.
The power saving multiplier M_ps is essentially the ratio of
MN_cycle_time to MN_RX_time. A higher M_ps means a higher ratio of
sleep time to activation time, resulting in more power savings.
[0080] The entire length of the preamble is thus determined as
follows:
preamble_length=CN_hop*MN_cycle_time=CN_hop*M.sub.--ps*2*L.sub.--pu,
where CN_hop is the maximum number of channels in the hopping
sequence of a node on the network having a maximum
MN_sleep_time.
[0081] Typically, CSMA (Carrier Sense Multiple Access) listening is
conducted before the preamble is output in order to make sure there
is currently no other preamble in the same frequency channel. If
there is already a preamble in that frequency channel, the node can
decide to wait until an on-going search process is over or to jump
to another frequency channel that has no on-going search process,
and begin its search process immediately in that other frequency
channel, i.e., send out the preamble in that other frequency,
etc.
[0082] The preamble may include timing information. More
specifically, following the preamble, the neighboring nodes
transmit their sequence data, wake-up data, and duty cycle data to
device 40. The neighboring nodes therefore need to know when the
preamble will end so that they can begin transmission. The preamble
therefore may include timing data, which identifies the end of the
preamble. For example, in one implementation, the preamble is a
sequence of data packets transmitted in a stream. Each data packet
may contain a countdown time, which indicates the remaining length
of the preamble. In other implementations, only select data packets
may contain the countdown time.
[0083] Following the preamble, device 40 transmits (56) its search
(or hello) packet 57 at 58a. The search packet includes the
sequence data, wake-up data, and duty cycle data for device 40.
That is, the search packet include the sequence of frequencies in
which device 40 listens for communications, the times at which
device 40 listens in those frequencies, and the listening duration.
This information is stored in frequency hopping list(s) in
neighboring nodes, and enables those nodes to send communications
to device 40.
[0084] After transmission of the search packet, device 40 listens
(59) in one or more frequencies at 58b for communications from its
neighboring nodes, specifically, the sequence data, wake-up data,
and duty cycle data for the neighboring nodes (e.g., nodes 42 and
43, assuming, for this example, that node 43 is already a member of
the network and within the transmission range of node 40). That is,
the sender goes into receive mode and waits to receive search
responses (if sending a broadcast packet, the sender will resume
its sleep/wake cycle). In order to transmit frequency hopping data,
such as sequence data, wake-up data, and duty cycle data, to device
40, the neighboring nodes need to know the frequency(ies) that
device 40 will be listening in following the preamble. In this
regard, device 40 may listen in a single frequency or in multiple
frequencies. In other words, device 40 may frequency hop during the
listening phase 59 following the preamble. In any case, the
frequency(ies) that device 40 will be listening in should be known
to the neighboring nodes.
[0085] In one implementation, device 40 listens in the same
frequency channel in which the preamble was sent. Thus, a
neighboring node identifies device 40's listening frequency(ies) by
the frequency of the preamble (i.e., it knows beforehand that the
two are the same). In this case, as shown in FIG. 5, a neighboring
node (e.g., node 42) locks onto that channel (in FIG. 5, channel
(Ch) 8 at 63) after the preamble is detected, and continues to wake
in that channel. It is noted, however, that the node's internal
channel hopping clock continues to run so that the node can resume
its frequency hopping sequence 60 at 58c (after sending its
frequency hopping data to device 40) as if that sequence were never
interrupted.
[0086] In another implementation, the neighboring nodes use timing
data in the preamble to schedule a communication to device 40. That
is, a neighboring node keeps track of the amount of time left in
the preamble and, following the preamble, sends its frequency
hopping data to device 40 in the same frequency channel as the
preamble. Prior to sending its frequency hopping data (e.g., time
61), the node hops frequencies, and performs transmitting and
receiving operations, in accordance with its usual schedule.
[0087] In still another implementation, device 40 sends the same
preamble packet without timing data and, when it is near the end of
preamble period, device 40 sets a flags in the preamble packet
indicating the upcoming end of the preamble (sometimes referred to
as its "tail"). The neighboring nodes may check between wake-up
times in order to determine whether the preamble is near its end.
Prior to this, the node hops frequencies, and performs transmitting
and receiving operations, in accordance with its usual schedule. In
this implementation, the preamble may contain data, such as timing
data, indicating that that the preamble is near its end. In this
case, when a node determines that the preamble is near its end
(e.g., within a predefined time of its end), the node locks onto
the channel that the preamble is transmitted in, and wakes only in
that channel. It is noted, however, that the node's internal
channel hopping clock continues to run so that the node can resume
its frequency hopping sequence (after sending its frequency hopping
data to device 40) as if that sequence were never interrupted.
[0088] In a case where device 40 listens in multiple frequencies,
device 40 provides the neighboring nodes with the sequence data,
wake-up data, and duty cycle data for the listening period. This
information can be provided in a search packet which is broadcast
following the preamble, and in the same channel as the preamble
(thereby ensuring that the neighboring nodes will receive the
search packet), but before the listening period. Alternatively,
this information may be provided in the preamble itself.
[0089] During the listening period 59 of device 40, the neighboring
nodes transmit their frequency hopping data, such as their sequence
data, wake-up data, and duty cycle data to device 40. Device 40
receives (62) the frequency hopping data for each node, and stores
it in memory in association with a node identifier and the clock
for the node. Thus, device 40 knows the frequency hopping data for
its neighboring nodes, and vice versa. Device 40 is thus able to
keep track of the current frequency channel, wake-up time, and duty
cycle of all of its neighboring nodes, and the neighboring nodes
can do the same for device 40.
[0090] During the listening mode, a number of neighboring nodes may
attempt to transmit frequency hopping data to device 40. For
example, first, second, third, etc. neighbors may all attempt to
transmit their sequence data, wake-up data, and duty cycle data to
device 40. If all devices attempt to transmit at the same time,
this can result in data collision, and the required data may not
reach device 40. Therefore, the listening mode of device 40 is
structured to be long enough so that numerous nodes can send their
data at different times. For example, the listening mode may be 100
ms, 500 ms, or longer.
[0091] In one implementation, the nodes of network 10 are each
programmed to send their frequency hopping data at a random time
during the listening mode of device 40. For example, each
neighboring node may contain an algorithm that picks a random time
during the listening mode to send its frequency hopping data.
Sending data at random times reduces the possibility of a collision
between frequency hopping data from different nodes.
[0092] In the event of a collision, a node will not receive an ack
packet back from device 40. In this case, the node re-sends its
frequency hopping data at a different time. If device 40 is
frequency hopping during the listening mode, its neighboring nodes
must also take this into account when sending their frequency
hopping data. This contingency is shown in FIG. 7, where node 42
sends its frequency hopping data at a first time 64 in channel 6
and node 43 sends its frequency hopping data at a second time 65 in
channel 1.
[0093] It is noted that if all nodes use the same hopping sequence,
wake-up times, and/or duty cycles, then-all frequency hopping data
need not be transmitted between nodes. For example, if all nodes on
the network have the same duty cycle (and every node or potential
is node knows this), there is no need to transmit duty cycle data
between nodes.
[0094] In order to reduce the amount of information that is
exchanged between device 40 and its neighbors, device 40, and its
neighboring nodes, may each store an algorithm, which receives a
single "seed" number and which processes that seed number to
generate a frequency hopping sequence. For example, device 40 may
receive a seed number from neighboring node 42 and process that
seed number to determine the sequence of frequencies through which
node 42 hops to receive data As a result, only a single number is
transmitted instead of a whole sequence, thereby reducing network
traffic.
[0095] In this regard, the frequency hopping channel should be
selected from a specified set of channels as randomly as possible
with a uniform distribution so that, on average, the device will
spend about the same amount of time on each channel. A calculation
may be performed iteratively to produce a new random number from
each iteration. A seed number will generate the first random
number, which corresponds to the first frequency channel. This
first random number is fed into the calculation as the new seed
number, which then generates a second random number, which
corresponds to the second frequency channel. This process can
continue indefinitely to select new channel during frequency
hopping. In one implementation, the calculation is used to generate
a random number that is 32-bits long in a binary representation. In
this implementation, a single number between 1 and 16 is generated,
and only the last 4-bits of this 32-bit random number are used. (In
decimal representation the 32-bit long number can be divided by 16
and the remainder used--a modulus operation.) The 32-bit long
random number is used as the new seed number I, every iteration so
that it is less likely that the random sequence becomes
periodical.
[0096] Numerous processes may be used to produce sequence of random
numbers based on a seed number. The following code is an example
from the book "The C Programming Language," by Kernighan and
Ritchie for generating a sequence from a seed number:
TABLE-US-00001 int rand( ) { random_seed = random_seed * 1103515245
+12345; return (unsigned int)(random_seed / 65536) % 32768; }
This function will return a random number between 0 to 32767,
depending on the seed number. If the total number of channels is
16, it is possible to use the least significant four bits to
generate a random number between 0 and 15. As described above, the
first random number obtained from the seed number can replace the
seed number in the foregoing process, which is then used to
generate a next random number. Via this method, it is possible
generate a full frequency hopping sequence of 16 channels using one
seed number.
[0097] There may be a situation that multiple nodes powers-up
exactly at the same time. For example, if the nodes are
line-powered, and line power is restored after a power outage, all
nodes connected to the same line power will turn on at the same
time. In this case, there is a chance that the wake-up times and
frequency channels may be exactly the same for all nodes. Although
this scenario may be acceptable for certain applications, this is
typically undesirable because it counteracts the asynchronous
behavior of the proposed frequency hopping protocol to increase the
overall communication capacity of the network. So, whenever a node
powers up, the node may have a certain amount of random "dead time"
before the node starts running. This way, it is unlikely that many
nodes will share the same wake-up schedule and frequency
channel.
[0098] It is advantageous if all neighboring nodes occupy different
timing slots during frequency hopping. For example, frequency
hopping can become more difficult when multiple listening slots of
neighboring nodes are close. This is because some pre- and
post-processing time may be necessary to send a packet to one node.
For example, if node A's wake-up schedule is too close to that of
node B, a neighboring node may have difficulty sending data packets
to both nodes A and B. After a neighboring node (e.g., node C)
sends a data packet to node A, it may need processing time to
complete the transmission and to prepare a next data packet to send
to node B. If the listening slots of node A and node B are too
close in time, node C may not have enough time to prepare for the
data transmission to node B. Furthermore, if two listening slots
are too close together, it may not be practical for node C to enter
its dormant mode and then wake up again in the short time between
slots. In this case, it may be better for the node C to remain
awake after sending the data packet to node A, and to wait for the
listening slot of node B while awake.
[0099] One way to reduce conflict caused by close proximity of
listening slots of multiple neighboring nodes is to adjust the
initial wake-up schedules of those nodes following their activation
(power-up). For example, when anode powers-up, it will send a
search packet and collect information about its neighboring nodes,
such as wake-up times, wake-up durations, and channel sequences.
Accordingly, a node that is entering the network may select its own
wake-up time after collecting information from neighboring nodes,
and choose a wake-up time that reduces (e.g., minimizes) conflicts
with neighboring nodes.
[0100] A device can consume a relatively large amount of power due
to the length of its preamble. As a result, it is advantageous to
reduce the amount of times a device must initiate communication
using the preamble. One way of doing this is to "piggy-back" node
frequency hopping (and other) data. Referring to FIG. 2, node 42
stores the frequency hopping data for its neighbors, including
those that are out of the transmission range of device 40, such as
node 28. Accordingly, when transferring its frequency hopping data
during the listening mode of device 40, node 42 may also transfer,
to device 40, the frequency hopping data of node 28, along with the
identity of node 28 (which device 40 then stores in memory). Thus,
device 40 will also receive the frequency hopping data of node 28.
This is particularly advantageous if device 40 or node 28 can come
within transmission range of each other (e.g., if one or both is
mobile). This concept can be extended. For example, node 28 stores
the frequency hopping data for its neighbors, including those that
are out of the transmission range of node 42, such as node 26. Node
28 may transmit this frequency hopping data to node 42, along with
its own frequency hopping data. Node 42 may then transfer, to
device 40, the frequency hopping data of two other nodes, one of
which (node 26) is outside the transmission range of node 42.
[0101] The use of preambles may be extended to normal node
communication in order to increase the duty cycle of network node
(e.g., by increasing the waking durations of the nodes). Referring
to FIGS. 1 and 8, if the receiving node (e.g., node 15) has a
relatively short waking duration, it may be difficult for the
sending node (e.g., node 12) to match its transmission to that
short waking duration. For example, a relatively short time shift
in the internal clock of the sending node may effectively prevent
the sending node from communicating with the receiving node (if the
receiving node's waking duration is sufficiently short). To address
this problem, the sending node may output a relatively short
preamble (e.g., 5 ms or 10 ms) in the appropriate channel prior to
the expected waking duration of the receiving node.
[0102] In the example shown in FIG. 8, the sending node outputs a
preamble 70 in channel 6 following the receiver's waking time in
channel 10. The receiving node detects the preamble during its
receiving duration in channel 6 and extends its receiving duration
71, as shown. During this extended receiving duration, sending node
12 sends a data packet 72 to the receiving node in channel (Ch) 6.
When the receiving node detects an "end-of-frame" indication in the
data packet, the receiving node ends the extended receiving
duration 71 at 68a. Thereafter, the sending node goes into
receiving mode 74 to receive an ack packet 75 from the receiving
device. It is noted, however, that the internal channel hopping
clocks of both devices continue to run so that the devices can
resume their frequency hopping sequence 76 (after the extended
receiving duration) as if that sequence were never interrupted.
[0103] An alternative to the timing shown in FIG. 8 is to make a
whole data packet fit into a single MN_sleep_time of the receiving
device. The preamble is thus long enough to cover an MN_RX_time.
Accordingly, the frequency hopping sequence need not be
disturbed.
[0104] As noted above, in 900 MHz radio operation, the FCC limits
communication in a single channel to 400 ms with a narrow-band
radio. There is no limit on the duration of a continuous
communication in a single channel if DSSS (Direct Sequence Spread
Spectrum) is used. However, there are very few 900 MHz commercial
radios that have this capability. By contrast, in 2.4 GHz
operation, there are many standard-based radios (e.g. IEEE
IEEE802.15.4) available with DSSS capabilities. Therefore, if a
preamble is to be longer than 400 ms during 900 MHz operation, then
the preamble may be transmitted in multiple frequencies. This is
referred to herein as "channel grouping".
[0105] As noted above, there is an FCC regulation that prohibits a
multi-channel narrow-band (e.g., 900 MHz) radio from staying in one
frequency channel for longer than 400 ms using FHSS (Frequency
Hopping Spread Spectrum). Accordingly, in this limited
circumstance, it is not legal for a preamble to stay in one channel
for more than 400 ms. To deal with this issue, a preamble that is
longer than 400 ms may be divided into multiple sections and
transmitted in different frequency channels in the 900 MHz
spectrum. For example, a 1.5 s preamble may be split into four (or
more) segments, which may be of equal or non-equal length. When one
segment, of the preamble completes transmission in a first
frequency channel, a node transmitting the preamble changes the
frequency channel and continues transmission of the preamble in a
new frequency channel. This is done until the entire 1.5 s preamble
is transmitted, without violating FCC regulations.
[0106] For the reasons explained above, at least one listening slot
of each neighboring node should be covered by the preamble, even
though the preamble is being transmitted multiple channels. To
achieve this, a process called "channel grouping" may be used.
[0107] By way of example, a preamble may be transmitted in N
different channels so that each channel includes no more than 400
ms of preamble. In this case, existing frequency channels may be
divided into N groups. For example, if the preamble 1.4 s, it may
be divided into four segments. In this case, N=4. If there are 16
channels available, the channels may be divided into four groups
with four channels per group. In another example, if the preamble
is 700 ms, the preamble may be divided into two segments, and there
may be two channels per group (assuming eight channels are
available).
[0108] While a node is hopping frequency channels to listen for
communications, the channel sequence of the node follows a group
order. In other words, once the node starts hopping in a channel
that is in a first group, group 1, a next hopping channel may be
selected from the same group 1 until group 1 runs out of channels.
After the node has hopped through all channels in group 1, the node
starts hopping among the channels in a second group, group 2, and
so on. For example, group 1 may include channel 1 to channel 10 and
group 2 may include channel 11 to channel 20. In this case, the
sequence should be defined so that channels 1 through 10 are
selected first, followed by channels 11 to 20.
[0109] The preamble, however, hops only inside one group. For
example, if the node transmitting the preamble selects group 2, the
preamble should hop only inside group 2. (e.g., among channel 11 to
20). Since the preamble include N segments, each group should
include at least N channels. This means that total number of
available RF channels should be larger than N.times.N. This is
because a preamble with N segments needs N groups of channels, and
each group needs to include minimum of N channels.
[0110] Using the foregoing process, there is a good chance that a
preamble, although transmitted in multiple channels, can be
captured by at least one listening channel of a neighboring node.
But, there is a non-negligible chance (approximately (N-1)/total
number of channels) that the preamble will be missed by one or more
neighboring nodes. However, the probability of a neighboring node
failing to capture the preamble decreases as the total number of
channels increases. For example, if the preamble is at least 1.5 s,
the preamble will hop through at least 4 channels (N=4) so that
each channel is occupied no more than 400 ms. Hence, the total
number of available channels should be greater than or equal to 16
(N.times.N=4.times.4=16). Of course, this may only apply when each
channel is a narrow-band (e.g., 900 MHz) channel. In the case of
2.4 GHz DSSS radios, such as IEEE802.15.4 standard-based radios,
channel grouping is not necessary, but may also be used.
[0111] One exemplary implementation of the frequency hopping
protocol uses the CC2420 transceiver from Chipcon Products. As
described on the Chipcon Products Web site, "[t]he CC2420 is a
low-cost transceiver designed specifically for low-power,
low-voltage RF applications in the 2.4 GHz unlicensed ISM band." In
this example, a whole packet (not just a byte) constitutes a unit
of the preamble. In this case, each preamble unit is approximately
1 ms. Accordingly MN_RX_time is at least 2.times. the preamble unit
size, or 2 ms. To achieve 100.times. power savings, MN_sleep_time
should be about 200 ms. The CC2420 provides sixteen channels. The
minimum preamble length that can cover all sixteen channels is 3.2
s (200 ms*16 channels). Since the CC2420 operates in the 2.4 GHz
range with DSSS capability, there is no limit on the maximum time
the CC2420 can stay in one channel. So, the preamble can stay in
the same channel for any length of time. Alternatively, the
preamble may be split among two to four channels. That is, the
preamble may be transmitted in different frequency channels.
[0112] For ordinary packet transmission, where there is one known
packet destination, if the time synchronism is sufficiently
accurate (accuracy of .about.1 ms), the start of a data packet can
be fit into the MN_RX_time (.about.2 ms) window of a receiving
device. The SFD will be detected by the receiving device before
MN_RX_time expires, and the MN_RX_time will be elongated until
packet reception is complete. If the time synchronism is less
accurate or the SFD is not available from the radio, several short
preamble packets (e.g., IEEE802.15.4 packets) may precede the real
data packet. This stream of short preamble packets may include five
to ten short IEEE802.15.4 packets (.about.1 ms/packet). The result
will be a preamble of about 10 ms, which will enable the 2 ms
MN_RX_time window to capture at least one preamble packet even if
there is .+-.5 ms clock drift. Once the receiver identifies the
preamble packet(s) in the MN_RX_time window, the receiver will
continue listening in the same frequency Channel (as described
above) in order to receive the real (i.e., non-preamble) data
packet. In this example, ack packets are sent on a next hop channel
after regular packet transmission. Alternately, ack packets may be
sent on the same channel as regular data packet transmission.
[0113] Clock drift may be an issue when the interval between packet
exchanges becomes quite long. To alleviate clock drift, each node
of a wireless network sends a local "heartbeat packet" to each of
its parent nodes (primary, and secondary, and so on) in a given
interval. For example, if the clocks of neighboring nodes must be
adjusted every 40 seconds to keep the drift in the acceptable
range, the local heartbeat may be generated, and sent to each
parent node, at least every 40 seconds. If a data packet is sent
out to a parent nod; the data packet may act as a replacement for
the local heartbeat packet.
[0114] The frequency hopping protocol may be used when a device
enters into a network and when a device broadcasts Over the
network. That is, during broadcast, a device needs to get the
attention of all of its neighbors. Since the neighbors may be in
different frequencies, sending a preamble and proceeding in the
manner described above enables a device to broadcast information to
its neighbors, who then may propagate that information to their
neighbors using the frequency hopping protocol, and so on until
every node on the network has received the information.
Network Device
[0115] Examples of network devices that may be used as nodes of
wireless network 10, and that may implement the processes described
herein, are described in U.S. Pat. No. 6,804,790, which issued on
Oct. 12, 2004, the contents of which are hereby incorporated by
reference into this application. FIGS. 9A to 9C show block diagrams
of one example of a network device 80 that may implement the
processes described herein. Network device 80 is a self-contained,
miniaturized computer. As shown in FIGS. 9A and 9B, network device
80 includes first processing unit 82, RF transceiver 84, second
processing unit 86, low clock frequency crystal 88, high clock
frequency crystal 90, and I/O connector 92, all mounted on circuit
board 94. As shown in FIG. 9C, a power source 96, such as a
battery, may be attached to the back of circuit board 94. A memory
containing instructions to be executed by each processing unit may
be included inside each processing unit or one or more such
memories (not shown) may be mounted on circuit board 94.
[0116] The small size and low power consumption of network device
80 allows network device 80 to operate from battery 90. In this
implementation, first processing unit 82 operates at a clock
frequency of 32 kHz, and second processing unit 86 operates at a
clock frequency of 4 MHz. A coordinating protocol operates so that
network device 80 may perform signal processing and RF transmission
with increased power efficiency.
[0117] The coordinating protocol is used to control the operation
of network device 80 by assigning tasks and operations to the
processing units based upon the speed required to perform a given
task of function. The coordinating protocol is designed to assign
tasks to the various processing units with the result being
increased power efficiency on network device 12. For example, the
coordinating protocol will allow CPU 82 to assign a given task or
operation (such as joining or establishing a presence in wireless
network 10) to itself or to CPU 86 based upon the speed
requirements of the task or operation and the clock, frequencies of
the processing units. Tasks and operations which require lower
clock frequencies will be assigned to CPU 86 with the lower clock
frequency. Because CPU 86 operates at lower clock frequency, the
power efficiency of the system as a whole is increased. When the
task load of the system is low enough, the CPUs may be shut-off or
placed into low-power mode to further increase the power efficiency
of the system.
Other Implementations
[0118] The processes described herein including, but not limited
to, the frequency hopping protocol (hereinafter referred to
collectively as "the processes") may find applicability in any
computing or processing environment. The processes may be
implemented using hardware, software, or a combination thereof. The
processes are not limited to use with the hardware and software
described herein; they may find applicability in any computing,
processing or networking environment, and with any type of machine
that is capable of executing machine-readable instructions.
[0119] The processes may be implemented using digital electronic
circuitry, or in computer hardware, firmware, software, or in
combinations thereof. The processes can be implemented via a
computer program product, i.e., a computer program tangibly
embodied in an information carrier, e.g., in one or more
machine-readable storage devices/media or in a propagated-signal,
for execution by, or to control the operation of, one or more data
processing apparatus, e.g., a programmable processor, a computer,
or multiple computers. A computer program can be written in any
form of programming language, including compiled or interpreted
languages, and it can be deployed in any form, including as a
stand-alone program or as a module, component, subroutine, or other
unit suitable for use in a computing environment. A computer
program can be deployed to be executed on one computer or on
multiple computers at one site or distributed across multiple sites
and interconnected by a communication network.
[0120] Actions performed by the processes can be performed by one
or more programmable processors executing one or more computer
programs to perform the functions of the processes. The actions can
also be performed by, and the processes can be implemented via,
special purpose logic circuitry, e.g., one or more FPGAs (field
programmable gate array) or ASICs (application-specific integrated
circuit).
[0121] Processors suitable for execution of a computer program
include, e.g., both general and special purpose microprocessors,
and any one or more processors of any kind of digital computer.
Generally, a processor will receive instructions and data from a
read-only memory or a random access memory or both. Elements of a
computer include a processor for executing instructions and one or
more memory devices for storing instructions and data. Generally, a
computer will also include, or be operatively coupled to receive
data from, or transfer data to, or both, one or more mass storage
devices for storing data, e.g., magnetic, magneto-optical disks, or
optical disks. Information carriers suitable for embodying computer
program instructions and data include all forms of non-volatile
memory, including by way of example, semiconductor memory devices,
e.g., EPROM, EEPROM, and flash memory devices; magnetic disks,
e.g., internal hard disks or removable disks; magneto-optical
disks; and CD-ROM and DVD-ROM disks. The processor and the memory
can be supplemented by, or incorporated in, special purpose to
logic circuitry.
[0122] The processes can be implemented via a computing system that
includes one or more back-end components, e.g., a data server, or
that includes one or more middleware components, e.g., an
application server, or that includes one or more front-end
components, e.g., a client computer. The components of the system
can be interconnected by any form or medium of digital data
communication, e.g., a communication network. Examples of
communication networks include a local area network ("LAN") and a
wide area network (WAN"), e.g., the Internet.
[0123] The computing system can include clients and servers. A
client and server are generally remote from each other and
typically interact through a communication network. The
relationship of client and server arises by virtue of computer
programs running on respective computers and having a client-server
relationship to each other.
[0124] The processes are not limited to the implementations
described. For example, the processes can be used with network
devices other than those shown in 9A to 9C. Any computer, router,
server, wireless device, or similar machine may implement the
processes. The processes can be used on homogeneous networks as
well. The processes can be used with networks having configurations
other than those shown in FIGS. 1 and 2, including networks that
have both wired and wireless portions. The processes are not
limited to use with the protocols and data transmission methods
described herein, but rather are universally adaptable.
[0125] The processing described herein may be performed in a
different order and/or portions thereof may be omitted. As such,
operations performed in furtherance of the processes are not
limited to the flows described herein.
[0126] Elements of different implementations may be combined to
form another implementation not specifically set forth above. Other
implementations not specifically described herein are also within
the scope of the following claims.
* * * * *