U.S. patent application number 11/499120 was filed with the patent office on 2008-02-07 for embedded parametric monitoring of optoelectronic modules.
Invention is credited to Clay E. Hudgins.
Application Number | 20080031576 11/499120 |
Document ID | / |
Family ID | 39029257 |
Filed Date | 2008-02-07 |
United States Patent
Application |
20080031576 |
Kind Code |
A1 |
Hudgins; Clay E. |
February 7, 2008 |
Embedded parametric monitoring of optoelectronic modules
Abstract
An optoelectronic module for converting and coupling an
information-containing electrical signal with an optical fiber
including a housing having an electrical input for coupling with an
external electrical cable or information system device and for
transmitting and receiving information-containing electrical
signals over such input, and a fiber optic connector adapted for
coupling with an external optical fiber for transmitting and
receiving an optical signal; an electro-optic subassembly coupled
to the information containing electrical signal and converting it
to and/or from a modulated optical signal corresponding to the
electrical signal; and parametric storage means disposed in said
housing for storing environmental and/or operational data
associated with the module.
Inventors: |
Hudgins; Clay E.; (Edgewood,
NM) |
Correspondence
Address: |
EMCORE CORPORATION
1600 EUBANK BLVD, S.E.
ALBUQUERQUE
NM
87123
US
|
Family ID: |
39029257 |
Appl. No.: |
11/499120 |
Filed: |
August 4, 2006 |
Current U.S.
Class: |
385/92 |
Current CPC
Class: |
H04B 10/075 20130101;
G02B 6/4246 20130101; G02B 6/4292 20130101 |
Class at
Publication: |
385/92 |
International
Class: |
G02B 6/36 20060101
G02B006/36 |
Claims
1. An optical transmitter for converting and coupling an
information-containing electrical signal with an optical fiber
comprising: a housing including an electrical input for coupling
with an external electrical cable or information system device and
for transmitting information-containing electrical signals over
said input, and a fiber optical connector adapted for coupling with
an external optic fiber for transmitting an optical signal; an
electro-optic subassembly disposed in said housing coupled to the
information containing electrical signal and converting it to a
modulated optical signal corresponding to the electrical signal at
a predetermined wavelength; and parametric storage means disposed
in said housing for storing operational data associated with the
transmitter.
2. The optical transmitter of claim 1, further comprising means in
the housing for recording and storing the total operating time of
the optical transmitter.
3. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing a temperature of the
housing.
4. The optical transmitter of claim 3, further comprising means
within the housing for recording and storing a time rate of change
of the temperature of the housing.
5. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing a temperature of a
laser or light emitting diode within the housing.
6. The optical transmitter of claim 5, further comprising means
within the housing for recording and storing a time rate of change
of the temperature of the laser or light emitting diode.
7. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing a laser bias current
and/or a digital setting that controls the bias current.
8. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing a laser modulation
current and/or a digital setting that controls the modulation
current.
9. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing a light emitting diode
drive current and/or a digital setting that controls the drive
current.
10. The optical transmitter of claim 2, further comprising means
within the housing for recording and storing indications of bit
errors on incoming electrical signals that simulate an optical
transmission function.
11. An optical receiver for converting and coupling an optical
signal from an optical fiber to an information-containing
electrical signal comprising: a housing including an optical input
for coupling with an external fiber optic cable and information
system device and for transmitting information-containing optical
signals received over said input, and an electrical connector
adapted for coupling with an external electrical cable for
transmitting an electrical signal; an electro-optic subassembly
disposed in said housing coupled to the information containing
optical signal and converting it to an electrical signal; and
parametric storage means disposed in said housing for storing
operational data associated with receiver.
12. The optical receiver of claim 11, further comprising means
within the housing for recording and storing the total operating
time of the optical receiver.
13. The optical receiver of claim 12, further comprising means
within the housing for recording and storing bit errors detected on
the external fiber optic cable.
14. The optical receiver of claim 12, further comprising means
within the housing for recording and storing a power level of
optical signals detected on the external fiber optic cable.
15. An optical transceiver for converting and coupling an
information-containing electrical signal with an optical fiber
comprising: a housing including an electrical input for coupling
with an external electrical cable or information system device and
for transmitting and receiving information-containing electrical
signals over said input, and a fiber optic connector adapted for
coupling with an external optical fiber for transmitting and
receiving an optical signal; an electro-optic subassembly disposed
in said housing coupled to the information containing electrical
signal and converting it to or from a modulated optical signal
corresponding to the electrical signal; and parametric storage
means disposed in said housing for storing operational data
associated with the transceiver.
16. The optical transceiver of claim 15, further comprising means
in the housing for recording and storing the total operating time
of the optical transceiver.
17. The optical transceiver of claim 16, further comprising means
within the housing for recording and storing the temperature of the
housing at periodic intervals.
18. The optical transmitter of claim 16, further comprising means
within the housing for recording and storing the time rate of
change of the temperature of the housing when the rate of change
exceeds a predetermined threshold.
19. The optical transmitter of claim 16, further comprising means
within the housing for recording and storing a temperature of the
light source within the housing.
20. The optical transmitter of claim 19, further comprising means
within the housing for recording and storing a time rate of change
of the temperature of the laser or light emitting diode within the
housing.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The invention relates to optical communications devices,
such as transmitters, receivers, and transceivers used in high
throughput fiber optic communications links in local and wide area
networks and storage networks, and in particular to parametric
monitoring of the performance of such devices.
[0003] 2. Description of the Related Art
[0004] Communications networks have experienced dramatic growth in
data transmission traffic in recent years due to worldwide Internet
access, e-mail, and e-commerce. As Internet usage grows to include
transmission of larger data files, including content such as full
motion video on-demand (including HDTV), multi-channel high quality
audio, online video conferencing, image transfer, and other
broadband applications, the delivery of such data will place a
greater demand on available bandwidth. The bulk of this traffic is
already routed through the optical networking infrastructure used
by local and long distance carriers, as well as Internet service
providers. Since optical fiber offers substantially greater
bandwidth capacity, is less error prone, and is easier to
administer than conventional copper wire technologies, it is not
surprising to see increased deployment of optical fiber in data
centers, storage area networks, and enterprise computer networks
for short range network unit to network unit interconnection.
[0005] Such increased deployment has created a demand for
electrical and optical transceiver modules that enable data system
units such as computers, storage units, routers, and similar
devices to be optionally coupled by either ran electrical cable or
an optical fiber to provide a high speed, short reach (less than 50
meters) data link within the data center.
[0006] A variety of optical transceiver modules are known in the
art to provide such interconnection that include an optical
transmit portion that converts an electrical signal into a
modulated light beam that is coupled to a first optical fiber, and
a receive portion that receives a second optical signal from a
second optical fiber and converts it into an electrical signal. The
electrical signals are transferred in both directions over an
electrical connectors that interface with the network unit using a
standard electrical data link protocol.
[0007] The optical transmitter section includes one or more
semiconductor lasers and an optical assembly to focus or direct the
light from the lasers into an optical fiber, which in turn, is
connected to a receptacle or connector on the transceiver to allow
an external optical fiber to be connected thereto using a standard
SC, FC or LC connector. The semiconductor lasers are typically
packaged in a hermetically sealed can or similar housing in order
to protect the laser from humidity or other harsh environmental
conditions. The semiconductor laser chip is typically a distributed
feedback (DFB) laser with dimensions a few hundred microns to a
couple of millimeters wide and 100-500 microns thick. The package
in which they are mounted typically includes a heat sink or
spreader, and has several electrical leads coming out of the
package to provide power and signal inputs to the laser chips. The
electrical leads are then soldered to the circuit board in the
optical transceiver. The optical receive section includes an
optical assembly to focus or direct the light from the optical
fiber onto a photodetector, which in turn, is connected to a
transimpedance amplifier/limiter circuit on a circuit board. The
photodetector or photodiode it typically packaged in a hermetically
sealed package in order to protect it from harsh environmental
conditions. The photodiodes are semiconductor chips that are
typically a few hundred microns to a couple of millimeters wide and
100-500 microns thick. The package in which they are mounted is
typically from three to six millimeters in diameter, and two to
five millimeters tall and has several electrical leads coming out
of the package. These electrical leads are then soldered to the
circuit board containing the amplifier/limiter and other circuits
for processing the electrical signal.
[0008] Optical transceiver modules are therefore packaged in a
number of standard form factors which are "hot pluggable" into a
rack mounted line card network unit or the chassis of the data
system unit. Standard form factors set forth in Multiple Source
Agreements provide standardized dimensions and input/output
interfaces that allow devices from different manufacturers to be
used interchangeably. Some of the most popular MSAs include XENPAK
(see www.xenpak.org), X2 (see www.X2 msa.org), SFF ("small form
factor"), SFP ("small form factor pluggable"), XFP ("10 Gigabit
Small Form Factor Pluggable", see www.XFPMSA.org), and the 300-pin
module (see www.300pinmsa.org).
[0009] Customers and users of modules are interested in such
miniaturized transceivers in order to increase the number of
interconnections or port density associated with the network unit,
such as, for example in rack mounted line cards, switch boxes,
cabling patch panels, wiring closets, and computer I/O
interfaces.
[0010] The reliability of such modules is an important
consideration in the selection of specific designs for many
commercial applications. When a module fails, users typically wish
to know why. This interest is not a mater of idle curiosity; it is
based on a generally accepted belief that knowing the root cause of
a failure is an essential step in preventing recurrence of that
failure. Thus, producers of optical modules expend much effort in
failure analysis to determine the root cause of any failures that
occur and to subsequently improve products to eliminate similar
failures in the future. Such failure analysis is often very labor
intensive, requiring technologists referred to as reliability
analysts. Such failure analysis often requires an extensive suite
of test equipment, which also represents a significant expense to
the manufacturer.
[0011] The failure analysis process is relatively expensive to the
manufacturer, and consequently one has an interest in conducting
the analysis as efficiently as possible. Additionally, sometimes
the failure is covered by a warranty, and honoring the warranty
represents an expense to the manufacturer.
[0012] Conversely, sometimes the failure is a result of unduly
harsh use by the user outside of normal environmental conditions,
or some other circumstance that voids the warranty (such as use
with incompatible equipment), and proving that the terms of the
warranty have been violated can enable the manufacturing to avoid
expense, particularly when the user chooses to return large
quantities of product for purported warranty service.
[0013] If more information regarding the actual operational history
of the optical module could be made available during the failure
analysis, then determining the validity of the warranty status
would be made more straightforward, and aspects of the failure
analysis would be simpler. Failure analysis involves a search for
information about the causes and circumstances of product failure,
often by exploiting very subtle clues, and any information that can
be made explicitly available to the reliability analyst has the
potential to make their task more productive, effective, and
efficient.
[0014] This process of failure analysis is extendable to cases in
which no failure has occurred. In this case, a reliability analyst
may analyze a used optical module to observe parametric shifts in
performance, and the information gained may be used to define
design and manufacturing process improvements to make the product
even more reliable in the future. Because of the importance of
reducing failures and parametric shifts in product performance, a
need exists for better methods of tracking the environmental and
operational conditions under which an optoelectronic module or
product is used in the field, i.e. in actual customer
installation.
SUMMARY OF THE INVENTION
1. Objects of the Invention
[0015] It is an object of the present invention to provide an
optoelectronic module in a small pluggable standardized form factor
with an embedded parametric monitoring circuit.
[0016] It is also another object of the present invention to
provide a module for use in an optical fiber transmission system
with means for storing total elapsed operating time, temperature,
temperature change, and data transmission error rate.
[0017] It is still another object of the present invention to
provide an optical transceiver with a storage unit for collecting
and storing environmental and operational data at periodic time
intervals.
[0018] It is still another object of the present invention to
provide a feature in an optoelectronic module for storing and
reporting excursions from normal environmental and operational
conditions.
[0019] It is also another object of the present invention to
provide an optical transceiver for use in an optical transmission
system with electrical and mechanical sensors to detect, record and
report abnormal operating conditions, and
2. Features of the Invention
[0020] Briefly, and in general terms, the present invention
provides an optical transceiver for converting and coupling an
information-containing electrical signal with an optical fiber
including a housing including an electrical connector for coupling
with an external electrical cable or information system device and
a fiber optic connector adapted for coupling with an external
optical fiber, at least one electro-optical subassembly in the
housing for converting between an information containing electrical
signal and a modulated optical signal corresponding to the
electrical signal including either a transmitter or a receiver, or
both, and means for detecting and recording environmental and/or
operation data associated with the module.
[0021] Additional objects, advantages, and novel features of the
present invention will become apparent to those skilled in the art
form this disclosure, including the following detailed description
as well as by practice of the invention. While the invention is
described below with reference to preferred embodiments, it should
be understood that the invention is not limited thereto. Thos of
ordinary skill in the art having access to the teachings herein
will recognize additional applications, modifications and
embodiments in other fields, which are within the scope of the
invention as disclosed and claimed herein and with respect to which
the invention could be of utility.
BRIEF DESCRIPTION OF THE DRAWINGS
[0022] These and other features and advantages of this invention
will be better understood and more fully appreciated by reference
to the following detailed description when considered in
conjunction with the accompanying drawings, wherein:
[0023] FIG. 1 is an exploded perspective view of an optoelectronic
module in accordance with an illustrated embodiment of the
invention.
[0024] Additional objects, advantages, and novel features of the
present invention will become apparent to those skilled in the art
from this disclosure, including the following detailed description
as well as by practice of the invention. While the invention is
described below with reference to preferred embodiments, it should
be understood that the invention is not limited thereto. Those of
ordinary skill in the art having access to the teachings herein
will recognize additional applications, modifications and
embodiments in other fields, which are within the scope of the
invention as disclosed and claimed herein and with respect to which
the invention could be of utility.
DESCRIPTION OF THE PREFERRED EMBODIMENT
[0025] Details of the present invention will now be described
including exemplary aspects and embodiments thereof. Referring to
the drawings and the following description, like reference numbers
are used to identify like or functionally similar elements, and are
intended to illustrate major features of exemplary embodiments in a
highly simplified diagrammatic manner. Moreover, the drawings are
not intended to depict every feature of the actual embodiment nor
the relative dimensions of the depicted elements, and are not drawn
to scale.
[0026] The present invention relates generally to parametric
monitoring for optical communications modules such as transmitters,
receivers, and transceivers used in fiber optic communications
systems.
[0027] Referring now to FIG. 1, there is shown an exploded view of
an exemplary pluggable optical transceiver module 100 according to
a preferred embodiment of the present invention. In this particular
embodiment, the module 100 is compliant with the IEEE 802.3ae
10GBASE-LX4 Physical Media Dependent sub-layer (PMD) and is
implemented in the XFP form factor having a length of 78 mm, a
width of 18.35 mm, and a height of 8.5 mm. It is to be noted,
however, that in other embodiments the transceiver module 100 may
be configured to operate under various other standard protocols
(such as Fibre Channel or SONET) and be manufactured in various
alternate form factors such as XENPAK, X2, etc. The module 100 is
preferably a 10 Gigabit Wide Wavelength Division Multiplexed (WWDM)
transceiver having four 3.125 Gbps distributed feedback lasers that
enable 300 meter transmission of an optical signal at least 300
meters over a single legacy installed multimode fiber or a distance
from 10 to 40 km over a single standard single mode fiber.
[0028] The transceiver module 100 includes a two-piece housing 102
including a base 104 and a cover 106. In addition, contact strips
(not shown) may be provided to ground the module to an external
chassis ground as well. The housing 102 is constructed of die-case
or milled metal, preferably die-cast zinc, although other materials
also may be used, such as specialty plastics and the like.
Preferably, the particular material used in the housing
construction assists in reducing EMI.
[0029] The front end of the housing 102 includes a faceplate 131
for securing a pair of receptacles 124, 126. The receptacles, 124,
126 are configured to receive fiber optic connectors (not shown)
which mate with optical plugs 128, 130 respectively. In the
preferred embodiment, the connector receptacles 124, 126 are
configured to receive industry standard LC duplex connectors. As
such, keying channels 132, 134 are provided to ensure that the LC
connectors are inserted into the receptacles 124, 126 in their
correct orientation. Further, as shown in the exemplary embodiment
and discussed further herein, the connector receptacle 124 is
intended for an LC transmitter connector, and the connector
receptacle 126 receives an LC receiver connector.
[0030] In one embodiment, the housing 102 holds three subassemblies
or circuit boards, including a transmit board 108, a receive board
110, and a physical coding sublayer (PCS)/physical medium
attachment (PMA) board 112, which is used to provide an electrical
interface to external computer or communications units (not shown).
The transmit subassembly includes four distributed feedback (DFB)
semiconductor lasers mounted which may be mounted in a single,
hermetically sealed enclosure 415, which interfaces to a fiber
coupling subassembly 416. The transmit board 108 is secured in
place at the bottom of the housing using a brace 418 attached to
the coupling subassembly 416. The brace also functions as a heat
sink for dissipating heat from the metallic fiber coupling
subassembly 416. In addition, the transmit board 108 and receive
board 110 are connected to the PCS/PMA board 112 by respective flex
interconnects 120, or other board-to-board electrical connectors or
cables. Thermally conductive gap pads may be provided to transmit
the heat generated by the lasers or other components in the
transmitter subassembly to the base 104 or cover 106 of the
housing, which acts as a heat sink. The receiver subassembly 110 is
directly mounted on the housing base 104 using a thermally
conductive adhesive to achieve heat dissipation. Different
subassemblies therefore dissipate heat to different portions of the
housing for a more uniform heat dissipation. The output optical
signal from the four lasers is multiplexed and input into a single
optical fiber 420 which coils and reverses direction, and is
preferably attached or mounted on a flexible substrate 140. The
flexible material may be an optical flexible planar material such
as FlexPlane.TM. available from Molex, Inc. of Lisle, Ill.,
although other flexible substrate may be used as well. The optical
fiber 420 originating from the transmitter subassembly is thereby
routed to the transmit optical connector plug 130, which is
attached to the housing 102. The fiber is routed and attached in
such a manner as to minimize sharp bends in the optical fiber to
avoid optical loss and mechanical failure.
[0031] The flexible substrate 140 may include an opening 142 or
hole in a portion of the material that is located directly above
the retimer IC or other heat generating components mounted on the
PCS/PMA board 112. The opening 142, which is substantially an area
the size of the unused portion of the substrate 140, enables the
heat sink on the cover 106 to contact a heat transmission gap pad
160, so as to provide access and a heat conductive path to the
mounted components on the board 112. This area on the board 112
normally would be inaccessible if not for the opening 142. For
example, a heat sink may be installed without interfering with the
routing of the optical fibers on the substrate 140 and without
removing the mounted substrate 140 to allow access to the PCS/PMA
board 112.
[0032] Although the embodiment described above is a pluggable 10
Gigabit WWDM transceiver, the same principles are applicable in
other types of optical transceivers suitable for operating over
both multimode (MM) and single mode (SM) fiber using single or
multiple laser light sources, single or multiple photodetectors,
and an appropriate optical multiplexing and demultiplexing system.
The design is also applicable to a single transmitter or receiver
module, or a module as either a transmitter, receiver, or
transceiver to communicate over different optical networks using
multiple protocols and satisfying a variety of different range and
distance goals.
[0033] Although in the depicted embodiment, the transceiver 100 is
manufactured in a modular manner using three separate subassemblies
mounted in the housing--a transmitter subassembly, a receiver
subassembly, and a protocol processing board, with each subassembly
or board having dedicated functions and electrically connected to
each other using either flex circuitry or mating multipin
connectors, land grid arrays, or other electrical interconnect
devices, the invention may also be implemented in a transceiver
having a single board or subassembly mounted inside the
housing.
[0034] The modern optical transceiver or module, such as that of
FIG. 1, as well as those complying with the XENPAK, X2, XPAK, or
300 pin form factors, possesses underutilized resources for
collecting information and nonvolatile memory in which to store it.
Optical modules 100 typically employ synchronous digital circuitry,
which gives them the capability to measure Elapsed Run Time (ERT)
in and Elapsed Run Time Meter (ERTM) as has long been done in the
field of industrial-grade internal combustion engines. The present
invention implements an ERTM in a digital logic circuit 209 or
other means within the housing of the module 100 for recording and
storing the total operating time of the optical module without a
microprocessor or alternately may implement the ERTM using a
clock-counter resource within an embedded microprocessor 210,
depending on the hardware resources available within the optical
module.
[0035] The purpose of the ERTM within the module 100 is to count
and to store the number of time units the module has run since
first being built and provided to the user, and to make that
information available to the producer if the user should ever
return the module to the producer for warranty service, or for any
other reason. For the ERTM to be useful to the producer's
reliability analyst, the ERTM must by default retain the ERT value
in nonvolatile memory when power is removed from the optical
module. Furthermore, the ERTM must by default continue counting ERT
when power is once again applied to the optical module, resuming
with the previous value rather than starting over from zero. In
addition to the default behaviors of the ERTM, which enable
autonomous operation without intervention by the user, the producer
must have a means to override default ERTM behaviors for
maintenance and diagnostic purposes, including the ability to reset
the ERTM to zero, to read the ERTM's value, to set the ERTM to
non-zero values, and to freeze (i.e., halt or stop) and restart the
ERTM.
[0036] The ERTM's demands on nonvolatile storage space within the
module 100 are quite modest. For example, a one-minute resolution
meter with thirty years duration may be implemented in only 24 bits
of nonvolatile memory (i.e., 30 years.times.365.25
days/year.times.24 hours/day.times.60 minutes/hour=15,778,800
minutes. It may be noted that 15,778,800 minutes is less than the
maximum value of a 24 bit register, which may represent a maximum
of 2.sup.24-1, which equals 16,777,215).
[0037] In practice, the ERTM is not constrained to measure time in
conventional time units (e.g., hours, minutes, seconds). The ERTM
may measure elapsed time in some local timebase, which may be a
convenient multiple of the period of any synchronous digital clock
employed within the optical module or some multiple of the
signaling period of data passing through the module. (Often optical
modules recover a digital clock signal from the received
pseudorandom optical data waveform.) The necessary general purpose
processing to convert the ERTM's timebase to conventional time
units need not be present within the optical module, since general
purpose processing capability will be conveniently available to the
producer's reliability analyst.
[0038] Typically, when a user decides that a module has failed from
the user's viewpoint, the failure will be related to some parameter
of direct interest to the user, and that failure will likely not
involve the ERTM, which is of only indirect interest to the user.
When the user removes the optical module from operation, the
optical module will no longer receive the electrical power required
to operate, and so the ERTM will cease to run. And so, when the
producer receives a failed module returned by the user, the ERTM
will provide the total elapsed run time experienced by the module
up to the point of failure, which is something that otherwise would
impossible to know and difficult to estimate.
[0039] In general, a number of parameters of interest (POIs) may be
collected. For example a temperature, as well as the time rate of
change of temperature, of the module body or case may be collected.
Similarly a temperature, as well as a time rate of change, of the
laser or LED within the body or case may be collected.
[0040] Other POIs may include a laser bias or modulation current
and/or the digital setting that controls it. In the case of LEDs, a
POI may include the LED drive current and/or the digital setting
that controls it.
[0041] Error signals may also be collected and saved as POIs.
Indications of bit errors on the fiber in the optical receive
function may be retained. For example, coding errors detected in
received blocks data, such as 8b10B or 64b/66b blocks of data.
Indications of bit errors on incoming electrical signals that
stimulate the module's optical transmission function may also be
retrained as POIs.
[0042] Indications of lower optical power in the optical receive
function may be retained as a POI. For example, optical power below
the receiver's specified capability or an optical power so low as
to indicate a "dark" optical fiber, which sometimes referred to as
a loss of signal (LOS) condition.
[0043] All other things being equal, semiconductor integrated
circuits age more quickly at higher temperatures, as do organic
materials. So, the history of Module Operating Temperature (MOT)
will be of interest to the reliability analyst. Typically, some
knowledge of temperature is available within an optical module,
because, as is well known, most optical modules employ a form of
temperature compensation to adjust the drive settings of the
optical emitter, which is typically a laser diode or an LED (light
emitting diode). Either this existing resource or some additional
specialized resource may be used to measure temperature.
[0044] The knowledge of time and temperature may together be used
to measure the time-rate-of-change of temperature, which may be of
interest since rapid temperature changes tend to be more stressful
than slow changes. Additionally if desired, sensors may be added to
measure mechanical stresses such as mechanical shock or mechanical
strain. Additional Parameters Of Interest (POI) may include
internal settings specific to the optical module's design, such as
digital potentiometer settings, laser drive settings, operational
modes, et cetera.
[0045] Also of interest are the issues surrounding how to
efficiently store the measured data within the module 10. The
nonvolatile memory available in the optical module is a finite
resource, and to make the most efficient use of the memory, the
stored data should be compressed in some way. Lossless data
compression techniques are well known, and certainly may be
employed. Additionally, these data are highly suitable for lossy
compression, because the measurements are not all of equal
value.
[0046] For example, the maximum and minimum values of a given POI
(e.g., temperature, shock, etc.) represent particularly valuable
information, and dedicated storage locations may be allocated for
those values. The time values at which such maxima and minima occur
are important information, and so it is desirable to store ERT and
these values as ordered pairs, such as {ERT, POI-maxima} and {ERT,
POI-minima}, and these time-measurement pairs are called
time-tagged data. If available memory is severely limited, then
these time-tagged pairs of maxima and minima for each POI together
with overall elapsed run time will provide a great deal of useful
information for reliability analysis. More generally, dedicated
storage registers may be allocated for time-tagged pairs of the N
highest values (N=1, 2, 3, . . . ) and M lowest values (M=1, 2, 3,
. . . ) of some or all of these POI. Table 1 depicts a data
structure showing such time-tagged data. The POI entries are called
Type-1 for generality, and could be any POI as previously
discussed, and the producer may choose to record several POI types
in practice.
TABLE-US-00001 TABLE 1 Example of Data Structure Containing
Time-tagged POI Type-1 Values POI Type-1 Value Time-Tag Highest
Observed Value ERT Value at Observation Time 2nd Highest Observed
ERT Value at Observation Time Value 3rd Highest Observed ERT Value
at Observation Time Value . . . . . . Nth Highest Observed ERT
Value at Observation Time Value Lowest Observed Value ERT Value at
Observation Time 2nd Lowest Observed ERT Value at Observation Time
Value 3rd Lowest Observed ERT Value at Observation Time Value . . .
. . . Mth Lowest Observed ERT Value at Observation Time Value
Recording an exhaustive history of all measured values of POI could
require so much memory as to be uneconomical to store within an
optical module. However, for the purposes of failure analysis, the
POI values in the time period closest to failure are of special
interest. So, dedicated registers may be allocated to store a
series of L (L=1, 2, 3, . . . up to some maximum set by the design)
time-sequential measurements of POI, with the oldest measurement
being discarded. These L measurements may be stored at equal time
intervals in the local timebase (e.g., once per minute), or stored
in a manner corresponding to predefined unequal time intervals, as
will be explained below. These measurements may be time-tagged in a
similar fashion to that of Table 1, but such time-tagging is not
necessary for every element of the series of measurements. The
final ERT value provides a single time-tag that, when combined with
the producer's a priori knowledge of the storage sequence, provides
sufficient information for the producer's reliability analyst to
use external general purpose processing capabilities to calculate
the ERT value associated with each individual measurement in the
sequence.
[0047] As an example implementation of predefined unequal storage
intervals, let T.sub.o be a time interval in the optical module's
local timebase, and let P, Q, R, and S be positive nonzero
integers. Then, one implementation is to store a POI value once
every T.sub.o interval into the first one of a set of P memory
registers, shifting the older values in the remaining P memory
registers on a first-in first-out (FIFO) basis, discarding the
oldest measurements such that only a quantity of P of the POI
measurements are retained.
[0048] Then once every P of these T.sub.o time intervals, store the
average of the values currently in the set of quantity-P registers
into the first element of a set Q registers, shifting the older
values in the remaining Q memory registers on a FIFO basis,
discarding the oldest value so that only a quantity of Q of such
values are retained.
[0049] Then once every P.times.Q of these T.sub.o time intervals,
store the average of the values currently in the set of quantity-Q
registers into the first element of a set of R registers, shifting
the older values in the remaining R memory registers on a FIFO
basis, discarding the oldest value so that only a quantity of R of
such values are retained.
[0050] Then once every P.times.Q.times.R of these T.sub.o time
intervals, store the average of the values currently in the set of
quantity-R registers into the first element of a set of S
registers, shifting the older values in the remaining S memory
registers on a FIFO basis, discarding the oldest value so that only
a quantity of S of such values are retained.
[0051] Of course, this progressive averaging may be extended to
arbitrarily many steps, and instead of using the individually named
indexes P, Q, R, S, one would establish an index series such as
P.sub.1, P.sub.2, P.sub.3, . . . P.sub.N for arbitrarily large N.
Note that an average taken over several composite averages is equal
to the average of the individual elements taken separately, as
indicated in Equation 1, below, for the specific case of a
three-level composite average. So, the averages-of-averages stored
in the composite tables (e.g., in Tables 2c, 2d, and 2e, below)
simply represent the average value of the POI for the corresponding
time period.
k = 1 P 3 1 P 3 [ 1 P 2 j = 1 P 2 [ 1 P 1 i = 1 P 1 x i ] j ] k = m
= 1 P 1 P 2 P 3 x m P 1 P 2 P 3 Equation 1 ##EQU00001##
As a numeric example, an optical module transferring data at 10
gigabits per second will typically have an internal digital clock
signal at some sub-multiple of the data rate, for example at 50
MHz, and a T.sub.o of one minute may then be conveniently measured
by counting three billion cycles of that 50 MHz clock. Then,
setting P=11, Q=7, R=5, and S=3, and calculating the various time
values produces the example of Tables 2a, 2b, 2c, 2d, and 2e.
[0052] Note: Tables 2b-2e display time rounded-off to the nearest
minute. Movements and manipulation of the data within the tables
are synchronized to the 50 MHz digital clock, which has a period of
one three billionth of a minute, and the negligible rounding error
imposed by ignoring the data transition time between registers is
of that order of magnitude.
TABLE-US-00002 TABLE 2a Example Variable Values Variable Value Unit
P 11 A unitless integer Q 7 A unitless integer R 5 A unitless
integer S 3 A unitless integer T.sub.o 1 Minutes
TABLE-US-00003 TABLE 2b An Example P-Table. Index Max- (Ranges imum
from 1 Age in to P) Minutes Entry Note 1 1 Most recent POI
measurement, The most 0 to 1 minutes old. recent POI 2 2 POI
measurement, 1 to 2 minutes old. value enters at 3 3 POI
measurement, 2 to 3 minutes old. the top of the 4 4 POI
measurement, 3 to 4 minutes old. table, and all 5 5 POI
measurement, 4 to 5 minutes old. entries are 6 6 POI measurement, 5
to 6 minutes old. pushed down 7 7 POI measurement, 6 to 7 minutes
old. the table once 8 8 POI measurement, 7 to 8 minutes old. every
T.sub.o = 1 9 9 POI measurement, 8 to 9 minutes old. minute period,
10 10 POI measurement, 9 to 10 minutes with the very old. oldest
entry 11 11 POI measurement, 10 to 11 minutes being old.
discarded.
TABLE-US-00004 TABLE 2c Example of the Q-Table Index (Ranges
Maximum from Age in 1 to Q) Minutes Entry Note 1 11 Average of the
11 P-Table Entries, Every P .times. T.sub.o = created 0 to 11
minutes ago. 11 minutes, 2 22 Average of the 11 P-Table Entries,
the values of created 11 to 22 minutes ago. the P-Table 3 33
Average of the 11 P-Table Entries, are averaged created 22 to 33
minutes ago. and placed at 4 44 Average of the 11 P-Table Entries,
the top of this created 33 to 44 minutes ago. Q-Table, and 5 55
Average of the 11 P-Table Entries, all older created 44 to 55
minutes ago. values are 6 66 Average of the 11 P-Table Entries,
pushed down, created 55 to 66 minutes ago. with the very 7 77
Average of the 11 P-Table Entries, oldest entry created 66 to 77
minutes ago. being discarded.
TABLE-US-00005 TABLE 2d Example of the R-Table Index (Ranges
Maximum from 1 Age in to R) Minutes Entry Note 1 77 Average of the
7 Q-Table Every P .times. Q .times. T.sub.o = Entries, created 0 to
77 77 minutes, the minutes ago. values of the Q- 2 154 Average of
the 7 Q-Table Table are averaged Entries, created 77 to 154 and
placed at the top minutes ago. of this R-Table, and 3 231 Average
of the 7 Q-Table all older values are Entries, created 154 to 231
pushed down, with minutes ago. the very oldest entry 4 308 Average
of the 7 Q-Table being discarded. Entries, created 231 to 308
minutes ago. 5 385 Average of the 7 Q-Table Entries, created 308 to
385 minutes ago.
TABLE-US-00006 TABLE 2e Example of the S-Table Index (Ranges
Maximum from 1 Age in to S) Minutes Entry Note 1 385 Average of the
5 R-Table Every P .times. Q .times. R .times. Entries, created 0 to
385 T.sub.o = 385 minutes, minutes ago. the values of the R- 2 770
Average of the 5 R-Table Table are averaged Entries, created 385 to
770 and placed at the top minutes ago. of this S-Table, and 3 1155
Average of the 5 R-Table all older values are Entries, created 770
to 1155 pushed down, with minutes ago. the very oldest entry being
discarded.
This progressive averaging of averages must ultimately be
terminated based on the amount of memory the producer chooses to
allocate to storage of POI values and their averages. However, this
lossy compression technique may be extended indefinitely in
principle, and is limited only by considerations of economy.
[0053] In this way, the reliability analyst will have access to
specific POI measurements in the time periods closest to time of
failure, and will also have access to less specific averaged
information for older POI measurements. The necessary general
purpose processing to convert these measured and averaged POI data
from the optical module's timebase to minutes need not be present
within the optical module, since general purpose processing
capability will be conveniently available to the producer's
reliability analyst.
[0054] Additionally, memory registers may be allocated to store
additional statistical information about each physical parameter as
observed over the life of the device. For example using the
nomenclature established above, dedicated registers may be
allocated to store the statistical moments about the mean of the
data, including variance, skew, kurtosis, and higher-order moments
about the mean of each POI over quantity-P, -Q, -R, -S (and so on)
of measurements. In this way, the optical module may store
historical statistical information for each POI, with storage
preference being given to more recent data.
[0055] For the EPMS data (including ERT and POI data structures)
described above to be useful to the producer, the producer must
have access to read the data after optical modules return from the
user, and also must be able to write and clear the EPMS data for
diagnostic purposes. For the optical module to be useful to the
user, the addition of this EPMS data must not compromise or violate
pre-existing standards or conventions.
[0056] Modern optical modules, such as those complying with the
XENPAK, X2, XPAK, or 300 pin form factors, do possesses digital
maintenance interfaces that in most cases may be used by the
producer for access to the EPMS data. Typically, the user has
pre-existing agreements with the producer allocating a set of
addresses in the maintenance interface's address space for use by
the producer. If the producer's selection of EPMS data will fit
with this allocated address space, then the producer may use the
existing digital maintenance interface for EPMS purposes in a
straightforward manner.
[0057] However, if the producer's selection of EPMS data will not
fit with this allocated address space, then the producer may
potentially still use the limited maintenance address space to set
up command and control registers by which the producer may command
the optical module into a maintenance mode, outside the scope of
the user's interests, to transfer EPMS data by direc memory access
(DMA). Once the optical module is in maintenance mode, the producer
may perform the DMA over the digital maintenance interface, or may
even perform the DMA over the optical module's optical port or the
high-bandwidth electrical port commonly paired with the optical
port.
[0058] Conversely, the producer may choose to make EPMS data
available only by partially disassembling the optical model (e.g.,
only by opening the optical module case). This approach has the
advantage of eliminating any foreseeable user concern about EPMS
impact on pre-existing understandings over use of the optical
module's standard digital maintenance interface, and the
disadvantage of requiring the producer to partially disassemble the
optical module to access EPMS information.
[0059] The producer might also choose to implement a low-power
wireless interface to access EPMS data, such as a low power radio
frequency interface or a free-space optical interface such as
optical infra-red. This approach would eliminate potential user
concerns about pre-existing understandings about use of the
standard digital maintenance interface, while also allowing the
producer to access EPMS data without partial module
disassembly.
[0060] And so, several approaches are available to the producer to
achieve the required EPMS data access, and the producer is free to
choose the option that best suits their circumstances.
[0061] Various modifications and improvements of the present
invention may also be apparent to those of ordinary skill in the
art. Thus, the particular combination of sensors and data storage
elements described and illustrated herein is intended to represent
only certain embodiments of the present invention, and is not
intended to serve as limitations of alternate devices within the
spirit and scope of the invention. Various aspects of the
techniques and apparatus associated with the sensing, processing,
storing, and reporting aspects of the invention may be implemented
in a digital circuitry, or in computer hardware, firmware,
software, or in combinations of them. Apparatus of the invention
may be implemented in dedicated digital logic circuitry tangibly
embodied in the module, or in a machine-readable storage device for
use by a programmable processor, or in software located in memory
along with communications processing software used in operation of
the module. The foregoing techniques may be performed, for example,
single central processor, a multiprocessor, on one or more digital
signal processors, gate arrays of logic gates, or hardwired logic
circuits for executing a sequence of signals or program of
instructions to perform functions of the invention by operating on
input data and generating output. The methods may advantageously be
implemented in one or more computer programs that are executable on
a programmable system including at least one programmable processor
coupled to receive data and instructions from, and to transmit data
and instructions to, a data storage system, at least one in/out
device, and at least one output device. Each computer program may
be implemented in a high-level procedural or object-oriented
programming language, or in assembly or machine language if
desired; and in any case, the language may be compiled or
interpreted language. Suitable processors include by way of
example, both general and special purpose microprocessors.
Generally, a processor will receive instructions and data from
read-only memory and/or random access memory. Storage devices
suitable for tangibly embodying computer program instructions and
data include all forms of non-volatile memory, including by way of
example, semiconductor devices, such as EPROM, EEPROM, and flash
memory devices; magnetic disks such as internal hard disks and
removable disks; magneto-optical disks; and CD-ROM disks. Any of
the foregoing may be supplements by or incorporated in,
specifically designed application-specific integrated circuits
(ASICS).
[0062] It will be understood that each of the elements described
above, or two or more together, also may find a useful application
in other types of constructions differing from the types described
above.
[0063] While the invention has been illustrated and described as
embodied in a transceiver for an optical communications network, it
is not intended to be limited to the details shown, since various
modifications and structural changes may be made without departing
in any way from he spirit of the present invention.
[0064] Without further analysis, the foregoing will so fully reveal
the gist of the present invention that others can, by applying
current knowledge, readily adapt it for various applications
without omitting features that, from the standpoint of prior art,
fairly constitute essential characteristics of the generic or
specific aspects of this invention and, therefore, such adaptations
should and are intended to be comprehended within the meaning and
range of equivalence of the following claims.
* * * * *
References