U.S. patent application number 14/174142 was filed with the patent office on 2015-03-12 for multi-level coding and distortion compensation.
This patent application is currently assigned to Avago Technologies General IP (Singapore) Pte. Ltd. The applicant listed for this patent is Avago Technologies General IP (Singapore) Pte. Ltd. Invention is credited to Georgios Asmanis, Faouzi Chaahoub, David W. Dolfi, Michael Allen Robinson.
Application Number | 20150071651 14/174142 |
Document ID | / |
Family ID | 52625740 |
Filed Date | 2015-03-12 |
United States Patent
Application |
20150071651 |
Kind Code |
A1 |
Asmanis; Georgios ; et
al. |
March 12, 2015 |
MULTI-LEVEL CODING AND DISTORTION COMPENSATION
Abstract
An optical communication system, a transmitter, a receiver, and
methods of operating the same are provided. In particular, a
transmitter is disclosed as being configured to encode optical
signals in accordance with a multi-level coding scheme. The
receiver is configured to provide receive and decode to the optical
signals received from the transmitter. One or both of the receiver
and transmitter are configured to compensate for non-idealities or
non-linearities introduced into the communication system by optical
components of the system.
Inventors: |
Asmanis; Georgios; (Lake
Forest, CA) ; Chaahoub; Faouzi; (San Jose, CA)
; Robinson; Michael Allen; (Fremont, CA) ; Dolfi;
David W.; (Los Altos, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Avago Technologies General IP (Singapore) Pte. Ltd |
Singapore |
|
SG |
|
|
Assignee: |
Avago Technologies General IP
(Singapore) Pte. Ltd
Singapore
SG
|
Family ID: |
52625740 |
Appl. No.: |
14/174142 |
Filed: |
February 6, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14020399 |
Sep 6, 2013 |
|
|
|
14174142 |
|
|
|
|
Current U.S.
Class: |
398/141 ;
398/200; 398/214 |
Current CPC
Class: |
H04B 10/54 20130101;
H04B 10/697 20130101; H04B 10/524 20130101; H04B 10/58
20130101 |
Class at
Publication: |
398/141 ;
398/200; 398/214 |
International
Class: |
H04B 10/25 20060101
H04B010/25; H04B 10/61 20060101 H04B010/61; H04B 10/2507 20060101
H04B010/2507; H04B 10/564 20060101 H04B010/564 |
Claims
1. A data transmission system, comprising: a transmitter comprising
an encoder configured to encode a signal into an multi-level-coded
signal; and a receiver configured to receive the multi-level-coded
signal, the receiver further comprising a decoder being configured
to sample a first level of the multi-level-coded signal, sample a
second level of the multi-level-coded signal, and output two or
more separate digital output signals, wherein at least one of the
encoder and decoder are configured to compensate for at least one
of non-linearities and non-idealities introduced into the
multi-level-coded signal by an optical component of the data
transmission system.
2. The system of claim 1, wherein the optical component that
introduces the at least one of non-linearities and non-idealities
into the multi-level-coded signal corresponds to at least one of a
light source and a light detector used to communicate the
multi-level-coded signal from the transmitter to the receiver.
3. The system of claim 2, wherein the at least one of a light
source and a light detector comprises at least one of a laser, a
laser diode, a Vertical Cavity Surface-Emitting Laser (VCSEL), a
photodiode, and an Integrated Circuit comprising an optical
component.
4. The system of claim 1, wherein the transmitter is configured to
compensate for the at least one of non-linearities and
non-idealities by implementing a non-uniform encoding of the
multi-level-coded signal such that a first voltage difference
between a first pair of adjacent levels in the multi-level-coded
signal is different from a second voltage difference between a
second pair of adjacent levels in the multi-level-coded signal.
5. The system of claim 4, wherein the first pair of adjacent levels
is higher than the second pair of adjacent levels and wherein the
first voltage difference is greater than the second voltage
difference.
6. The system of claim 4, wherein the first pair of adjacent levels
is lower than the second pair of adjacent levels and wherein the
first voltage difference is smaller than the second voltage
difference.
7. The system of claim 1, wherein the receiver is configured to
compensate for the at least one of non-linearities and
non-idealities by implementing at least one of: (i) a
level-dependent timing offset and (ii) a level-dependent amplitude
offset.
8. The system of claim 7, wherein the receiver is configured to
implemented both of: (i) the level-dependent timing offset and (ii)
the level-dependent amplitude offset.
9. The system of claim 7, wherein the receiver implements the
level-dependent timing offset by delaying the sample time of the
first level from the sample time of the second level.
10. The system of claim 8, wherein the sample time of the first
level is delayed from the sample time of the second level by
delaying at least one of a clock signal provided to a comparator of
the first level and delaying the multi-level-coded signal from
being provided to the comparator of the first level.
11. The system of claim 8, wherein the sample time of the first
level is advanced from the sample time of the second level by
advancing at least one of a clock signal provided to a comparator
of the first level and advancing the multi-level-coded signal that
is provided to the comparator of the first level.
12. The system of claim 8, wherein the receiver implements the
level-dependent amplitude offset by sampling each level of the
multi-level-coded signal at sample levels that are not equidistance
from one another.
13. The system of claim 1, wherein the transmitter and receiver are
both configured to compensate for the at least one of
non-linearities and non-idealities introduced into the
multi-level-coded signal by the optical component of the data
transmission system.
14. The system of claim 1, wherein the multi-level-coded signal
comprises a Pulse Amplitude Modulated N-level (PAM-N) signal and
wherein the two or more separate digital output signals comprise
two or more Non-Return-to-Zero (NRZ) signals.
15. A method of communicating across a data transmission system,
the method comprising: receiving one or more input signals at an
encoder; encoding the one or more input signals into a
multi-level-coded signal with the encoder; causing a light source
to transmit the multi-level-coded signal across an optical
communications link; receiving the multi-level-coded signal at a
light detector of a receiver connected to the optical
communications link; decoding the multi-level-coded signal with a
decoder, wherein decoding the multi-level-coded signal comprises
sampling a first level of the multi-level-coded signal and sampling
a second level of the multi-level-coded signal; outputting two or
more separate digital output signals based on the decoding of the
multi-level-coded signal; and compensating for at least one of
non-idealities and non-linearities introduced into the
multi-level-coded signal by at least one of the light source, the
light detector, a driver of the light source, and a driver of the
light detector.
16. The method of claim 15, wherein the at least one of
non-idealities and non-linearities are introduced into the
multi-level-coded signal as a result of directly modulating the
light source and wherein the light source comprises at least one of
a laser, a laser diode, and a Vertical Cavity Surface-Emitting
Laser (VCSEL).
17. The method of claim 15, wherein the transmitter compensates for
the at least one of non-linearities and non-idealities by
implementing a non-uniform encoding of the multi-level-coded signal
such that a first voltage difference between a first pair of
adjacent levels in the multi-level-coded signal is different from a
second voltage difference between a second pair of adjacent levels
in the multi-level-coded signal.
18. The method of claim 17, wherein the first pair of adjacent
levels is higher than the second pair of adjacent levels and
wherein the first voltage difference is greater than the second
voltage difference.
19. The method of claim 17, wherein the first pair of adjacent
levels is lower than the second pair of adjacent levels and wherein
the first voltage difference is smaller than the second voltage
difference.
20. The method of claim 15, wherein the receiver is configured to
compensate for the at least one of non-linearities and
non-idealities by implementing at least one of: (i) a
level-dependent timing offset and (ii) a level-dependent amplitude
offset.
21. The method of claim 20, wherein the receiver implements the
level-dependent timing offset by delaying the sample time of the
first level from the sample time of the second level.
22. The method of claim 20, wherein the receiver implements the
level-dependent timing offset by advancing the sample time of the
first level from the sample time of the second level.
23. The method of claim 20, wherein the receiver implements the
level-dependent amplitude offset by sampling each level of the
multi-level-coded signal at sample levels that are not equidistance
from one another.
24. A receiver adapted for use in an optical communication system
comprising at least one optical component, the receiver comprising:
a first decoder element configured to sample a first level-specific
component of a multi-level-coded signal at a first time and compare
the value sampled at the first time with a first reference voltage;
and a second decoder element configured to sample a second
level-specific component of the multi-level-coded signal at a
second time and compare the value sampled at the second time with a
second reference voltage, wherein the receiver compensates for
non-ideal behavior of the at least one optical component by: (i)
delaying the first time relative to the second time; (ii) causing
the first reference voltage to be offset relative to the second
reference voltage by an amount different than a voltage difference
between levels of the multi-level-coded signal; (iii) advancing the
first time relative to the second time; and/or (iv) causing the
first reference voltage to be offset relative to the second
reference voltage by an amount different than a voltage difference
between levels of the multi-level-coded signal.
25. The receiver of claim 24, wherein the first time is either
delayed or advanced relative to the second time by at least one of
a predetermined amount and an adjustable amount.
26. The receiver of claim 24, wherein the first decoder element
comprises a first comparator and wherein the second decoder element
comprises a second comparator and a delay element that delays the
second comparator with respect to the first comparator.
27. The receiver of claim 24, wherein the multi-level-coded signal
is amplitude modulated by direct laser modulation.
28. The receiver of claim 24, further comprising: a third decoder
element configured to sample a third level-specific component of
the multi-level-coded signal at a third time and compare the value
sampled at the third time with a third reference voltage, wherein
at least one of the following conditions is true: (a) the third
time is delayed with respect to the first time and with respect to
the second time; and (b) a first difference between the first
reference voltage and second reference voltage is different from a
second difference between the second reference voltage and the
third reference voltage.
29. A transmitter adapted for use in an optical communication
system, the transmitter comprising: a light source; and an encoder
configured to receive two or more digital signals, encode the two
or more digital signals into a multi-level-encoded signal, and
cause the light source to transmit the multi-level-encoded signal
across an optical communications link, wherein the encoder is
further configured to compensate for non-ideal behavior of the
light source by implementing a non-uniform encoding of the
multi-level-coded signal.
30. The transmitter of claim 29, wherein the two or more digital
signals comprise a first signal, DA, and a second signal, DB,
wherein the encoder produces a first voltage, V00, when both the
first and second signal are `0`, wherein the encoder produces a
second voltage, V01, when the first signal is `0` and the second
signal is `1`, wherein the encoder produces a third voltage, V10,
when the first signal is `1` and the second signal is `0`, and
wherein the encoder produces a fourth voltage, V11, when both the
first and second signal are `1`.
31. The transmitter of claim 30, wherein a first difference between
the first voltage, V00, and the second voltage, V01, is different
from a second difference between the second voltage, V01, and the
third voltage, V10.
32. The transmitter of claim 31, wherein the second difference is
different from a third difference between the third voltage, V10,
and the fourth voltage, V11.
33. The transmitter of claim 32, wherein the first difference is
less than the second difference and wherein the second difference
is less than the third difference.
34. The transmitter of claim 33, wherein the light source comprises
at least one of a laser, a laser diode, and a Vertical Cavity
Surface-Emitting Laser (VCSEL) that produces more noise at the
fourth voltage than the first voltage.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This Application is a Continuation-in-Part and claims the
benefit of U.S. patent application Ser. No. 14/020,399, filed Sep.
6, 2013, entitled "MULTI-LEVEL DECODER WITH SKEW CORRECTION," the
entire disclosure of which is hereby incorporated herein by
reference.
FIELD OF THE DISCLOSURE
[0002] The present disclosure is generally directed toward data
transmission and reception and, in particular, toward mechanisms
for use in fiber optic-based data transmission systems.
BACKGROUND
[0003] Data transmissions in enterprise optical communication
systems have not relied on overly-complicated encoding and/or
decoding schemes because the technology has been more than
sufficient to support desired data transmission rates. However, as
computing devices become faster and the need for increased data
transmission rates is realized, the physical limits of optical
devices will become a limiting factor. Accordingly, optical
communication systems will begin heading toward the use of more
complicated encoding and decoding schemes.
[0004] Pulse-Amplitude Modulation (PAM) is a form of signal
modulation where the message information is encoded in the
amplitude of a series of signal pulses. It is pulse modulation
scheme in which the amplitudes of a train of carrier pulses are
varied according to the sample value of the message signal.
Demodulation of a PAM-encoded signal is performed by detecting the
amplitude level of the carrier at every symbol period.
[0005] In a PAM-4-based optical link, two Non-Return-to-Zero
(NRZ)-coded two-level signals are combined together in a PAM-4
encoder to create a single PAM four-level signal. The PAM-4 signal
is the signal that is ultimately transmitted across a communication
network (e.g., through fiber optics). An advantage of a PAM-4
encoding scheme is that the four-level code utilizes the same baud,
or symbol rate, of either of the two NRZ codes while containing
twice the information of either. This is an attractive solution
when the components of the link are baud rate limited, as is often
the case for very high-speed fiber links.
[0006] Traditional PAM-4 signaling has a strict linearity
requirement. Specifically, any PAM signal (e.g., PAM-N, where N is
an integer greater than or equal to 4) has been traditionally
constrained by the requirement that each signal level is uniformly
spaced apart from adjacent signal levels. A conventional PAM-4
encoder translates two NRZ signals (DA and DB) to a PAM-4 signal
via a table as shown in FIG. 1A, where a constant value, Ds, is
used to define the spacing between adjacent levels. In other words,
as shown in FIG. 1B, the base level where DA and DB are both `0`
results in a PAM-4 signal of V0. The next level, where DA is `0`,
but DB is `1` results in a PAM-4 signal of V0+Ds. Each subsequent
level is greater than the previous level by the constant, Ds.
[0007] A PAM-4 receiver then decodes the PAM-4 signal received from
the encoder and recovers the original two NRZ (DA and DB) data
streams. The receiver samples the PAM-4 signal at N-1 (e.g., 3
points in a PAM-4 signal) at a common sample time (e.g., ts) and
performs an inverse mapping to the encoder table of FIG. 1A. In
this way, the PAM-4 encoder behaves much like a Digital-to-Analog
converter and the PAM-4 receiver behaves much like an
Analog-to-Digital converter.
[0008] The strict linearity of encoding and sampling works well for
pure electronic systems whose behavior is relatively linear, but
the strict linearity presents a number of problems in optical
systems due to non-idealities and the non-linear behavior of
optical components.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The present disclosure is described in conjunction with the
appended figures, which are not necessarily drawn to scale:
[0010] FIG. 1A depicts a conventional PAM-4 encoding table;
[0011] FIG. 1B depicts signals output by encoding according to the
table of FIG. 1A;
[0012] FIG. 2A depicts a PAM-4 encoding table in accordance with
embodiments of the present disclosure;
[0013] FIG. 2B depicts levels of a PAM-4 signal generated by
optical components whose behavior is non-linear and/or non-ideal at
higher levels in accordance with embodiments of the present
disclosure;
[0014] FIG. 3 is a block diagram depicting a data transmission
system in accordance with embodiments of the present
disclosure;
[0015] FIG. 4 depicts components of an encoder in accordance with
embodiments of the present disclosure;
[0016] FIG. 5 depicts components of a decoder in accordance with
embodiments of the present disclosure;
[0017] FIG. 6 depicts an improved sampling technique for decoding a
PAM-4 encoded signal in accordance with embodiments of the present
disclosure;
[0018] FIG. 7 depicts a first receiver structure to achieve the
decoding of FIG. 6;
[0019] FIG. 8 depicts a second receiver structure to achieve the
decoding of FIG. 6;
[0020] FIG. 9 depicts a third receiver structure to achieve the
decoding of FIG. 6;
[0021] FIG. 10 is a flow chart depicting an encoding method in
accordance with embodiments of the present disclosure; and
[0022] FIG. 11 is a flow chart depicting a decoding method in
accordance with embodiments of the present disclosure.
DETAILED DESCRIPTION
[0023] Various aspects of the present disclosure will be described
herein with reference to drawings that are schematic illustrations
of idealized configurations.
[0024] While certain examples of systems and methods will be
described with reference to a particular multi-level digital
encoding scheme, embodiments of the present disclosure are not so
limited. More specifically, while certain embodiments of the
present disclosure are depicted and described in connection with a
PAM-4 signal, it should be appreciated that embodiments of the
present disclosure are not limited to systems utilizing PAM-4
encoding. Rather, embodiments of the present disclosure have
applicability to more complicated coding schemes. The examples of a
PAM-4 encoding scheme are intended to provide an easy-to-understand
example or set of examples. It should be appreciated that the
concepts disclosed herein can be applied to any type of multi-level
encoding scheme (e.g., PAM-5, PAM-6, PAM-8, . . . , PAM-32,
etc.).
[0025] Moreover, embodiments of the present disclosure are not
necessarily limited to systems that employ a laser to modulate the
signals. The embodiments described herein reference the use of
lasers for signal modulation since it has been observed that lasers
and laser diodes have different non-linear behavior and/or
non-ideal behavior at different levels (e.g., different
transmission voltages). It should, however, be apparent to those of
ordinary skill in the art that embodiments of the present
disclosure are not limited to communication systems employing
lasers for signal modulation and/or demodulation.
[0026] While multi-level encoding is useful for increasing the data
transmission rate of a communication system, the price paid for
this encoding technique is the additional complexity of introducing
the multi-level encoded signal at the transmit end and accurately
decoding the signal at the receiving end. More specifically, as a
non-limiting example, complications arise in decoding when the
optical modulation which creates the optical PAM-4 signal is
supplied by directly modulating a laser. Due to the inherent
non-linear and/or non-ideal behavior of semiconductor lasers, the
laser's modulation speed is dependent on the drive level of the
current that is supplying the modulation signal.
[0027] In addition to non-idealities introduced by the laser of an
optical communication system, other components are known to further
introduce non-idealities and/or non-linearities into the system,
thereby making the traditional techniques of encoding and decoding
(where strict linearity is required) less desirable. Examples of
such non-idealities and/or non-linearities include the linearity
(or lack thereof) of the laser driver LI curves, the asymmetry (or
lack thereof) of the dynamic response of the laser diode that leads
to different and level-dependent rise and fall times, the
non-linearity of the photodetector and Transimpedance Amplifier
(TIA), the level-dependent noise to the transmitted PAM-4 signal,
etc.
[0028] In a PAM-4 situation, this implies that the response of the
laser is fastest at the upper levels and slowest at the lower
levels. This creates both a distortion in the optical eyes as well
as a skew in the arrival time at the receive end of the
communications link. Additionally, as shown in FIG. 2B, the noise
introduced into the PAM-4 signal is level-dependent. More
specifically, the voltage produced for the highest level of the
PAM-4 signal has more noise than the lower level voltages produced
by the same laser diode.
[0029] Accordingly, embodiments of the present disclosure propose
the ability to account for and, in some situations, neutralize the
non-linear and/or non-idealistic behavior of the components in an
optical system utilizing PAM-N signals.
Signal Encoding and Transmission
[0030] A first aspect of the present disclosure is to compensate
for the increasing noise produced at the higher voltages produced
by a laser (or any other component exhibiting similar behavior).
FIG. 2A shows an encoding table that can be used in accordance with
embodiments of the present disclosure. More specifically, the
encoding scheme proposed herein relaxes the strict linear
constraints previously imposed on PAM-4 systems. Even more
specifically, the PAM-4 signal produced for the various DA and DB
values utilize non-uniform encoding. As an example, the lowest
level of the PAM-4 signal corresponds to a condition where DA and
DB are both `0`. Under this condition, an encoder of the present
disclosure will produce a PAM-4 signal having a voltage equal to
V00. The next level of the PAM-4 signal corresponds to a condition
where DA is `0`, but DB is `1`. Under this condition, an encoder of
the present disclosure will produce a PAM-4 signal having a voltage
equal to V01. Continuing with the table of FIG. 2A, the next level
of the PAM-4 signal corresponds to a condition where DA is `1`, but
DB is `0`. The voltage produced by the encoder under this condition
corresponds to V10. Finally, the highest level of the PAM-4 signal
corresponds to a condition where both DA and DB are `1`. Under this
condition, an encoder of the present disclosure will produce a
PAM-4 signal having a voltage equal to V11.
[0031] In contrast to PAM-4 encoders of the prior art and the
encoding table of FIG. 1A, the difference between the voltage
produced for the lowest level and its adjacent level (e.g., the
second lowest level) is not necessarily the same as the difference
between the voltage produced for the highest level and its adjacent
level (e.g., the second highest level). Likewise the difference
between the voltage produced for the second and third levels of the
PAM-4 signal are not necessarily the same as the differences
between voltage for any other pair of levels in the PAM-4 signal.
More specifically, embodiments of the present disclosure propose a
PAM-N encoder that increases voltages at each successive level by
more than a constant amount. Referring back to the example of FIGS.
2A and 2B, the difference between V00 and V01 may correspond to a
first difference (e.g., DS1), the difference between V01 and V10
may correspond to a second difference (e.g., DS2), and the
difference between V10 and V11 may correspond to a third difference
(e.g., DS3). In a non-limiting example, the PAM-4 encoder may
encode the various levels of the PAM-4 signal according to the
following rule: DS3>DS2>DS1. By increasing the difference
between each successive level in the PAM-N signal, the encoder can
improve the signal-to-noise ratio (SNR) of the PAM-N signaling.
Following such an encoding scheme leads to a reduction of the
transmitted levels of the least noisy symbols to maintain a
constant peak-to-peak swing.
[0032] In other words, embodiments of the present disclosure
propose to utilize a non-uniform PAM-N encoder that translates
input digital signals (e.g., two NRZ signals in a PAM-4 situation)
to a PAM-N signal having different gaps between at least two
adjacent levels. Following the proposed encoding behavior of
relaxing the linearity constraints of the laser driver's Integrated
Circuit (IC), the system's SNR and bit error rate (BER) can be
improved dramatically. Accordingly, the non-idealities and/or
non-linear behavior of the system can be compensated for with a
non-uniform encoding scheme.
[0033] In addition to improving system performance by implementing
a new encoding scheme, it may also be possible to improve system
performance at the transmission side by utilizing an improved
driver configuration. Accordingly, with reference now to FIGS. 3-5,
details of an improved system configuration having an improved
driver will be described in accordance with embodiments of the
present disclosure. It should be appreciated that the improved
system configuration described herein can be utilized alone or in
combination with the non-uniform encoding behavior described above
with reference to FIGS. 2A and 2B.
[0034] Referring now to FIG. 3, components of a data transmission
system 300 will be described in accordance with embodiments of the
present disclosure. The data transmission system 300 is shown to
include a transmitter 304 and receiver 308 connected by an optical
fiber link 312. The length of the optical fiber link 312 may be as
small as a few meters or as long as several kilometers. The
transmitter 304 and/or receiver 308 may be associated with a common
computer network or may be separated by one or several
communication networks. In some embodiments, the transmitter 304
and/or receiver 308 may be operating in a signal boosting station
rather than being incorporated into a computing network.
[0035] In some embodiments, the transmitter 304 receives an input
signal 324 from some computing device or from another fiber link.
The input signal 324 may be in the form of a digital signal or a
plurality of digital signals. For instance, the input signal 324
may correspond to two or more NRZ signals representing two
different pieces of information. The transmitter 304 includes an
encoder 316 that is configured to encode the input signal 324 and
prepare the signal for transmission across the optical fiber link
312. In some embodiments, the encoder 316 comprises a laser (e.g.,
semiconductor laser) or similar source of coherent light. The
encoder 316 may be driven by an input current and, in some
embodiments, the encoder 316 may be configured to encode the input
signal into a multi-level encoded signal, such as a PAM-N signal.
However, the encoder 316 may inherently skew the multi-level
encoded signal or introduce other non-idealities or non-linearities
into the PAM-N signal.
[0036] The multi-level encoded signal may then be transmitted by
the encoder 316 across the optical fiber link 312 where the encoded
signal is received at the receiver 308. The receiver 308 may be
configured to employ a decoder 320 to decode the multi-level
encoded signal and produce a corresponding output signal 328, which
may correspond to the digital signals of the input signal 324
(e.g., two or more NRZ signals). In some embodiments, the decoder
320 is configured to account for or otherwise correct the skew
introduced into the signal by the encoder 316. In some embodiments,
the decoder 320 may comprise a plurality of discrete decoding
elements that are each adapted to sample different levels of the
multi-level encoded signal transmitted by the encoder 316.
[0037] With reference now to FIG. 4, additional details of
components that may be included in the transmitter 304 and, more
specifically, may be included in the encoder 316 will be described
in accordance with embodiments of the present disclosure. The
transmission system 404, in some embodiments, may directly
correspond to the encoder 316. In other embodiments, the
transmission system 404 may correspond to a sub-component of the
encoder 316.
[0038] The transmission system 404 is depicted as having two sets
of sub-components, namely a gearbox IC 408 and a laser driver 412.
While the remainder of this example will be described in connection
with circuitry configured to drive a laser, laser diode, or a
Vertical Cavity Surface-Emitting Laser (VCSEL), it should be
appreciated that embodiments of the present disclosure are not so
limited. In fact, the light source used to transmit the signal
across the optical fiber link 312 may correspond to any one or
collection of devices capable of transmitting light. Furthermore,
although the gearbox IC 408 is depicted as only being configured to
condition two separate input signals for transmission to the laser
driver 412, the components of the transmission system 404 can be
multiplied to accommodate a larger number of signals. The following
example will be described in connection with encoding two NRZ
signals into a PAM-4 signal for convenience of understanding.
[0039] In the depicted example, the gearbox IC 408 is shown to
include a pair of equalizers 416, a pair of coders 420, a de-skew
and lane adjust module 424, and a line driver 428. The laser driver
412 is shown to include a laser driver and level control module 432
and a light source 436 (e.g., laser, laser diode, VCSEL, etc.).
[0040] In accordance with at least some embodiments, the proposed
architecture of the gearbox IC 408 enables the gearbox IC 408 to
align and encode multiple data streams (e.g., a first NRZ data
stream for DA and a second NRZ data stream for DB). The equalizer
416 and coder 420 condition the separate data streams to comparable
amplitudes. The NRZ coders 420 then output the separate data
streams to the de-skew and lane adjust module 424, which basically
aligns the phases of the two data streams. The output of the
de-skew and lane adjust module 424 is then provided to the line
driver 428, which outputs two separate data streams 414a, 414b to
the laser driver 412.
[0041] In accordance with some embodiments of the present
disclosure, the gearbox IC 408 is a single IC having its components
incorporated therein. The laser driver 412 may be connected to the
gearbox IC 408 via two separate leads, one of which carries the
first NRZ data stream output by the line driver 428 and the other
of which carries the second NRZ data stream output by the line
driver 428. This means that the interface trace bandwidth of each
interface trace between the gearbox IC 408 and the laser driver 412
can be reduced. In other words, the traces on a Printed Circuit
Board (PCB) or the like that carry the separate data streams 414a,
414b can be significantly smaller than if a single trace was used
to carry a PAM-4 signal from the gearbox IC 408 to the laser driver
412 as is currently done in existing systems. This reduction of
interface trace bandwidth leads to a lower cost and power-optimal
solution for the entire transmission system 404. Specifically, even
though two traces are used to connect the gearbox IC 408 to the
laser driver 412 instead of one, the total costs of the two traces
required for high bandwidth signals (e.g., two 10 Gbit/sec NRZ data
streams) is less than the cost to support a PAM-4 signal of double
the bandwidth requirement.
[0042] Upon receiving the two separate data streams from the
gearbox IC 408, the laser driver and level control module 432 can
generate the necessary PAM-N encoded signal, which is used to drive
the light source 436. The light source 436 subsequently transmits
the PAM-N encoded signal across the link 312. It should be
appreciated that the transmission system 404 may be configured to
accommodate any number of signals have any data rate frequency. As
a non-limiting example, the transmission system 404 may be
configured to receive two or more 10G NRZ data streams and transmit
a single 20G PAM-4 signal. When operating at these high
frequencies, the non-linearities and/or non-idealities of the light
source 436 and the other components in the laser driver 412 may
arise; thus, it may be desirable to utilize the non-uniform
encoding techniques described above in combination with the
proposed transmission system 404, although such a construction is
not required. That is, the transmission system 400 may also be
utilized in connection with a traditional linear encoding
scheme.
[0043] In some embodiments, the gearbox IC 408 and laser driver 412
configuration enables the gearbox IC 408 and the PCB connecting the
gearbox IC 408 with the laser driver 412 to utilize less expensive
and more power efficient components. If the gearbox IC 408 were
required to output a PAM-N signal to the laser driver, the
modifications to the IC would be non-trivial and the components
needed to carry the PAM-N signal to the laser driver 412 would be
much more expensive and consume significantly more power than the
proposed system 404 configuration.
Signal Receiving and Decoding
[0044] In addition to the improved encoding schemes described
above, additional improvements can be implemented in a PAM-N
receiver to improve the system's overall SNR and BER performance.
Such receiver solutions can be implemented independent to the
above-described encoding solutions or encoder configurations, but
could also be applied to a system utilizing the non-uniform
encoding behavior and/or improved gearbox IC 408 to further
increase the system's performance.
[0045] With reference now to FIG. 5, additional details regarding
the receiving side of a data transmission system will be described
in accordance with embodiments of the present disclosure.
Specifically, an improved set of receiver components 504 will be
described. The receiver components 504 may constitute some or all
of the receiver 308 or the decoder 320 contained within the
receiver 308. The receiver components 504 may comprise a receiver
IC 508 having a light detector 512, an Analog Front End (AFE) 516,
and a decoder and phase detector module 520.
[0046] The light detector 512 may correspond to any device or
collection of devices configured to convert light energy into an
electrical signal. Non-limiting examples of a suitable light
detector 512 include a photodetector, a photo diode, a photo
resistor, or the like. Moreover, the light detector 512 may or may
not be mounted on or integrated into the receiver IC 508.
[0047] The light detector 512 provides an electrical output signal
to the AFE 516, which forwards the signal to the decoder and phase
detector module 520. In some embodiments, the signal provided to
the decoder and phase detector module 520 may correspond to a PAM-N
signal; thus, although a PAM-4 decoder is depicted in FIG. 5, it
should be appreciated that any type of decoder may be utilized
without departing from the scope of the present disclosure.
[0048] The decoder and phase detector module 520 samples the
received signal and produces two or more separate digital output
signals. As an example, the decoder and phase detector module 520
may output two or more NRZ signals, which may correspond to DA and
DB as shown in FIGS. 2A and 2B.
[0049] As mentioned above, in some embodiments, the receiver
components 504 may simply behave like a traditional PAM-N receiver
and decoder. However, in other embodiments, the receiver components
504 and particularly the decoder and phase detector module 520 may
correspond to a modified non-uniform sampling receiver. In some
embodiments, the receiver components 504 may be configured to
introduce a code-dependent amplitude offset so that the system's
SNR is optimized. More specifically, and with reference to FIG. 6,
a modified non-uniform receiver architecture may be configured to
sample the PAM-N signal at levels V0, V1, and V2, which are
adjustable and/or not necessarily equidistance apart. Said another
way, the receiver may be configured to sample the PAM-4 signal at
voltages that are selected to minimize the increased noise
introduced at the higher levels of the signal. Thus, V2-V1 is not
necessarily equal to V1-V0.
[0050] Furthermore, the receiver may set levels V0, V1, and/or V2
at positions other than equidistance between the adjacent levels.
Said another way, V0 may be closer to V00 than V01 so as to avoid
the higher amount of noise on V01 as compared to V00. Likewise, V1
may be closer to V01 than V10 and V2 may be closer to V10 than V11.
As a non-limiting example, since the higher levels may comprise a
higher amount of noise, it may be desirable to sample at levels V0,
V1, and V2 such that the distance between V0 and V1 is different
than the distance between V1 and V2. Such a sampling scheme can
help to increase the system's overall SNR and BER.
[0051] In some embodiments, the receiver may also selectively
adjust the values of V0, V1, and/or V2 depending upon current
operating conditions of the system, thereby providing a dynamic
optimization. Moreover, since the higher levels of the PAM-N signal
contribute more noise than the lower levels, it may be desirable or
sufficient to only measure the noise at the highest level (e.g.,
V11) to determine what adjustments should be made to some or all of
the sampling levels V0, V1, and/or V2.
[0052] In addition to or as an alternative optimization to the
proposed code-dependent amplitude offset, the decoder and phase
detection module 520 may be configured to perform a code-dependent
(also referred to as a level-dependent) timing offset. As shown in
FIG. 6, it may be desirable to sample the different levels of the
PAM-N signal at offset sample times. This code-dependent timing
offset may be desirable since the optical components of the data
transmission system may cause different levels of the PAM-N signal
to be skewed or delayed relative to one another. In particular, it
may be desirable to sample the highest levels of the signal earlier
in time than the lower levels. A number of different receiver
architectures may be employed to achieve the code-dependent timing
offset.
[0053] With reference now to FIGS. 7-9, examples of the various
receiver architectures that can achieve the code-dependent timing
offset will be described in accordance with embodiments of the
present disclosure. As with other improvements described herein, it
should be appreciated that this code-dependent timing offset can be
used alone or in any combination with the other improvements to
increase the system's SNR and BER.
[0054] The first example of such a receiver architecture 704
implementing a code-dependent timing offset is shown in FIG. 7. The
illustrated receiver architecture 704 comprises a decoder 708 and a
plurality of comparators 712a, 712b, 712c, which may correspond to
sampling flip-flops or the like. Each comparator 712a, 712b, 712c
receives the input encoded signal (e.g., the PAM-N or PAM-4
signal). The first comparator 712a samples the input signal at a
sample time delayed by ts2 and compares the sampled value with V2.
The second comparator 712b samples the input signal at a sample
time delayed by ts1 and compares the sampled value with V1. The
third comparator 712c samples the input signal at sample time
delayed by ts0 and compares the sampled value with V0. In the
depicted embodiment, V2>V1>V0 and ts2>ts1>ts0.
[0055] The particular architecture 704 shows that the timing delay
for each comparator 712a, 712b, 712c is controlled by a clock value
received from the decoder 708, but delayed by different amounts
from one level to the next. In other words, receiver architecture
704 introduces timing skew on the clock phases to the receiver's
sampling flip-flops. The timing skew is introduced by controlling
the delay of the receiver's clock to the comparators 712a, 712b,
712c. Moreover, the code-dependent delay can be adaptive and
controlled by the receiver's phase detector or can be digitally
pre-programmed and set to system-specific parameters that
characterize the light source's asymmetric rise and fall times, the
system's non-stationary noise components, and the IC's
non-linearities. In the case of an adaptive delay, the phase
detector can lock on any of the eyes in the PAM-N signal.
[0056] While the illustrative architecture 704 shows the concept of
delaying sample times, it should be appreciated that the
architecture 704 could be alternatively configured to advance the
sample time of one level relative to another. More specifically,
the sample time of a first comparator could be advanced relative to
the sample time of another comparator by advancing the clock signal
provided to the first comparator as compared to the other
comparators. Thus, while examples described herein utilize the
concept of sample delay, it should be appreciated that embodiments
of the present disclosure are not so limited and sampling can be
advanced as opposed to delayed to compensate for the
non-linearities and/or non-idealities of the system.
[0057] FIG. 8 depicts a second receiver architecture 804 capable of
implementing a code-dependent timing offset in accordance with
embodiments of the present disclosure. The second receiver
architecture 804 is similar to the first receiver architecture 704
in that the second receiver architecture 804 comprises a decoder
808, a plurality of comparators 812a, 812b, 812c (e.g., sampling
flip-flops), and a clock signal 816. The second receiver
architecture 804, however, introduces the timing skew to the input
signal itself via delays 820a, 820b, 820c rather than the clock 816
phases to the receiver's comparators 712a, 712b, 712c. Similar to
the first architecture 704, the second architecture 804 can have
the code-dependent delay be adaptive and controlled by the
receiver's phase detector or the delay can be digitally
pre-programmed and set to system-specific parameters that
characterize the light source's asymmetric rise and fall times, the
system's non-stationary noise components, and the IC's
non-linearities. In the case of an adaptive delay, the phase
detector can lock on any of the eyes in the PAM-N signal.
[0058] As with the first architecture, the second architecture 804
can alternatively be modified to utilize signal advancement instead
of signal delay. For instance, the PAM-N signal can be advanced to
one comparator as compared to other comparators rather than
delaying the PAM-N signal.
[0059] FIG. 9 shows a third receiver architecture 904 where the
adaptive delay from one level to the next is achieved by use of an
eye monitor 908. Specifically, the third receiver architecture 904
is similar or identical to the second receiver architecture 904
except that the eye monitor 908 is used to dynamically control the
delay values 820a, 820b, 820c imparted on the inputs signal. Of
course, the eye monitor 908 could also be incorporated into the
first receiver architecture 704 to dynamically control the delays
to the clock signal. In some embodiments, the code-dependent delay
and the code-dependent amplitude sampling points Vi (e.g., V0, V1,
V2, etc.) can both be adaptive and controlled by the receiver's
phase detector or the eye monitor 908. Specifically, a phase
detector could be utilized to optimize delays ts0, ts1, and/or ts2
based on current system conditions. Alternatively or additionally,
the eye monitor 908 could be used to adaptively optimize the
amplitude thresholds Vi (e.g., V0, V1, and/or V2).
[0060] By using some or all of the techniques described herein, a
data transmission system's performance can be greatly enhanced
and/or the costs associated with building and implementing such a
system can be reduced.
[0061] With reference now to FIG. 10, a method of encoding and
transmitting a signal will be described in accordance with
embodiments of the present disclosure. The method begins when an
encoder 316 or components thereof receives an input signal 324 for
transmission across a communication network (step 1004). The
encoder 316, in some embodiments, may be configured to encode the
input signal 324 using a multi-level amplitude modulation scheme,
such as PAM-N. As a non-limiting example, the encoder 316 may be
configured to encode the signal by modulating a laser, whose
response or encoding speed may be dependent upon the drive level of
the current which is supplying the modulation signal. In a PAM-4
situation, for example, the response of the laser will be the
fastest at the upper levels and slowest at the lower levels.
Moreover, in a PAM-4 situation the upper levels may have more noise
than the lower levels.
[0062] Accordingly, the method continues with the encoder 316
determining the worst performing level combination (step 1008) and
optimizing its encoding to compensate for the worst performing
level combination (step 1012). Continuing the PAM-4 example, the
encoder 316 may determine that the higher levels of the PAM-4
signal will have more noise than the lower levels of the PAM-4
signal and may implement a non-uniform PAM-4 encoding scheme to
counteract the increased noise introduced at the higher levels by
the non-linearities and/or non-idealities of the system. It should
be appreciated, however, that the implementation of the non-uniform
PAM-4 encoding scheme may correspond to an optional step.
[0063] The method continues with the encoder 316 encoding the
signal according to the determined encoding scheme (e.g., the
optimize encoding scheme) (step 1016). The encoded signal is then
transmitted across the optical fiber link 312 (step 1020). In some
embodiments, the transmitted signal may correspond to a PAM-N
signal that is either uniformly or non-uniformly encoded on a
per-level basis.
[0064] With reference now to FIG. 11, a method of receiving and
decoding a signal will be described in accordance with embodiments
of the present disclosure. The steps of the receiving method may or
may not be performed in combination with the optimization steps
described in the transmitting/encoding method.
[0065] The method begins when an encoded signal is received at the
receiver 308 (step 1104). The receiver 308 then employs its decoder
320 to introduce one or both of code-dependent amplitude offset
and/or timing skew at the receiver's sampling points (step 1108).
The code-dependent timing skew may be imparted in a number of
different ways (e.g., using any of the architectures described
herein above) and the code-dependent amplitude offset may be
implemented with or without the code-dependent timing skew.
[0066] The decoder 320 then samples each level-specific component
of the signal at the desired amplitude offset and/or time (e.g., by
introducing a delay on a per-level-basis) (step 1112). Using the
sampled values, the decoder 320 is then able to decode the received
signal into two or more digital signals (e.g., two or more NRZ data
streams) that can be output to a separate receiving circuit or the
like (step 1116).
[0067] This output decoded signal may be provided to a computer or
computer network for processing or may be re-encoded for
re-transmission across another optical fiber link 312. By following
the above method and optimizing the signal output by the encoder
316, the proposed system and method can enhance the SNR and BER of
the overall data transmission system and improve the performance
and robustness of multi-level-coding based communication links.
[0068] In the foregoing description, for the purposes of
illustration, methods were described in a particular order. It
should be appreciated that in alternate embodiments, the methods
may be performed in a different order than that described. It
should also be appreciated that the methods described above may be
performed by hardware components or may be embodied in sequences of
machine-executable instructions, which may be used to cause a
machine, such as a general-purpose or special-purpose processor
(GPU or CPU) or logic circuits programmed with the instructions to
perform the methods (e.g., Application Specific Integrated Circuits
(ASICs), Field-Programmable Gate Arrays (FPGAs), or the like).
These machine-executable instructions may be stored on one or more
machine readable mediums, such as CD-ROMs or other type of optical
disks, floppy diskettes, ROMs, RAMs, EPROMs, EEPROMs, magnetic or
optical cards, flash memory, or other types of machine-readable
mediums suitable for storing electronic instructions.
Alternatively, the methods may be performed by a combination of
hardware and software.
[0069] Furthermore, embodiments may be implemented by hardware,
software, firmware, middleware, microcode, hardware description
languages, or any combination thereof. When implemented in
software, firmware, middleware or microcode, the program code or
code segments to perform the necessary tasks may be stored in a
machine readable medium such as storage medium. A processor(s) may
perform the necessary tasks. A code segment may represent a
procedure, a function, a subprogram, a program, a routine, a
subroutine, a module, a software package, a class, or any
combination of instructions, data structures, or program
statements. A code segment may be coupled to another code segment
or a hardware circuit by passing and/or receiving information,
data, arguments, parameters, or memory contents. Information,
arguments, parameters, data, etc. may be passed, forwarded, or
transmitted via any suitable means including memory sharing,
message passing, token passing, network transmission, etc.
[0070] Specific details were given in the description to provide a
thorough understanding of the embodiments. However, it will be
understood by one of ordinary skill in the art that the embodiments
may be practiced without these specific details. In other
instances, well-known circuits, processes, algorithms, structures,
and techniques may be shown without unnecessary detail in order to
avoid obscuring the embodiments.
[0071] While illustrative embodiments of the disclosure have been
described in detail herein, it is to be understood that the
inventive concepts may be otherwise variously embodied and
employed, and that the appended claims are intended to be construed
to include such variations, except as limited by the prior art.
* * * * *