Using FEC statistics to tune serdes
Serializer-Deserializer (SerDes) operation is optimized for signals based on signal error statistics. Forward Error Correction (FEC) may provide feedback of error statistics or error correction statistics to a SerDes tuner, which uses the statistics to selectively tune or adjust SerDes operating parameters, such as vertical and horizontal sampling or slicing offsets, gain and equalization, to decrease the bit error rate (BER). Statistics report which bits and patterns are corrected and to what values. Knowledge of expected and actual signals is leveraged to correlate detected errors with underlying problems and solutions to optimize SerDes operation. Each node in a network, such as a Ethernet Passive Optical Network (EPON), is enabled to fine tune its operation independently for each logical or physical channel.
1. A device comprising:
a Serializer-Deserializer (SerDes) tuner configured to tune operation of a SerDes based on a plurality of different error statistics generated from a plurality of different types and/or patterns of errors tracked in the operation of the SerDes.
2. The device of claim 1 , wherein the at least one of the plurality of different error statistics is generated from at least one error detected by a Forward Error Correction (FEC) component.
3. The device of claim 2 , wherein the at least one of the plurality of different error statistics comprises at least one error correction statistic.
4. The device of claim 2 , further comprising:
an FEC component;
a SerDes component; and
a feedback comprising the SerDes tuner.
5. The device of claim 4 , wherein the feedback comprises at least one of:
an amplifier feedback configured to tune operation of an amplifier component of the SerDes based on the at least one of the plurality of different error statistics;
an equalizer feedback configured to tune operation of an equalizer component of the SerDes based on the at least one of the plurality of different error statistics;
a sampler feedback configured to tune operation of a sampler component of the SerDes based on the at least one of the plurality of different error statistics; and
a CDR feedback configured to tune operation of a clock data recovery (CDR) component of the SerDes based on the at least one of the plurality of different error statistics.
6. The device of claim 1 , the SerDes tuner comprising:
a statistics generator configured to generate the at least one of the plurality of different error statistics; and
a tuning logic configured to correlate a setting of at least one operating parameter of the SerDes with the at least one of the plurality of different error statistics.
7. The device of claim 6 , wherein an 8 bit symbol converted from a received 10 bit encoded symbol is converted back to the 10 bit encoded symbol prior to the tuning logic correlating the setting of the at least one operating parameter with the at least one of the plurality of different error statistics, the at least one of the plurality of different error statistics including an error detected in the 10 bit encoded symbol.
8. The device of claim 6 , wherein the at least one of the plurality of different error statistics generated by the statistics generator comprises a quantity of error corrected one bits and a quantity of error corrected zero bits; and
wherein the tuning logic is configured to receive the quantity of error corrected one bits and the quantity of error corrected zero bits and to generate a tuning signal to adjust the at least one operating parameter when one of the quantity of the error corrected one bits reaches a first threshold or the quantity of error corrected zero bits reaches a second threshold.
9. A method comprising:
processing a Serializer-Deserializer (SerDes) communication;
tracking a plurality of different types and/or patterns of errors in the SerDes communication;
generating a plurality of different error statistics from the tracked plurality of different types and/or patterns of errors; and
tuning at least one operating parameter or component of the SerDes using at least one of the plurality of different error statistics.
10. The method of claim 9 , further comprising:
correlating the at least one of the plurality of different error statistics with the at least one operating parameter or component of the SerDes.
11. The method of claim 10 , wherein the tuning of the at least one operating parameter of the SerDes using the at least one of the plurality of different error statistics comprises:
selecting a setting of the correlated operating parameter or component; and
tuning the at least one operating parameter or component of the SerDes using the selected setting.
12. The method of claim 9 , further comprising:
providing the at least one of the plurality of different error statistics as feedback to a SerDes tuner.
13. The method of claim 9 , further comprising:
providing a plurality of tuning feedbacks to the SerDes based on the at least one of the plurality of different error statistics.
14. The method of claim 9 , wherein the generating of the at least one of the plurality of different error statistics comprises:
generating a first error statistic for a first type of detected error; and
generating a second error statistic for a second type of detected error.
15. The method of claim 14 , wherein the tuning of the at least one operating parameter or component of the SerDes using the at least one of the plurality of different error statistics comprises:
tuning a first operating parameter or component with a first adjustment when the first error statistic reaches a first threshold;
tuning the first operating parameter or component with a second adjustment when the second error statistic reaches a second threshold; and
tuning a second operating parameter or component with a third adjustment when the first error statistic reaches a third threshold and the second error statistic reaches a fourth threshold.
16. The method of claim 9 , further comprising:
receiving a plurality of signals, respectively, from a plurality of communication channels; and
for each signal of the plurality of signals, separately:
detecting at least one error in the signal;
generating at least one error statistic from the at least one error detection; and
tuning at least one operating parameter of the SerDes using the at least one error statistic.
17. A communication channel, comprising:
an amplifier having a gain parameter that amplifies a received signal into an amplified signal;
an equalizer having an equalization parameter that modifies at least one frequency component of the amplified signal into a frequency-modified signal;
a sampler having a sampling offset parameter that samples the frequency-modified signal to generate a first digital signal;
a serial-to-parallel converter that converts the first digital signal to a second digital signal;
a FEC (forward error correction) unit that detects a plurality of different types of errors in the second digital signal and performs error correction on the second digital signal based on the plurality of different types of detected errors; and
a feedback comprising:
a statistics generator that generates a plurality of different error statistics based on the plurality of different types of detected errors, and
correction logic that generates at least one feedback signal configured to adjust at least one of the gain, equalization and sampling offset parameters based on at least one of the plurality of different types of error statistics.
18. The communication channel of claim 17 , wherein the statistics generator comprises a plurality of counters that count the plurality of different types of detected errors.
19. The communication channel of claim 18 , wherein the correction logic generates the at least one feedback signal to adjust the at least one parameter when at least one of the counted plurality of different types of detected errors reaches a predetermined threshold.
20. The communication channel of claim 17 , wherein the correction logic generates the least one feedback signal to adjust the gain parameter, sampling offset parameter and equalization parameter.