System and method for interference cancellation in optical transmission

Patent metadata

Field	Value
Publication number	US-11558115-B1
Application number	US-202117375831-A
Country	US
Kind code	B1
Filing date	Jul 14, 2021
Priority date	Jul 14, 2021
Publication date	Jan 17, 2023
Grant date	Jan 17, 2023

Abstract

Methods and systems for low complexity interference cancellation in multichannel optical transmission. Local or self-iteration is performed one or more times between an expected propagation decision feedback equalizer and a soft demapper. Following local iteration, a soft decision forward error correction decoder determines bit log-likelihood ratios, which are fed back to the expected propagation decision feedback equalizer and soft demapper for further self-iteration. Global iteration involving the decoder can also be performed one or more times before a bitstream is decoded.

First claim

The invention claimed is:

1. A method of interference cancellation in a signal received over a communication channel, the method comprising:
for at least one i-th global iteration, wherein i is an integer incremented from 1 to m:
for at least one j-th local iteration, wherein j is an integer incremented from 1 to n:
filtering the signal using a filter to generate a j-th equalized signal of the i-th global iteration; and
generating soft symbols from the j-th equalized signal of the i-th global iteration using a demapper to produce a j-th estimated soft symbol probability block of the i-th global iteration, wherein an n-th estimated soft symbol probability block of the i-th global iteration becomes an i-th local estimated soft symbol probability block;

converting the i-th local estimated soft symbol probability block to an i-th block of forward error-encoded bit log-likelihood ratios;
soft error decoding based on the i-th block of forward error-encoded bit log-likelihood ratios to generate an i-th error-decoded block of bit log-likelihood ratios; and
when i is less than m, generating an i-th global estimated soft symbol probability block based on the i-th local estimated soft symbol probability block and the i-th error-decoded block of bit log-likelihood ratios; and

following completion of the global iteration, decoding a bit stream from a m-th error-decoded block of bit log-likelihood ratios.

2. The method of claim 1 , further comprising, for the at least one j-th local iteration, updating a plurality of coefficients of the filter based on the j-th estimated soft symbol probability block of the i-th global iteration.

3. The method of claim 2 , further comprising, for the at least one i-th global iteration, when i is less than m, updating the plurality of coefficients of the filter based on the i-th global estimated soft symbol probability block.

4. The method of claim 1 , wherein m is at least 2.

5. The method of claim 1 , wherein n is at least 2.

6. The method of claim 1 , wherein generating the soft symbols further comprises computing a mean and covariance of a distribution of symbols in the respective estimated soft symbol probability block, and wherein the estimating is based on a measure of the soft symbol probabilities.

7. The method of claim 6 , further comprising computing a noise covariance matrix based on the measure of the soft symbol probabilities, wherein computing the noise covariance matrix comprises slicing each respective symbol in the respective equalized signal to generate a respective sliced symbol and subtracting each respective symbol from the respective sliced symbol.

8. The method of claim 7 , further comprising updating a plurality of coefficients of the filter based on the noise covariance matrix.

9. The method of claim 1 , wherein generating the i-th global estimated soft symbol probability block comprises combining, using a maximum a posteriori principle, the estimated soft symbol probability block with the error-decoded block of bit log-likelihood ratios as decoded.

10. The method of claim 1 , wherein the signal comprises a first and second polarized signal, further comprising performing the global iteration and the local iteration for each of the first and second polarized signal, respectively.

11. The method of claim 1 , wherein prior to the global iteration, the method further comprises estimating inter-symbol interference by computing a plurality of coefficients of the filter.

12. An apparatus comprising:
an equalizer configured to:
repeatedly filter a signal using a filter to generate a plurality of equalized signals;

a soft demapper configured to:
generate soft symbols from each of the plurality of equalized signals to produce a plurality of local estimated soft symbol probability blocks;

a decoder configured to:
convert one or more of plurality of local estimated soft symbol probability blocks to one or more blocks of forward error-encoded bit log-likelihood ratios;
soft error decode based on each of the one or more blocks of forward error-encoded bit log-likelihood ratios to generate one or more error-decoded blocks of bit log-likelihood ratios;
generate one or more global estimated soft symbol probability blocks based on the local estimated soft symbol probability block and the error-decoded block of bit log-likelihood ratios; and
decode a bit stream from at least one of the error-decoded block of bit log-likelihood ratios.

13. The apparatus of claim 12 , wherein each of the plurality of equalized signals is filtered using a plurality of filter coefficients of the filter, and wherein the plurality of filter coefficients of the filter are based on the plurality of local estimated soft symbol probability blocks.

14. The apparatus of claim 13 , wherein at least a subset of the plurality of filter coefficients are based on the one or more global estimated soft symbol probability blocks.

15. The apparatus of claim 12 , wherein the equalizer and the soft demapper generate the plurality of equalized signals and the soft symbols at least twice.

16. The apparatus of claim 12 , wherein the decoder generates the one or more global estimated soft symbol probability blocks at least twice.

17. The apparatus of claim 12 , wherein generating the soft symbols comprises computing a mean and covariance of a distribution of symbols in the respective estimated soft symbol probability block, and wherein the estimating is based on a measure of the soft symbol probabilities.

18. The apparatus of claim 17 , wherein a plurality of filter coefficients are updated by computing a noise covariance matrix based on the measure of the soft symbol probabilities, wherein computing the noise covariance matrix comprises slicing each respective symbol in the respective equalized signal to generate a respective sliced symbol and subtracting each respective symbol from the respective sliced symbol.

19. The apparatus of claim 12 , wherein generating the one or more global estimated soft symbol probability blocks comprises combining, using a maximum a posteriori principle, the estimated soft symbol probability block with the error-decoded block of bit log-likelihood ratios as decoded.

20. The apparatus of claim 12 , further comprising an inter-symbol interference channel estimator configured to estimate inter-symbol interference and update a plurality of filter coefficients.

Assignees

• Huawei Tech Co Ltd

Inventors

• Mitra Jeebak
• Jana Mrinmoy
• Lampe Lutz Hans-Joachim
• Wang Jin
• Li Chuandong

Classifications

• H04J14/0224

  Irregular wavelength spacing, e.g. to accommodate interference to all wavelengths · CPC title

• H04B10/2543 Primary

  due to fibre non-linearities, e.g. Kerr effect · CPC title

• H04B10/58

  Compensation for non-linear transmitter output · CPC title

• H04J14/0298

  with sub-carrier multiplexing [SCM] · CPC title

• H04B10/2569

  due to polarisation mode dispersion [PMD] · CPC title