Statistics adaptive soft decision forward error correction in digital communication

Patent metadata

Field	Value
Publication number	US-9762351-B2
Application number	US-201414220049-A
Country	US
Kind code	B2
Filing date	Mar 19, 2014
Priority date	Mar 20, 2013
Publication date	Sep 12, 2017
Grant date	Sep 12, 2017

Abstract

A digital communication receiver uses a maximum likelihood sequence estimation stage to recover symbols from digitized sample values of a received signal. A probability density function is calculated and used to improve a soft decision forward error correction calculation. The results of error decoding, which represent error corrected data bits, are further used to improve the probability density function calculation.

First claim

What is claimed is what is described and illustrated, including:

1. A method of generating data bits from a received signal, comprising:
processing the received signal to generate a sequence of signal values, wherein the received signal comprises a duobinary signal;
converting the sequence of signal values to soft data value estimates using a soft decision maximum likelihood sequence estimation technique in which an estimation probability is associated with each data value estimate;
estimating a probability density function of data values numerically based on a histogram of the soft data value estimates, the probability density function including two peaks showing a spread of probability values for different combinations of signal values from the duobinary signal; and
forward error correcting, using the soft data value estimates and the probability density function, the data values by a soft decision forward error correction technique to generate data bits, wherein
the estimated probability density function is fed into the forward error correcting for a log likelihood ratio (LLR) calculation, and
the data bits are fed back to the estimating the probability density function to iteratively improve accuracy of the probability density function estimation.

2. The method of claim 1 , wherein the received signal is an optical signal.

3. An apparatus generating data bits from a received signal, comprising:
a receive chain that processes the received signal to generate a sequence of signal values,. wherein the received signal comprises a duobinary signal;
a maximum likelihood sequence estimation (MLSE) module that converts the sequence of signal values to soft data value estimates using a soft decision MLSE technique in which an estimation probability is associated with each data value estimate;
a probability density function (PDF) module that estimates a probability density function of data values numerically based on a histogram of the soft data value estimates, the probability density function including two peaks showing a spread of probability values for different combinations of signal values from the duobinary signal; and
a forward error decoding (FEC) module that performs soft-decision forward error correction, using the soft data value estimates and the probability density function, the data values to generate data bits, wherein
the probability density function is fed into the forward error decoding module for a log likelihood ratio (LLR) calculation, and
the data bits are fed back to the PDF module to iteratively improve accuracy of the probability density function estimation.

4. The apparatus of claim 3 , wherein the received signal is an optical signal.

5. A data reception apparatus comprising:
a memory for storing instruction code; and
a processor that executes the instruction code to implement a method of generating data bits from a received signal that comprises a duobinary signal, the method comprising:
processing the received signal to generate a sequence of signal values;
converting the sequence of signal values to soft data value estimates using a soft decision maximum likelihood sequence estimation technique in which an estimation probability is associated with each data value estimate;
estimating a probability density function of data values numerically based on a histogram of the soft data value estimates, the probability density function including two peaks showing a spread of probability values for different combinations of signal values from the duobinary signal; and
forward error decoding, using the soft data value estimates and the probability density function, the data values by a soft decision forward error correction technique to generate data bits, wherein
the estimated probability density function is fed into the forward error decoding for a log likelihood ratio (LLR) calculation, and
the data bits are fed back to the estimating the probability density function to iteratively improve accuracy of the probability density function estimation.

6. The data reception apparatus of claim 5 , wherein the received signal is an optical signal.

7. An optical communication system comprising:
an optical signal transmitter configured to transmit an error correction coded optical signal; and
an optical signal receiver configured to:
receive the error correction coded optical signal;
process the received signal to generate a sequence of signal values, wherein the received signal comprises a duobinary signal;
convert the sequence of signal values to soft data value estimates using a soft decision maximum likelihood sequence estimation technique in which an estimation probability is associated with each data value estimate;
estimate a probability density function of data values numerically based on a histogram of the soft data value estimates, the probability density function including two peaks showing a spread of probability values for different combinations of signal values from the duobinary signal; and
forward error decode, using the soft data value estimates and the probability density function, the data values by a soft decision forward error correction technique to generate data bits, wherein
the probability density function is fed into the forward error decode for a log likelihood ratio (LLR) calculation, and
the data bits are fed back to estimate the probability density function to iteratively improve accuracy of the probability density function estimation.

8. The optical communication system of claim 7 , wherein the received signal is an optical signal.

Assignees

• Zte Usa Inc

Inventors

• Cai Yi
• Jia Zhensheng

Classifications

• H04L25/03286

  with channel-decoding circuitry · CPC title

• H04L1/0054 Primary

  Maximum-likelihood or sequential decoding, e.g. Viterbi, Fano, ZJ algorithms · CPC title