Frequency domain equalizer for optical communications

Patent metadata

Field	Value
Publication number	US-9917712-B2
Application number	US-201615068888-A
Country	US
Kind code	B2
Filing date	Mar 14, 2016
Priority date	Mar 14, 2016
Publication date	Mar 13, 2018
Grant date	Mar 13, 2018

Abstract

A method decodes an optical signal transmitted over an optical channel from a transmitter to a receiver. The receiver receives the transmitted optical signal to produce a digital signal which is filtered in the frequency domain for compensating static effects and/or dynamic effects. The filtering is performed in the frequency domain, while the frequency coefficients of the filter are updated in the time domain by updating at least some of time coefficients of the filter and transforming the time coefficients into the frequency domain.

First claim

We claim:

1. A method for decoding an optical signal transmitted over an optical channel from a transmitter to a receiver, comprising:
transforming the optical signal received in a time domain over the optical channel into a frequency domain to produce a discrete spectrum;
updating at least some of time coefficients of a filter for filtering in the time domain;
transforming the time coefficients into the frequency domain to produce frequency coefficients of the filter for filtering in the frequency domain;
filtering the discrete spectrum in the frequency domain using the frequency coefficients of the filter;
transforming the filtered discrete spectrum into a digital signal in the time domain; and decoding symbols of the digital signal, wherein steps of the method are performed using a processor of the receiver, wherein the filter includes a static filter for filtering static effects of the optical channel and a dynamic filter for filtering dynamic effects of the optical channel, wherein the frequency coefficients of the static filter are constant, and wherein the frequency coefficients of the dynamic filter are updated in the time domain.

2. The method of claim 1 , wherein the updating comprises:
transforming the discrete spectrum into the time domain to produce a first signal;
transforming the filtered digital spectrum into the time domain to produce a second signal;
determining time coefficients of the filter in the time domain based on a function of the first and the second signals; and
transforming at least part of the time coefficients in the frequency domain to update the frequency coefficients.

3. The method of claim 2 , wherein the transforming into the frequency domain is performed using a Fourier transform, and wherein the transforming into the time domain is performed using an inverse Fourier transform.

4. The method of claim 2 , wherein the transforming the time coefficients comprises:
updating the time coefficients by replacing a subset of time coefficients with zeros; and
transforming the updated time coefficients into the frequency domain.

5. The method of claim 2 , wherein the discrete spectrum includes a set of overlapping blocks of samples having more than one but less than two samples per symbol, further comprising:
resampling the filtered discrete spectrum to an integer number of samples per symbol;
transforming the resampled and filtered discrete spectrum in the time domain to produce the second signal;
determining an error between at least some samples in the second signal and a set of predetermined values; and
updating the time coefficients using the first signal adjusted according to the error.

6. The method of claim 1 , wherein the optical signal includes a set of pilot symbols, and wherein the updating comprises:
transforming the discrete spectrum into the time domain to produce a first signal;
transforming the filtered discrete spectrum into the time domain to produce a second signal;
determining time coefficients of the filter in the time domain based on a function of the first signal, values of the pilot symbols in the second signal, and predetermined values of the pilot symbols;
updating the time coefficients by replacing a subset of time coefficients with zeros; and
transforming the updated time coefficients into the frequency domain.

7. A method for decoding an optical signal transmitted over an optical channel from a transmitter to a receiver, comprising:
receiving the optical signal transmitted over the optical channel to produce a digital signal in a time domain;
partitioning the digital signal in the time domain into a set of overlapping blocks of samples;
transforming a block of samples into a frequency domain to produce a discrete spectrum;
filtering the discrete spectrum with a static filter to produce a first filtered spectrum;
filtering the first filtered spectrum with a dynamic filter to produce a second filtered spectrum;
transforming the first filtered spectrum into the time domain to produce a first signal;
transforming the second filtered spectrum into the time domain to produce a second signal;
determining time coefficients of the dynamic filter in the time domain based on a difference between at least some elements of the first and the second signals;
updating the time coefficients by replacing a subset of time coefficients with zeros;
transforming the updated time coefficients into the frequency domain;
updating frequency coefficients of the dynamic filter in the frequency domain with the updated time coefficients transformed into the frequency domain; and
reconstructing the optical signal in the time domain using a set of second digital signals corresponding to the set of overlapping blocks of samples, wherein steps of the method are performed using a processor of the receiver.

8. The method of claim 7 , wherein the block of samples includes more than one but less than two samples per symbol, further comprising:
resampling the second filtered discrete spectrum to an integer number of samples per symbol;
transforming the resampled and filtered second discrete spectrum in the time domain to produce the second signal;
determining an error between at least some samples in the second signal and a set of predetermined values; and
updating the time coefficients using the first signal adjusted according to the error.

9. The method of claim 7 , wherein the optical signal includes pilot symbols, and wherein the time coefficients are determined based on a function of the first signal, values of the pilot symbols in the second signal, and predetermined values of the pilot symbols.

10. A receiver for receiving and decoding an optical signal transmitted over an optical channel, comprising:
a frond end including an optic and an electronic for receiving the optical signal in a time domain over the optical channel and transforming the optical signal into a frequency domain to produce a discrete spectrum, wherein the optical signal includes pilot symbols;
a digital signal processor for filtering the discrete spectrum in the frequency domain using frequency coefficients of a filter and for transforming the filtered discrete spectrum into a digital signal in the time domain, wherein the processor
updates at least some of time coefficients of the filter for filtering in the time domain and transforms the time coefficients into the frequency domain to produce the frequency coefficients of the filter;
transforms the discrete spectrum into the time domain to produce a first signal;
transforms the filtered digital spectrum into the time domain to produce a second signal;
determines time coefficients of the filter in the time domain based on a function of the first signal, values of the pilot symbols in the second signal, and predetermined values of the pilot symbols;
updates the time coefficients by replacing a subset of time coefficients with zeros; and
transforms the updated time coefficients into the frequency domain; and
a decoder for decoding symbols of the digital signal, wherein the filter includes a static filter for filtering static effects of the optical channel and a dynamic filter for filtering dynamic effects of the optical channel, wherein the frequency coefficients of the static filter are constant, and wherein the frequency coefficients of the dynamic filter are updated in the time domain.

Assignees

• Mitsubishi Electric Res Laboratories Inc

Inventors

• Millar David
• Paskov Milen
• Koike Akino Toshiaki
• Parsons Kieran
• Kojima Keisuke

Classifications

• H04B10/6162

  Compensation of polarization related effects, e.g., PMD, PDL · CPC title

• H04B10/6161 Primary

  Compensation of chromatic dispersion · CPC title

• H04B10/61

  Coherent receivers · CPC title

• H04L25/03891 Primary

  Spatial equalizers (MIMO diversity systems H04B7/0413) · CPC title