Method and apparatus for residual phase noise compensation

Patent metadata

Field	Value
Publication number	US-9537683-B1
Application number	US-201514876246-A
Country	US
Kind code	B1
Filing date	Oct 6, 2015
Priority date	Oct 6, 2015
Publication date	Jan 3, 2017
Grant date	Jan 3, 2017

Abstract

A method and apparatus for performing residual phase noise compensation is described. A coarse carrier compensation of a received modulated signal is performed to obtain a coarse carrier compensated signal and a trellis-based residual carrier recovery is performed to estimate a residual phase noise of the coarse carrier compensated signal. The coarse carrier compensated signal is compensated based on the estimated residual phase noise.

First claim

What is claimed is:

1. A method of residual carrier recovery of a received modulated signal performed by a digital signal processing apparatus implemented at least partially in hardware, the method comprising:
performing a coarse carrier compensation of the received modulated signal to obtain a coarse carrier compensated signal;
performing a trellis-based residual carrier recovery to estimate a residual phase noise of the coarse carrier compensated signal; and
compensating the coarse carrier compensated signal based on the estimated residual phase noise, wherein performing the trellis-based residual carrier recovery comprises:
constructing a trellis comprising L trellis stages and M states corresponding to M possible residual phase noise values of the coarse carrier compensated signal; and
traversing the L trellis stages of the trellis to trace a most likely residual phase noise as the residual phase noise of the coarse carrier compensated signal.

2. The method as defined in claim 1 wherein the most likely residual phase noise is calculated based on a survived path and at least one competitive path, wherein the survived path is traced back from a state of a state metric having an extremal value in a L th trellis stage, and the competitive path is traced back from a state of a state metric having a second or higher order extremal value in the L th trellis stage.

3. The method as defined in claim 1 , wherein the received modulated signal comprises a plurality of symbols and performing the trellis-based residual carrier recovery further comprises:
grouping the plurality of symbols into a plurality of super-symbols, each super-symbol being made up of a group of consecutive symbols; and
performing a residual carrier recovery to estimate a residual phase noise for each of the super-symbols, the estimated residual phase noise of each super-symbol being used as an estimated residual phase noise for each of the consecutive symbols making up the super-symbol.

4. The method as defined in claim 1 , wherein constructing the trellis further comprises, for each state in the trellis,
phase-rotating the coarse carrier compensated signal according to the possible residual phase noise corresponding to the state to obtain a phase-rotated signal;
mapping the phase-rotated signal to a constellation point in a modulation constellation of the received modulated signal;
calculating a normalized Euclidean distance between the phase-rotated signal and the constellation point;
calculating a probability of a phase noise jump corresponding to a branch leading to the state; and
calculating a branch metric of the branch based on a sum computation of the normalized Euclidean distance and the probability of the phase noise jump.

5. The method as defined in claim 4 , wherein the received modulated signal is modulated with a plurality of symbols and the plurality of symbols is grouped into a plurality of super-symbols, each super-symbol being made up of a group of N consecutive symbols, and calculating the normalized Euclidean distance further comprises:
calculating an intermediate normalized Euclidean distance between the phase-rotated signal of each of the N symbols making up the super-symbol and a corresponding constellation point; and
calculating an average of the calculated intermediate normalized Euclidean distances.

6. The method as defined in claim 5 , wherein calculating the average comprises calculating a weighted average.

7. The method as defined in claim 4 , wherein the calculated normalized Euclidean distance is averaged over a plurality of neighboring symbols.

8. The method as defined in claim 5 , wherein the calculated normalized Euclidean distance is averaged over a plurality of neighboring super-symbols.

9. The method as defined in claim 4 , wherein calculating the normalized Euclidean distance between the phase-rotated signal and the constellation point further comprises:
calculating an intermediate normalized Euclidean distance between the phase-rotated signal and each of a plurality of constellation points closest to the phase-rotated signal; and
calculating the normalized Euclidean distance based on the calculated intermediate normalized Euclidean distances.

10. The method as defined in claim 1 , wherein the possible residual phase noises corresponding to the states of the trellis have uniformly distributed values.

11. The method as defined in claim 1 , wherein the possible residual phase noises corresponding to the states of the trellis have uniformly distributed arctan values.

12. The method as defined in claim 1 , wherein the residual carrier recovery is performed in a dual-polarized coherent optical system and performing the trellis-based residual carrier recovery comprises estimating a residual phase noise of each of two orthogonal polarizations of the coarse carrier compensated signal separately.

13. The method as defined in claim 12 , wherein performing the trellis-based residual carrier recovery comprises:
calculating a first estimated residual phase noise of a first polarization of the coarse carrier compensated signal and a second estimated residual phase noise of a second polarization of the coarse carrier compensated signal; and
estimating the residual phase noise of the coarse carrier compensated signal based on an average of the first and the second estimated residual phase noises.

14. The method as defined in claim 1 , wherein the trellis-based residual carrier recovery is configured to propagate the trellis from a trellis stage to a next trellis stage based on a reduced set of states, wherein the reduced set of states in each stage is selected based on state metric values.

15. A digital signal processing apparatus implemented at least partially in hardware, the digital signal processing apparatus comprising:
a coarse carrier recovery module for performing a coarse carrier compensation of a received modulated signal; and
a trellis-based processing module for estimating and compensating a residual phase noise of the received modulated signal,
wherein the trellis-based processing module is configured to utilize a trellis comprising L trellis stages and M states corresponding to M possible residual phase noise values of a received symbol, and wherein the trellis-based processing module is configured to traverse the trellis to trace a most likely residual phase noise as the residual phase noise of the received modulated signal.

16. The digital signal processing apparatus as defined in claim 15 , wherein each state of the trellis representing a possible residual phase noise of a super-symbol and each super-symbol being made up of N consecutive symbols, and wherein the trellis-based processing module is configured to traverse the trellis to trace a most likely residual phase noise for each super-symbol.

17. The digital signal processing apparatus as defined in claim 15 , wherein the trellis-based processing module is configured to, for each state in the trellis,
phase-rotate an output of the coarse carrier recovery module according to the possible residual phase noise corresponding to the state;
map the phase-rotated output of the coarse carrier recovery module to a constellation point in a modulation constellation of the received modulated signal;
calculate a normalized Euclidean distance between the phase-rotated output of the coarse carrier recovery module and the constellation point;
calculate a probability of a phase noise jump corresponding to a branch leading to the state; and
calculate a branch metric of the branch based on a sum computation of

Assignees

• Zamani Mahdi
• Najafi Hossein
• Yao Demin
• Mitra Jeebak
• Li Chuandong
• Zhang Zhuhong
• Huawei Tech Co Ltd

Inventors

• Zamani Mahdi
• Najafi Hossein
• Yao Demin
• Mitra Jeebak
• Li Chuandong
• Zhang Zhuhong

Classifications

• H04L25/03203

  Trellis search techniques · CPC title

• H04L25/03273 Primary

  with carrier recovery circuitry · CPC title

• H04L27/2659

  Coarse or integer frequency offset determination and synchronisation · CPC title

• H04L27/3836

  in which the carrier is recovered using the received modulated signal or the received IF signal, e.g. by detecting a pilot or by frequency multiplication · CPC title

• H04L27/0014

  Carrier regulation (of chaotic carriers H04L27/001; for multicarrier receivers H04L27/2657) · CPC title