Signal decision equalization method and apparatus

Patent metadata

Field	Value
Publication number	US-12388686-B2
Application number	US-202318397782-A
Country	US
Kind code	B2
Filing date	Dec 27, 2023
Priority date	Jun 30, 2021
Publication date	Aug 12, 2025
Grant date	Aug 12, 2025

Abstract

A signal decision equalization method and apparatus are provided. The method includes: obtaining an input signal; determining a decision circuit of the input signal; obtaining a first group of decision thresholds and a first group of equalization expectations of the decision circuit; determining a decision value of the input signal based on the first group of equalization expectations, the first group of decision thresholds, and the input signal, and outputting the decision value; updating, based on the decision value and the input signal, a first equalization expectation that is in the first group of equalization expectations and that corresponds to the decision value to a second equalization expectation to obtain a second group of equalization expectations; and updating at least one decision threshold in the first group of decision thresholds based on the second equalization expectation to obtain a second group of decision thresholds.

First claim

What is claimed is:

1. A signal decision equalization method, comprising:
obtaining an input signal, wherein the input signal is a pulse amplitude modulation (PAM) signal or a pre-equalized signal that is output after pre-equalization is performed on the PAM signal;
determining a decision equalization of a decision circuit of the input signal;
obtaining a first group of decision thresholds and a first group of equalization expectations of the decision circuit;
determining a decision value of the input signal based on the first group of equalization expectations, the first group of decision thresholds, and the input signal;
outputting the decision value;
updating a first equalization expectation in the first group of equalization expectations to a second equalization expectation based on the decision value and the input signal to obtain a second group of equalization expectations, wherein the first equalization expectation is an equalization expectation corresponding to the decision value; and
updating at least one decision threshold in the first group of decision thresholds based on the second equalization expectation to obtain a second group of decision thresholds, wherein the second group of decision thresholds and the second group of equalization expectations are used for next decision equalization of the decision circuit.

2. The method according to claim 1 , wherein the updating the first equalization expectation in the first group of equalization expectations to the second equalization expectation comprises:
determining a decision error of the decision circuit based on the first equalization expectation and the input signal; and
updating the first equalization expectation to the second equalization expectation based on the decision error.

3. The method according to claim 2 , wherein the decision error is equal to the first equalization expectation minus the input signal.

4. The method according to claim 2 , wherein the updating the first equalization expectation to the second equalization expectation based on the decision error comprises:
updating the first equalization expectation to the second equalization expectation based on the decision error and a first formula, wherein the first formula is:
EV new =EV old −sign( e k ),

wherein EV new indicates the second equalization expectation, EV old indicates the first equalization expectation, e k indicates the decision error, and k indicates a current moment; or
updating the first equalization expectation to the second equalization expectation based on the decision error and a second formula, wherein the second formula is:
EV new =EV old −μ*( e k ),

wherein EV new indicates the second equalization expectation, EV old indicates the first equalization expectation, e k indicates the decision error, μ indicates a preset constant, and k indicates a current moment.

5. The method according to claim 1 , wherein the updating the first equalization expectation in the first group of equalization expectations to the second equalization expectation comprises:
obtaining a decision error based on the input signal and the first equalization expectation;
recording the decision error to generate a decision error set, wherein the decision error set comprises a decision error corresponding to the first equalization expectation; and
based on a quantity of decision errors in the decision error set reaching a preset quantity:
determining a final decision error based on the decision errors in the decision error set; and
updating the first equalization expectation to the second equalization expectation based on the final decision error.

6. The method according to claim 5 , wherein the final decision error is equal to a sum of a plurality of decision errors.

7. The method according to claim 1 , wherein the updating at least one decision threshold in the first group of decision thresholds comprises:
updating the at least one decision threshold in the first group of decision thresholds based on the second equalization expectation and a third equalization expectation to obtain the second group of decision thresholds, wherein the third equalization expectation is an equalization expectation that is in the second group of equalization expectations and that is adjacent to the second equalization expectation.

8. The method according to claim 1 , wherein the method further comprises:
receiving an initial convergence parameter output by a decision feedback equalizer (DFE) circuit, wherein the initial convergence parameter is used as an initial equalization expectation of the input signal during decision equalization.

9. The method according to claim 1 , wherein the method further comprises:
receiving an adaptive parameter output by a feed forward equalizer circuit, wherein the adaptive parameter is used as an initial equalization expectation of the input signal during decision equalization.

10. The method according to claim 1 , wherein the method further comprises:
determining a quantity of decision circuits and a quantity of decision thresholds based on a quantity N of levels of the PAM signal and a quantity M of equalized signals in a decision value set, wherein N is a positive integer and M is a positive integer.

11. The method according to claim 10 , wherein the determining the quantity of decision circuits and the quantity of decision thresholds comprises:
determining the quantity of decision circuits and the quantity of decision thresholds based on the quantity N of levels of the PAM signal and the quantity M of equalized signals in the decision value set by using a third formula, wherein the third formula is:
S=N M
P=N M *( N− 1)′

wherein S indicates the quantity of decision circuits, P indicates the quantity of decision thresholds, N indicates the quantity of levels of the PAM signal, and M indicates the quantity of equalized signals in the decision value set.

12. The method according to claim 1 , wherein the determining the decision value of the input signal comprises:
determining the decision value of the input signal based on the first group of decision thresholds, the first group of equalization expectations, and the input signal through maximum likelihood sequence estimation.

13. The method according to claim 1 , wherein the determining the decision circuit of the input signal comprises:
determining the decision equalization of the decision circuit of the input signal based on a decision value set, wherein the decision value set comprises at least one decision value, and an occurrence moment of the at least one decision value is earlier than an occurrence moment of the input signal.

14. A decision equalization apparatus, comprising:
a non-transitory memory storing instructions; and
a processor coupled to the non-transitory memory, wherein the instructions, when executed by the processor, cause the decision equalization apparatus to be configured to:
obtain an input signal, wherein the input signal is a pulse amplitude modulation (PAM) signal or a pre-equalized signal that is output after preprocessing is performed on the PAM signal;
determine a decision equalization of a decision circuit of the input signal;
obtain a first group of decision thresholds and a first group of equalization expectations of the decision circuit;
determine a decision value of the input signal based on the first group of equalization expectations, the first group of decision thresholds, and the input signal;
output the decision value;
update a first equalization expectation in the first group of equalization expectations to a sec

Assignees

• Huawei Tech Co Ltd

Inventors

• Wang Weiyu
• Li Huanlu
• Huang Zhilei
• Lu Yuchun
• Zhang Haiyang
• Zhou Qinyu

Classifications

• H04L25/4917 Primary

  using multilevel codes · CPC title

• H04L25/03057 Primary

  with a recursive structure (H04L25/03031 takes precedence) · CPC title

• H04L27/06

  Demodulator circuits; Receiver circuits · CPC title

• H04L25/03006 Primary

  Arrangements for removing intersymbol interference · CPC title