Polar Code Soft-Bit Decoding from SC-List Paths

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current polar code decoding methods, such as SCL and CA-SCL, primarily produce hard bit outputs, limiting their application in scenarios requiring soft bit outputs, where existing soft bit decoding methods like BP and SCAN offer average performance, necessitating a method that achieves high decoding performance while providing soft bit outputs.

Innovation Solution

A polar code decoding method that performs successive cancellation-list (SC-list) decoding to generate multiple paths, calculates soft bit values based on these paths, and applies iterative decoding to enhance soft bit output, allowing for better adaptability in various application scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If SCL decoding or CA-SCL decoding is used, then decoding performance is greatly improved, but soft bit output capability is lost

Engineering Contradiction:
Improvedecoding performanceVSAvoidsoft bit output capability
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent segments the SCL decoding process into multiple independent paths, each maintaining its own state information. By preserving multiple candidate paths instead of selecting a single hard decision, the system can extract soft bit values from the path metrics and likelihood ratios while maintaining high decoding performance.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from hard bit domain to soft bit domain by introducing a new dimension of information representation. Instead of outputting binary decisions, the system outputs log-likelihood ratios that capture the confidence level of each bit decision, enabling both high performance and soft bit output capability.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If BP decoding or SCAN decoding is used, then soft bit output is achieved, but decoding performance is only average

Engineering Contradiction:
Improvesoft bit output capabilityVSAvoiddecoding performance
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent introduces an intermediary mechanism that bridges SCL hard bit decoding and soft bit output requirements. By using path metrics and likelihood ratios as intermediaries, the system translates the high-performance SCL decoding results into soft bit format without performance degradation.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If hard bit decoding is used, then decoding speed is fast, but adaptability to scenarios requiring soft bit output is limited

Engineering Contradiction:
Improvedecoding speedVSAvoidapplication scenario compatibility
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent makes the SCL decoder universal by enabling it to perform both traditional hard bit decoding and soft bit output functions. The same decoding framework supports multiple output formats, making it adaptable to various application scenarios including concatenated coding and interference cancellation while maintaining fast decoding speed.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3720023B1Polar code decoding method and communication device
Publication Date: 2024.02.07 HUAWEI TECH CO LTD
  • EP3720023B1 patent drawingFigure 1
  • EP3720023B1 patent drawingFigure 2
  • EP3720023B1 patent drawingFigure 3~4

AI summary

This application discloses a polar polar code decoding method and a communications device. The method includes: obtaining a to-be-decoded sequence, where a length of the to-be-decoded sequence is N, and N is a positive integer; performing successive cancellation-list SC-list decoding on the to-be-decoded sequence, to obtain L first paths and a weight value of each of the L first paths, where L is a positive integer, each of the L first paths includes M bits, and M is a positive integer less than or equal to N; calculating soft bit values of a first-phase decoding sequence based on the L first paths and the weight value of each of the L first paths, where a length of the first-phase decoding sequence is P, and P is a positive integer less than or equal to M; and decoding the soft bit values of the first-phase decoding sequence to obtain a decoding result. Implementation of this application can satisfy a soft bit output, and ensure high decoding performance.