Structured LDPC Base Matrix for HARQ and High-Throughput Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional structured LDPC codes struggle to support incremental redundancy HARQ, provide ultra-high throughput, and achieve performance close to turbo codes while maintaining sufficient flexibility and low complexity.
Innovation Solution
The design of a structured LDPC code with a specific base matrix structure that includes multiple submatrices, such as upper-left submatrices Hb1 and Hb2, and an expansion factor Z, which supports various code rates and lengths, enabling efficient encoding and decoding operations with improved girth values for better performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional structured LDPC codes are used, then encoding and decoding operations can be performed, but they cannot support incremental redundancy HARQ and achieve ultra-high throughput with performance close to turbo codes
Solution Approach 1:
The base matrix is divided into multiple submatrices (first submatrix, second submatrix, third submatrix, fourth submatrix) with specific structural characteristics. This segmentation allows different parts of the code to serve different functions, enabling both high reliability through optimized error correction and adaptability for incremental redundancy HARQ operations.
Solution Approach 2:
The patent employs dynamic code rate adjustment capability where the same base matrix structure can support multiple code rates (e.g., 1/3, 2/3, 3/4) by selectively using different submatrices or combinations thereof. This dynamic adaptability enables the system to switch between different operating modes to support incremental redundancy HARQ while maintaining ultra-high throughput.
2Reliability
If complex base matrix structures are used to achieve better performance, then decoding performance improves, but encoding and decoding complexity increases
Solution Approach 1:
Different submatrices within the base matrix have different local structures optimized for specific purposes. For example, certain submatrices are designed with properties that improve girth values and decoding convergence, while other submatrices maintain sparsity to reduce complexity. This local optimization allows the overall system to achieve high decoding performance without uniformly increasing complexity across the entire matrix.
Solution Approach 2:
The base matrix structure employs a nested arrangement where submatrices are organized in a hierarchical pattern. This nesting allows the decoder to process information in a structured manner, potentially exploiting the hierarchical structure to reduce computational complexity while maintaining the performance benefits of the complex overall structure.
3Reliability
If the base matrix includes multiple submatrices with specific structures, then girth values improve and performance increases, but the device complexity and implementation difficulty increase
Solution Approach 1:
The base matrix employs asymmetric submatrix structures where different submatrices have different dimensions and properties. This asymmetry is deliberately designed to maximize girth values and improve decoding performance. The asymmetric structure is implemented through systematic construction rules that, while complex in appearance, follow deterministic patterns that can be pre-computed and stored, reducing implementation difficulty.
Data Source
AI summary
Provided is an encoding method and device and a decoding method and device for structured LDPC. The encoding method includes: determining a base matrix used for encoding and performing an LDPC encoding operation on a source information bit sequence according to the base matrix and an expansion factor Z corresponding to the base matrix to obtain a codeword sequence, where Z is a positive integer. The base matrix includes multiple submatrices and the submatrices include an upper-left submatrix Hb1 and an upper-left submatrix Hb2, and the upper-left submatrix Hb1 is an upper-left submatrix of the upper-left submatrix Hb2.