LDPC Parity Check Matrix Layout for Fewer Decoding Iterations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MoCA 2 parity check matrices have a lower triangular structure, leading to reduced connections with the codeword, increased decoding errors in parity bits, and potentially more decoding iterations.
Innovation Solution
A robust LDPC parity check matrix is designed with a systematic portion and a parity portion, where the weights of each column of the parity portion are the same, allowing for efficient data access, communication, and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a lower triangular structure is used for the parity check matrix, then the matrix structure is simple and systematic, but the connections with the codeword are reduced leading to increased decoding errors
Solution Approach 1:
The parity check matrix is divided into distinct blocks: a systematic portion with identity submatrices and a parity portion with structured submatrices. This segmentation allows the matrix to maintain systematic simplicity while incorporating connection patterns that improve decoding performance by ensuring sufficient connections between parity bits and codeword bits.
Solution Approach 2:
Different regions of the matrix are assigned different structures optimized for their specific functions. The systematic portion uses identity matrices for straightforward data bit placement, while the parity portion uses structured submatrices with specific connection patterns to ensure robust decoding connections without complicating the overall systematic structure.
2Ease of manufacture
If a lower triangular structure is used for the parity check matrix, then the encoding process is straightforward, but decoding iterations increase reducing efficiency
Solution Approach 1:
The matrix structure segments encoding and decoding functions into different regions. The systematic portion enables simple encoding operations, while the parity portion's structured submatrices are designed to facilitate efficient iterative decoding by creating favorable connection patterns that reduce the number of iterations required for convergence.
Solution Approach 2:
The matrix structure parameters are optimized to balance encoding simplicity with decoding efficiency. By carefully designing the submatrix dimensions, connection patterns, and weight distributions in the parity portion, the structure achieves fast decoding convergence while preserving the systematic encoding advantages.
3Reliability
If conventional randomly generated sparse matrices are used, then capacity-approaching performance is achieved, but the structure lacks systematic organization reducing ease of implementation
Solution Approach 1:
The matrix combines locally optimized structures with globally systematic organization. Each submatrix region is designed with specific connection patterns that contribute to capacity-approaching performance, while the overall arrangement maintains systematic organization with identity matrices in the systematic portion, enabling easy implementation.
Solution Approach 2:
The parity check matrix is constructed as a composite structure combining different types of submatrices (identity matrices, structured submatrices, and parity submatrices) with specific connection patterns. This composite structure achieves the performance benefits of randomly generated matrices while maintaining the implementation advantages of systematic organization.
Data Source
AI summary
An LDPC parity check matrix includes a systematic portion having a plurality of systematic elements and a parity portion having a plurality of parity elements. The value of each systematic element determines a cyclic shift to be applied to rows of an identity submatrix corresponding to that element. The value of each parity element determines a cyclic shift to be applied to rows of an identity submatrix corresponding to that element. The weights of two or more columns of the parity portion are the same.