Adaptive Parity Check Matrix for Iterative Block Code Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current soft decision decoding techniques for linear block codes, such as Reed Solomon (RS) codes, face limitations in achieving good performance-complexity trade-offs and generating reliable soft outputs, especially for high density parity check codes, due to computational complexity and the non-sparse nature of parity check matrices, which leads to error propagation and saturation of checks.
Innovation Solution
An iterative message passing algorithm is employed with an adaptive binary image of the parity check matrix, where the matrix is reduced to a sparse form based on bit reliabilities before each iteration, focusing on less reliable bits to prevent saturation and improve message passing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a standard iterative message passing algorithm is used with the original parity check matrix, then the decoding process can be implemented, but the non-sparse nature of the matrix leads to error propagation and saturation of checks, degrading decoding performance
Solution Approach 1:
The parity check matrix is dynamically adapted during the iterative decoding process. In each iteration, the matrix is modified based on the reliability information from previous iterations, specifically by zeroing out entries corresponding to reliable bits. This dynamic adaptation transforms the fixed, non-sparse matrix into an iteration-dependent sparse matrix, preventing error propagation while maintaining decoding effectiveness
Solution Approach 2:
The structure of the parity check matrix is changed as a parameter during decoding. The matrix transitions from its original dense form to a sparse form where certain entries are set to zero based on bit reliability metrics. This parameter change (sparsity) directly addresses the error propagation problem while the adaptation mechanism ensures decoding performance is maintained
2Reliability
If the parity check matrix is made sparse to prevent error propagation, then error correction capability is improved, but the adaptive modification increases computational complexity
Solution Approach 1:
The parity check matrix structure is modified by changing its parameters (entries) based on bit reliability. Reliable bits identify which matrix entries should be zeroed out, transforming the matrix into a sparse form that prevents error propagation. This parameter change approach achieves sparsity adaptively
Solution Approach 2:
The decoding process uses feedback from previous iterations to guide matrix adaptation. Reliability information from prior decoding steps feeds into the matrix modification process, creating an adaptive loop where past decoding performance informs future matrix structure. This feedback mechanism enables intelligent sparsification without exhaustive search
3Reliability
If iterative decoding is applied to high density parity check codes, then soft decision capability is achieved, but the non-sparse parity check matrix causes saturation of checks and error propagation
Solution Approach 1:
The parity check matrix is made dynamic and iteration-dependent rather than static. In each iteration, the matrix adapts its structure based on the reliability of bits from previous iterations. This dynamic sparsification prevents the saturation of checks that occurs with dense matrices, thereby eliminating error propagation while preserving soft decision capabilities
Solution Approach 2:
Different parts of the parity check matrix are treated differently based on local bit reliability characteristics. Entries corresponding to reliable bits are zeroed out, while entries corresponding to unreliable bits are maintained. This local differentiation creates a sparse matrix structure tailored to the specific error patterns in the received signal
Data Source
AI summary
A method of decoding linear block code uses an iterative message passing algorithm with a binary image of a parity check matrix of the linear block code, wherein the parity check matrix is adapted from one iteration to another based on the reliabilities of bits in the linear block code. The adaptation involves reducing a submatrix corresponding to the less reliable bits in the linear block code to a sparse nature before applying the message passing algorithm in each iteration. An apparatus that performs the method is also provided and several variations of the algorithm are also provided.