Non-Binary LDPC Decoding With Partial Hard Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-binary Low Density Parity Check (LDPC) codes face high decoding complexity, which hinders their commercialization despite offering better performance than binary LDPC codes, especially in next-generation communication systems with evolving channel environments.
Innovation Solution
A method and apparatus for decoding non-binary LDPC codes that involves hard-decisioning all messages except one, combining these hard-decided values with the unhard-decided message to update the final output, reducing complexity while maintaining performance through the Extended Iterative Soft-reliability Based Majority-logic (EISBM) decoding algorithm.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-binary LDPC code is used to achieve better error correction performance and shorter code length, then decoding complexity increases significantly
Solution Approach 1:
The decoding process is segmented into two distinct phases: hard decision phase where all messages are hard-decided except one, and combination phase where hard-decided values are combined with the unhard-decided message. This segmentation reduces computational complexity while maintaining the error correction performance of non-binary LDPC codes.
Solution Approach 2:
The invention changes the parameter of message processing from full soft-decision to partial hard-decision. By hard-deciding all messages except one and then combining with the unhard-decided message, the system transforms the decoding parameter strategy to reduce complexity while preserving performance.
2Reliability
If full soft-decision decoding is used for non-binary LDPC code, then decoding performance is optimized but computational complexity becomes prohibitive for commercialization
Solution Approach 1:
Instead of performing full soft-decision decoding on all messages, the invention applies partial action by hard-deciding all messages except one. This partial approach maintains sufficient decoding performance while dramatically reducing computational complexity to enable commercialization.
Solution Approach 2:
The invention uses a simplified hard-decision approach for most messages (treating them as disposable approximations) and only applies the more complex soft-decision processing to a single message. This strategy achieves near-optimal performance at a fraction of the computational cost.
3Reliability
If non-binary code mapping to signal constellation is implemented to maximize performance, then error correction capability improves but decoding process becomes more complicated
Solution Approach 1:
The invention extracts and isolates the single message that requires soft-decision processing, while all other messages are processed through the simpler hard-decision path. This extraction approach maintains error correction capability while removing unnecessary complexity from the majority of the decoding process.
Data Source
AI summary
A method of decoding a non-binary Low Density Parity Check (LDPC) code is provided. The method includes a plurality of messages to perform hard decision for all messages except for one message, and combines the hard-decided values with the one message that is not hard-decided, to update a final output message.