Non-Binary Codeword Decoding with Compressed Reliability Messages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-binary error correction code decoding is complex and requires further reduction in complexity and performance improvement, as existing algorithms like Extended Min-Sum and Trellis Extended Min-Sum do not fully address the need for reduced communication load and latency in decoding processes.
Innovation Solution
A method is proposed where a non-binary constraint node sends messages to a variable node comprising measures of reliability for selected candidate symbols, reducing communication load by requesting and including only the most reliable symbols, known as nR requested symbols and nB best symbols, which are selected using compression and decompression blocks, and setting default values for remaining symbols, allowing for reduced message size and complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional non-binary decoding algorithms (Extended Min-Sum, Trellis Extended Min-Sum) are used, then decoding capability is provided, but communication load and latency remain high
Solution Approach 1:
The patent extracts and transmits only the most relevant information from check node to variable node: specifically, the indices and reliability values of the nB most reliable candidate symbols and nR requested candidate symbols. This selective extraction reduces the communication load between nodes while maintaining decoding accuracy, directly addressing the contradiction between throughput and complexity.
Solution Approach 2:
The patent changes the parameter of message representation from transmitting full reliability information for all candidate symbols to transmitting only critical parameters (indices and reliability values of selected symbols). This parameter optimization reduces message size and processing complexity while preserving essential decoding functionality.
2Reliability
If full message information is transmitted between nodes, then decoding accuracy is maintained, but message size and processing load increase
Solution Approach 1:
The patent extracts only the most critical message components: the indices and reliability values of nB most reliable symbols and nR requested symbols. This selective extraction maintains decoding accuracy by preserving the most important information while reducing message size by transmitting only essential data elements.
Solution Approach 2:
The patent transmits a partial set of message information (only the most reliable and requested symbols) rather than complete information for all candidate symbols. This partial action is sufficient to maintain decoding accuracy while significantly reducing the quantity of transmitted data.
Data Source
Figure 1A~1C
Figure 2~3A
Figure 3B~4
AI summary
The invention relates to a method for decoding a codeword, obtained by implementing a code (such as a non-binary LDPC, a non-binary polar code or a non-binary turbo code) comprising at least one non-binary constraint node C, wherein said decoding method implements, for at least one connection between a non-binary variable node Vj and the non-binary constraint node C, at least one decoding iteration comprising sending (21) a message from the non-binary constraint node C to the non-binary variable node Vj , said message comprising: - nR measures of reliability obtained for nR candidate symbols selected by the non-binary variable node Vj or by a first compression block implemented between the non-binary variable node Vj and the non-binary constraint node C, - nB measures of reliability obtained for nB candidate symbols selected by the non-binary constraint node C or by a second compression block implemented between the non-binary constraint node C and the non-binary variable node Vj, with 1 ≤ nB < q, 1 ≤ nR < q and nR + nB < q.