Turbo Code List Decoding with LLR-Guided Candidate Tracing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional turbo decoders face challenges with high complexity and inconsistent path metrics due to segmentation/windowing of the trellis, leading to inefficiencies in identifying the most likely codewords and increased sorting complexity, especially in parallelized decoding settings.
Innovation Solution
The Approximate Maximum-Log List Algorithm (AMLLA) method identifies the least reliable bits based on decoder soft output, generates candidate codewords by flipping bits in the trellis representation, and performs forward and backward traces to determine candidate codewords, reducing the need for global list sorting and comparisons across all trellis stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional Maximum-Log List (MLL) decoding is applied to segmented trellis, then path metrics become inconsistent across segments, but the sorting complexity increases significantly
Solution Approach 1:
The patent applies segmentation by dividing the trellis decoding into independent stages, where each stage processes a subset of path metrics without requiring global sorting across the entire trellis. This allows parallel processing of segments while maintaining path metric consistency through localized selection operations.
Solution Approach 2:
Instead of performing complete global sorting of all path metrics across the entire trellis, the patent performs partial sorting operations on subsets of paths at each stage. This partial action reduces computational complexity while still identifying the most likely paths sufficient for decoding accuracy.
2Reliability
If list decoding generates multiple candidate codewords, then decoding error probability decreases, but computational complexity increases
Solution Approach 1:
The patent performs preliminary identification of the most likely path at each trellis stage before generating the full list of candidate codewords. This preliminary action prunes the search space early, allowing multiple candidates to be generated with reduced computational effort compared to exhaustive list decoding.
Solution Approach 2:
The patent applies different processing quality to different paths: the most likely path receives full processing attention at each stage, while less likely paths are processed with reduced complexity. This local quality differentiation maintains decoding reliability for the dominant path while reducing overall computational complexity.
Data Source
AI summary
A method, system, and non-transitory computer-readable recording medium of decoding a signal are provided. The method includes receiving signal to be decoded, where signal includes at least one symbol; decoding signal in stages, where each at least one symbol of signal is decoded into at least one bit per stage, wherein Log-Likelihood Ratio (LLR) and a path metric are determined for each possible path for each at least one bit at each stage; determining magnitudes of the LLRs; identifying K bits of the signal with smallest corresponding LLR magnitudes; identifying, for each of the K bits, L possible paths with largest path metrics at each decoder stage for a user-definable number of decoder stages; performing forward and backward traces, for each of the L possible paths, to determine candidate codewords; performing a Cyclic Redundancy Check (CRC) on the candidate codewords; and stopping after a first candidate codeword passes the CRC.