LLR Biasing in Non-Binary LDPC Decoding for Near-Codewords
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional non-binary iterative decoders for LDPC codes face challenges in efficiently detecting and correcting near-codewords, leading to high power consumption and low throughput, making them unsuitable for low-power applications like NAND flash controllers.
Innovation Solution
The proposed solution involves using a soft output Viterbi algorithm to generate incremental hard decision values and LLR amplitude values, which are selectively transmitted to an LDPC decoder, allowing for reduced power consumption by calculating the complete hard decision value only once and updating it during subsequent iterations, and utilizing a memory mask to minimize power access during LLR biasing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If LLR values for all LDPC symbols are calculated and set for every LLR biasing iteration, then near-codeword detection capability is improved, but power consumption increases and throughput decreases
Solution Approach 1:
The patent extracts only the necessary LLR values (those corresponding to unsatisfied check nodes) from the full set of LDPC symbols, rather than processing all symbols. This selective extraction reduces computational load and power consumption while maintaining the ability to detect and correct near-codeword errors.
Solution Approach 2:
The patent segments the LLR calculation process by identifying and processing only specific symbols that require correction based on syndrome check results. Instead of uniformly processing all LDPC symbols, the system divides the work into necessary corrections only, improving efficiency.
2Reliability
If LLR values for all LDPC symbols are calculated and set for every LLR biasing iteration, then near-codeword detection capability is improved, but execution time increases
Solution Approach 1:
The patent extracts only the necessary LLR values (those corresponding to unsatisfied check nodes) from the full set of LDPC symbols, rather than processing all symbols. This selective extraction reduces computational load and power consumption while maintaining the ability to detect and correct near-codeword errors.
Solution Approach 2:
The patent performs preliminary syndrome checks to identify which check nodes are unsatisfied before proceeding with LLR calculations. This preliminary action allows the system to pre-determine which symbols need correction, avoiding unnecessary calculations and reducing execution time.
3Reliability
If conventional iterative decoding is used for deeply biased near-codewords, then decoding completeness is improved, but power consumption increases
Solution Approach 1:
The patent implements a self-service mechanism where the syndrome check results directly identify which symbols need correction, and the system automatically focuses computational resources only on those specific symbols. This self-directed approach eliminates wasted computation on already-correct symbols, reducing power consumption while maintaining decoding completeness.
Solution Approach 2:
The patent applies local quality by providing different treatment to different symbols based on their correction needs. Symbols corresponding to unsatisfied check nodes receive full LLR calculation and correction attention, while other symbols are skipped. This localized processing maintains decoding effectiveness while reducing overall power consumption.
Data Source
AI summary
Systems, methods, and apparatus are provided for iteratively decoding a codeword. Once a codeword is received, the codeword is processed to generate an incremental hard decision value and a log likelihood ratio amplitude value. These values are generated by processing the codeword using a soft output Viterbi algorithm. A faulty symbol in the codeword is identified. A complete hard decision value is generated using the incremental hard decision value. The LLR amplitude value and complete hard decision value corresponding to the identified faulty symbol are selectively provided to a decoder and the decoder uses these values to decode the codeword.