Irregular LDPC Column Processing with Convergence Zone Skipping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Irregular low density parity check (LDPC) codes used in non-volatile memory systems, such as NAND flash memory, face challenges in efficient decoding due to varying column weights, leading to increased decoding iterations and reduced data integrity.
Innovation Solution
A method for column processing of irregular LDPC codes involves grouping bits by their degrees of convergence, classifying these groups based on decoding iteration metrics, dividing the convergence time into zones for selective processing, and skipping decoding in non-converging zones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If irregular LDPC codes are used to improve data integrity in non-volatile memory systems, then error correction capability is improved, but decoding complexity and time increase due to varying column weights
Solution Approach 1:
The patent segments the decoding process into multiple zones based on the convergence behavior of different degree groups. Bits are classified into degree groups (e.g., degree 1, degree 2, degree 3) and processed in different zones during decoding iterations. This segmentation allows the decoder to handle different bit types with different processing strategies, reducing overall decoding complexity while maintaining error correction capability.
Solution Approach 2:
The patent performs preliminary classification of bits into degree groups before decoding begins. The convergence characteristics of each degree group are analyzed in advance, and decoding zones are pre-defined based on these characteristics. This preliminary action enables the decoder to efficiently navigate through the decoding process by knowing which zones to process and when to skip them, reducing decoding time and complexity.
2Reliability
If traditional LDPC decoding is applied to all bits uniformly, then decoding completeness is maintained, but decoding time increases due to processing all bits through all iterations
Solution Approach 1:
The patent applies partial action by skipping decoding in certain zones where bits are not converging. Instead of uniformly processing all bits through all decoding iterations, the decoder identifies zones with non-converging bits and skips them after a predetermined number of iterations. This partial processing approach maintains decoding completeness for converging bits while significantly reducing decoding time by avoiding redundant processing of non-converging bits.
Solution Approach 2:
The patent introduces dynamic zone skipping based on the convergence status of different degree groups during decoding iterations. The decoder dynamically determines which zones to process and which to skip based on real-time convergence metrics. This dynamic approach allows the decoding process to adapt to the actual convergence behavior of different bit groups, optimizing decoding time while maintaining reliability.
3Measurement precision
If all degree groups are processed in each decoding iteration, then convergence accuracy is improved, but power consumption increases due to continuous processing of all bits
Solution Approach 1:
The patent segments the bit population into degree groups with different convergence characteristics and processes them in different zones during decoding iterations. This segmentation enables the decoder to concentrate processing power on zones with converging bits while skipping zones with non-converging bits, thereby maintaining convergence accuracy for important bits while reducing overall power consumption.
Solution Approach 2:
The patent applies different processing qualities to different zones based on their convergence status. Converging zones receive full processing attention to maintain high convergence accuracy, while non-converging zones are skipped to reduce power consumption. This local quality approach ensures that power resources are allocated efficiently to where they are most needed, maintaining accuracy where it matters while conserving energy.
4Reliability
If column weights are made irregular to improve error correction performance, then data reliability is improved, but decoding throughput decreases due to increased number of iterations required
Solution Approach 1:
The patent segments the decoding process into zones based on degree groups and their convergence behavior. This segmentation allows the decoder to process different degree groups in parallel or in an optimized sequence, improving throughput by avoiding sequential processing of all bits through all iterations. The irregular column weights are preserved for error correction performance while the segmented processing recovers throughput.
Solution Approach 2:
The patent performs preliminary analysis of degree group convergence characteristics before decoding and pre-defines processing zones and iteration limits. This preliminary action enables the decoder to efficiently navigate through the decoding process by knowing in advance which zones to process and when to terminate processing for each zone, thereby improving throughput without sacrificing the error correction benefits of irregular LDPC codes.
Data Source
AI summary
Decoding method and memory system which group bits in irregular LDPC codes having similar degrees of convergence into respective degree groups, classify the degree groups according to a metric indicative of a number of decoding iterations for convergence, divide a time period for convergence of the decoding iterations into different zones for the processing of selected degree groups within each zone, and skip decoding of the bits in a non-converging zone where the bits are not converging.