Layered Non-Binary LDPC Decoder for Faster Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transfer systems face challenges in efficiently detecting and correcting errors in digital data due to the limitations of traditional error checking systems, particularly in handling errors that are not uniformly distributed, which can overwhelm LDPC decoders.
Innovation Solution
The implementation of a multi-level LDPC layer decoder that performs min-sum based layered decoding of non-binary LDPC codes, utilizing a quasi-cyclic LDPC code structure and a hardware-friendly architecture that processes circulant sub-matrices in layers, reducing the complexity and time to converge on accurate data values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional error checking systems are used, then error detection capability is provided, but the system cannot efficiently correct errors that are not uniformly distributed and overwhelms LDPC decoders
Solution Approach 1:
The patent divides the LDPC decoding process into multiple levels or stages, where different error correction strategies are applied at different levels. This segmentation allows the system to handle non-uniformly distributed errors more effectively without overwhelming the decoder, as the problem is broken down into manageable portions that can be processed sequentially.
Solution Approach 2:
The patent modifies decoding parameters and strategies based on the characteristics of errors detected at different stages. By dynamically adjusting parameters such as check node processing intensity, variable node update rules, and convergence thresholds, the system adapts to handle non-uniform error distributions while maintaining decoder efficiency and avoiding overload.
2Reliability
If LDPC decoders process all errors uniformly, then comprehensive error correction is attempted, but processing time and computational resources are excessively consumed
Solution Approach 1:
The patent implements a multi-level decoding approach where not all check nodes are processed with equal intensity at every iteration. Instead, the system applies partial processing strategies, focusing computational resources on the most critical error patterns while using lighter processing for less severe cases. This allows comprehensive error correction capability to be maintained while significantly reducing average processing time.
Solution Approach 2:
The patent employs periodic or iterative decoding cycles with varying processing intensities. Different levels of decoding scrutiny are applied in successive iterations, allowing the system to quickly identify and correct obvious errors in early iterations, then apply more intensive processing only when necessary in later iterations, thereby reducing overall decoding time while maintaining completeness.
3Measurement precision
If detailed min-sum based layered decoding is performed, then accurate data values are achieved, but memory requirements and processing complexity increase
Solution Approach 1:
The patent segments the LDPC code structure into quasi-cyclic components with circulant sub-matrices, allowing the min-sum based layered decoding to be applied in a structured, hierarchical manner. This segmentation enables accurate decoding by maintaining proper message passing relationships while organizing computations into repeatable patterns that reduce overall processing complexity and improve hardware implementability.
Solution Approach 2:
The patent adjusts decoding parameters such as the scaling factor in min-sum calculations and the layer processing order to optimize the balance between accuracy and complexity. By carefully selecting and modifying these parameters, the system achieves high data value accuracy while keeping processing complexity at manageable levels through standardized operational patterns.
4Reliability
If standard LDPC decoding is applied to non-binary codes over large Galois Fields, then robust error correction is provided, but memory requirements and processing time increase significantly
Solution Approach 1:
The patent utilizes the quasi-cyclic structure of non-binary LDPC codes, segmenting the large Galois Field operations into smaller, manageable circulant sub-matrices. This segmentation allows memory-intensive operations to be broken down into smaller chunks that can be processed with reduced memory footprints, while maintaining the robust error correction capabilities provided by the full non-binary code structure.
Solution Approach 2:
The patent optimizes memory usage by carefully managing the representation of Galois Field elements and the storage of intermediate messages during decoding. Through parameter optimizations such as message precision levels, storage formats, and processing batch sizes, the system achieves robust error correction for non-binary codes while significantly reducing the memory requirements compared to standard implementations.
Data Source
AI summary
Various embodiments of the present invention are related to methods and apparatuses for decoding data, and more particularly to methods and apparatuses for multi-level layered LDPC decoding. For example, in one embodiment an apparatus includes a variable node processor and a check node processor. The variable node processor is operable to generate variable node to check node messages and to calculate perceived values based on check node to variable node messages. The check node processor is operable to generate the check node to variable node messages and to calculate checksums based on variable node to check node messages. The check node processor includes a min finder circuit operable to identify a minimum, a next minimum and an index of minimum value in the variable node to check node messages. The variable node processor and check node processor are operable to perform layered multi-level decoding.