Sparse Graph Code Decoding with Alternating Trivial and Gauss Elimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current decoding techniques for sparse graph codes, particularly in error correction for communication systems, face challenges in achieving maximum likelihood decoding at low computational complexity, as methods like Gauss elimination increase operational complexity and reduce applicable data size, while trivial decoding methods do not fully exploit the potential of sparse graph codes.
Innovation Solution
A method that alternately and repeatedly uses trivial decoding and Gauss elimination methods for maximum likelihood decoding, where trivial decoding treats independent lost data as restored data and Gauss elimination is applied to reduce fill-in occurrences and select pivot columns efficiently, thereby reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Gauss elimination method is used for maximum likelihood decoding, then decoding accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the lost data into two categories: trivially decodable lost data and non-trivial lost data. The trivial decoding unit handles lost data that can be decoded through simple operations, while the Gauss elimination unit handles the remaining non-trivial lost data. This segmentation allows the system to achieve maximum likelihood decoding accuracy while minimizing computational complexity by avoiding Gauss elimination for all lost data.
2Measurement precision
If Gauss elimination method is used for maximum likelihood decoding, then decoding accuracy is improved, but applicable data size is reduced
Solution Approach 1:
The patent divides lost data into trivial and non-trivial portions, enabling the system to handle larger data sizes. The trivial decoding unit can efficiently process trivially decodable lost data without the computational burden of Gauss elimination, thereby increasing the applicable data size while maintaining decoding accuracy through the complementary Gauss elimination unit.
3Device complexity
If trivial decoding method is used, then computational complexity is reduced, but decoding accuracy is not fully optimized
Solution Approach 1:
The patent merges two decoding approaches: the trivial decoding unit and the Gauss elimination unit. The trivial decoding unit first attempts to decode lost data through simple operations, and the Gauss elimination unit supplements by decoding the remaining non-trivial lost data. This combination achieves both low computational complexity for trivial cases and high decoding accuracy for all lost data.
4Device complexity
If trivial decoding method is used, then computational complexity is reduced, but error correction capability is insufficient
Solution Approach 1:
The patent combines trivial decoding and Gauss elimination decoding in a unified error correction system. The trivial decoding unit provides fast, low-complexity error correction for easily decodable lost data, while the Gauss elimination unit ensures complete error correction capability for non-trivial lost data. This merged approach maintains low computational complexity while achieving reliable error correction for all types of lost data.
Data Source
AI summary
This method and device makes it possible to implement maximum likelihood decoding of a sparse graph code at low computational complexity in the maximum likelihood decoding of the sparse graph code. This is, in the maximum likelihood of decoding of the sparse graph code, a lost data decoding process by a trivial decoding method and a lost data decoding process by a Gauss elimination method are performed repeatedly and alternately.