Non-Binary LDPC Parity Node Candidate Selection for Faster Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of non-binary LDPC code decoding in elementary parity nodes is high due to the need for multiple multiplications and comparisons, which can be reduced by filtering and sorting incoming messages, but further optimization is necessary to minimize calculations.
Innovation Solution
A method for controlling elementary parity nodes that receives sorted input lists and generates output lists with reduced calculations by selecting candidate elements based on predetermined criteria, using a flag to manage replacement rules and optimize candidate selection, thereby reducing the number of operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all elements from two input lists are processed to generate output elements, then the completeness of decoding is improved, but the number of calculations (multiplications and comparisons) increases significantly
Solution Approach 1:
The patent extracts only the most relevant elements from the input lists based on predetermined criteria (such as magnitude thresholds or significance metrics) rather than processing all elements. This selective extraction maintains decoding reliability by focusing on critical elements while reducing the calculation complexity by excluding less significant elements from the processing pipeline.
Solution Approach 2:
The patent implements partial action by processing only a subset of elements from the input lists according to predetermined criteria. Instead of exhaustively processing all possible element combinations, the method performs calculations on selected elements that meet specific conditions, thereby achieving sufficient decoding performance with reduced computational burden.
2Productivity
If the number of elements processed in elementary parity nodes is reduced, then the calculation complexity is decreased, but the decoding performance may deteriorate
Solution Approach 1:
The patent applies preliminary action by pre-establishing selection criteria and identifying significant elements from the input lists before the main decoding calculations are performed. This preliminary filtering and sorting of elements ensures that only relevant data proceeds to the calculation stage, maintaining decoding performance while improving productivity through reduced computational workload.
Solution Approach 2:
The patent changes parameters by introducing predetermined selection criteria (such as magnitude thresholds, significance metrics, or ordering conditions) that transform the input data into a filtered set of relevant elements. This parameter-based filtering maintains the essential information needed for accurate decoding while reducing the volume of data requiring intensive calculation, thus balancing performance and speed.
3Measurement precision
If multiple candidate elements are evaluated to ensure accuracy, then the precision of output elements is improved, but the number of comparisons required increases
Solution Approach 1:
The patent applies partial action by evaluating only a limited number of candidate elements that meet predetermined selection criteria rather than exhaustively comparing all possible candidates. This selective evaluation maintains output accuracy by focusing on the most promising candidates while reducing processing time by excluding less likely options from the comparison process.
Solution Approach 2:
The patent performs preliminary sorting, filtering, or ranking of candidate elements based on predetermined criteria before the final selection and comparison stages. This preliminary organization of data ensures that the most accurate candidates are identified early, allowing the system to achieve high measurement precision with fewer comparisons by avoiding exhaustive evaluation of all possibilities.
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method for controlling a basic parity node of a decoder for decoding non-binary LDPC codes or a code decoder using at least one non-binary parity constraint, and to the corresponding basic parity node. The basic parity node receives first and second input lists (U1, U2) of nm elements sorted in ascending or descending order, nm being greater than 1, and outputs an output list (Uout) of nm' elements sorted in said ascending or descending order, nm' being greater than 1, each element from the output list (Uout) being the result of a calculation operation f between an element from the first input list (U1) and an element from the second input list (U2). According to the invention, a reduced number of candidates is selected for each element from the output list to be generated in order to reduce the number of operations to be carried out in the basic parity node.