LDPC Decoder Early Stopping for Undecodable DVB-S2 Codewords
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LDPC code decoders in systems like DVB-S2 often waste energy and processing time by completing the full number of iterations for undecodable codewords, which are identified late in the decoding process.
Innovation Solution
A method to control the decoding of LDPC encoded codewords by calculating a second sum of absolute values of messages exchanged between variable and check nodes, stopping the process if this sum remains unchanged or decreases within two successive iterations and meeting a predetermined threshold condition, allowing early termination for undecodable codewords.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the full number of iterations is performed for all codewords, then the decoding reliability is improved, but the energy consumption and processing time increase significantly for undecodable codewords
Solution Approach 1:
The patent applies preliminary action by performing early detection of undecodable codewords before completing all iterations. The detection mechanism monitors the sum of absolute values of variable node messages during the decoding process, and when this sum fails to increase over consecutive iterations, the codeword is identified as undecodable. This allows the system to terminate decoding early for undecodable codewords, avoiding waste of energy and processing time, while maintaining full iteration count for decodable codewords to ensure decoding reliability.
2Measurement precision
If the full number of iterations is performed for all codewords, then the decoding accuracy is improved, but the processing time is excessively increased for undecodable codewords
Solution Approach 1:
The patent implements preliminary detection of undecodable codewords by monitoring the evolution of the sum of absolute values of variable node messages during iterations. When this sum does not increase over two or more consecutive iterations, the codeword is detected as undecodable and processing is terminated early. This preliminary detection mechanism significantly reduces processing time for undecodable codewords while preserving full iteration processing for decodable codewords, thereby maintaining decoding accuracy.
3Loss of energy
If early termination is implemented for undecodable codewords, then the energy efficiency is improved, but the complexity of the decoding control mechanism increases
Solution Approach 1:
The patent employs feedback by continuously monitoring the sum of absolute values of variable node messages during the decoding iterations and using this information to control the decoding process. The detector compares the current sum with previous iteration sums to determine whether to continue or terminate decoding. This feedback mechanism enables early termination for undecodable codewords, improving energy efficiency while adding controlled complexity to the decoding system.
Solution Approach 2:
The decoding system performs self-service by automatically detecting undecodable codewords through intrinsic properties of the decoding process itself. The sum of absolute values of variable node messages naturally diverges for decodable codewords and remains stagnant for undecodable ones, allowing the system to self-identify undecodable cases without external intervention. This self-service approach reduces the need for complex external control mechanisms.
4Productivity
If early termination is implemented for undecodable codewords, then the productivity is improved, but the complexity of detecting undecodable codewords increases
Solution Approach 1:
The patent uses feedback by monitoring the evolution of the sum of absolute values of variable node messages during decoding iterations. The detector continuously compares current iteration sums with previous iteration sums to detect stagnation patterns indicative of undecodable codewords. This feedback-based detection mechanism improves decoding throughput by enabling early termination while managing detection complexity through a relatively simple comparison operation.
Solution Approach 2:
The patent replaces complex mechanical or algorithmic detection mechanisms with a simpler mathematical observation: monitoring whether the sum of absolute values of variable node messages increases over iterations. This substitution transforms the detection problem into a straightforward numerical comparison task, improving productivity while keeping detection complexity manageable.
Data Source
AI summary
This is a method for controlling the decoding of a LDPC encoded codeword composed of several digital data, said LDPC code being represented by a bipartite graph between check nodes (CN1) and variable nodes (VNi). Said method comprises updating messages exchanged iteratively between variable nodes (VN1) and check nodes (CN1). Said method comprises, at each iteration, calculating for each variable node a first sum (Λn) of all the incident messages (λi) received by said variable node and the corresponding digital data (λch) and calculating a second sum (VNRnew) of all the absolute values of the first sums (Λn), and stopping the decoding process if the second sum (VNRnew) is unchanged or decreases within two successive iterations and if a predetermined threshold condition is satisfied.