Convolutional LDPC Decoder Early Termination Using LLR Confidence

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Convolutional low-density parity-check (LDPC) decoders face challenges in early termination due to time constraints and dependencies in convolutional coding, making it difficult to reduce power consumption effectively.

Innovation Solution

The implementation of methods using check fail counters and log likelihood ratio (LLR) sign change to determine when to terminate decoder processing in convolutional LDPC decoders, optimizing power consumption while maintaining performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If iterative decoding is performed to improve bit determination accuracy and reduce errors, then reliability is improved, but power consumption increases

Engineering Contradiction:

Improvebit determination accuracyVSAvoidpower consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements early termination by monitoring the confidence values of decoded bits during iterative processing. When all bits achieve maximum confidence values, the decoding process terminates early, preventing unnecessary additional iterations. This feedback mechanism allows the system to stop processing when reliability requirements are met, thereby reducing power consumption while maintaining bit determination accuracy.

Inventive Principle:

Principle #23Feedback

2Use of energy by moving object

If early termination is implemented to reduce power consumption, then power consumption is reduced, but reliability may deteriorate due to time constraints in convolutional coding

Engineering Contradiction:

Improvepower consumptionVSAvoidbit error rate

Core Design Contradiction:

Use of energy by moving objectVSReliability

Solution Approach 1:

The patent replaces traditional time-based termination constraints with a confidence-value-based termination criterion. Instead of forcing a fixed number of iterations regardless of actual decoding quality, the system uses confidence values (log-likelihood ratios) to dynamically determine when termination is appropriate. This substitution ensures that early termination does not compromise reliability, as termination only occurs when all bits have achieved sufficient confidence levels.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If confidence values are monitored for all bits to determine termination timing, then reliability is improved, but device complexity increases

Engineering Contradiction:

Improvedecoding accuracyVSAvoiddecoder circuit complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts and monitors only the confidence values (log-likelihood ratios) of decoded bits to determine termination timing, rather than implementing complex real-time error detection and correction mechanisms. By focusing specifically on confidence value monitoring, the system achieves reliable early termination with minimal additional circuit complexity, as confidence values are already computed during normal decoding operations.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS12244414B1Early termination in convolutional low density parity check decoding

Publication Date: 2025.03.04 INFINERA CORP

AI summary

Systems and methods for more efficiently decoding convolutional LDPC encoded communication signals are described. A decoder circuit iteratively determines the values of received data bits and confidence values corresponding to the determined data bit values. Using the confidence values and one or both of check fail counters and log likelihood ratio (LLR) sign changes, the decoder circuit can determine that a desired level of confidence has been satisfied and can subsequently terminate decoding operations to save power.