LDPC Decoder Threshold Locking for Low-Power Message Passing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional low density parity check (LDPC) decoding systems are power hungry, consuming excessive energy due to unnecessary calculations and message passing in iterations where nodes have already settled.
Innovation Solution
Implementing a low-power LDPC decoding method that locks variable and check nodes once their probabilities exceed certain thresholds, reducing energy expenditure by avoiding further message calculations and passing between locked nodes, and dynamically determining thresholds based on Quality of Service (QoS) requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional LDPC decoding continues message passing for all iterations regardless of node convergence, then decoding accuracy is maintained, but power consumption increases excessively
Solution Approach 1:
The patent implements dynamic node locking where variable nodes and check nodes are selectively locked or unlocked based on their convergence status during decoding iterations. This dynamic adaptation allows the system to maintain decoding accuracy for unconverged nodes while saving power by stopping computations for converged nodes, directly resolving the contradiction between reliability and energy consumption.
Solution Approach 2:
The patent applies different processing states to different nodes within the LDPC decoder. Instead of uniformly processing all nodes in each iteration, the system identifies and locks specific nodes that have converged, while continuing to process only those nodes that have not yet converged. This localized differentiation maintains overall decoding accuracy while reducing total power consumption.
2Reliability
If LDPC decoding performs calculations for all nodes in each iteration, then decoding completeness is ensured, but unnecessary calculations increase power consumption
Solution Approach 1:
The patent implements partial action by performing message passing only for nodes that require further processing. Instead of executing complete iterations for all nodes, the system identifies converged nodes and excludes them from subsequent calculations, performing only the necessary partial computations required to achieve decoding completeness, thereby reducing energy loss.
Solution Approach 2:
The patent extracts and removes converged nodes from the active computation set by locking them once their probability thresholds are exceeded. This extraction eliminates unnecessary calculations for these nodes while maintaining decoding completeness for the remaining unconverged nodes, directly reducing energy expenditure without sacrificing reliability.
3Reliability
If message passing continues between all nodes in each iteration, then decoding thoroughness is maintained, but power consumption increases
Solution Approach 1:
The patent dynamically controls message passing by locking variable nodes and check nodes when their probabilities exceed predetermined thresholds. This dynamic message passing mechanism ensures thorough decoding for active nodes while eliminating energy-consuming message exchanges for locked nodes, resolving the contradiction between decoding thoroughness and energy consumption.
Solution Approach 2:
The patent applies local quality by enabling message passing only between unconverged nodes that require further processing. Converged nodes are locked and excluded from message passing, creating a localized computation pattern that maintains decoding thoroughness where needed while reducing energy consumption in converged regions.
Data Source
AI summary
In an example implementation of this disclosure, a message passing low density parity check (LDPC) decoder may, during decoding of a first group of bits, lock a first variable node upon a bit-value probability of the first variable node reaching a determined threshold, and lock a first check node upon all variable nodes connected to the first check node being locked. The LDPC decoder may cease decoding the first group of bits upon all variable nodes of the LDPC decoder being locked, all check nodes of the LDPC decoder being locked, reaching a maximum number of iterations, or reaching a timeout. During a particular iteration of the decoding of the first group of bits in which the first variable node is locked, the LDPC decoder may refrain from generating a bit-value probability for the locked first variable node.