LDPC Decoder Post-Processing for Hard Error Trapping Sets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low-density parity check codes suffer from error floors caused by trapping sets in flash memory applications, particularly due to hard errors that lead to saturated log-likelihood ratios, which hinder effective decoding.
Innovation Solution
A post-processing method that utilizes known hard error locations to flip and saturate log-likelihood ratios, allowing for re-decoding until convergence, and combines this information with unsatisfied check nodes after decoding failure to improve error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If iterative decoding algorithms are used for low-density parity check codes, then soft decision decoding capability is achieved, but error floors are caused by trapping sets
Solution Approach 1:
The patent applies preliminary action by flipping log-likelihood ratios at known hard error locations before the iterative decoding process begins. This pre-correction of hard errors prevents them from triggering trapping sets during iteration, allowing the decoder to converge correctly without being trapped in error floors.
Solution Approach 2:
The patent converts the harmful effect of hard errors into a beneficial signal by using the knowledge of hard error locations to flip log-likelihood ratios. Instead of treating hard errors as complete failures, the system uses this information to correct them, transforming the harmful trapping set effect into a successful error correction mechanism.
2Reliability
If hard errors are present in flash memory, then log-likelihood ratios become saturated or almost saturated in the wrong direction, but conventional decoding cannot correct these errors
Solution Approach 1:
The patent applies inversion by flipping the sign of log-likelihood ratios at hard error locations. Instead of trying to correct saturated values directly, the system inverts the incorrect saturation direction, transforming the wrong-direction saturation into the correct direction, thereby enabling successful decoding.
Solution Approach 2:
The patent changes the parameter state of log-likelihood ratios by flipping their signs at hard error locations. This parameter transformation converts the saturated wrong-direction values into correct-direction values, enabling the decoder to process them accurately and correct the underlying hard errors.
3Reliability
If post-processing is applied to flip log-likelihood ratios at hard error locations, then error floor is reduced, but additional processing steps are required
Solution Approach 1:
The patent performs the log-likelihood ratio flipping operation before the main iterative decoding process, treating it as a preliminary action. This approach consolidates the complexity into a pre-processing step that prepares the data for successful decoding, rather than adding complexity during the critical iteration process.
Solution Approach 2:
The patent makes the decoding process self-service by using the decoder's own knowledge of hard error locations to automatically flip the appropriate log-likelihood ratios. This self-correction mechanism eliminates the need for external intervention or complex additional processing systems, reducing overall device complexity while maintaining high reliability.
Data Source
AI summary
A low-density parity-check decoder utilizes information about hard errors in a storage medium to identify bit locations to flip log-likelihood ratios while attempting to decode codewords. The decoder iteratively flips and saturates log-likelihood ratios for bits at hard error locations and re-decodes until a valid codeword is produced. The decoder also identifies variable nodes associated with trapping sets for iterative log-likelihood ratio bit flipping.