Decoder-Assisted LLR Calculation for Corrupted Memory Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing soft decoding methods fail when original data stored in memory is corrupted due to errors, leading to unsuccessful decoding.
Innovation Solution
A multi-stage decoding process using a first hard decoder followed by a second soft decoder, where the second decoder utilizes soft information derived from the first decoder's output to improve error correction capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If soft decoding is based on original data stored in memory, then decoding can be performed using original data, but if that original data is corrupted then soft decoding fails to decode the original data stored in memory
Solution Approach 1:
The decoding process is divided into two distinct stages: a first decoding attempt using original data, and a second decoding attempt using synthesized data that incorporates information from the first decoding. This segmentation allows the system to handle corrupted data more effectively by providing an alternative decoding path when the first attempt fails.
Solution Approach 2:
The first decoding is performed as a preliminary action before the second decoding. The results and information from this first decoding are used to synthesize the data for the second decoding attempt. This preliminary action prepares the system by identifying errors and patterns that inform the subsequent more robust decoding process.
2Reliability
If a single decoder is used for decoding, then the device complexity is low, but the decoding success rate decreases when data is corrupted
Solution Approach 1:
The system merges the outputs of two different decoding processes: the first decoding result and the second decoding result. By combining these results, the system achieves higher reliability in decoding corrupted data while managing device complexity through the integration of multiple decoding paths.
Solution Approach 2:
The first decoding provides feedback information that is used to synthesize the data for the second decoding. This feedback mechanism allows the system to adapt the second decoding process based on the outcomes of the first decoding, improving overall decoding success rates for corrupted data.
3Reliability
If hard decoding is used first then soft decoding is used when first decoding fails, then error correction capability is improved, but the processing time increases
Solution Approach 1:
The decoding process is structured as periodic action with two distinct phases: a first decoding phase followed by a second decoding phase only when needed. This periodic structure allows the system to perform error correction efficiently by activating the more time-consuming soft decoding only when the hard decoding fails, rather than always using it.
Solution Approach 2:
The system applies partial action by using hard decoding for the initial decoding attempt, which is faster but less comprehensive. Only when this partial action proves insufficient does the system engage in the more extensive soft decoding process, thereby balancing error correction capability with processing time requirements.
Data Source
AI summary
A method for decoding data in a memory device includes attempting to decode, using a first decoder, initial bits based on original data read from memory, determining that the first decoding was not successful, and, in response to determining that the first decoding was not successful, attempting to decode, using a second decoder, the initial bits based on the original data and information derived during the first decoding. The first decoder may be a hard decoder, and the second decoder may be a soft decoder. The information derived during the first decoding may include soft information corresponding to each of the initial bits. The soft information may include LLR values corresponding to each of the initial bits. Alternatively, the information derived during the first decoding may include a syndrome weight, and the soft information may be based on the syndrome weight.