Memory Controller Error Floor Mitigation via Dual Read Logical Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Rewritable non-volatile memory modules face the challenge of error floor phenomenon during iterative decoding, where error bits are not reduced despite multiple iterations, leading to incorrect data reading.
Innovation Solution
A data reading method that involves sending a first read command sequence to obtain a data string, performing decoding, and if errors are present, sending a second read command sequence to obtain another data string, performing logical operations to adjust the decoded data string, and re-performing decoding to improve data correctness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iterative decoding is performed multiple times to improve data reading correctness, then the decoding accuracy is improved, but the error floor phenomenon occurs where error bits are no longer reduced
Solution Approach 1:
The patent applies preliminary action by performing a first read operation and decoding before the error floor becomes problematic. When the decoding result indicates remaining errors, the system performs a second read operation and combines the results through logical operations (XOR) to adjust the decoded data. This preliminary combination of multiple read results prevents the error floor from limiting further improvement in data reading correctness.
2Measurement precision
If multiple read command sequences are sent and decoding is re-performed to address error bits, then data reading correctness is improved, but the reading time and processing complexity increase
Solution Approach 1:
The patent implements feedback by checking the decoding results after the first read operation. When error bits are detected in the decoded data, the system triggers a second read operation and performs logical operations to adjust the decoded result. This feedback mechanism ensures that additional reading operations are only performed when necessary, thereby improving data reading correctness while minimizing unnecessary time loss.
3Reliability
If error correcting code is used with iterative decoding to protect data, then data protection capability is improved, but the error floor limits the correcting capability
Solution Approach 1:
The patent applies merging by combining the results from multiple read operations through logical operations (XOR). The first read result and second read result are combined to generate an adjusted decoded data string. This merging approach enhances data protection capability by leveraging multiple independent read operations to overcome the error floor limitation of traditional iterative decoding, thereby improving the effective correcting capability.
Data Source
AI summary
A data reading method is provided. The data reading method includes receiving a read command from a host system; sending a first read command sequence to obtain a first data string from memory cells of a rewritable non-volatile memory module; performing a decoding procedure on the first data string to generate a decoded first data string; and, if there is an error bit in the decoded first data string, sending a second read command sequence to obtain a second data string from the memory cells, performing a logical operation on the decoded first data string and the second data string to obtain an adjusting data string, adjusting the decoded first data string according to the adjusting data string to obtain an adjusted first data string, and using a data string obtained after re-performing the decoding procedure on the adjusted first data string as the decoded first data string.