Memory Controller Data Validation via Threshold Voltage Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory management systems often misjudge valid data as invalid due to errors in non-volatile memory modules, particularly after power-off conditions, leading to incorrect data validation.
Innovation Solution
A memory management method that determines the storage state of data based on a default bias range and threshold voltage distribution, performing different operations depending on whether the data conforms to a first or second condition, thereby reducing the misjudgment of valid data as invalid.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data validation is performed based on conventional error detection methods, then invalid data can be identified, but valid data may be misjudged as invalid due to errors occurring after power-off
Solution Approach 1:
The patent changes the validation parameter from simple error bit counting to threshold voltage distribution analysis. By examining the statistical distribution of threshold voltages across memory cells and comparing it against expected distributions for valid data, the system can distinguish between actual errors and normal variations in memory cell characteristics, thereby reducing false positives in data validation
Solution Approach 2:
The patent replaces conventional mechanical/error-based detection methods with a statistical analysis approach. Instead of relying on error correction codes and bit counting, the system uses threshold voltage distribution patterns as a fingerprint to validate data, substituting physical error detection with statistical pattern recognition
2Measurement precision
If strict error detection is applied to identify invalid data, then data integrity is improved, but the probability of misjudging valid data as invalid increases
Solution Approach 1:
The patent implements a feedback mechanism where the threshold voltage distribution of previously validated valid data is stored as a reference. During validation, the current threshold voltage distribution is compared against this reference feedback, allowing the system to adaptively distinguish between actual errors and normal variations, thereby maintaining high detection precision while preserving valid data
Data Source
AI summary
A memory management method, a memory storage device and a memory controlling circuit unit are provided. The method includes: programming data into a plurality of memory cells of a rewritable non-volatile memory module; determining whether a storage state of the data conforms with a first condition or a second condition based on a default bias range and a threshold voltage distribution of the memory cells storing the data; performing a first operation if the storage state of the data conforms with the first condition; and performing a second operation if the storage state of the data conforms with the second condition. Accordingly, the probability of misidentifying the valid data as the invalid data may be reduced.