Multi-Level Flash Decoding with Lee Distance Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional error correction codes, such as binary BCH codes, are inadequate for multi-level flash memory devices as they do not effectively handle the rare occurrence of multiple bit errors, which can lead to increased error rates and reduced reliability over time.
Innovation Solution
The implementation of Lee distance-based error correction codes that encode data over a prime field, binarize non-binary redundancy data, and combine it with binary redundancy data to generate restored redundancy data, which is then stored in multi-level cells, allowing for improved error correction capabilities compared to Hamming distance-based codes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional binary BCH error correction codes are used in multi-level flash memory devices, then the implementation is simple and compatible with existing systems, but the error correction capability is insufficient for handling multiple bit errors in multi-level cells
Solution Approach 1:
The patent changes the fundamental parameter of the error correction code from binary (BCH) to non-binary (Lee metric over prime fields), enabling the code to directly handle multi-level cell errors. This parameter change allows the system to correct multiple bit errors in MLCs by treating them as symbol errors in a larger alphabet, thereby improving reliability without requiring complex multi-stage decoding procedures
Solution Approach 2:
The patent segments the error correction process into distinct functional blocks: non-binary syndrome computation, error location identification, and error correction. This segmentation allows each module to be optimized independently and facilitates hardware implementation with reduced complexity compared to conventional approaches that require multiple binary decoding stages
2Reliability
If more redundancy cells are allocated to increase error correction capability, then more errors can be corrected, but the storage capacity for actual data decreases
Solution Approach 1:
By changing from binary to non-binary coding, the patent achieves higher error correction capability with fewer redundancy symbols. The Lee metric code over prime fields can correct multiple bit errors per cell as single symbol errors, reducing the overhead ratio compared to binary BCH codes that would require multiple redundancy bits for the same protection level
Solution Approach 2:
The patent designs a universal error correction code that can handle various error patterns in multi-level cells through a unified non-binary framework. This universal approach eliminates the need for different coding schemes for different error scenarios, allowing efficient use of redundancy resources across diverse operating conditions and extending device lifespan
Data Source
AI summary
Apparatus and methods for operating a flash device characterized by use of Lee distance based codes in a flash device so as to increase the number of errors that can be corrected for a given number of redundancy cells, compared with Hamming distance based codes.