NAND Flash Soft Information Adjustment for Error Rate Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In NAND Flash memory, the error rate cannot be effectively reduced due to inaccurate soft information input into error correction mechanisms, as the soft information is not adjusted or corrected after being inputted.
Innovation Solution
An information processing method that determines a reliable page in the memory cell based on coding combinations, obtains a parameter from inter-page relationships, and adjusts the probability values of bits in adjacent pages to correct soft information, thereby reducing the error rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If soft information is input into the soft decoder without adjustment or correction, then the decoding process can be completed, but the error rate cannot be effectively reduced due to inaccurate soft information
Solution Approach 1:
The patent applies preliminary action by adjusting the soft information before it is input into the soft decoder. Specifically, the method adjusts the probability values of bits in pages other than the target page based on the target page's data, and performs preliminary inversion correction on bits that are determined to be inverted. This preliminary adjustment ensures that the soft information fed into the decoder is more accurate, thereby effectively reducing the error rate without requiring additional decoding iterations.
2Quantity of substance
If the number of pages N increases to store more data in a memory cell, then the storage capacity increases, but the complexity of determining inter-page relationships and adjusting probability values increases
Solution Approach 1:
The patent applies segmentation by dividing the N pages into a target page and N-1 other pages, and further dividing the N-1 pages into L pages with adjusted probability values and N-1-L pages without adjustment. This segmented approach allows the method to systematically handle inter-page relationships by focusing adjustments on specific pages (L pages) that have direct relationships with the target page, while avoiding unnecessary processing of other pages, thus managing complexity even as N increases.
Solution Approach 2:
The patent applies local quality by selectively adjusting probability values only for bits in L pages that have direct inter-page relationships with the target page, rather than uniformly processing all N-1 other pages. The method determines specific pages that need adjustment based on their relationship with the target page and applies correction only where needed. This localized approach reduces overall processing complexity while maintaining accuracy for the most critical pages.
Data Source
AI summary
An information processing method and an electronic device are described where the method includes determining, based on coding combination in a memory cell, an M-th page among N pages in the memory cell; obtaining, based on N−1 inter-page relationships between the M-th page and N−1 pages among the N pages except the M-th page, a first parameter; and adjusting, based on the first parameter, a current probability value of probability of that each bit of L bits corresponding to L pages among the N−1 pages is represented as 0 or 1, as a first probability value, L being a positive integer not greater than N−1.