Overlapping ECC Codewords for Sector-Aligned Memory Reads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile semiconductor memory devices face a trade-off between enhanced error correction coding (ECC) strength and read performance, as multi-sector encoding improves ECC but can lead to decreased read performance due to misaligned data requests across encoded words.
Innovation Solution
Implementing overlapping error correction operations where data is encoded into overlapping codewords, ensuring each sector is aligned with a codeword boundary, thereby enhancing ECC strength while maintaining read performance by allowing single-read operations across sector boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple sectors are encoded together to improve ECC strength, then error correction capability is improved, but read performance deteriorates due to misaligned data requests
Solution Approach 1:
The data stream is segmented into multiple sectors, each independently encodable. The patent allows encoding of multiple sectors together when beneficial for ECC strength, but maintains the ability to decode individual sectors independently. This segmentation enables flexible encoding strategies that can adapt to different read request patterns, resolving the contradiction between improved ECC strength and maintained read performance.
2Reliability
If multi-sector encoding is used to enhance ECC capability, then more errors can be corrected, but read complexity increases requiring full reading and decoding of multiple encoded words
Solution Approach 1:
The patent implements a dynamic encoding strategy where the encoding approach adapts based on the read request characteristics. When a read request is aligned with encoded word boundaries, the system can decode only the necessary portion. When misaligned, the system dynamically determines whether to decode multiple full words or use alternative strategies. This dynamic behavior reduces average read complexity while maintaining enhanced ECC capability.
Solution Approach 2:
The patent applies different error correction strategies to different portions of data based on local requirements. Instead of uniformly applying multi-sector encoding to all data, the system can apply stronger ECC only where needed while using simpler encoding elsewhere, or decode only the specific sectors required by the read request rather than always decoding full multi-sector words. This localized approach reduces overall complexity while maintaining error correction capability.
Data Source
AI summary
Systems and methods of overlapping error correction operations are disclosed. A method at an encoder device includes receiving data to be encoded, including a first data group and a second data group. The first data group includes a first subgroup that contains a first plurality of bits and a second subgroup that contains a second plurality of bits. The second data group includes the second subgroup and a third subgroup that includes a third plurality of bits. A first encode operation encodes the first data group to generate a first codeword and a second encode operation encodes the second data group.