A Reliability Enhancement Design under the Flash Translation Layer for MLC-Based Flash-Memory Storage Systems

Although flash memory has gained very strong momentum in the storage market, the reliability of flash-memory chips has been dropped significantly in the past years. This article presents a reliability enhancement design under the flash management layer (i.e., flash translation layer) to address this concern so as to reduce the design complexity of flash-memory management software/firmware and to improve the maintainability and portability of existing and future products. In particular, a log-based write strategy with a hash-based caching policy is proposed to provide extra ECC redundancy and performance improvement. Strategies for bad block management are also presented. The failure rate of flash-memory storage systems is analyzed with the considerations of bit errors. The proposed design is later evaluated by a series of experiments based on realistic traces. It was shown that the proposed approach could significantly improve the reliability of flash memory with very limited system overheads.