钱搜索是与非型闪速(NAND flash)存储器中BCH译 码器的重要组成部分,并行钱搜索延迟较小并可高速运行,但过高的复杂度制约了其的应用 。为了降低并行钱搜索的复杂度,提 出一种并行钱搜索的改进和优化方法。首先对传统并行钱搜索以及有关文献进行了分析和研 究;然后对公共子表达式的搜索范 围进行了扩展,并合并了相关计算;最后对迭代匹配算法进行了改进,提出一种基于二维搜 索的改进迭代匹配算法。实验结果 表明,本文方法取得了较好的优化效果,有效地降低了并行钱搜索的复杂度;在对BCH(2047,1926,1)的 32bit并行钱搜索 优化后,与传统并行钱搜索以及有关文献的两种并行钱搜索相比,本文方法的 优化率分别达 到了85.4%、38.7%和29.2%,并可以更好地适应不同并行度和不同纠错能力的并行钱搜索结构。
The Chien search process is an important component of the BCH decoder in NAND flash memory.Parallel Chien search has lower latency and can run at high speed,but higher complexity restricts its app lication.In order to reduce the complexity of parallel Chien search,an improvement and optimization method of parallel Chien search is proposed.Firstly,the traditional parallel Chien search and related references are analyzed and studied.Then,the search range of common subexpression is extended,and the relevant calculation is merged.Finally,iterative matching al gorithm is improved,and an improved iterative matching algorithm based on two-dimensional search is proposed.The experimenta l results show that this method can achieve good optimization results,and effect ively reduce the complexity of parallel Chien search.Compared with the tradition al parallel Chien search and two kinds of the parallel Chien search in the related references,the optimization rates of this method approximate to 85.4%,38.7% and 29.2% respectively after the 32-bit pa rallel Chien search of BCH (2047,6,11) has been optimized.This method can better adapt to parallel Chien search archite ctures with various parallel factors and various error correcting capability.