An adaptive allocation of blocks strategy for garbage collection in NAND flash memory
In the garbage collection of NAND flash memory,hot-cold separation can improve the garbage collection efficiency.However,the existing algorithms divide the data into a fixed number of classes and do not take into account the clustering change of the data heat during the writing process.In order to solve this problem,the adaptive allocation of blocks strategy for garbage collec-tion algorithm is studied.Adjusting the block allocation pool size by monitoring the valid page proportion of the victim block and the write frequency of the allocation block can avoid the block idle problem caused by inaccurate data heat classification,and achieve more efficient hot-cold separation.The experimental results demonstrate that the algorithm can reduce the overhead of gar-bage collection and improve the degree of wear leveling.
万方数据
维普
