Dynamic wear leveling addresses the problem of a limited number of Erase cycles for a given block. 动态平均读写解决了块的Erase周期的次数限制。
An organizing approach of data based on pyramid model using simplification, leveling and block processing technologies is put forward. 通过对地形数据简化分层、分块等数据处理建立金字塔数据模型来组织地形数据。
For wear leveling, data will be divided into two types, cold data and hot data. They are alternately stored in different types of free block, to achieve the purpose of the erase block wear leveling. 本文通过将数据分成冷热数据两类,把它们交替地储存在不同类型的空闲块上,从而达到擦除块间的磨损均衡。
Due to the physical characteristics of NAND flash, NAND flash facing garbage collection, wear leveling and bad block handling three major challenges. 由于NANDFlash存储介质的物理特性,NANDFlash面临垃圾收集、损耗均衡和坏块处理三大挑战[1]。