期刊
IEEE TRANSACTIONS ON CONSUMER ELECTRONICS
卷 69, 期 3, 页码 431-440出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCE.2022.3228404
关键词
Flash memory; garbage collection; logical interval; wear leveling
In NAND flash-based consumer electronics, garbage collection can negatively impact performance and lifespan. Hot-cold data separation is crucial for efficient garbage collection. This paper presents a novel algorithm that accurately separates hot-cold data by introducing the T-P factor. Experimental results demonstrate its superiority in terms of garbage collection overhead, wear leveling, and extra memory consumption.
In NAND flash-based consumer electronics, garbage collection can degrade consumer electronics performance. And the extra page writes caused by garbage collection reduce the lifespan of consumer electronics. Hot-cold data separation has a critical impact on garbage collection. Therefore, inaccurate hot-cold data separation significantly affects garbage collection overhead and the performance of wear leveling. In addition, accurate data separation depends greatly on heat calculation of each page in NAND flash and extra memory is required to record page heat information. In this paper, we propose an efficient and novel hot-cold data separation garbage collection algorithm. The proposed algorithm introduces T-P factor which combines update time and page sequence number to calculate the heat of different logical interval rather than page, thus accurate hot-cold data separation can be achieved. Based on T-P factor, garbage collection strategy is improved accordingly. Experimental results show that the proposed algorithm is superior to existing algorithms in terms of garbage collection overhead, wear leveling and extra memory consumption in NAND flash-based consumer electronics.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据