Article Details

Abstract

In consumer electronics products based on NAND flash memory, garbage collection significantly affects device performance, and cold-hot data separation is the key to optimizing the garbage collection algorithm. This paper proposes an innovative method that uses the K-means clustering algorithm to automatically determine the data heat intervals and achieve efficient cold-hot data separation. Additionally, an adaptive influence factor adjustment strategy is designed to dynamically balance the influence of historical data heat and recent data heat. On this basis, the garbage collection strategy is optimized. The simulation results show that compared with the existing algorithms, this method has a significant effect on improving the read-write performance and achieving wear leveling, which is helpful in prolonging the service life of the equipment and improving the overall performance.

Keywords

flash memory; garbage collection; cold-hot separation; K-means; clustering

References

1. [1] CHANG L P，KUO T W. Efficient Management for Large-scale Flashmemory Storage Systems with Resource Conservation [J].ACM Transactions on Storage，2005，1（4）：381-418.
2. [2] XIE W，CHEN Y. An Adaptive Separation-Aware FTL for Improving the Efficiency of Garbage Collection in SSDs [C]//2014 14th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing.Chicago：IEEE，2014:552-553.
3. [3] GU J Q，WU C T，LI J. HOTIS: A Hot Data Identification Scheme to Optimize Garbage Collection of SSDs [C]//2017 IEEE International Symposium on Parallel and Distributed Processing with Applications and 2017 IEEE International Conference on Ubiquitous Computing and Communications (ISPA/IUCC).Guangzhou：IEEE，2017：331-337.
4. [4] YANG J，PEI S Y. Thermo-GC: Reducing Write Amplification by Tagging Migrated Pages during Garbage Collection [C]//2019 IEEE International Conference on Networking, Architecture and Storage (NAS).EnShi：IEEE，2019：1-8.
5. [5] LUO Q W，FANG X X，SUN Y C，et al. Self-Learning Hot Data Prediction: Where Echo State Network Meets NAND Flash Memories [J].IEEE Transactions on Circuits and Systems I: Regular Papers，2020，67（3）：939-950.
6. [6] LEE J，KIM J S. An empirical study of hot/cold data separation policies in solid state drives (SSDs) [C]//Proceedings of the 6th International Systems and Storage Conference.Haifa：ACM，2013：1-6.
7. [7] SHAFAEI M，DESNOYERS P，FITZPATRICK J. Write Amplification Reduction in Flash-Based SSDs Through Extent-Based Temperature Identification [C]//Proceedings of the 8th USENIX Conference on Hot Topics in Storage and File Systems.Denver：USENIX Association，2016：106-110.
8. [8] YAN H，HUANG Y，ZHOU X Z，et al. An Efficient and Non-Time-Sensitive File-Aware Garbage Collection Algorithm for NAND Flash-Based Consumer Electronics [J].IEEE Transactions on Consumer Electronics，2018，65（1）：73-79.
9. [9] MACQUEEN J. Some Methods for Classification and Analysis of Multivariate Observations [C]//Proceedings of the fifth Berkeley Symposium on Mathematical Statistics and Probability.California：University of California Press，1967：281-297.
10. [10] GAGNIUC P A. Markov Chains: From Theory to Implementation and Experimentation [M].Germany：Wiley，2017.
11. [11] CHUNG J Y，GULCEHRE C，CHO K H，et al. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling [J/OL].arXiv:1412.3555 [cs.NE].[2024-10-03].https://doi.org/10.48550/arXiv.1412.3555.
12. [12] LIN M W，CHEN S Y. Efficient and Intelligent Garbage Collection Policy for NAND Flash-based Consumer Electronics [J].IEEE Transactions on Consumer Electronics，2013，59（3）：538-543.

About the authors

Fu Xiaoyu (1999-), male, Hui ethnicity, Tianjin native, currently pursuing a master's degree in Nand Flash garbage collection algorithm; Corresponding author: Yan Hua (1971-), male, Han ethnicity, from Qu County, Sichuan Province, Professor, PhD, research direction: Intelligent Information Systems.