Journal of Internet of Things and Convergence (사물인터넷융복합논문지)

The Korea Internet of Things Society (한국사물인터넷학회)
권호 표지
DOI QR코드

DOI QR Code

Wear Leveling Technique using Random Selection Method in Flash Storage

플래시 스토리지에서 랜덤 선택 방법을 활용한 마모도 평준화 기법

  • Jung Kyu Park (Department of Computer Engineering, Changshin University) ;
  • Eun Young Park (Rinascita College of Liberal Arts and Sciences, Shinhan University)
  • 박정규 (창신대학교 컴퓨터전공) ;
  • 박은영 (신한대학교 리나시타교양대학)
  • Received : 2024.04.29
  • Accepted : 2024.05.27
  • Published : 2024.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, reliability has become more important as flash-based storage devices are actively used in cloud servers and data centers. Flash memory chips have limitations in reading/writing, so if writing is concentrated in one location, the chip can no longer be used. To solve this problem and improve reliability, it is necessary to equalize the wear of flash memory chips. However, in order to equalize the wear of flash memory with increasing capacity, the workload increases proportionally. In particular, when searching for a block with the maximum/minimum number of deletions for all blocks of a flash memory chip, the cost increases depending on the capacity of the storage device. In this paper, a random selection method of blocks was applied to solve the previous problem. When k is the randomly selected block, actual experimental results confirmed that searching all blocks with an k value of 4 or more yields similar results.

최근에는 클라우드 서버, 데이터센터 등에서 플래시 기반의 저장장치가 활발히 활용되면서 신뢰성이 더욱 중요해지고 있다. 플래시 메모리 칩은 읽기/쓰기에 제한이 있어 한곳에 쓰기가 집중되면 칩을 더 이상 사용할 수 없게 된다. 이와 같은 문제를 해결하고 신뢰성을 향상시키기 위해서는 플래시 메모리 칩의 마모를 균등화하는 것이 필요하다. 그러나 대용량이 되어가는 플래시 메모리의 마모 균등화를 위해서는 작업 부하가 비례적으로 증가한다. 특히, 플래시 메모리 칩의 전체 블록의 삭제 횟수가 최대/최소인 블록 블럭을 검색할 때 저장장치의 용량에 따라 비용이 증가한다. 본 논문에서는 앞의 문제를 해결하기 위해서 블럭의 무작위 선택 방법을 적용하였다. 무작위로 선택하는 블록을 k 라고 할 때 실제 실험 결과를 통해 k 값이 4 이상 전체 블록을 검색하는 것과 비슷한 결과를 보여주는 것을 확인하였다.

Keywords

Acknowledgement

This work was supported by Changshin University Research Fund of 2024-011

References

  1. H.Uchigaito, S.Miura and T.Nito. "Efficient Data-Allocation Scheme for Eliminating Garbage Collection During Analysis of Big Graphs Stored in NAND Flash Memory," IEEE Transactions on Computers, Vol.67, No.5, pp.646-657. 2018. https://doi.org/10.1109/TC.2017.2775624
  2. J.K.Park, Y.Seo and J.Kim, "A Flash-Based SSD Cache Management Scheme for High Performance Home Cloud Storage," IEEE Transactions on Consumer Electronics, Vol.65, No.3, pp.418-425. 2019. https://doi.org/10.1109/TCE.2019.2897686
  3. F.Wang, Y.Feng, X.Zhan, B.Chen and J.Chen, "Implementation of Data Search in Multi-Level NAND Flash Memory by Complementary Storage Scheme. IEEE Electron Device Letters," Vol.41, No.8, pp.1189-1192, 2020. https://doi.org/10.1109/LED.2020.3004989
  4. J.K.Park and E.Y.Park, "Algorithm Performance Comparison for Encryption of Big Data Stored in Cloud," Journal of Next-generation Convergence Technology Association, Vol.7, No.10, pp.1577-1584, 2023.
  5. J.K.Park, "Machine Learning Algorithm to Predict Wildfire Occurrence Based on IoT Sensor Data. Journal of Next-generation Convergence Technology Association, Vol.6, No.6, pp.967-973, 2022. https://doi.org/10.33097/JNCTA.2022.06.06.967
  6. Y.Cho, "An Exploratory Study on Consumers' Perception towards Upcycling through Big Data Analysis," Journal of Next-generation Convergence Technology Association, Vol.6, No.2, pp.172-185, 2022. https://doi.org/10.33097/JNCTA.2022.06.02.172
  7. C.Huang, F.Liu, Q.Wang and Z.Huo. Low Power Program Scheme With Capacitance-Less Charge Recycling for 3D NAND Flash Memory. IEEE Transactions on Circuits and Systems II: Express Briefs, Vol.68, No.7, pp.2478-2482, 2021.
  8. E.K.Jang, I.J.Kim, C.A.Lee, C.Yoon and J.S.Lee, "Analysis of Residual Stresses Induced in the Confined 3D NAND Flash Memory Structure for Process Optimization," IEEE Journal of the Electron Devices Society, Vol.10, pp.104-108, 2023. https://doi.org/10.1109/JEDS.2022.3140774
  9. J.Kim and J.K.Park, "Building Reliable Massive Capacity SSDs through a Flash Aware RAID-Like Protection," Appliled Science, Vol.10, pp.9149, 2020.
  10. H.S. Lee, "A Design of SSD Dedicated RAID System for Efficient Resource Management", Journal of Korea Internet of Things Society, Vol.10, No.2, pp.109-114, 2024.
  11. S.Wu, C.Du, W.Zhu, J.Zhou, H.Jiang, B.Mao and L.Zeng. "EaD: ECC-Assisted Deduplication With High Performance and Low Memory Overhead for Ultra-Low Latency Flash Storage," IEEE Transactions on Computers, Vol.72, No.1, pp.208-221, 2023. https://doi.org/10.1109/TC.2022.3152665
  12. J.Cui, C.Liu, J.Liu, J.Huang and L.T.Yang, "Exploiting Uncorrectable Data Reuse for Performance Improvement of Flash Memory," IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol.41, No.6, pp.1785-1798, 2022. https://doi.org/10.1109/TCAD.2021.3100310
  13. Y. Song, Y. Lv and L. Shi, "Adaptive Differential Wearing for Read Performance Optimization on High-Density nand Flash Memory," IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol.43, No 1, pp.380-393, 2024. https://doi.org/10.1109/TCAD.2023.3297971
  14. M.C.Yang, Y.H.Chang, T.W.Kuo and F.H.Chen. "Reducing Data Migration Overheads of Flash Wear Leveling in a Progressive Way," IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Vol.24, No.5, pp.1808-1820, 2016. https://doi.org/10.1109/TVLSI.2015.2495252
  15. M.Raquibuzzaman, A.Milenkovic and B. Ray, "Intrablock Wear Leveling to Counter Layer-to-Layer Endurance Variation of 3-D NAND Flash Memory," IEEE Transactions on Electron Devices, Vol.70, No.1, pp.70-75, 2023. https://doi.org/10.1109/TED.2022.3224420
  16. M. Mitzenmacher, "The power of two choices in randomized load balancing. IEEE Transactions on Parallel and Distributed Systems," Vol.12, No.10, pp.1094-1104, 2001. https://doi.org/10.1109/71.963420