Flash memory has been revealing its massive potential since 1990s, it has becoming widely to use this nonvolatile storage to replace magnetic disk in many commercial portable device and the INTEL corporation produce the Z68 chipset supporting the solid-state disk as a disk cache to promote faster access to user data than the hard disk recently. The portable device not only become lighter and thinner but provide more battery usage due to without no moving part inside of the storage. Even the flash memory provide the high performance, physical stability, shockproof, low power consumption and portability, but one disadvantage the flash memory suffered is that cannot be overwritten on place. Before this overwrite operation process, the time-consuming erase operation must be needed, which takes an order of significantly latency than a write operation because the block unit erase operation differ to the page-unit read or write operation. One block would compose of a range from 64 to 256 pages in the modern flash memory-based product. To address this technical obstacle, device could put a firmware called flash translation layer (FTL) between the file system and low level physical flash memory unit to postpone even eliminate numbers of erase operation. However, among many proposed FTL schemes, their performance depends on the trace pattern. Hence, we propose a revised FTL scheme much suitable networking server trace and multi-programming system. Furthermore, in order to simplify our work, we will focus on translate logic address of write request to physical address more efficiently in this paper. The read request, as we have well known that the cache could reduce the read request latency on the basis of program spatial locality and temporal locality and the flash memory take advantage of its faster random read characteristic than the disk.

自1990年起，快閃記憶體逐漸發展其驚人的潛力，並廣泛取代傳統硬碟成為可攜帶商業產品的非揮發性記憶儲存體。近來，IT產業大廠英特爾更推出Z68晶片可支援固態硬碟為傳統硬碟的快取單元以提升整體電腦存取資料效能。使用快閃記憶較傳統磁性硬碟的好處是讓裝置變得更輕、更薄、可靠度增加、存取速度快，最重要的是其省電特性延長電池單次的使用時間。基於快閃記憶體內部使用電晶體作為存取的單元，而非傳統硬碟透過同時移動讀寫頭與不停轉動的磁盤，因而減少了物理能量的消耗。雖然快閃記憶體擁有很多的優勢像是高效能、資料穩定度高、抗震力強、低耗電以及體積小等等，但是資料無法直接重寫在相同位置上為其所苦的唯一重大缺點。要在一個已寫入的頁面上覆蓋下一筆資料，必須要先對區塊做整體抹除後，才能將資料寫入，並且每次抹除的操作時間通常為寫入與讀出這兩項操作十數倍，因此抹除操作相對性的耗時，根據文獻與產品的調查，在固態硬碟的設計方針上，通常使用64到256頁面作為一個區塊單位的大小。為了降低這個缺點對整體效能的影響，我們透過在裝置中放置一個韌體被稱作快閃記憶體位址轉譯層(Flash Translation Layer, FTL)，該項韌體可以用來重新轉譯檔案系統要求的寫入位置到實體快閃記憶中的存放位置以降低抹除操作次數或延後抹除時間。縱觀文獻所提出的方案當中，我們可以發現每一個方案的效能將受到工作量軌跡格式的不同而有所影響。 對此，我們針對透過網路方式連結的裝置儲存體和多工操作使用的電腦裝置儲存體提供一個可動態調整SW日誌區塊大小的FAST快閃記憶體位址轉譯層策略，並透過模擬方式表現改善後的效能。為簡化我們的研究工作，我們將研究的主旨放在如何有效的轉譯檔案系統所要求寫入的位置到實體快閃記憶體中的存放位置，而從檔案系統的讀取行為，從作業系統原理當中，我們已知，因為程式的空間區域性和時間區域性能使記憶體中的快取策略，有效降低讀取延遲時間，除此之外，快閃記憶題本身具備快速隨機讀取特性，因此，往後篇幅將不提讀取方式的改善。

[1] M. Rosenblum, J. K. Ousterhout, ”The Design and Implementation of a Log-Structured File System.” ACM Transactions on Computer Systems., vol. 10, no.1, p.26-52, 1992.
[2] J. Kim, J. M. kim, S. Noh, S. Min, Y. Cho, ”A Space-Efficient Flash Translation
Layer for Compactflash System.” IEEE Transactions on Consumer Electronics., vol. 48, no.2, p.366-375, 2002.
[3] J. Kang, H. Jo, J. Kim, J. Lee, ”A Superblock-based Flash Translation Layer for
NAND Flash Memory.” In Proceedings of the International Conference on Embedded Software., p161-170, 2006.
[4] S. Lee, D. Park, T. Chung, D. Lee, S. Park, H. Song, ”A Log Buffer-based Flash Translation Layer Using Fully-Associative Sector Translation.” IEEE Transac- tions on Embedded Computing System., vol. 63, no. 3, Article 18, 2007.
[5] S. Lee, D. Shin, Y. Kim, J. Kim, ”LAST:Locality-Aware Sector Translation for NAND Flash Memory-Based Storage Systems.” ACM Gigops Operating Systems Review., vol. 42, no. 6, p.63-42, 2008.
[6] A. Gupta, Y. Kim, B. Urqaonkar, ”DFTL:A Flash Translation Layer Employing
Demand-Based Selective Caching of Page-Level Address Mappings.” In Proceedings of the International Conference on Architectural Support for Programming Lauguages and Operating System., p.229-240, 2009.
[7] Y. Kim, B. Tauras, A. Gupta, B. Urqaonkar, ”FlashSim:A Simulator for NAND
Flash-Based Solid-State Drives” In Proceedings of the International Conference on System Simulation., p125-131, 2009.
[8] G. Ganger, B. Worthington, Y. Patt, “The DiskSim Simulation Environment Version 3.0 Reference Manual.”
[9] INTEL Corporation, ”Intel Z68 Express Chipset Product,”
http://www.intel.com/Assets/PDF/prodbrief/325516.pdf.
[10] CompactFlash Association, ”Information About the CompactFlash,” http://www.compactash.org.
[11] White Paper: ”Datacenter SSDs: Solid Footing for Growth,”
http://www.samsung.com/global/business/semiconductor/products/Products.htm.
[12] Intel Corporation, ”Understanding the Flash Translation Layer Specification,” http://www.scribd.com/doc/7168824/Understanding-Flash-Translation-Layer.
[13] A. Silberschatz, P. B. Galvin, G. Gagne, ”Operating System Principles, 7th ed,” 2006
[14] Samsung Electronics, “1G x 8 Bit/ 2G x 8 Bit NAND Flash Memory Product: K9F8G08X0M Specification,” http://www.samsung.com/global/system/business/ semiconductor/product/2007/6/11/NANDFlash/SLC_LargeBlock/8Gbit/K9F8G08U0M/ds_k9f8g08x0m_rev10.pdf, 2007
[15] M. -L. Chiang, P. C. Lee, R.-C. Lee, “Using Data Clustering to Improve Cleaning Performance for Flash Memory,” Software-Practice and Experience., vol. 29, no. 3, p.267-290, 1999
[16] A. Ban, “Flash File System.” United States Patent, no. 5,404,485, 1995
[17] C. Park, W. Cheon, J. Kang, K. Roh, W. Cho, “A Reconfigurable FTL(Flash Translation Layer) Architecture for NAND Flash-Based Applications,” ACM Transactions on Embedded Computing Systems, vol. 7, no. 38, 2008.
[18] X. -Y. Hu, E. Eleftheriou, R. Haas, I. Iliadis, R. Pletka, “Write Amplification Analysis in Flash-Based Solid State Drives,” In Proceedings of the Israeli Experimental Systems Conference on SYSTOR, no. 10, 2009
[19] OLTP Trace from UMass Trace Repository.
http://traces.cs.umass.edu/index.php/Storage/Storage.
[20] FlashSim Test Trace from PSU Computer Systems Lab.
http://csl.cse.psu.edu/?q=node/322