EF-Greedy: A Novel Garbage Collection Policy for Flash Memory Based Embedded Systems

  • Authors:
  • Ohhoon Kwon;Jaewoo Lee;Kern Koh

  • Affiliations:
  • School of Computer Science and Engineering, Seoul National University,;School of Computer Science and Engineering, Seoul National University,;School of Computer Science and Engineering, Seoul National University,

  • Venue:
  • ICCS '07 Proceedings of the 7th international conference on Computational Science, Part IV: ICCS 2007
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Flash memory is becoming increasingly important for embedded systems because it has many attractive features such as small size, fast access speeds, shock resistance, and light weight. Although flash memory has attractive features, it should perform garbage collection, which includes erase operations. The erase operations are very slow, and usually decrease the performance of the system. Besides, the number of the erase operations allowed to each block is also limited. To minimize the garbage collection time and evenly wear out, our proposed garbage collection policy focuses on minimizing the garbage collection time and wear-leveling. Trace-driven simulations show that the proposed policy performs better than existing garbage collection policies in terms of the garbage collection time and the endurance of flash memory. Specifically, we have shown that the performance improvement of our proposed policy against the greedy policy in terms of the endurance of flash memory is as much as 90.6%.