Data-centric garbage collection for NAND flash devices

Chundong Wang,Qingsong Wei,Mingdi Xue,Jun Yang,Cheng Chen
DOI: https://doi.org/10.1109/NVMSA.2015.7304360
2015-01-01
Abstract:Garbage collection has been concerned for NAND flash devices for years. The ever-increasing utilization of flash device demands more effective and efficient garbage collection strategies. This paper proposes a novel approach, namely Data-centrIc Garbage collection (DIG). DIG online forecasts update intervals for data and clusters them accordingly into groups in a lightweight way. Data with similar update intervals form a group and are stored together. Obsolete data and valid data are hence prevented from being mixed. Moreover, DIG takes advantage of clustering to further separate data and select promising victims for reclamations. Experiments show that DIG can significantly reduce the overheads of garbage collection by 94.3% and 73.5% on average, respectively, compared to two state-of-the-art algorithms.
What problem does this paper attempt to address?