A Novel Chunk Coalescing Algorithm for Data Deduplication in Cloud Storage

Data deduplication has been widely attracted by researchers in cloud storage. Existing works usually adopt variable-size chunking and merge some subchunks into superchunk so as to improve storage and bandwidth utilization, but the overheads of chunk coalescing process are not considered in these algorithms. Due to this situation, we propose a novel chunk coalescing algorithm in this paper, which concerns both the maximum and minimum number of subchunks, that should be coalesced into superchunks. Experimental results show that our algorithm indeed reduces the overheads of chunk coalescing process and accelerates the whole procedure during the data deduplication.