FTCG: Fine-to-Coarse Multi-Granularity Grouping for Migratory Compression

With the exponential growth of data volume, lossless compression technology has become a crucial technique for alleviating storage pressure. Existing Migratory Compression (MC) algorithms exhibit inefficient utilization of the compression sliding window due to excessively large groups and distant similar data blocks. To address this issue, this paper proposes FTCG, a Fine-to-Coarse Multi-Granularity grouping migratory compression scheme. FTCG employs Hierarchical Super-Features (HSF) to achieve progressive similarity quantification from fine to coarse granularity, thereby prioritizing the aggregation of highly similar data blocks to enhance local redundancy elimination. Simultaneously, an efficient sorting-based grouping strategy, HSFRank, is designed to replace the traditional greedy algorithm with linear complexity, significantly reducing computational overhead. Our evaluation results show that FTCG improves the average compression ratio by 42.50% across nine datasets, with compression ratio gains of 58.20% and 51.57% on the Bash and LKT datasets, respectively, and a maximum throughput improvement of 60.60%.