In the literature of distributed database system (DDBS), several methods sought to meet the satisfactory reduction on transmission cost (TC) and were seen substantially effective. Data Fragmentation, site clustering, and data distribution have been considered the major leading TC-mitigating influencers. Sites clustering, on one hand, aims at grouping sites appropriately according to certain similarity metrics. On the other hand, data distribution seeks to allocate the fragmented data into clusters/sites properly. The combination of these methods, however, has been shown fruitful concerning TC reduction along with network overheads. In this work, hence, a heuristic clusteringbased approach for vertical fragmentation and data allocation is meticulously designed. The focus is directed on proposing an influential solution for improving relational DDBS throughputs across an aggregated similaritybased fragmentation procedure, an effective site clustering and a greedy algorithm-driven data allocation model. Moreover, the data replication is also considered so TC is further minimized. Through the delineatedbelow evaluation, the findings of experimental implementation have been observed to be promising.

No abstract