A novel fuzzy approach to improve near neighborhood allocation algorithm in DDB

Since accurate migrations of data fragments in distributed database systems, known dynamic fragment allocation, play an important role in amendment of distributed database performance, several algorithms each of which shows different performance in various conditions, have been proposed to improve dynamic fragment allocation in distributed database systems. In this paper, we are going to propose a novel algorithm which is able to detect oscillation conditions and provides a solution to prevent redundant fragment migration. Proposed algorithm, an improved version of the previous near neighborhood allocation (NNA) algorithm, uses a fuzzy inference engine to detect oscillations in fragment request and ignore fragment migrations under these conditions. Our results indicate that, fuzzy based clearly improved the performance of distributed database systems in which oscillation conditions are more probable. This algorithm, providing data clustering, is very suitable for DDBS in the networks, with heavy loads, and frequent requests for data fragments coming from different sites.