A Genetic Algorithm for Horizontal Fragmentation To Reduce Network Traffic

In order to minimize the number of network locations to be accessed in query processing an appropriate data distribution is required. Additionally data distribution should allow equitable distribution of system load to all network locations. This process is NP-Hard and requires the optimization of conflicting objectives. In this study an improved and practical genetic algorithm was designed and verified based on multi-niche crowding GA to achieve load balancing. The power of this method stems from its ability to find multiple optima over multimodal landscape.

Keywords:

Distributed database , Fragmentation , Multi-Niche Crowding Genetic Algorithms