Teacher-learner & Multi-objective Genetic Algorithm Based Query Optimization Approach for Heterogeneous Distributed Database Systems

Growing database demands more technological developments and computing paradigms, as Grid and Cloud computing, unleashed new developments in the database technology sector. Query Optimization is essentially a complex search task to obtain the best possible plan from the enormously increasing databases. Heterogeneous Distributed database management systems (DDBMS) are amongst the most important and successful software developments where the query processing is more difficult since large number of parameters effect the performance of the queries. Thus, the author attempted to introduce a new approach for Query Optimization in Heterogeneous DDBMS both for local and global optimization separately. In this paper, two stochastic approaches such as multi-objective genetic algorithm for local optimization and teacher-learner based optimization for global optimization is employed. The local optimization approach deals with optimization within the local sites whereas global optimization works with the sites at different locations globally. Join ordering cost (JOC), Total Local Processing Cost (TLPC) and Total Communication Cost (TCC) are used to obtain the optimal query plans amongst the relation between the query sites. The Experimental Analysis of the proposed approach showed that it has better performance i.e. less cost when compared with other heterogeneous DDBMS and has more cost when compared with the other existing homogeneous approaches.