Sambit Kumar Mishra^1,, Srikanta Pattnaik², Dulu patnaik³

The Main idea of multiple query processing is to optimize a set of queries together and execute the common operations once. Major tasks in multiple query processing are common operation or expression identification and global execution plan construction. Query plans are generally derived from registered continuous queries. They are composed of operators, which perform the actual data processing, queries which buffer data as it moves between operators to hold state of operators. The complex part is to decompose queries and query plans and rearrange the sub queries and query plans on the network. The main functions to achieve an optimal query distribution are usually minimizing network usage and minimizing response time of queries. While dealing with query distribution problem, the challenges like modeling topology of the network, decomposing queries into some sub queries and sub query placement may be occurred. Operators are the basic data processing units in a query plan. An operator takes one or more streams as input and produces a stream as output. As in the traditional database management system, a plan for query connects a set of operators in a tree. The output of a child operator forms an input of its parent operator. In this paper it is aimed to retrieve the cost of query plans as well as cost of particles of swarm in multiple query processing environments by applying particle swarm optimization techniques.

query plan, swarm, NP hard, particle, SMT, personal best, global best