Distribution of Jobs Based on Communication Time and Cost in Service Oriented Distributed Computing

Abstract

Computation time and communication cost increase continuously when the requests for services increase. To keep systems offering a good level of computation, we need to deal with these two parameters in distributed systems. Service oriented computing can be used to participate and divide jobs between computers in a network to improve computing speed. This article introduces a service-oriented approach, which shows a simultaneous improvement for main service's elapsed time. We discuss the importance of division of jobs in distributed computing and deal with important aspects of SOA, which are used and implemented in our research. A communication and formal model are given. Finally, a case study, which contains three different approaches following the same goal, is presented. Then, elapsed times by each approach is compared and conclusions are discussed.