Optimization of reconfiguration time during client movement on Siena
Kumar S. D Madhu, Babitha Balachandran · 2013
An event based system is characterized by loosely coupled heterogeneous entities which interact asynchronously. In general, the event based systems can effectively support mobile clients. The mobile clients can be the host associated with the event based system that can publish events or subscribe for events. Expressiveness and Scalability are the two key features of an event based system. This paper addresses the issue of the client movement on Siena, the event based system with expressiveness and scalability. Optimization is performed to reduce the reconfiguration time during the client movement. Standard Algorithm is the basic algorithm to support client mobility. The performance of the Standard Algorithm is enhanced using optimization techniques. The optimization techniques presented are Proxy Based Optimization, Locality Of Reference Based Optimization and Prefetch with Delayed Teardown Optimization. Analysis of the algorithms has been performed. The algorithms were simulated on simple discrete event simulator using hybrid topology of Siena. From the analysis and experimental evaluation, we conclude that Prefetch with Delayed Teardown Optimization has a finer reconfiguration time than Proxy Based Optimization and Locality Of Reference Based Optimization.