ENHANCED QoS BASED SERVICE SELECTION AND SERVICE ORCHESTRATION USING MODIFIED PSO WITH ENHANCED LOCAL SEARCH

Web service selection and orchestration have played major roles in enhancing the utility of the available web services. Due to the availability of large number of web services, it becomes mandatory to select appropriate services in order to obtain the best results. However, the huge number of web services available for a single requirement has made the process of selection very complex. The major objective of this work is to propose a fast and efficient service selection mechanism to operate on the service repository to identify services that satisfy the user's requirements. This paper proposes a variant of the metaheuristic optimization method PSO, called the DPSOSA. PSO has the major problem of getting struck in local optima. The proposed technique modifies PSO by incorporating Simulated Annealing in its local search module to eliminate local optima. Further, improvement in service selection is brought about by incorporating dynamicity into the selection process. Experiments were conducted on QWS 1.0 and QWS 2.0 datasets. It has been observed from the experiments that the proposed technique of DPSOSA exhibits lesser latency and accurate QoS when compared to the regular methods. PSO in general operates on continuous space. This is modified and the PSO used in the proposed approach is discretized to suit the current needs. Further, improvements in the working of PSO is brought about by incorporating Simulated Annealing, which is another metaheuristic technique. The proposed DPSOSA operates effectively on the data to provide results faster and also satisfying the required QOS.