Research On Web Service Selection Based On Improved BPSO Algorithm

With the extensive application of the web services on the internet, a single webservice cannot meet the complicated requirements of service requester, so servicecomposition has been raised to solve the problem. The composite web service is madeof the previous single web services, and it offers a new and more powerfulperformance. However, there are so many web services with different quality ofservice and similar or identical functions on internet. How to select a set of servicesamong a lot of services in order to make the web service compositions achieve thebetter user satisfaction and better quality has been an important problem. But theexisting selection technologies of the composition web services based on functionaldescription can not solve the problem. So the web service selection technology baseon quality of service has became a focus of attention for both industry and academia.To overcome the drawbacks of sub-optimization involve in standard particleswarm optimization (PSO) algorithm, proposed the improved binary particle swarmoptimization algorithm (BPSO).This paper applied the improved algorithm to theselection of web service composition based on QoS.Firstly, this paper introduced the related technologies of web service, includingthe concept、features and data integration technologies of system structure. For theexisting advantages and disadvantages of the emerging Web service Compositionselection algorithm has been detailed analyzed. Deep research on service quality ofservice (QoS) quantization process technology and a service selection model withmultiple QoS constraints has been proposed. We have defined the QoS attributemeaning, and a set of service quality evaluation function. According to the user’sapplication environment and personal preference to set the QoS attribute weight, theweb services composition more meet the needs of users.Secondly, considering the QoS based web service selection problem is a NP hardproblem, by analyzing the principle of particle swarm optimization algorithm and it’sthe advantages and disadvantages, this paper introduce mutation operation in iterationprocess, enhanced the ability to break away from the local optimum. According to thepopulation’s fitness variance value adaptively to adjust the mutation probability. Theinertial weight reduced by the regular can enhance the optimization capability of thealgorithm and greedy strategy can modify the infeasible particle. At last, experiments on QoS-aware web services selection by using improvedbinary particle swarm optimization algorithm (BPSO) have been designed andimplemented. By comparing with genetic algorithm according to simulationexperiments, the results indicate that the superiority of the algorithm on the solvespeed and the average of fitness.