P2p Environment Based On Improved Particle Swarm Optimization (pso) Algorithm Of Web Service Selection Research

With the growing number of Web services, most of the publish services on the network have defects of simple structure and single function at present. And the needs of users for a business are dynamic. So how to select Web services according to the requirement of users and effectively combine various services distributed on the network to realize the seamless integration among services, have become an important step in the process of Web service development. The traditional centralized Web service system frame has encountered some problems such as single point of failure, performance bottleneck and service invalidation and so on. And P2P (Peer-to-Peer) technology has the advantage of efficient integration of services and autonomy of resources. So establishing Web services selection mechanism on P2P network becomes an ideal Web services implementation scheme.In this paper, it mainly combined Web services selection mechanism, particle swarm optimization algorithm and P2P technology. And Web services selection based on improved particle swarm algorithm in P2P environment was studied. The main research work is as follows:(1)An improved P2P network model was put forward. And P2P network was structural processed according to the layering thought and consistency problem of network topology. The whole P2P network was divided into three-layer Chord structure of field community, service community and regional community. This can effectively organize all kinds of Web service resources on the network and establish good network foundation for the latter research on services selection, thus improve efficiency of services selection.(2)In order to further improve the efficiency of services selection, node intelligent selection mechanism was established for user priority to select the nodes whose geographical position is the closest to users’to provide the necessary Web services. The mechanism firstly obtained users’ geographical position information by analyzing users’IP address, secondly coded every areas of our country based on the geographical position information, thirdly calculated the access distance between user nodes and service nodes using the coding information, at last chose the node which is the closest to the user geographically to provide Web services.(3)The existing service selection algorithms have defects such as inefficient, non-global optimization. Moreover there are some shortages of particle swarm optimization used in complex combinatorial optimization problems. Web services selection which based on user’s preference and grouping particle swarm optimization was presented by analysis of user’s specific preference in Web services selection. This algorithm in use of particle clustering, improves the global search capability in Web services and avoids algorithm from rapid convergence or premature. At the same time, the user’s preferences are transformed into a fuzzy constraint satisfaction problem. So the users can get humanized service when they select services.(4)A simulation experiment of Web services selection method based on improved particle swarm algorithm in P2P environment was made in the OpenChord platform. Experiments show that this method can effectively solve the bottleneck problems of the centralized Web services selection in the process of the service-oriented computing and has higher efficiency of services selection.