Application And Research On The Improved Ant Colony Optimization In Web Services

With the rapid development of internet,Web service is accepted by customers as a new form of Web application.Besides it brings a huge business opportunity to Web service providers.However,there are many fine-grained services on the Web which have same or similar function.How to combine and reuse these Web services to meet users' requirements becomes the application requirement and the research focus of the Web service combination.During the process of service combination,customers value quality more than function of Web services.But at present most Web service combination can' t solve global optimization of Web service because of the lack of global consideration.Aim at the difficulty and hot issue of Web service and based on overseas and domestic research,research the problem of Web service combination and Web service selection based on QoS in depth.First,summarize roundly the definition and calculation method of QoS in common use.On the basic of existing Web service architecture and combining the customers' non-functional requirements,present the Web service composition model based on QoS by adding the modules of users' customization and of estimate and feedback of QoS.Standardize the index of QoS for comparing and evaluating the good and bad among them.Secondly,focused research Ant Colony Optimization in Web service selection.Using basic ACO during the service selection of Web service composition model based on Qos,traditional ACO based on QoS takes shape.For improving the low and unstable success rate of traditional ACO,present MAACO.MAACO alters the rule of updating phenomone and brings in the list of optimal path in every single loop.The improvement makes MAACO have better ability of globle search and faster rate of convergence in the guarantee of success rate.At last,implement these two algorithm on VS and select the Web service among simulative sets of Web service.Through analysing the success rate of MAACO with different number of loops and ants,learn the condition on what MAACO can output optimal combination steadily.Compared with the result of traditional ACO,verify that MAACO has better performance than traditional ACO.