Optimizing The QoS Composition Process Based On The Improved Cuckoo Search Algorithm

With the fast development of the cloud computing, the research on the WebService has faced more challenges. Generally, traditional Web service compositionalgorithms only take the consideration of the functional requirements, but care littleabout the non-functional requirements. The composition algorithm is the key elementwhich can affect the final QoS of the service composition.This paper has summarized the research about the web services composition, andclassified from multiple research directions. For the direction of the optimizationmethod people choose, most of the researches are based on the PSO algorithm,simulated annealing algorithm, skyline algorithm etc. Because of many experimentshave proved that the Cuckoo Search (CS) algorithm performed better than the otheralgorithm at solving multi-objective function problems, and compared with othermeta-heuristic optimize algorithm, it has wider versatility. In this paper, we appliesCS algorithm to the web service composition problem. Firstly, this paper establishedthe mathematical models of the web services composition problem based on QoS,transformed the problem from multi-objective problem to single-objective problem.Secondly, this paper established a abstract service composition model, through usingthe similar web services choosing from the QWS database. For verifying thefeasibility of the proposed method, this paper also implemented the method which ismost widely used on the web service composition problem, based on themathematical model. By comparison, this paper proved the feasibility of solving theweb service composition problem with Cuckoo Search algorithm. Finally, this paperdo some improved for the process of random value, through experimenting on theservices sets, proved the improved CS algorithm are more efficiency in the webservice composition optimization problem.