Research On Service Composition In Cloud Computing Environment

The service-oriented computing model combines services with single functions as a new service to satisfy the complex needs of users.With the widespread applications of cloud computing,service composition is facing new challenges.The open and dynamic cloud environment has brought more uncertainty to the evaluation of services.Massive cloud services also present new challenges to the effectiveness and efficiency of the existing service composition methods.With the increasing diversity and scale of cloud services,it is an important issue to obtain the optimal cloud service composition solution in a short period of time in the research filed of service computing.In view of the above problems,based on the analysis of the existing service evaluation and composition technologies,this dissertation makes an in-depth analysis and research on the cloud service trust evaluation,the improvement of the service composition efficiency,and the service composition under complex requirements.The main innovations of this dissertation can be summarized as follows:(1)A fuzzy evaluation method based on consistency intensity for cloud service is proposed.Under the cloud service evaluation framework,different trusted indicators are designed for the infrastructure service and the application service respectively.A novel fuzzy evaluation method based on consistency intensity is proposed.According to the linguistic discount factor and the consistency intensity index,the certain value can be obtained from the fuzzy evaluation information,to solve the problem of the cloud service trust evaluation under the uncertain environments.A serial of experiments on NetLogo show that the proposed method is both practical and efficient.(2)A service composition method based on the discrete gbest-guided artificial bee colony algorithm is proposed.Time attenuation function is added in the service composition model to increase the time weights of recent scores,making the quality of service more consistent with the current characteristics.The service composition problem is formalized as a nonlinear integer programming problem.To solve this problem,a discrete gbest-guided artificial bee colony algorithm(DGABC)is proposed,which simulates the search for the optimal service composition solution by the exploration of bees for food.Experimental results show that the DGABC algorithm is particularly effective and efficient for large-scale data.(3)A novel multi-objective service composition model based on cost-effective optimization is proposed.According to the complex requirements of service composition,the service composition problem is as formalized as a multi-objective nonlinear integer programming problem,maximizing the QoS and minimizing the cost.Furthermore,the elite-guided multi-objective artificial bee colony(EMOABC)algorithm is proposed by adding fast non-dominated sorting method,population selection strategy,elite-guided discrete solution generation strategy,and multi-objective fitness calculation strategy into the original ABC algorithm.EMOABC has an advantage on both solution quality and efficiency comparing with other algorithms,thus is better applicable to the cloud services composition.(4)A nonlinear service composition method based on skyline is proposed.Firstly,by filtering the redundant services in each service group by skyline,the search space can be reduced to increase the efficiency of service composition.Then,the service composition problem is formulated as a 0-1 nonlinear integer programming by a modeling language AMPL,and is solved by the Bonmin solver.The experiment results show that our method can significantly improve the efficiency of service composition,while guaranteeing the service quality.