Multiobjective Optimization Algorithm Of Service Composition Based On Cloud Evolutionary

Currently, the global information technology industry is a massive wave of "Cloud Computing". The cloud-computing resources stored in the cloud providers cluster virtual machines on the local computer needs only through the Internet to send a task request, cloud on the needs of the user information through the combination of resources and optimize the composition of different forms of cloud applications. Cloud Computing philosophy of "services on demand" concept, which is the second personal computers, the Internet after the third wave of information technology.To achieve the goal of on-demand services, cloud providers through virtualization, parallel computing, distributed storage key technologies such as the application of traditional patterns of migration to the cloud, through virtualization technology form various features of the virtual resource pool allows users to access computing resources on demand. Under the pay-on-demand business model, users are more concerned with how to accomplish more tasks with minimum cost requirements. Service composition through dynamic optimization technology encapsulates the cloud resources, has formed a variety of forms and functions of new services to meet diverse user needs. Therefore, cloud environment service composition execution optimization problem has always been one o f the hotspots.In a cloud environment, service providers need to offer virtual machine hosting services to users, while providing many optimized to form case applicable to different user types to choose from. Thesis abstract the cloud model, users are most concerned about two of the objective function, that it took to complete the task of running time and costs, apply the NSGA? method and MOEA/D method to the model, service portfolio optimizations performed, under the tasks of different sizes, better solutions are obtained. The main tasks are as follows:(1) The architecture and service model of cloud services is introduced, and the problem of task scheduling in cloud environment is described, and the mathematical model of service composition optimization problem is analyzed.(2) Based on the analysis of cloud computing scheduling model, the optimization problem is modeled as a graph mapping problem, and the objective function is based on running time and running cost.(3) The optimization model based on the user task execution time and operation cost is established, and the optimization method and algorithm of service composition based on NSGA? algorithm and MOEA/D algorithm are discussed. Based, using the improved C hebyshev decomposition method and increase regio nal preference of MOEA/D is improved. In order to achieve better effect.Finally, through the C++ simulation platform to compile, through NSGA? and MOEA/D two kinds of algorithms, it shows that the MOEA/D algorithm can better meet the needs of the user tasks, and determine the optimal service composition, and verify that the proposed method has good performance in time span and cost.