Research Of QoS-aware Web Services Selection

Since web services, due to web service is a platform independent, low coupling,interactive convenience, easy to expand and other characteristics, making web services the rapid increase in the number of deployed on the Internet. The number of Web services increased, making the Internet appear many functions similar to the same Web service.These feature the same web services with service quality difference, also due to the different users and web service is located between the network performance, resulting for different users of web service quality feel different. As a result, users need a way to help them choose to use the quality of the better Web services. This problem is more obvious in the composition of service composition of multiple Web services, so Web service selection is the key problem in the process of service composition construction. This thesis mainly studies the Qo S aware method and the Qo S based Web service selection algorithm.At present, the selection of Web services is based on Qo S data, which make the accuracy of Qo S become very important. The accuracy of Qo S depends on the method of Qo S aware. Most of the existing Qo S aware methods only monitor the running of the Web service, and then calculate the average of the parameters. Or, directly by the service provider to provide Qo S metrics. Through the average Web service quality can only provide a Web service quality, the lack of attributes to measure the stability of Web services. On the other hand, simply Web service all the running data onto Qo S perception,it can not reflect the actual situation of different users for the quality of Web services feels different from the actual situation. And the direct provision of Qo S indicators by service providers is no accuracy at all. In order to solve these problems, this thesis proposes a cloud based method of the uncertain Qo S perception.The traditional Web based Qo S service selection algorithm is mostly single objective optimization algorithm, or multi-objective optimization problem is transformed into a single target by weighted way. This method requires the user to provide the Qo S attribute weight relationship of the service selection. This is often difficult for the user to determine, but it simplifies the service selection algorithm. Since the user is not really sure the weight relationship of the attributes of Qo S, the results obtained from the service selection algorithm are not determined to be consistent with the user’s needs. In this thesis, the multi-objective optimization method is used, and the existing research is less, and the problem of maintaining the complexity of Pareto solution set is also higher. Between the web service and the location of the user request the local logical network distance to optimize the Pareto solution set of the maintenance process, and reduces the computational complexity, and design the quantum genetic algorithm based on multi-objective optimization algorithms.In order to obtain a more close to web service the actual situation of the experimental environment, the design of the QWS data set, the real network performance data and logical distance based, using operation data method of cloud model in normal cloud generator simulation to generate the web service. In this thesis, the service selection method and the traditional NSGA-II algorithm and the multi target artificial bee colony algorithm is compared to this simulation experiment environment. The experimental results verify the feasibility of the proposed Qo S aware algorithm based on cloud model and the multi-objective optimization algorithm based on quantum inspired genetic algorithm. The experimental results show that it is better than the other two algorithms in most cases. In the end, the thesis designs a small Web service selection simulation system based on the above theory.