The Research Of Real-time Streaming Service Strategy To Meet The QoS Consistency

As an important web service, real-time streaming services have been widely used, but the inconsistency between the service behavior and service quality for the service has not a satisfactory solution. There are many real-time streaming services which have the same functions but different QoS attributes, the key issues is how the users to choose the service which consistent with their own preferences but also ensure the result of media quality is the best. In this thesis, firstly, we define the service behavior of the real-time streaming services on the basis of the existing solutions, and we have also done a detailed analysis of the interaction between the operation during the execution of the service behavior and service requestor and service provider, at the same time, we have used Petri network technology for service provider to formally describe a single service behavior and composed service behavior. On this basis, we have done a detailed analysis for the service which has the impact of QoS. Firstly, we have given the method to calculate the QoS parameters based on the different combinations and we also give the method of calculation of QoS parameters to be normalized.. Secondly, we define the QoS parameters for the client to describe the service behavior for requesters’QoS preferences, and done the analysis for the service behavior which have the impact of QoS based on the Petri net. On the above theory, we suggest a QoS-aware real-time streaming service architecture model based on service-oriented architecture, which add the QoS agent module to response for real-time streaming services’non-functional properties match. According to the given parameters of service behavior and quality, with reference to the client needs to match the three-match multi-objective global optimization to ensure that users can get the best service. Finally, we had proved the feasibility of the solution and superiority.