Virtual Resource Allocation and Composition for Data Stream Processing of Online Auditing in Cloud Computing

As an effective service mode, online auditing which serves the highest state audit institution is emerging as an important tool for national governance. However, as a crucial part during online auditing workflow, data stream processing has become one of the most important problems that restrains the development of online auditing. To address this problem, our study are primarily focusing on three research contents: Firstly, a hierarchal online auditing cloud framework applying to data stream processing based on SON (Service Overlay Network, SON) is proposed in this paper. Secondly, an efficient virtual resource scheduling and composition strategy called LOSPS (Local Optimal Service Path Selection, LOSPS) is also proposed, which adequately taking consideration of demands for underlying resources that satisfies upper application. Thirdly, a dynamic resource selection and composition algorithm in view of multiple attribute decision making and local optimization, which greatly supports the strategy we proposed above, is further presented. At last, various classic resource allocation algorithms including ours are discussed and their performances are compared in a simulated online auditing scenario. Simulation results show that the virtual resource scheduling and composition scheme in LOSPS outperforms the classic service composition algorithms either in static or in dynamic environments. The research findings can greatly improve the reliability and availability of online auditing via adopting the effective virtual resource allocation and composition way we proposed.