On Solving QoS-Aware Service Selection Problem with Service Composition
Runtime QoS-aware service selection is difficult because the number of web services providing similar functionality is large and constantly changing, and, even more challenging, due to the complicated structure of composite services. This paper proposes an efficient divide-and-conquer approach for service selection based on a high-level abstract model for service composition. The service composition model is recursively defined with (1) abstract control-flows of five general constructs (Sequence, Switch, While, Flow, and Pick), and (2) extended data-flows with constructs of AND and OR. Based on this service composition model, a QoS model, which supports non-linear QoS aggregation, is proposed. Therefore, complex control-flows are handled in an integrated way without separating and merging multiple execution paths, and interface constraints are captured in form of data-flow declaration. This approach allows to divide the original service to several smaller services, which are then solved separately by a recursive branchand- bound algorithm Experiments show that the proposed approach is efficient and scalable for runtime service selection, and better handles complicated structure of service composition than existing approaches.
QoS service modeling service selection service composition divide-and-conquer algorithm
Changlin Wan Carsten Ullrich Limin Chen Rui Huang Jiewen Luo Zhongzhi Shi
Key Lab of IIP, Institute of Computing Technology, CAS, Beijing, China Graduate University of Chines Shanghai Jiao Tong University, Shanghai, China Key Lab of IIP, Institute of Computing Technology, CAS, Beijing, China
国际会议
第七届网格与协同计算国际会议(Seventh International Conference on Grid and Cooperative Computing GCC 2008)
深圳
英文
467-474
2008-10-24(万方平台首次上网日期,不代表论文的发表时间)