Siping Liu, Yaoxue Zhang, Yuezhi Zhou, Hao Liu, Di Zhang, and Wei Hu
Web Service Selection, Quality of Service, Probe Cost, User Requirements, Optimal Stopping Theory
This paper demonstrates a distributed on-line service selection(probe/access) scheme: optimal stopping web service selection scheme based on the rate of return problem from optimal stopping theory. There are three differences between our scheme and the conventional schemes. Firstly, it does not need to probe all web services, and only probe a few web services. Secondly, our scheme focuses on maximizing the average QoS(Quality of Service) return per unit of cost over all stages of probe and access for a long period rather than maximizing QoS return per single stage of probe and access in usual schemes. Thirdly, our scheme develops a return function based on three factors: QoS return, user's requirement and probe cost which are seldom considered simultaneously before. Through theory analysis and computation, we demonstrate that compared with the conventional schemes our scheme has additional advantages while achieving same good performances.
Important Links:
Go Back
