Optimal Selection of Web Services with End-to-End Constraints

Yan Gao,Bin Zhang,Jun Na,Lei Yang,Yu Dai,Qiang Gong

DOI: https://doi.org/10.1109/imsccs.2006.246

2006-01-01

Abstract:Nowadays, Web services are growing very fast and are usually aggregated into a composite one to satisfy customer's more and more complex requirements. Generally, there may be several different candidate services to carry out one task in a composite service, so a choice needs to be made for helping users select the most suitable one to satisfy their end-to-end constraints. This paper addresses this issue as a multistage decision problem. Based on the quality and attribute value of services, this paper generates a weighted multistage graph for composite service, and transform the problem of services selection into the one of how to get a longest path under a given end-to-end constraints. Considering the problem of interface matching, this paper presents a 3-layer Web service organization model (WS3LM), which can help getting executable composite service with well matching. This paper describes and compares two of global optimal selection approaches: dynamic programming and genetic algorithm