Complexity Analysis for Petri Net-Based Business Process in Web Service Composition

Web services technology provides a way to integrate some distributed service units over the network into a coordinative system. Compared with the traditional enterprise application integration (EAI) techniques, it provides better interoperability for data exchange and application invocation. Therefore, it has been widely adopted for constructing distributed applications. Due to code invisibility and distributed execution of Web service unit, how to precisely measure the control complexity of Web service composition (WSC) is a very difficult task. In the paper, we mainly concern on the complexity measurement of Petri net-based business process in Web service composition. Two metric sets are presented through analyzing the WSC´s execution logics and dependency relations in workflow. The first one is count-based metric set, and includes seven metrics such as number of place, average degree of transition, transfer number per service and cyclomatic complexity. The second is an execution path-based metric set, in which the typical one is average execution path complexity (AEPC). In addition, The usability and effectiveness of our metric sets have been validated by a real-world Web service composition.