Service Composition And Validation Based On Hierarchical State Machine

In service-oriented software architecture,the function of software system is regarded as service it provides.With the complexity of software system increase,most classical software development method tends to realize new functions of the system using combination of existing components either directly from development platform or other developed components.Thus the problem of how to use existing services to satisfy more complex business logic arise,and the effectiveness of composite service needs to be fully validated.Through the analysis and comparison of existing service composition methods,state machine is used to model the service provided by the software system,thus the problem of service composition is transformed into state machine synthesis.In order to solve the problem of state space explosion caused by system complexity in state based modeling method,hierarchical method is adopted to divide the state space reasonably.Firstly,this thesis introduce the concept of non-deterministic and deterministic finite automata into the field of service,and put forward the concept of external and internal schema of service state machine based on regular expressions,which describe the service of the software system.Secondly,on the basis of service description,according to the existence of intersection among the sub service state space,a decision algorithm of service combination is proposed.Thirdly,according to the reachability analysis of state machines,the validity of the service composition is verified by calculating Cartesian product of the state machines.Fourthly,in order to relive state space explosion problem in complex system,by original sub service and corresponding composite service,get the corresponding hierarchical state machine,and properties invariant of composite service state machine can be verified by the proposed algorithm.Finally,by modeling,combination and validation of service,the proposed method is applied to an example of car rental system,which demonstrate that the proposed method can effectively solve the problem of service composition.