The Research On Feature Interaction Detection By Stages Technology In IN

The service interaction detecting technologies(FIDT) in Intelligent Network(IN) are the pivotal techniques in the field of IN research. It is great importance for the IN services that can run in IN with safety, stability and efficiency. Today, in the various IN services situation, how to find out the interactions among the IN services with accuracy and efficiency is a prominent problem.Firstly, starting with dynamic and static techniques belong to the FIDT, the thesis emphasizes on generalizing and analyzing the research of software engineering approach and formal method which is in the static techniques; secondly, the thesis studies the basic theory and the Basic Call State Model of IN, and realize that the IN services achieve it's functions mainly through modification of the executing program in the simplified plain ordinary telephone service (POTS), then also summarize the formalized definition of the service interaction that is service interaction appear as abnormal state in regular transition process; thirdly, based on the model of IN services lifecycle, and the character of software engineering approach and formal method, the thesis give a model of service interaction detection by stages. This model divides the entire detection process into two steps. In the services requirement stage, it uses the filtering approach to filter interactions, then in the services development stage, it uses the detecting approach to detect interactions. This model can integrate the two methods properly; fourthly, we study the visual scenario modeling tools use case map (UCM), and the call process UCM model. According to the model, it designs a modified interaction filtering method based on scenario, which can be used in the first step. The modified method is improved by adding precondition of filtering services combination and strengthen filtering theorem. Both the theoretical analysis and experiment prove that it can improve accuracy of filtering and reduce dubious conclusion, then promote the efficiency of filtering; fifthly, it studies the relationship between the formal language state transition rule (STR) and the modeling tools colored Petri net (CPN), and the character of P invariant. According to the POTS's CPN model, it designs a modified interaction detecting method based on validation database and P invariant reachability analysis, which can be used in the second step. The modified method is improved by modify user replacing theorem and import validation database. Both the theoretical analysis and experiment prove it can...