Uncertainty Detection Of Complex Systems With Controllable Networks

With the rapid development of computer network technology and the more and more in-depth network applications,network has been a important part of the complex systems.Because of the opening and the uncertainty of network environments,the behaviors of the complex systems could be beyond expectations.In some extreme situations,it could lead to irreversible loss in the fields of security.So,detection of uncertain behaviors of a system under test in network environments is very important to enhance the correctness and reliability of systems.In this paper,research work focuses on detection of uncertain behaviors of sys-tems under test in network environments,by applying the frameworks of model-based testing and search-based software engineering.By combining parameters of network environments and system inputs to construct the input space of testing,network en-virotnents can be controlled and uncertain behaviors of systems under test can be detected.The main work of this paper includes the following parts:1)Abstract the system under test with its network environments based on UML St te Machine to build a testing model considering network environments changing,which could ignore irrelevant information.2)Design test case generation criterion based on the testing model considering network environments changing,which could guide the search in the large-scale state space.3)Complete the search-based test cases generation based on the test case genera-tion criterion.4)Carry out a case study on the feasibility and validity of the method proposed by this paper.