Research On Reliability Test Case Generation Based On Markov Model

Software reliability is an important indicator to measure the quality of software.Software reliability testing can ensure and improve the reliability of software.Among all kinds of reliability models,Markov usage model has become a hot research topic.In the software reliability test based on Markov usage model,we need to build Markov usage model for software at first,and then generate test cases according to this model,measure the reliability of software at last.The test case generation is a key point during the phases of the testing process.In this paper,we study the existing reliability testing process,and optimize the related techniques for the test case generation.In the model building phase,an optimization method based on Consistency Ratio,Trajectory Entropy and the important degree of the model is proposed.In the test case generation stage,combined with the rapid mapping,an accelerate method of test cases generation based on basis test paths set is provided.In the stage of determining adequacy,combined with the existing method a hybrid stopping criteria is worked out,and an adaptive threshold scheme according to the reliability and complexity of model is presented.Finally,a test case generation tool MCTGEN is designed and implemented based on these theories.The experiments show the tool is simple and easy to use,which demonstrate the feasibility and effectiveness of the theories.