Research on generation algorithm of SOA-oriented integration test order

The generation of integration test order is an important issue in the research of software integration testing. A reasonable test order can improve the efficiency of integration testing and effectively reduce the cost of testing. The architectural complexity of the Service-Oriented Architecture and the complex coupling relationship between services make it difficult for traditional top-down and bottom-up integration test strategies to generate the best integration test order. Therefore, the paper proposes a test priority algorithm oriented to the TDG(test dependency graph) to represent the importance of integration testing between service software, and a genetic algorithm is designed to generate optimal integrated test order with the optimization goal of reducing test costs. Finally, through examples demonstrate the feasibility and validity of the method. The example results show that the method can generate integration test order that allows service software with a relatively high test priority to be integrated first and has a lower test cost.