Software Testing Resource Allocation Algorithm based on Improved Evolutionary Algorithm

Previous studies on software systems mostly focus on static models without fully considering dynamic changes. Aiming at the problem of the insufficient testing resources in the early stage, we proposed a software testing resource algorithm based on improved evolutionary algorithm. In the designed model, the multi-objective algorithm is constructed, the execution of the multi-objective optimization algorithm also needs to first establish the reliability model of the many-core software, and estimate the parameters of the reliability model from the failure data. In addition, it is necessary to establish a test cost model and determine its related parameters. At the same time, this paper chooses throughput, delay and packet loss rate as the performance measurement indicators of virtualized network functions and virtualized network function service chains. Through the test of controling the iteration times, the performance of the proposed model is tested.