Prediction of Software Performance Using Genetic Programming

Performance is a non-functional requirement for a software product. It is related to reliability, security and other non-functional requirements. Various approaches are available for software performance prediction. In this paper we present a novel method of using Genetic Programming in reverse engineering concept. Reverse Engineering is the process of analyzing software product with the aim of recovering its design. Genetic Programming is applied for components to predict its performance. Reusing of existing components is common in Component Based Software Engineering. The response time of an entire component is predicted from the duration of execution of individual component in it. In this paper, the time taken for compression of real time data is predicted and actual time is measured. This approach can be best utilized to reverse engineer parameterized behavior models consisting of black box components.