Predicting Performance of Multi-Agent systems during feasibility study

Agent oriented software engineering (AOSE) is a software paradigm that has grasped the attention of researchers/developers for the last few years. As a result, many different methods have been introduced to enable researchers/developers to develop multi agent systems. However Performance, a non- functional attribute have not been given that much importance for producing quality software. Performance issues must be considered throughout software project development. Predicting performance early in the life cycle during feasibility study is not considered for predicting performance. In this paper, we consider the data collected (technical and environmental factors) during feasibility study of Multi-Agent software development to predict performance. We derive an algorithm to predict the performance metrics and simulate the results using a case study on scheduling the use of runways on an airport.