Methodology for Performance Evaluation of Distributed Multi Agent System

Abstract

This paper presents a methodology for the assessment of the distributed multi agent system. It requires apt attention to check as if multi agent system is performing well. In recent research performance is taken as a rate, queuing system or some measure without capturing all characterization of the multi agent system, there is a lag for the proper evaluation of multi agent system. Our methodology takes the generic characteristics of the distributed multi agent system which can affect the overall efficiency of the system, and spotted some performance metrics. These metrics depict the clear image to measure the efficiency of the distributed multi agent systems. It is a step by step technique; in initial step, some performance metrics are derived from the generic properties of the multi agent system. In second step, the graphical interaction model of the real time multi agent system is developed. The Course Management System (CMS) and the Remote Patient Monitoring Systems (RPMS) are two distributed multi agent systems, their interaction models are generated using log files in Java Agent Development Environment. In the final step; the function from the graph theory is taken to validate the performance metrics for the both multi agent system CMS and RPMS.