Visualization and optimization of agent coordination strategies in Mobile Agent Network

This paper deals with the visualization of simulation results and the optimization of operation distribution used by agents to perform operations in the mobile agent network model. This model consists of agents capable of performing remote software management operations and of various network elements and nodes agents use during operation execution. For the purpose of testing different operation distributions which define how these operations are performed a mobile agent network simulator was developed. This simulator allows users to define operations agents must perform and the physical network they use in the process. The data from the simulator is gathered during simulation, visualized using a visualization framework and analyzed at a later time by a human user. The paper will also show an approach based on genetic algorithm that optimizes the distribution of operations to reduce network load and the total execution time.