Tackling trust issues in virtual organization load balancing

Abstract

Agent-based Virtual Organizations are complex entities where dynamic collections of agents agree to share resources in order to accomplish a global goal. Virtual Organizations offer complex services that require of the cooperation of distributed agents. An important problem for the performance of the Virtual Organization is distributing the agents across the computational resources so that the system achieves a good load balancing. In this paper, a solution for the agent distribution across hosts in an agent-based Virtual Organization is proposed. The solution is based on a genetic algorithm that is meant to be applied just after the formation of the Virtual Organization. The developed genetic strategy uses an elitist crossover operator where one of the children inherits the most promising genetic material from the parents with higher probability. This proposal differs from current works since it takes into account load balancing, software requirements of the agents and trust issues. In order to validate the proposal, the designed genetic algorithm has been succesfully compared to different heuristic methods that solve the same addresed problem. © 2010 Springer-Verlag.