Interdependence Between Agents in Multi Agent Systems : A comprehensive study on agents' relationship for enhancing cooperation rate and individual agents' utility

Interdependence relationships have defined the foundation of cooperation between agents but have been limited by existing protocols. As a consequence, the idle agents are not able to form joint collaboration and benefit from it. First, the inter-relationship has been studied which it secures agents' mutual gains. It shows the possibility of finding an ideal partner agent and further reduces the number of agents in relationship. Also, the cost required for forming collaboration between agents is noticeable reduced by introducing dependence relationship reduction. Next, the join coalition mechanism is proposed to help idle agents to join existing macroscopic and microscopic coalitions which are based on the goals compatibilities, budget and trust. The proposed methods enable the idle agent to join the coalition as long as it has compatible goals. Contrast to traditional protocol which is outsourcing agents for achieving certain goal, join coalition mechanism allows agent to be pro-active and hunt for their ideal coalition.