Self Organisation in Multi-Agent Systems: A Survey of State-of-the-Art and Future Challenges

Abstract

Background/Objectives: To analyze the performance of modern information technology using multi agent system. The complexity problems arise in economics, ecology, global computer, ensuring the safety of society and in many other fields. Methods/Statistical Analysis: A systematization and description of the self organization models and mechanisms implemented in the framework of the Multi-Agent is given, networks and biological self-organization mechanisms. By using techniques such as Decentralized Swarm Agents with behaviour - based intelligence, Environment and Agent model and the formulae for each particle is defined as a solution for the problem in particular task. Findings: In this research, an approach based on solving two problems in stochastic game such as coordination team by moving anywhere and finding the joint action. For dynamic partitioning we build a coordinate tree with coordinate agent and by choosing the individual agent with a low cost and by using the techniques such as checking the group among agents based on same type of trees to evaluate the same team for each agent and to calculate the fixed solution for the joint of muti agent cooperation. Applications/Improvements: Applications of self organizing in Multi-Agent systems are telecommunications, grid resource management and routing in computer. Some of the merits are each an individual agent in an environment will find the agent with low price and it is suitable for Markov or stochastic games and group of robots cooperation and group of robot playing football game. Multi agent system used to evaluate the complex problems and it will have more number of agents which are very intelligent and interactive to overcome any kind of problems.