I.J. Intelligent Systems and Applications, 2012, 5, 8-15 Published Online May 2012 in MECS (http://www.mecs-press.org/) DOI: 10.5815/ijisa.2012.05.02 Copyright © 2012 MECS I.J. Intelligent Systems and Applications, 2012, 5, 8-15 An Analysis of the Effect of Communication for Multi-agent Planning in a Grid World Domain Satyendra Singh Chouhan Department of Electronics & Computer Engineering, Indian Institute of Technology, Roorkee Roorkee-247667, Uttrakhand, INDIA Email:saty1pec@iitr.ernet.in Rajdeep Niyogi Department of Electronics & Computer Engineering, Indian Institute of Technology, Roorkee Roorkee-247667, Uttrakhand, INDIA Email: rajdpfec@iitr.ernet.in Abstract — Agent-based technology has generated a lot of attention in recent years because of its promise as a new paradigm for conceptualizing, designing, and implementing software systems. Some problems in real world cannot be handled by a single agent. Multiple agents work together to accomplish some task. Although multi-agent systems (MASs) provide many potential advantages, they also present many difficult challenges. This paper illustrates the importance of communication for planning in a multi-agent setting by considering a grid world domain that consists of obstacles at different locations. This paper provides a theoretical framework that is validated by the experimental results. Performance analysis with respect to plan size and execution time is also reported. Index Terms— Grid world Domain, Communication, Multi-agent system I. I NTRO DUC TIO N Research in multi-agent systems have led to its applicability in varied real world scenarios such as e- commerce [1], supply chain management [2], robotics [3], and also developing complex game application [4].It is widely being advocated for use in networking and mobile technologies, to achieve automatic and dynamic load balancing, high scalability, and self-healing networks. Such systems are becoming increasingly important as they draw together a number of important trends in modern technology [5]. A multi-agent system (MAS) consists of a collection of loosely-coupled interacting autonomous agents working in an environment. Here agents are usually software agents and can perform actions, have some computational abilities, and may communicate with each other. Multi-agent systems support modularization i.e., a large complex problem is handled by developing a number of functionally specific and modular components that are efficient to solve a specific problem aspect. This decomposition allows each agent to use the most appropriate paradigm for solving its particular problem. When dependent problems arise, the agents in the system coordinate with one another to ensure that interdependencies are properly managed. Cooperation refers to distributed and communicated group of agents that share a common interest and work together to achieve a common goal in an environment [6], [7]. The remaining paper is structured as follows. In Section 2 we discuss about agent communication and grid world domain. Section 3 describes a model used for communication. Section 4 describes a theoretical framework for multi agent communication. In Section 5 the experimental results are presented along with the analysis. Related work is discussed in Section 6. Section 7 concludes the paper. II. AGENT CO MMUNIC ATION The information exchange between agents is termed as communication between agents. Communication improves the behavior of agents and discourages any regard to other agent’s internal structure. The communication between agents may be peer to peer, broadcast or mediated. In recent years several techniques evolved for the communication between agents. Some of these include Blackboard system [8], message passing via communication standards like knowledge query manipulation language (KQML) [9], and FIPA-ACL [10]. Blackboard system is an indirect approach of communication between agents. In this technique there is a common blackboard which is shared between all agents. For peer to peer communication KQML and FIPA-ACL are two most popular languages. Since these languages are wrapper languages and have high level of abstraction they can be extended depending on the need