Distributed Algorithm for Multi-agent Environment

Traditional algorithms are suitable when steps of algorithm are to be executed in a sequence and at a single location. These algorithms have limitations, when used in multi-agent environment, where different agents of single application run on different nodes, thus emphasizing the need for distributed algorithms. Agents in multi-agent systems perform task independently and share the result with other agents. This motivates to use distributed algorithm, where tasks of the business process are distributed among agents, so that they can perform in parallel and independently. This paper presents a distributed algorithm for multi agent systems with case study of budget allocation in web-based environment. The algorithm is based on heuristic approach. Three agents of systems perform different tasks of budget allocation procedure like information gathering, filtering, evaluating and budget allocation at different locations and timings. This approach takes advantage of distributing and parallel computing.