Distributed decentralized architecture for autonomous cooperative operation of multiple agent system

Abstract

This paper presents a new distributed decentralized architecture for the organization, command, control and communication (C/sup 3/) of multiple agent system which has to execute a mission, cooperatively and autonomously. This architecture can be applied to multiple unmanned underwater vehicles (UUVs), to unmanned air vehicles (UAVs), as well as to unmanned ground vehicles (UGVs), unmanned space vehicles (USVs) or to any other types of autonomous cooperative multiple agent systems. The rationale for autonomous cooperative operation of a multiple agent system, stems from the need to execute critical missions under time, space resources and availability constraints, which are beyond the capability of a single agent to perform successfully. If the system agents are distributed geographically, cooperative operation is a reasonable approach. It supports share of information, share of resources, efficient resource allocation, context and situation driven responsiveness, robustness and flexibility under changing conditions, as well as redundancy. A key feature of the proposed architecture is that all the agents are identical in their sensing, information processing, decision making, communication and mission related capabilities. Thus, a failed agent can be functionally replaced by a peer agent. The architecture is both distributed and decentralized. It is implemented via four major building blocks, which are embedded within each agent.