Optimal Moves for Gossiping Among Mobile Agents

Abstract

Mobile-agent-based distributed systems are attracting widespread attention as the adaptive and flexible systems: mobile agents traverse the distributed system and carry out a task at each node. In such mobile-agent-based systems, gossip is the most fundamental scheme supporting cooperation among mobile agents. It requires to accomplish all-to-all information exchange over all agents so that each agent can obtain the all information each agent initially has. Rendezvous algorithms, which require that all the agents rendezvous on a node at a time, can achieve this requirement, however it takes excessive cost for our objective. In this paper, we newly introduce the mobile agent gossip problem. In this problem, an agent can obtain the information of another agent by meeting the agent itself or the agent that has already got the information. The gossip scheme is expected to accomplish the all-to-all information exchange with a smaller number of agents’ moves than the rendezvous algorithms. We propose mobile agent gossip algorithms on several network topologies, and prove that all proposed algorithms are asymptotically optimal in term of the number of moves.