Abstract

Mobile Agent System (MAS) is an open multi agent system, in which mobile agents can transport its state from one environment to another, with its data intact and capable of performing in new environment. This technology gave biggest opportunity to internet based applications. But fault detection and recovery is a vital issue in the deployment of MAS, whose purpose is to provide a distributed computing infrastructure for supporting applications in which components can move freely in heterogeneous environment without any hindrance. Despite the considerable efforts spent on researching and developing Multi-Agent Systems there is a noticeable absence of deployed systems. In the past the MAS research community ignored this problem – arguing that it is not a genuine MAS problem and consequently of lesser importance than other unsolved issues like cooperation, coordination, negotiation and communication. This paper presents a Watchdog approach to discover communication failure during MAS execution and recover it through parallel check pointing antecedence graph approach with low overhead. We believe that introducing fault identification and recovery in a mobile agent system through antecedence graphs is novel and provides a low overhead and effective solution for fault-tolerance in a multi-agent system.