Security Issues in Mobile Agents

Mobile agents offer a new paradigm for distributed computation, but their potential benefits must be weighted against the very security threats they pose. These threats originate not just in malicious agents but malicious hosts as well. Thus security is a very crucial issue in dealing with mobile agent systems without which their implementation in real life applications will be rendered useless. This paper highlights the problems encountered in a mobile agent system from a host’s point of view as well as from a mobile agents point of view, as well as the various approaches to dealing with those problems. The paper concludes with the implementation of two mobile agent systems: Aglets & D’Agents and describes their approaches to integrate security into the system. 1