Effects of Communication Disruption in Mobile Agent Trust Assessments for Distributed Security

Abstract. A mobile agent is an autonomous agent that is not bound to the system on which it begins execution. Unfortunately, the mobility and interaction inherent in mobile agent technology introduces significant security problems, such as malicious host attacks, which can disable the agent, alter its code, or disrupt its communication. This research explores distributed security solutions based on cooperating agents that communicate host trustworthiness assessments to protect each other from malicious hosts. In addition, trust-based strategies are examined by which mobile agents assist each other in avoiding malicious hosts and recovering from host attacks. Communication among agents is vital to robust soft security to ensure that agents can cooperate by sharing their host trustworthiness assessments. Since agent mobility inherently makes communication difficult, unreliable, or sometimes impossible, this research conducts experiments to examine the affect of communication link disruption on distributed trust assessment techniques to pave the way for future security solutions that accommodate inter-agent communication problems. The experiments demonstrate that agent communication can be used to ensure accurate trust assessments; only a small fraction of possible communication links is required for accurately identifying and avoiding malicious hosts.