Protection of agent data state and partial results in mobile agent systems continues to draw research interest. Current solutions to integrity attacks are geared at detection of malicious activity a posteriori. We propose multi-agent architecture that uses cooperating multi-hop and single-hop agents to prevent such attacks and discuss security features of our scheme. We also examine data protection services and the means for collecting agent data state in this context. Key Words: mobile agents, security, data integrity, partial result protection
J. Todd McDonald, Alec Yasinsac, Willard C. Thomps