ISSN: 0976-4860
Aarti Singh, Paru Ahuja
Although agent based frameworks are finding applications in critical and sensitive applications; however with these applications are originating concerns for security & privacy of such intelligent systems. Mobile agent frameworks must be able to counter new threats as agents may opt to masquerade, generate Denial-of –service and may try to access unauthorized resources. This paper focuses on security issues of mobile agents, particularly narrowing towards agent-to-platform issues of security. This work proposes a robust two layer algorithm for ensuring complete security of the host platform by overcoming the above mentioned threats, the proposed algorithm is robust since even on failure of one layer it continues to provide security to the platform. In the proposed framework agent is required to earn a reputation score by behaving well on a host platform and thus provides genuine means to judge it.