Science & Engineering Research Support soCiety (SERSC)
Mobile agents' paradigm wide acceptance depends on the ability to overcome its major problem of security. In this paper, the major concern is the subtle issue of malicious hosts. Many solutions had been proposed in this area, varying from tamper free hardware to pure software protocols. Nevertheless, neither of them is fully satisfactory. This paper investigates the problem on a novel basis. It develops an incremental protocol giving the opportunity to appreciate the benefit of making weaker hints available (.i.e., information about the lower suspicious character of malicious hosts).