Protecting agents from malicious hosts using TPM.
Antonio Muñoz, Antonio Maña, Daniel Serrano · 2009
Software agents represent a promising computing paradigm. They are an elegant technology to solve problems that can not be easily solved in other way. The Scientific Community has proved that the use of the software agents approach simplifies the solution of difierent type of traditional computing problems. A proof of this is that several important applications exist based on this technology. It is clear that software agents have so many benefits, but unfortunately, a lack of the appropriate security mechanisms for systems based on them represents a barrier for the widespread use of this paradigm in the industry. The application of the current security mechanisms is not trivial for agent based system developers, since they usually are not security experts and consequently do not count on the appropriate expertise. This paper presents a new protection infrastructure solving the problem known as malicious host in mobile agent systems. This protection infrastructure implements a secure protocol to migrate agents from host to host relying in hardware elements, particularly it is based on the recent advances in Trusted Platforms Modules (TPM) computing. In order to provide an easy way of using the proposed infrastructure we provide it by means of an extension to the Java Agent Development framework (JADE). Finally, the migrating protocol presented has been validated using