Protecting mobile agents against malicious hosts using trusted computing

Abstract

The concept of mobile agents was invented about a decade ago. Since then, plenty of innovative applications of mobile agents have been proposed. But it remains to be the fact that these applications have not been deployed to the desired extent. Security concerns associated with mobile agent systems are primary obstacles to their widespread adoption. Protecting mobile agent against malicious host is the unsolved aspect of these security concerns. Existing solutions to this problem are either not practical or not robust. We have provided a practical and robust solution to this problem using the concepts of Trusted Computing. We have designed security architecture for mobile agent systems. Our architecture uses the capabilities of the TCG (Trusted Computing Group) TPM (Trusted Platform Module) and TCG based integrity measurement architecture to provide the solution to the above mentioned problem. Since most of the mobile agent systems are Java based, we have provided an implementation of TCG aware Java Virtual Machine, which is one component of the overall architecture.