There are some security threats when mobile agent codes are loaded or running on destination host platforms. Those threats are detecting, draining or altering the agent's intention. In order to protect the code from these attacks, we discuss some measures to assure the original code security with agent code obfuscation, encryption, self-defining classloaders and rebuilding the JVM system classloader in this paper.
展开▼
