Skip to Main Content
The mobile agent (MA) paradigm advocates the migration of agent code to achieve computational goals. MAs require an executable environment on hosts where mobile code can be executed. The execution of foreign code on a host raises security concerns for both the agent and the host. In Corradi et al.  it has been recognized that most of the approaches for providing security in MA suffer from a limitation of expressing complex security requirements. Thus, approaches have been proposed that introduce the use of a policy language for specifying security policies to control MA's access to host resources. With this paper, we outline a framework where security policies can be uniformly specified for protecting both MAs and host resources.