|
Projects
The CoPS Lab has several on-going projects, briefly described here.
For more information on a project, follow the link to that project's
page.
Security for Mobile Agents
In this project, we investigate techniques for secure computation in
mobile code. Mobile agents are pieces of code that travel from host
to host, along with their state, to accomplish some task, such as
doing intelligent data searches, managing network configurations, or
acting as an intelligent bidding agent in e-commerce. In some
situations the agent originator and the remote execution hosts are
mutually distrusting, so it is vital to consider techniques for
protecting the host from a malicious agent, and for protecting the
agent from a malicious host. The former problem has been studied
extensively, and we focus on the less well-known problems of
protecting an agent from a malicious host. In particular, we are
designing protocols for computing remotely with private information
that is not revealed to the executing host, as well as protocols for
ensuring data integrity in the presence of malicious hosts.
Main investigators: Steve Tate, Ke Xu, Vandana Gunupudi
Tools for Managing SELinux
In this project, we investigate tools for managing SELinux (the Security Enhanced
Linux project from the National Security
Agency), and in particular we are designing tools for creating and
analyzing security policies for SELinux. SELinux is a very flexible
system, but with that flexibility comes the cost of very complex
security rules. For example, a recent release of SELinux from the NSA
had an example security policy which generated 37,217 basic rules for
the type-enforcement system. While they supply a policy compiler that
makes the task of generating such a policy easier, we are exploring
the possibility for more powerful tools, and tools which enable formal
analysis of security policies.
Main investigators: Steve Tate
|
