UMBC CSEE Computer Science & Electrical Engineering
University of Maryland Baltimore County
Baltimore Maryland 21250 USA
voice: 410-455-3500 fax: -3969

cs.umbc.edu web
Secure Agent COmmunication Languages

Secure Agent Communication Languages

INVESTIGATORS

Dr. Tim Finin (CO-PI)
Dr. Alan Sherman (CO-PI) Muhammed Rabi

SPONSORSHIP

DoD, 1997-98

ABSTRACT

Autonomous, asynchronous software agents are emerging as a powerful programming paradigm, especially well suited for applications in electronic commerce. Unfortunately, existing Agent Communication Languages (ACLs) lack adequate security mechanisms for such applications. We will design and develop a security architecture for defining software agents that can apply fundamental security concepts and reason about them. The security features will be an integral part of the ACL model; they will not simply be relegated to lower levels of the communications protocols. Specially, we will develop this security architecture by extending the Knowledge Query and Manipulation Language (KQML).

KQML is a communication language and protocol that enables autonomous and asynchronous software agents to share their knowledge and work towards cooperative problem solving. Developed as a part of the DARPA {Knowledge Sharing Effort}, KQML can be thought of as consisting of three layers: the content layer, the message layer, and the communication layer. The content layer bears the actual content of the message, in the program's own representation language. The communication level encodes a set of message features which describe the lower level communication parameters, such as the identity of the sender and recipient, and a unique identifier associated with the communication. The message layer forms the core of the KQML language, and determines the kinds of interactions one can have with a KQML--speaking agent. A primary function of the message layer is to identify the protocol to be used to deliver the message and to supply a speech act or performative which the sender attaches to the content (such as that it is an assertion, a query, a command, or any of a set of known performatives).

Preliminary steps toward adding security features to KQML have been carried out by Finin, Mayfield, and Thirunavukkarasu. We will begin by critically evaluating and expanding these preliminary designs. Our eventual goal is to implement these designs and to demonstrate their effectiveness in agent-based applications in electronic commerce. The project will benefit from the collaboration of Finin (an expert in intelligent agents) and Sherman (an expert in cryptography).

Our approach integrates smoothly with promising contemporary approaches to models of distributed trust, such as those proposed by Rivest and Lampson, and Blaze, Feigenbaum, and Lacy.

RELEVANT PAPERS AND LINKS

FOR MORE INFORMATION

Contact Tim Finin, finin@umbc.edu .

UMBC   |   CSEE   |   User pages   |   Search CSEE: