CAREER: Key Management for Secure Dynamic Group Communications

University Of Washington, Seattle WA

Investigators

Abstract

Many emerging Internet and broadband wireless network services such as real-time stock quotes and information services, pay per view, and dynamic coalition for collaborations are based on group communications. Innovative new networking specifications such as Bluetooth (1999), WAP (1998), and SWAP (1998) allow multiple devices with different capabilities to form groups and take advantage of each other's capabilities. Emerging from these innovations is the potential for developing group communication at the level of embedded systems, which has enormous ramifications both academically and commercially. The most relevant enabling network technology for group communications is the multicast which supports the requirements of the service providers as well as the end users. However, research into secure multicast communications has been emerging only recently. The extensive efforst addressing reliable data delivery, guaranteed quality of service provision, congestion control and resource allocation for conventional and multi-rate multicast indicate the importance of multicast for Internet applications. However, the adaptation of multicast communications into commercial and selective applications depends on the ability to secure the multicast communications so that only the intended receivers have access to the data. Due to the untrusted nature of the network medium used for multicast communications, securing multicast communications requires cryptographic methods. In particular, generating, storing, distributing, deleting and updating of the cryptographic keys used for secure group communications from the core of the key management problem. This career proposal contains the following three components: Key Distribution: Research into new key distribution schemes that will be significant improvements over the currently available methods and will be efficient with respect to user and/or sender key storage, and rekeying messages whil preventing user collusion with the focus of this effort. PI previously answered a conjecture related to the optimality of a key distribution scheme and demonstrated security flaw of a "secure" multicast key distribution scheme. Group Key Generation: Group key generation schemes allow a set of members to jointly generate cryptographic keys. Permitting group membership changes, preventing illegal collaboration among members trying to steal secrets or bias the final joint key are some of the features that need to be considered in designing new group key generation schemes. The proposed work will be based on the current work of PI in the area of distributed key generation. The past work has resulted in a pending national and international patent and is currently being commercialized by a collaboration with Lockheed Martin Global Telecommunications under Maryland Industrial partnership program. (Dr. Roger Mancuso, CTO, LMCO) Education: PI will develop and introduce two new undergraduate courses at University of Washington Seattle and an advanced network security laboratory that will house the course related projects and basic research. Since there is a strong presence of Information Technology companies in Seattle, the university-industry collaboration will help to provide strong research and educational background to UW students and lead to higher return on investment to the academia, industry and the NSF. PI has been in contact with the research and development groups of security at Microsoft to establish close ties. (Dr. Sekar Chandrasekaran, Lead NT security manager, MSFT)