A Distributed Trust Management Scheme in the Pervasive Computing Environment

Abstract

Pervasive computing allows a seamless interaction among users, devices, and applications, anytime and anywhere. Yet portable devices in pervasive computing are mainly powered by batteries and have limited computational and communication capability. Thus the open and dynamic environment in pervasive computing raises challenges in security and trust management. Without trust, pervasive devices cannot cooperate effectively, and the deployment of pervasive computing systems will be restricted to specific application scenarios. The traditional centralized security management schemes are not directly applicable in pervasive computing environments. Moreover, existing user authentication and access control schemes are inadequate to ensure security in pervasive computing. To overcome the limitation of centralized schemes, we need a distributed solution. In this paper, we propose a distributed trust management scheme to ensure security in pervasive computing environments. The main contributions of this paper are: (1) the employment of a simple, distributed trust computation and maintenance mechanism to reduce communication and computational overhead without compromising security; (2) the building of an aggregate trust metric that is based on direct observation and indirect observations obtained from neighbors' recommendations. The scheme gives more weight to direct observations and less weight to indirect observations. Every device computes and updates the trust value periodically in a distributed fashion. However, the exchange of trust information is carried out on demand to reduce communication overhead. The operation of the proposed scheme with varying parameter settings is illustrated, using an analytical approach.