Situational awareness uncertainty impacts on Dynamic Spectrum Access performance

This paper explores the value of situational awareness information to Dynamic Spectrum Access (DSA) systems, which access wireless spectrum in an ad hoc manner to meet user needs while avoiding harmful interference to other spectrum users. In general, DSA systems must make adaptation decisions with imperfect information in factors such as local propagation characteristics as well as the presence, location, and spectrum usage behaviors of other spectrum users. DSA methods for developing situational awareness (SA) include spectrum sensing, information sharing among networked users, and access to various databases. This paper derives the relationship between SA uncertainty and DSA system performance. The SA and decision process uses a probabilistic reasoning and decision-making model for DSA systems. The method uses a Bayesian Networking model with causality assumptions along with multiattribute utility theory. The BN is built as a functional causal model from common functional relationships and algorithms from communications theory. The associated analysis provides insight into the decision-making trades of a DSA system when acting on imperfect situational awareness and highlights some general findings and principles. The results demonstrate a direct impact on interference and capacity potential resulting from uncertainty. Uncertainty is quantified in terms of incorrect assessment of the mean and variance of the path loss probability distribution.