Probabilistic Reasoning and Risk-Constrained Dynamic Spectrum Access

Abstract

Uncertainties regarding wireless propagation environments pose challenges for spectrum management in general and specifically hinder the implementation of dynamic spectrum sharing systems. Without the ability to reliably evaluate interference risks, spectrum sharing policies specify spectrum access behaviors such as exclusion zones and maximum transmit powers based on risk thresholds applied to statistical results from propagation models and measurements. Because the models can contain significant levels of uncertainty, establishing behavior limits for low interference risk necessarily results in significant spectrum access inefficiencies. It is only by reducing the degree of uncertainty that risk thresholds can be maintained while increasing spectrum access efficiency. Probabilistic reasoning applied to dynamic spectrum sharing systems provides potential to increase spectrum sharing by reducing situational uncertainty. Further, probabilistic reasoning approaches enable risk-constrained spectrum access, a concept in which regulators and spectrum users establish spectrum access rules defining acceptable levels of interference and spectrum access risks. This paper develops the concepts and underlying theory of probabilistic reasoning and risk-constrained spectrum access for spectrum sharing. It further presents simulation results showing that situation-specific probabilistic reasoning combined with risk-constrained spectrum access potentially enables greater spectrum sharing. Specifically, probabilistic reasoning with real-time spectrum sensing is shown to greatly reduce situational uncertainty, which then results in better interference prevention and more effective spectrum sharing as measured by user capacity and overall network density.