Distributed Joint Resource Allocation in Primary and Cognitive Wireless Networks

Abstract

We develop a game theoretic framework for distributed resource allocation in the uplink of a cognitive radio network where secondary users (SUs) share the bandwidth with primary users (PUs). Each PU has a fixed data rate and applies power control to achieve its target SINR, while each SU jointly adjusts its data rate and transmit power by maximizing its utility. Since the SUs' total interference on the primary network is kept below a given threshold, there exist couplings between SUs' transmit power levels. Hence, the SUs' game belongs to the generalized Nash equilibrium (GNE) problems. We separately analyze the proposed resource allocation algorithm for SUs and the one utilized by PUs, and derive the condition under which the PUs' power control algorithm converges for SUs' fixed transmit power levels. We also derive the sufficient condition under which the SUs' joint data rate and power control algorithm converges for fixed transmit power levels of PUs. We then derive the sufficient condition for convergence of the algorithms when PUs and SUs simultaneously apply their resource allocation schemes. Simulations confirm our analysis and demonstrate that the proposed framework is energy efficient and provides PUs and SUs with their quality of service requirements.