In this paper, a centralized Power Control (PC)scheme and an interference channel learning method are jointlytackled to allow a Cognitive Radio Network (CRN) access tothe frequency band of a Primary User (PU) operating basedon an Adaptive Coding and Modulation (ACM) protocol. Thelearning process enabler is a cooperative Modulation and CodingClassification (MCC) technique which estimates the Modulationand Coding scheme (MCS) of the PU. Due to the lack ofcooperation between the PU and the CRN, the CRN exploitsthis multilevel MCC sensing feedback as implicit channel stateinformation (CSI) of the PU link in order to constantly monitorthe impact of the aggregated interference it causes. In this paper,an algorithm is developed for maximizing the CRN throughput(the PC optimization objective) and simultaneously learning howto mitigate PU interference (the optimization problem constraint)by using only the MCC information. Ideal approaches for thisproblem setting with high convergence rate are the cuttingplane methods (CPM). Here, we focus on the analytic centercutting plane method (ACCPM) and the center of gravity cuttingplane method (CGCPM) whose effectiveness in the proposedsimultaneous PC and interference channel learning algorithm isdemonstrated through numerical simulations.
展开▼
