Traffic-Aware Optimal Spectrum Sensing Policy in Wireless-Powered Cognitive Radio Networks

Abstract

We consider an orthogonal frequency division multiple access (OFDMA)-based primary user (PU) network model which offers different spectral access and energy harvesting opportunities in wireless-powered cognitive radio networks (CRNs). In this setting, we propose a traffic-aware optimal spectrum sensing policy for opportunistic spectrum access and energy harvesting under the energy causality and PU collision constraints in the wireless-powered CRNs. PU channels carry different traffic patterns and show distinct idle/busy frequencies, and hence the spectral access and energy harvesting opportunities are application specific. Secondary user (SU) obtains traffic pattern information through observation of PU channels and classifies the idle/busy period statistics for each channel using the variational inference approach. Based on the statistics, we adopt a stochastic model for evaluating SU capacity by which the energy detection threshold for spectrum sensing can be found with higher sensing accuracy. Towards this, we formulate the Markov decision process (MDP) model obtained by quantizing the amount of SU battery. The effectiveness of the proposed stochastic model is confirmed through comparison with the optimal one obtained from an exhaustive method.