Joint Spectrum Sensing and Transmission Strategy for Energy-Efficient Cognitive Radio Networks

In this paper, we consider a cognitive radio network in which a secondary user (SU) has opportunities to access one of multiple channels licensed by several primary users (PUs). Before transmission, SU should sequentially decide which channel to explore and in what sequence, as well as whether to access or to sleep. Our goal is to design a joint spectrum sensing and transmission strategy for SU to maximize energy efficiency, which is defined as the number of bits transmitted per unit energy consumption (including both transmission and operating energy). By modeling PUs' occupancy activities as a Markov process, we formulate the problem of designing optimal sensing and transmission strategy as a Partially Observable Markov Decision Process (POMDP). To solve it in a practical way, we establish the optimal threshold structure of myopic strategy and propose the algorithm to determine the optimal sensing order, which takes low computational complexity. Simulation results show that the energy efficiency loss of proposed approach is less than 8% compared with optimal solution. In addition, our approach can improve energy efficiency by roughly 15% to 35% compared to the traditional approaches, with less than 10% loss of capacity.