Power Allocation for Capacity Maximization in Sensing-Based Cognitive DF Relay Networks With Energy Harvesting

Green communications have been widely studied in the researches of cognitive radio networks (CRNs), which involve low power consumption, new and renewable energy, and some energy-saving technologies. In addition, the spectrum sensing uncertainties are inevitable errors from realistic factors, such as wireless channel fading, channel estimation, and signal measurement. In this paper, to maximize total capacity of secondary user (SU), we propose a power allocation (PA) strategy in a cognitive decode-and-forward (DF) relay network with the spectrum sensing uncertainties, in which the relay is powered by an energy harvesting (EH) device with a capacity-limited battery. While formulating the optimization problem, we consider the total capacity expressions of SU and the interference models in both the perfect and the imperfect sensing cases which affect actual PA of SU and the relay. Then, we transform this traditional multi-variable optimization with the imperfect spectrum sensing into single variable optimization according to the capacity maximization criteria under the DF protocol. Thereafter, we solve the optimization problem by the Lagrange dual decomposition method. The simulations in both single time slot and multiple time slots are given to verify that our proposed algorithm can efficiently improve the capacity performance of SU while protecting the communications of the primary user (PU).