Abstract
In this paper, we consider a cognitive unmanned aerial vehicle (UAV) to disseminate data to a group of Internet of Things (IoT) devices. The cognitive UAV shares the wireless spectrums of primary users (PUs) where it accesses the spectrums opportunistically when the corresponding channels are available. The UAV predicts the OFF periods of the PU, then it decides the number of consecutive transmission slots to be used for data transmission. To protect the PUs from harmful interference, the number of slots used for transmission should not exceed a certain threshold, i.e., maximum allowable interfering slots. We formulate our problem mathematically, where the objective is to maximize the minimum bits received by the IoT devices. The formulated problem is in a form of mixed integer non-linear program (MINLP). Therefore, we propose a successive convex approximation-based algorithm to solve this problem after approximating it to a convex problem. We show that the algorithm converges to a Karush-Kuhn-Tucker (KKT) point. In the simulation, we examine how the OFF period model of the PUs, maximum number of transmission slots, maximum allowable interference threshold affect the maximum minimum bits received by the IoT devices.