Statistical Performance Modeling of Solar and Wind-Powered UAV Communications

We develop novel statistical models of the harvested energy from renewable energy sources considering harvest-store-consume (HSC) architecture. We consider three renewable energy harvesting scenarios, i.e., (i) harvesting from the solar power, (ii) harvesting from the wind power, and (iii) hybrid solar and wind power. In this context, we first derive the closed-form expressions for the density functions and moments of the harvested power solar and wind power. Then, we calculate the probability of energy outage at UAVs and signal-to-noise ratio (SNR) outage at ground cellular users. The energy outage occurs when the UAV is unable to support the flight consumption and transmission consumption from its battery power and the harvested power. Due to the intricate distribution of the hybrid solar and wind power, we derive novel closed-form expressions for the moment generating function (MGF) of the harvested solar power and wind power. Then, we apply Gil-Pelaez inversion to evaluate the energy outage at the UAV and SNR outage at the ground users. In addition, we formulate the SNR outage minimization problem and obtain closed-form solutions for the transmit power and flight time of the UAV. Furthermore, we demonstrate the application of moments in computing novel metrics such as the probability of charging the UAV battery within the flight time, average UAV battery charging time, probability of energy outage at UAVs, and the probability of eventual energy outage (i.e., the probability of energy outage in a finite duration of time) at UAVs. Numerical results validate the analytical expressions and reveal interesting insights related to the optimal flight time and transmit power of the UAV as a function of the harvested energy.