I. Introduction
Mobile devices such as unmanned aerial vehicles (UAVs) or ground robots can optimize their location in order to maximize the signal strength to a basestation or user equipment on the ground. Various placement optimizations algorithms for signal strength have been proposed to solve for optimal placement of such UAVs or mobile robots. We can roughly divide the said placement optimization algorithms into model-free and model-based solutions. In model-free solutions the UAV does not aim to predict the signal strength across the entire optimization space but rather finds a path that maximizes the expected increase in signal strength [1], [2]. Model-based solutions rely on being able to predict the signal strength across the optimization space and use that to find the optimal path to maximize the signal strength [3], [4]. At the heart of model-based solutions is some type of spatial signal strength prediction algorithm, which relies on some prior information, such as transmitter location, previously collected signal strength measurements, estimated parameters of channel attenuation and shadowing. The use of spatial signal strength prediction algorithms extends beyond placement of mobile radio devices. They are widely used to enable dynamic spectrum access in cognitive radio networks, improve cellular coverage, and facilitate power control.