3-D Radio Map Estimation Based on Active Measurement Trajectory Selection

Three-dimensional (3D) radio maps characterize the spatial distribution of received signal strength (RSS) in 3D space, providing a pivotal instrument for managing spectrum resources and coordinating interference. The existing 3D radio map estimation algorithms either employ randomly distributed sensors or utilize unmanned aerial vehicles (UAVs) to conduct RSS measurements along predefined flight paths, without considering the optimization of measurement locations. In contrast, this letter proposes a novel 3D radio map estimation framework centered on active measurement trajectory selection. The UAVs can actively choose measurement positions to efficiently obtain high-quality map estimates. Specifically, a cutting-edge 3D convolutional autoencoder is employed, which possesses the capability to estimate radio maps and quantify the uncertainty metric at each location. Then, an uncertainty metric-based measurement trajectory planning algorithm is developed, guiding the UAVs to conduct measurements at locations rich in information. The experimental results demonstrate the efficacy of our proposed framework in enhancing 3D radio map reconstruction accuracy. Under a sampling ratio of 5%, an estimation error of 2.61 dB is achieved.