An aerial Radio Environment Map (REM) characterizes the spatial distribution of Received Signal Strength (RSS) across a geographic space of interest, which is crucial for optimizing wireless communication in the air. Aerial REM construction can rely on Unmanned Aerial Vehicles (UAVs) to autonomously select interesting positions for sampling RSS data, enhancing the quality of construction. However, due to the lack of prior information about the environment, it is challenging for UAVs to determine suitable sampling positions online. Additionally, achieving efficient exploration of the target area through collaboration among multiple UAVs is difficult. To address this issue, this paper proposes a multi-stage approach to data sampling and path planning with multiple UAVs. Specifically, the UAVs’ data sampling task over the target area is divided into multiple stages. By selecting an appropriate stage position, we use the RSS values at that position to determine whether additional data need to be sampled in a specific local area. At each stage, the area is divided into Voronoi diagrams based on the current position of each UAV, assigning each UAV its own region to explore. In our sampling strategy, the probability distribution for sampling is obtained by estimating the RSS and uncertainty of unsampled positions and then taking the weighted sum of these two values. To obtain the shortest flight path for selected sampling positions, we employ a network structure based on self-attention as the policy network, which is trained through the actor–critic framework to obtain an improvement heuristic strategy, replacing traditional manually designed strategies. Experimental results across three different scenarios indicate that the approach improves the quality of aerial REM construction while efficiently planning the shortest paths for UAVs between sampling positions.

Radio environment maps (REMs) integrate multidomain environmental data from real-world wireless scenarios, of which the high-quality construction is crucial for optimizing wireless communication and ensuring efficient spectrum utilization. Owing to the high mobility and flexibility, UAV-enabled radio sampling is highly effective for REM construction by informative path planning in information-rich positions. However, existing UAV sampling strategies rely solely on Gaussian process (GP) uncertainty reduction, which induces structural inaccuracies in REMs, particularly in environments with unknown obstacles. To address this issue, we first introduce a dual-information metric that combines uncertainty reduction with signal variation magnitude. Then we formulate an adaptive path planning problem for radio sampling under the energy constraint. Considering the NP-hardness of this problem, we propose a deep reinforcement learning-based adaptive sampling for UAV, termed DRL-ASU. To be specific, we propose an obstacle-aware probabilistic roadmap (PRM) generation mechanism that transforms the initial problem into a sequential graph-based decision problem. On this basis, we design an attention-based policy model for efficient resolution, optimized via multi-environment proximal policy optimization (PPO) training, which enables robust cross-scenario adaptability in diverse radio environments. Simulation results demonstrate that DRL-ASU achieves a 5.39% reduction in root mean square error (RMSE) and a 10.14% improvement in structural similarity index measure (SSIM) compared to single-metric baselines. Furthermore, we conduct real-world experiments to demonstrate the generalizability of our proposed DRL-ASU scheme.