RadioMapMotion: A Dataset and Baseline for Proactive Spatio-Temporal Radio Environment Prediction

TL;DR

The paper tackles proactive spatio-temporal radio environment prediction by defining the problem of forecasting a sequence of future radio maps from historical observations. It introduces RadioMapMotion, the first large-scale dataset that captures temporally continuous RM evolution driven by vehicle trajectories, and RadioLSTM, a ConvLSTM-based UNet baseline for multi-step RM forecasting. Empirical results show that explicit temporal modeling improves accuracy and structural fidelity while delivering low inference latency, making it suitable for real-time network management. This work lays a foundation for proactive, environment-aware networking in 6G and against dynamics such as V2X mobility, with avenues for extending to more scenarios and longer horizons.

Abstract

Radio maps (RMs), which provide location-based pathloss estimations, are fundamental to enabling proactive, environment-aware communication in 6G networks. However, existing deep learning-based methods for RM construction often model dynamic environments as a series of independent static snapshots, thereby omitting the temporal continuity inherent in signal propagation changes caused by the motion of dynamic entities. To address this limitation, we propose the task of spatio-temporal RM prediction, which involves forecasting a sequence of future maps from historical observations. A key barrier to this predictive approach has been the lack of datasets capturing continuous environmental evolution. To fill this gap, we introduce RadioMapMotion, the first large-scale public dataset of continuous RM sequences generated from physically consistent vehicle trajectories. As a baseline for this task, we propose RadioLSTM, a UNet architecture based on Convolutional Long Short-Term Memory (ConvLSTM) and designed for multi-step sequence forecasting. Experimental evaluations show that RadioLSTM achieves higher prediction accuracy and structural fidelity compared to representative baseline methods. Furthermore, the model exhibits a low inference latency, indicating its potential suitability for real-time network operations. Our project will be publicly released at: https://github.com/UNIC-Lab/RadioMapMotion upon paper acceptance.

Figures (7)

• Figure 1: Conceptual illustration of the structure of the RadioMapMotion dataset. Static environmental layouts (Env), dynamic vehicle trajectories (Traj), and transmitter locations (Tx) are combined to generate continuous radio map sequences, where signal propagation evolves causally over time in response to vehicle motion.
• Figure 2: Illustration of the SRM versus the DRM. The yellow heatmap represents pathloss intensity. Blue elements are static buildings. The DRM in (b) additionally includes vehicles (represented as red elements).
• Figure 3: The overall architecture of RadioLSTM. The model adopts a UNet-like encoder-decoder structure, where the core blocks in both the encoder and decoder are replaced by ConvLSTM cells to capture spatio-temporal dynamics. The right panel provides a detailed view of the internal structure of a single ConvLSTM block.
• Figure 4: PACF analysis of the RadioMapMotion dataset, averaged over all sequences.
• Figure 5: Qualitative comparison for dynamic generalization (seen environment). Each row illustrates the output of a specific method for an unseen vehicle trajectory within a familiar environment.