CoBA: Integrated Deep Learning Model for Reliable Low-Altitude UAV Classification in mmWave Radio Networks

TL;DR

CoBA addresses the challenge of reliably classifying low-altitude UAVs in dense mmWave networks by fusing spatial and temporal radio features through a CNN–BiLSTM–Attention architecture. The approach is validated on real 5G mmWave data collected at TalTech, outperforming traditional ML and fingerprinting benchmarks with end-to-end accuracy approaching or exceeding 99%. Key findings show that topology-related features, notably PCI and SSB Idx, drive most of the discriminative power, and that CoBA maintains strong performance even with a drastically reduced feature set. The work demonstrates the potential for real-time, edge-deployable airspace monitoring in regulated UAV operations, and points to future multiclass and deployment-focused extensions.

Abstract

Uncrewed Aerial Vehicles (UAVs) are increasingly used in civilian and industrial applications, making secure low-altitude operations crucial. In dense mmWave environments, accurately classifying low-altitude UAVs as either inside authorized or restricted airspaces remains challenging, requiring models that handle complex propagation and signal variability. This paper proposes a deep learning model, referred to as CoBA, which stands for integrated Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (BiLSTM), and Attention which leverages Fifth Generation (5G) millimeter-wave (mmWave) radio measurements to classify UAV operations in authorized and restricted airspaces at low altitude. The proposed CoBA model integrates convolutional, bidirectional recurrent, and attention layers to capture both spatial and temporal patterns in UAV radio measurements. To validate the model, a dedicated dataset is collected using the 5G mmWave network at TalTech, with controlled low altitude UAV flights in authorized and restricted scenarios. The model is evaluated against conventional ML models and a fingerprinting-based benchmark. Experimental results show that CoBA achieves superior accuracy, significantly outperforming all baseline models and demonstrating its potential for reliable and regulated UAV airspace monitoring.

Figures (5)

• Figure 1: System Architectur
• Figure 2: 5G Advanced mmWave setup at TalTech featuring RUs, BBU, Ericsson 5G SA core, UAV, and radio scanner
• Figure 3: Performance of the CoBA with individual features.
• Figure 4: Performace of the CoBA using PCI and SSB Idx
• Figure 5: Algorithmic Complexity of the Models