Efficient Malicious UAV Detection Using Autoencoder-TSMamba Integration

2025-05-14

AI Summary

Malicious drones pose a severe security threat to next-generation networks. To address this, we propose a lightweight and efficient detection framework tailored for edge deployment. Our method introduces the novel four-layer, three-directional spatial Mamba (TSMamba) autoencoder—designed to compactly model spatiotemporal features of drone signals—and an AE-ResNet residual-driven cascaded architecture, where residual features extracted by the autoencoder explicitly guide ResNet-based classification, balancing representational capacity and computational efficiency. Evaluated on both binary and multi-class drone detection tasks, our framework achieves a 99.8% recall rate—3.1 percentage points higher than state-of-the-art baselines—while significantly reducing model complexity (e.g., parameter count and FLOPs). This enables scalable, real-time inference on resource-constrained edge devices.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Malicious Unmanned Aerial Vehicles (UAVs) present a significant threat to next-generation networks (NGNs), posing risks such as unauthorized surveillance, data theft, and the delivery of hazardous materials. This paper proposes an integrated (AE)-classifier system to detect malicious UAVs. The proposed AE, based on a 4-layer Tri-orientated Spatial Mamba (TSMamba) architecture, effectively captures complex spatial relationships crucial for identifying malicious UAV activities. The first phase involves generating residual values through the AE, which are subsequently processed by a ResNet-based classifier. This classifier leverages the residual values to achieve lower complexity and higher accuracy. Our experiments demonstrate significant improvements in both binary and multi-class classification scenarios, achieving up to 99.8 % recall compared to 96.7 % in the benchmark. Additionally, our method reduces computational complexity, making it more suitable for large-scale deployment. These results highlight the robustness and scalability of our approach, offering an effective solution for malicious UAV detection in NGN environments.

Problem

• Detect malicious UAVs threatening next-generation networks.
• Improve detection accuracy and reduce computational complexity.
• Enable large-scale deployment for robust UAV surveillance.

Innovation

• Autoencoder-TSMamba integration for UAV detection
• ResNet-based classifier with residual values
• Low complexity and high accuracy solution