Search for a command to run...
Unmanned Aerial Vehicles (UAVs) are increasingly integrated into critical applications ranging from logistics and agriculture to defence and security operations, surveillance and emergency response. At the core of these systems lies the communication link between the UAV and its ground control station (GCS), which serves as the backbone for command, control and data exchange. However, communications links remain highly vulnerable to cyber-threats, including eavesdropping, signal falsification, radio frequency interference (RFI) and hijacking. These risks highlight the urgent need for secure communication protocols and effective defence mechanisms capable of protecting data confidentiality, integrity, availability and authentication. This study performs a comprehensive survey of secure UAV-GCS communication protocols and artificial intelligence (AI)-driven intrusion detection techniques. Initially, we review widely used communication protocols, examining their security features, vulnerabilities and existing countermeasures. Accordingly, a taxonomy of UAV-GCS security threats is proposed, structured around confidentiality, integrity, availability and authentication and map these threats to relevant attacks and defences. In parallel, our study examines state-of-the-art intrusion detection systems for UAVs, while particular emphasis is placed on emerging methods such as deep learning, federated learning, tiny machine learning and explainable AI, which hold promise for lightweight and real-time threat detection. The survey concludes by identifying open challenges, including resource constraints, lack of standardised secure protocols, scarcity of UAV-specific datasets and the evolving sophistication of attackers. Finally, we outline research directions for next-generation UAV architectures that integrate secure communication protocols with AI-based anomaly detection to achieve resilient and intelligent drone ecosystems.