Anomaly Detection based Security Solution in Internet of Drones

The Internet of Drones (IoD), which refers to the coordinated usage of drones across various industries, presents a significant cybersecurity threat as it is highly vulnerable to exploitation by hackers. With the increasing reliance on drones for critical applications such as surveillance, delivery, and disaster management, ensuring their security has become a top priority. However, existing anomaly detection systems for IoD suffer from several limitations. They rely heavily on supervised learning, which requires labeled datasets that are often unavailable or expensive to obtain. Additionally, current systems struggle to handle the dynamic and unpredictable nature of drone networks, leading to high false positive rates and delayed threat detection. Moreover, many existing models lack scalability, making it difficult to adapt to large-scale drone networks with hundreds or thousands of drones. To address these challenges, this paper introduces the application of advanced deep learning models, specifically an autoencoder, to detect and alert against potential cybersecurity threats during IoD operations. The proposed autoencoder model is highly scalable and can efficiently handle large volumes of drone data, making it suitable for real-world deployment in large-scale IoD systems. By training the autoencoder with traffic data generated by the IoD, it can effectively detect anomalous patterns indicative of potential attacks. These detected outliers are flagged as cybersecurity threats, enabling the implementation of an automated alert system that notifies the IoD security team and triggers appropriate countermeasures. Real-time traffic data of drones is scanned using tools such as Wireshark filters, which extract and analyze the live traffic data of IoD networks. This project holds significant potential for enhancing the security of drone swarms and ensuring the safe operation of IoD systems across industries.