X-CID: Exhaustive Ensemble Technique for Cyber Invasion Detection in Drone Transportation Network
Simeon Okechukwu Ajakwe, Kazeem Lawrence Olabisi, Dong-Seong Kim · 2024
Drones, with their mobility and low deployment costs, play a vital role in time-sensitive logistics. However, ensuring a secure network is crucial to protect them from cyber threats such as exploitation and interception. This study presents X-CID, an AI-driven approach to intrusion detection that leverages optimization techniques. Using the random forest algorithm as its baseline classifier, the model employs randomized search cross-validation for hyperparameter tuning and Pearson correlation for dimensionality reduction, enabling faster detection of cyberattacks. Trained and validated on three cybersecurity datasets, X-CID was compared to four traditional machine learning models. It achieved an impressive average accuracy of 99.28%, a latency of 74 seconds, and a low false alarm rate, demonstrating its efficiency and reliability.