A Hybrid Deep Learning Model for Intrusion Detection in Aerospace Vehicles

In today's linked world, aircraft vehicles need advanced communication technologies to operate. However, this dependency makes them susceptible to cyber dangers such intrusions into communication networks. In this research, we develop a hybrid deep learning model that enhances aerospace vehicle Intrusion Detection Systems (IDS). Our cascading LSTM and GRU network model handles time-series data well, solving MIL-STD-1553 communication traffic issues. Quantitative analyses surpass machine learning in detection metrics. The model can correctly detect complex infiltration attempts with few false negatives, with accuracy and recall of 99.33% and 99.17%, respectively.