Malware Attack Detection in Vehicle Cyber Physical System for Planning and Control Using Deep Learning

Cyber-Physical Systems (CPS), which comprise smart health, smart transportation, smart grids, etc., are designed to turn traditionally separated automated critical infrastructure into modernized linked intelligent systems by interconnecting human, system, and physical resources. CPS is also expected to have a significant positive impact on the economy and society. Complexity, dynamic variability, and heterogeneity are the features of CPS, which are produced as an outcome of relationships between cyber and physical subsystems. In addition to the established and crucial safety and reliability criteria for conventional critical systems, these features create major obstacles. Within these cyber-physical systems and crucial infrastructures, for instance, connected autonomous vehicles (CAVs) may be considered. By 2025, it is anticipated that 95 per cent of new vehicles will be equipped with vehicle to vehicle (V2V), vehicle to infrastructure (V2I), and other telecommunications capabilities. To prevent CAVs on the road against unintended or harmful intrusion, innovative and automated procedures are required to ensure public safety. In addition, large-scale and complicated CPSs make it difficult to monitor and identify cyber physical threats. Solutions for CPS have included the use of Artificial Intelligence (AI) and Machine Learning (ML) techniques, which have proven successful in a wide range of other domains like automation, robotics, prediction, etc. This research suggests a Deep Learning (DL) -based Convolutional Neural Network (CNN) model for attack detection and evaluates it using the most recent V2X dataset, According to the simulation results, in this research CNN exhibits superior performance compared to the most advanced ML approaches such as Random Forest (RF), Adaptive Boosting (AdaBoost), Gradient Boosting (GBoost), Bagging and Extreme Gradient Boosting (XGBoost) and achieves an outstanding level of accuracy in the application of anomaly detection.