Adaptive IDS for Cooperative Intelligent Transportation Systems Using Deep Belief Networks

The adoption of cooperative intelligent transportation systems (cITSs) improves road safety and traffic efficiency. Vehicles connected to cITS form vehicular ad hoc networks (VANET) exchange messages. Like other systems, cITSs are targeted by attackers intent on compromising disrupting system integrity availability. They can repeatedly spoof false information causing bottlenecks, jams even accidents. existing security infrastructure assumes that the network topology and/or attack behavior is static. However, inherently dynamic in nature. Moreover, may have ability resources change their continuously. Assuming a static IDS model for VANETs not suitable lead low detection accuracy high alarms. Therefore, this paper proposes an adaptive solution based deep learning contextual references cope with nature topologies increasingly common behaviors. In study, belief (DBN) modeling was used train model. Binary cross entropy as loss function measure prediction error. Two activation functions were used, Relu Softmax, input–output mapping. hidden layers, while Sigmoid last layer map real vector output between 0 1. adaptation mechanism incorporated into using moving average monitors predicted values within time window. way, readjust classification thresholds on-the-fly appropriate. proposed evaluated Next Generation Simulation (NGSIM) dataset, which commonly such related works. result improved accuracy, demonstrating study effective.