Robust SDN-assisted vehicle network intrusion detection model with stealthy beacon identification based on blockchain-enabled DFTLaLURDA2RN

In order to ensure traffic efficiency and network security, an accurate intrusion detection system (IDS) is pivotal in the software-defined networking (SDN)-enabled vehicular ad-hoc network (VANET) environment. However, the traditional methodologies failed to identify the presence of stealthy beacons (SBs) in the VANET environment, thus degrading the system’s reliability. Hence, an enhanced SDN-powered vehicle IDS with SB identification based on blockchain (BC)-enabled deep federated transfer collapsing linear unit R-drop alpha regularized-recurrent neural network is proposed in this article. Initially, the vehicles register in the data plane, followed by application programming interface (API) key generation and digital signature creation (DSC). Afterward, the sensed data are transferred while mitigating the collision. The digital signature verification (DSV) is done in the south-bound APIs. Once the verification is successful, the data are fed into the control plane via the switch. The control messages are protected in the control plane. In addition, the DSV is done in the north-bound APIs. By utilizing the fuzzy bimodal triangular membership inference system (FBTMIS), the presence of an SB is identified in the application plane. If the SB is present, then the transmitted data are terminated; otherwise, the trained local IDS model is utilized for categorizing the data into benign or malignant. Furthermore, to track the attacker’s behavior, the trapdoors are utilized. Thereafter, the detected events are stored in the BC and further broadcast to the vehicles. Similarly, through the router and firewall, the outcomes are updated in the global model (GM) on the server. Therefore, the proposed approach had high reliability with an accuracy of 99.9742%.