Adaptive Machine Learning in Roadside Units for Blackhole-Resistant Secure Routing in VANETs: RSML

Abstract

Vehicle traffic on roads is the major problem in the whole world. Vehicles are moving on the roads but are not aware of the vehicle’s density in a particular area. The role of attacker presence in the network makes it more complicated to recognize the traffic-free route. In a Vehicular Ad hoc Network (VANET), all the vehicles are transferring traffic status information to other vehicles to ignore traffic congestion. The route formation process in vehicular communication is critical, as a malicious node may divert the path or disrupt the network for personal advantage during this phase. Consequently, it is imperative to choose a suitable routing method that guarantees secure communication. This research proposed an RSML approach for securing the traffic information dropping from the blackhole attack in VANET. This proposed approach facilitates the quickest communication route but requires enhancement through a security mechanism capable of detecting and preventing the VANET from blackhole attacks. A vehicle acquires the communication channel from the roadside unit (RSU) when attempting to communicate with another vehicle. Machine learning (ML) is an algorithmic framework designed to analyse extensive traffic information to discern patterns and facilitate decision-making or forecasting. The RSML method uses machine learning to effectively detect and mitigate black hole threats in vehicle ad hoc networks. The machine learning concept is incorporated with the AODV routing protocol to ascertain dependable paths and provide secure communication. This RSU functions as a centralized controller system, which offers real-time roadside information, manages any congestion, and gathers data from every vehicle within its range for security purposes. The RSML approach shows better data receiving and throughput as compared to the previous T-AODV and ANN-AODV approaches.