Hybrid optimization enabled trust-based secure routing with deep learning-based attack detection in VANET

Abstract

Vehicular ad-hoc network (VANET) model is easy to arrange and inexpensive answer for clever traffic control and traffic failure protective measures. Generally, the nodes present in VANET utilize broadcast protocols for effectively transmitting safety information, although nodes do not perform based on routing protocols. The misbehavior can produce through targeted attack in which aggressive vehicle can purposely send malevolent packets to damage. Moreover, due to the reason of dynamic behavior of nodes in VANET plus routing difficulties, unexpected misconduct also occurs owing to the software or hardware failures in vehicle. The challenges faced by existing techniques are provided here. VANET becomes a most significant concern to enhance the expediency, mobility, and security application in ITS. Precautions are further tough owing to its rapid mobility, wireless communication and large coverage of nodes. Also, the computational time required for executing independent operations are conspicuously maximum when, the number of vehicles is maximized. In order to overcome these challenges, in this paper, hybrid optimization-based Deep Maxout Network (DMN) is developed for attack classification in VANET. The Cluster Head (CH) selection and routing process is performed using designed hybrid optimization algorithm. The feature selection process is more significant in order to perform effective classification process. In addition, the attack classification is carried out using DMN and it is taught by introduced optimization algorithm. The developed optimization-based DMN model obtained better classification performance with precision and recall of 0.9395, and 0.9462 as well as routing performance with energy and trust of 0.2454J, and 0.4402.