Jellyfish Search Chimp Optimization Enabled Routing and Attack Detection in SDN based VANETs

Abstract

In recent times, Vehicular Ad hoc Network (VANET) has been the focal point of the research community to devise efficient smart transportation systems. VANET provides the key advantage of providing cautionary measures and safety to passengers and drivers. With the evolution of fifth-generation (5G) network technology and rapid growth in vehicles, it becomes challenging for conventional VANET to manage large-scale dynamic heterogeneous networks due to their limited flexibility and scalability features. Moreover, the dynamic nature of VANET makes it vulnerable to malicious attacks. Software Defined Networking (SDN) is a technology that provides an integrated improvement over the conventional VANETs. SDN architecture is flexible, programmable, scalable, and provides globally the knowledge of the network. However, its centralized nature makes SDN based VANETs a prime target of attackers, which may adversely impact the VANETs causing life-threatening consequences. To address these issues, this paper presents two novel schemes. Firstly, this paper presents a trusted routing scheme named Jellyfish Chimp Optimization Algorithm (JChOA) for SDN based VANETs. JChOA is designed by amalgamation of the Jellyfish Search Optimization algorithm (JS) and Chimp Optimization algorithm (ChOA). Secondly, this paper presents an attack detection and mitigation scheme named JChOA_RideNN for SDN based VANETs. This attack detection scheme utilizes the Rider Optimization Algorithm based neural network (RideNN) architecture at the SDN controller, where the weighting parameters of RideNN tunned through the use of JChOA. The effectiveness of JChOA routing is evaluated based on the metrics energy and trust value where the performance of JChOA_RideNN is assessed using precision and recall. Moreover, the JChOA routing algorithm attained greater performance with a maximum of 0.947 J energy and 0.462 trust value and JChOA_RideNN attained with a maximum of 93.9% precision, and 93.1% recall than other traditional approaches. The results of the experiments clearly show the effectiveness of the proposed defensive schemes for SDN based VANETs.