Abstract
The wireless nature of the Vehicular Ad Hoc Network (VANET), a technology that offers facilities such as traffic management and safety services, makes it vulnerable to distributed denial-of-service (DDoS) attacks that exploit network communications and reduce network reliability and performance. This paper proposes a design of a secure VANET architecture using a Software-Defined Networking (SDN) controller and Neural Network Self-Organizing Maps (SOMs). In the proposed design, we adopt the SDN architecture by using its separation of the control plane from the data plane and adding intelligent capabilities to the VANET. To resolve the drawbacks of standard SOMs and to enhance the SOM’s efficiency, a Multilayer Distributed SOM (MSOM) model based on two levels of clustering and classification is used. Experimental results show that our solution can efficiently detect malicious traffic, prevent and mitigate DDoS attacks, and increase system security and recovery speed from the attacking traffic. Moreover, the proposed scheme achieves a high accuracy rate (99.67%). Simulation results demonstrate the effectiveness and efficiency of the MSOM regarding detection accuracy and other studied metrics.
Highlights
Vehicular Ad Hoc Network (VANET), an innovative network that uses different technologies for next-generation Intelligent Transportation Systems (ITS), offers an array of functional services to vehicles and roadside infrastructure, all of which result in better efficiency and safety of transport operations [1]
The results suggested that the detection rate with the Multilayer Distributed SOM (MSOM) model was higher than with both the Self-Organizing Maps (SOMs) and the distributed SOM (DSOM) models
This finding can be explained by the fact that the proposed model has more layers, which results in a stronger feature learning ability as compared to the other tested models; in addition, the proposed model can obtain an extensive feature presentation and higher detection ability
Summary
Vehicular Ad Hoc Network (VANET), an innovative network that uses different technologies for next-generation Intelligent Transportation Systems (ITS), offers an array of functional services to vehicles and roadside infrastructure, all of which result in better efficiency and safety of transport operations [1]. In this environment, vehicles communicate with each other through an ad hoc network—using the Vehicle-to-Vehicle (V2V) mode. The main idea of this type of attack is to turn the victim node inaccessible or to reduce the correspondence all the way by the network, thereby adversely affecting availability Attacks of this kind result in network overloading.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
