A Hybrid Price Auction-Based Secure Routing Protocol Using Advanced Speed and Cosine Similarity-Based Clustering against Sinkhole Attack in VANETs

Abstract

In ad-hoc vehicle networks (VANETs), the random mobility causes the rapid network topology change, which leads to the challenge of the reliable data transmission. In this paper, we propose a hybrid-price auction-based secure routing (HPA-SR) protocol using advanced speed and cosine similarity-based (ASCS) clustering to establish a secure route to avoid sinkhole attacks and improve connectivity between nodes. The main features and contributions of the proposed HPA-SR protocol are as follows. First, the HPA-SR protocol is employed by the first- and second-price auctions to avoid sinkhole attacks. More specifically, using the Markov decision process (MDP), each node can select a kind of auction method to establish the secure route by avoiding the sinkhole attack. Second, the advanced speed cosine similarity clustering protocol that is considered as underlying structure is presented to improve the connectivity between nodes. The ASCS is constructed based on the cosine similarity and distance between nodes using the speed and direction of the nodes. The results of the performance show that the proposed HPA-SR protocol can establish the secure route avoiding the sinkhole attack while the proposed ASCS clustering can support the strong connectivity. Besides, the HPA-SR with ASCS protocol can show better performance than the benchmark protocol in terms of the routing delay, packet loss ratio, number of packet loss, and control overhead.

DLSMR: Deep Learning-Based Secure Multicast Routing Protocol against Wormhole Attack in Flying Ad Hoc Networks with Cell-Free Massive Multiple-Input Multiple-Output

“…The performance metrics used in this paper for performance evaluation purposes are the following [29]:…”

“…The node degree of its nodes can be written as where can be defined as stands for the node degree of the node , stands for the communication range of the node i , and stands for distance between node i and j . Then, the average node degree can be expressed as The self-adaptive node degree variance is calculated by subtracting the node measure from its average measure, which can be expressed as: The second factor is cosine similarity between two nodes which can be defined as [ 29 , 37 ] where and are the i -th and j -th nodes’ vector information, respectively. Each node is related with a mobility vector information metric value (i.e., speed, direction, and position) , ,⋯, , where constitutes the vector values which indicate link information between nodes.…”

“…However, the DAWA protocol has a high control overhead because the DAWA protocol broadcasts the packet discovery to all nodes in the network. Meanwhile, the authors of [ 29 ] proposed a hybrid price auction (HPA)-based routing protocol to avoid sinkhole attacks. In addition to that, the HPA protocol can improve the security, routing delay, and scalability performance.…”

Accelerated univariate sequential clustering

“…The similarity metrics facilitate obtaining temporal information. Pramitarini et al used a cosine similarity-based algorithm for network protocols [24]. In [25], authors designed a clustering algorithm for IoT applications based on a similarity measure.…”

RC-LAHR: Road-Side-Unit-Assisted Cloud-Based Location-Aware Hybrid Routing for Software-Defined Vehicular Ad Hoc Networks

“…Traditional VANETs employ a conventional network architecture where each node serves as a router, combining forwarding and routing functions into a single unit [ 15 , 16 , 17 , 18 , 19 ]. Nevertheless, this conventional approach has notable drawbacks, such as its inability to accommodate heterogeneity, scalability issues, limited programmability, and other concerns [ 20 ].…”