Detecting Sybil Attacks in VANET: Exploring Feature Diversity and Deep Learning Algorithms with Insights into Sybil Node Associations

Vehicular ad hoc networks (VANET) facilitate vehicle to everything (V2X) communication between vehicles and road side units (RSU) to exchange safety and alert messages required for the successful implementation of VANET applications. However, VANET is exposed to Sybil attacks, where malicious vehicl...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of network and systems management Jg. 32; H. 3; S. 51
Hauptverfasser: Sultana, Rukhsar, Grover, Jyoti, Tripathi, Meenakshi, Sachdev, Manhar Singh, Taneja, Sparsh
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York Springer US 01.07.2024
Springer Nature B.V
Schlagworte:
Algorithms
Artificial neural networks
Associativity
Communications Engineering
Computer Communication Networks
Computer Science
Computer Systems Organization and Communication Networks
Cybersecurity
Deep learning
Euclidean geometry
Information Systems and Communication Service
Learning
Machine learning
Mobile ad hoc networks
Networks
Neural networks
Operations Research/Decision Theory
Roadsides
Short term memory
Vehicles
Deep learning
VANET
Traffic flow
Sybil attack detection
Attacks
Dynamic time warping
ISSN:1064-7570, 1573-7705
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Vehicular ad hoc networks (VANET) facilitate vehicle to everything (V2X) communication between vehicles and road side units (RSU) to exchange safety and alert messages required for the successful implementation of VANET applications. However, VANET is exposed to Sybil attacks, where malicious vehicles generate multiple false identities to seize control of network resources or create non-existent traffic situations. While several Sybil attack detection mechanisms based on deterministic and learning methods exist, their performance is restricted to particular scenarios. This work proposes a deep learning-based Sybil Attack Detection mechanism. It identifies the Sybil nodes by detecting the associativity between senders at a time instant using common vehicle characteristics, including Euclidean distance, speed, flow, and computed similarity between senders by using the dynamic time warping (DTW) method. Among the five applied deep learning models, the proposed solution achieves improved detection performance through convolutional neural network (CNN) and a combination of CNN with long short-term memory (LSTM) models in varying network scenarios.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1064-7570
1573-7705
DOI:10.1007/s10922-024-09827-7