Jul 2026· International Conference on Ubiquitous and Future Networks· pp. 906-910· 0 citations· 16 references
Abstract
Vehicular Ad Hoc Networks (VANETs) play a critical role in Intelligent Transportation Systems (ITS) by enabling real-time vehicle communication for safety and traffic management. However, the open and decentralized nature of VANETs makes them vulnerable to False Information Attacks (FIA), where malicious vehicles disseminate fabricated data such as fake congestion alerts or incorrect speed information. This paper presents a lightweight and infrastructure-free framework for detecting FIA using an unsupervised machine learning approach based on the Isolation Forest algorithm. Unlike existing methods that require roadside units (RSUs), labeled datasets, or computationally intensive network simulators, the proposed framework operates using a small set of behavioral features extracted from vehicle beacon messages. The system is implemented entirely in Python and evaluated on a synthetically generated dataset designed to emulate realistic VANET conditions. Experimental results demonstrate an accuracy of 94.2%, precision of 86.1%, recall of 77.5%, and an F 1 -score of 81.6%. The results show that the proposed framework achieves competitive detection performance while maintaining low computational overhead, making it suitable for deployment on resource-constrained onboard units (OBUs) in real-world vehicular networks.
Vehicular Ad-Hoc Networks (VANETs) enable realtime communication for safety-critical applications including collision avoidance and traffic control. Their decentralized, dynamic architecture, however, makes them vulnerable to multiple attack classes, including Sybil, spoofing, Denial-of-Service (DoS), and other cyber threats. Existing defenses typically address cyber and physical layers independently, limiting their ability to capture the interplay between mobility patterns and attack propagation. This paper presents a cyber-physical simulation framework integrating vehicular mobility with the CybORG environment for multi-class attack mitigation. A Road Side Unit (RSU) acts as the infrastructure-based defender, monitoring vehicle behavior, maintaining trust scores, and executing defense actions via a Dueling Double Deep Q-Network with Prioritized Experience Replay (D3QN-PER). The agent learns optimal policies through environment interaction rather than static labeled data. Evaluation against two unsupervised baselines, Exponentially Weighted Moving Average (EWMA) and Trust-Gated Isolation Forest, demonstrates perfect detection performance (Recall = 100%, $\mathbf{F} \mathbf{1} \boldsymbol{=} \mathbf{1. 0 0 0 0})$ with zero false positives and zero false negatives, compared to 95.12% recall (EWMA) and 84.95% recall (Isolation Forest). The framework handles up to six concurrent attackers within the RSU's 200 m range with sub-millisecond latency, establishing a foundation for intelligent, adaptive security in vehicular networks.
Fasna Nadeera Irumpidamkandiyil Pocker, Farsana Ansari, Alexandre dos Santos Roque et al.· International Conference on...· 0 citations
Vehicular Ad Hoc Networks (VANETs) rely on Basic Safety Messages (BSMs) to support safety-critical applications such as collision avoidance and traffic awareness. However, BSMs can be exploited in Sybil attacks, where adversaries generate multiple ghost vehicles to manipulate traffic conditions. In this work, we introduce Phantom Jam, a motion-consistent Sybil attack designed to induce large-scale traffic disruptions while maintaining temporally consistent vehicle behavior. Unlike traditional Sybil attacks that rely on deterministic motion patterns, Phantom Jam combines map-aware trajectory replay with generative temporal modeling. Specifically, TimeGAN is used to synthesize plausible braking and acceleration dynamics, enabling ghost vehicles to emulate natural driving behavior during slowdown and recovery phases. We evaluate Phantom Jam using the F2MD simulation framework and the LuST Nano traffic scenario, with a recent deep-learning-based misbehavior detection system as the benchmark. Our experimental results show that Phantom Jam can reduce recall to as low as 0.66, indicating that a substantial portion of malicious vehicles remain undetected. Our work demonstrates that plausible temporal dynamics in Sybil attacks can pose significant challenges for modern VANET misbehavior detection systems.
Colby Cook, Ahmed Mohamed, Mengjun Xie· 2026 International Conferenc...· 0 citations
Vehicular Ad-hoc Networks (VANETs) is a very basic form of Intelligent Transportation Systems (ITS), which enables the information exchange in real time among heterogeneous entities that includes vehicles, roadside units (RSUs), basic pedestrians, and also emergency vehicles. However, the open wireless medium, high node mobility, and absence of centralized infrastructure make VANETs highly susceptible to routing attacks such as blackhole attacks, denial-of-service (DoS) attacks, and node impersonation. Existing approaches largely assume static adversarial models and homogeneous network composition, limiting their practical applicability in dynamic road environments. This paper proposes a secure multi-strategy adaptive routing framework for heterogeneous VANET environments under dynamic adversarial conditions. The framework models the road network as an undirected dynamic graph derived from the California road network dataset (roadNet-CA) obtained from the Stanford Network Analysis Project (SNAP) repository. Node heterogeneity explicitly includes classification of network by having four different types that is RSU, vehicles, Pedestrians and also emergency vehicles. This uses a degree centrality-based assignment. The framework employs a mutual authentication mechanism that is having prior route which is defined and will be followed by a multi strategy secure routing engine. This mechanism evaluates by using four path strategies: Direct Avoidance, K-Shortest Path, Weighted Random Routing, and Node-Disjoint Routing. The proposed dynamic routing mechanism continuously monitors all the possible active path and then it tries to trigger and ensure the re-compute is done where the attacker is present. While doing the same analysis and computation we understand that the packet delivery ratio has improved and also there is a lot of energy saving. This method ensures to have reduced delay end to end and final stage the through was also compared the proposed method higher throughput.
M. Shilpa, V. Shilpa, P. Karthik et al.· Discover Computing· 0 citations
Cache poisoning attacks in Vehicular Named Data Networking (V-NDN) pose a serious threat by injecting false content into the Content Store, compromising network integrity. This paper proposes a threshold-based reputation algorithm to detect and mitigate such attacks in V-NDN using ndnSIM with a V2V multi-hop topology of 31 nodes across three urban road segments. The algorithm assigns reputation values to all nodes, applying penalties for malicious behavior and blocking nodes below a threshold of 0.5. Two scenarios are evaluated: an attack scenario without mitigation and an attack scenario with the proposed reputation algorithm. Results show a Detection Rate (DR) of 75%, False Positive Rate (FPR) of 0%, and False Cache Poisoning Rate (FCP) of 5.87%, with Cache Hit Ratio (CHR) improving from 28.94% to 30.99% after mitigation, confirming the algorithm's effectiveness in detecting and mitigating cache poisoning attacks in dynamic vehicular network environments.
Zhikya Sekar Lutfi Purnomo, L. V. Yovita, Istikmal· International Seminar on Int...· 0 citations
The exchange of beacon messages plays an important role in the operation of vehicles in VANETs. The data transmitted through these messages can be subject to deliberate manipulation from cyberattacks or accidental distortion caused by faulty sensors. To address this challenge, a Multi-Criteria Voting mechanism was developed by integrating three independent algorithms: Threshold, Cooperative, and LA-DETECTS. The purpose of this mechanism is to improve the efficiency of detecting misbehavior. The proposed mechanism is based on the principle of consensus, whereby a vehicle is classified as malicious only if at least two algorithms confirm the presence of manipulation, thus reducing reliance on a single, potentially biased source. Simulations conducted using an integrated environment (OMNET++, Veins, and SUMO) demonstrate the efficiency of the developed mechanism in a complex urban scenario. It outperforms existing single-criterion algorithms by achieving an optimal balance between detection accuracy and false positive reduction, recording a peak F1-Score of 69.02 % and a Precision of 56.05%.
I. Shaleesh, Akram A. Almohammedi, Mohammed Balfaqih· 2026 6th International Confe...· 0 citations
Rogue Base Stations (RBS) remain a persistent security threat to fifth-generation (5G) and emerging sixth-generation (6G) cellular systems by impersonating legitimate infrastructure and exploiting vulnerabilities in pre-authentication signaling and mobility procedures. The risk is particularly critical in vehicular and Vehicle-to-Everything (V2X) environments, where high mobility and millisecond-scale handover operations tightly couple communication reliability with safety-critical control functions. Although prior surveys examine LTE identity catchers and general cellular security threats, they rarely evaluate RBS detection under vehicular mobility dynamics or within the latency and reliability constraints of Ultra-Reliable Low-Latency Communication (URLLC) services. In addition, the limited availability of realistic measurement report (MR) datasets have hindered reproducible benchmarking of data-driven detection methods. This article presents a vehicular-oriented survey of RBS detection in 5G and beyond networks, explicitly addressing mobility-constrained detection, handover-security interactions, and V2X safety requirements that are not systematically addressed in prior surveys, which primarily focus on pre-5G threat models, IMSI-catcher attacks, or general cellular security. We introduce a method-centric taxonomy that organizes existing approaches into five families based on their primary evidence sources and inference mechanisms: signal anomaly detection, protocol and traffic analytics, RF fingerprinting, network-level frameworks, and machine-learning-based detection. Using a PRISMA-compatible structured literature review across 102 included studies and a structured comparative evaluation framework, each family is analyzed across detection latency, computational overhead, robustness to mobility, false alarm susceptibility, and feasibility within quantified pre-handover decision windows. Direct cross-study quantitative comparison is precluded by heterogeneous reporting conventions across the surveyed literature; the framework, therefore, provides structured qualitative synthesis and indicative performance ranges rather than pooled empirical estimates. The analysis reveals that no individual technique satisfies vehicular URLLC constraints in isolation, motivating layered architectures combining lightweight UE-side detection with edge-assisted and operator-level analytics. A scenario-driven safety analysis links detection error rates to operational consequences across five V2X use cases under varying URLLC severity levels. The survey formalizes evaluation criteria for MR-driven detection and highlights realistic MR generation as a foundation for reproducible evaluation and cross-study comparison in next-generation vehicular communication systems.
Roland Lamptey, M. Saedi, V. Stankovic et al.· IEEE Open Journal of the Com...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
