Wireless vehicular communications are a promising technology. Most applications related to vehicular communications aim to improve road safety and have special requirements concerning latency and reliability. The traditional channel estimation techniques used in the IEEE 802.11 standard do not properly perform over vehicular channels. This is because vehicular communications are subject to non-stationary, time-varying, frequency-selective wireless channels. Therefore, the main goal of this work is the introduction of a new channel estimation and equalization technique based on a Semi-supervised Extreme Learning Machine (SS-ELM) in order to address the harsh characteristics of the vehicular channel and improve the performance of the communication link. The performance of the proposed technique is compared with traditional estimators, as well as state-of-the-art machine-learning-based algorithms over an urban scenario setup in terms of bit error rate. The proposed SS-ELM scheme outperformed the extreme learning machine and the fully complex extreme learning machine algorithms for the evaluated scenarios. Compared to traditional techniques, the proposed SS-ELM scheme has a very similar performance. It is also observed that, although the SS-ELM scheme requires the largest operation time among the evaluated techniques, its execution time is still far away from the latency requirements specified by the standard for safety applications.

The opportunistic exchange of information between vehicles can significantly contribute to reducing the occurrence of accidents and mitigating their damages. However, in urban environments, especially at intersection scenarios, obstacles such as buildings and walls block the line of sight between the transmitter and receiver, reducing the vehicular communication range and thus harming the performance of road safety applications. Furthermore, the sizes of the surrounding vehicles and weather conditions may affect the communication. This makes communications in urban V2V communication scenarios extremely difficult. Since the late notification of vehicles or incidents can lead to the loss of human lives, this paper focuses on improving urban vehicle-to-vehicle (V2V) communications at intersections by using a transmission scheme able of adapting to the surrounding environment. Therefore, we proposed a neuroevolution of augmenting topologies-based adaptive beamforming scheme to control the radiation pattern of an antenna array and thus mitigate the effects generated by shadowing in urban V2V communication at intersection scenarios. This work considered the IEEE 802.11p standard for the physical layer of the vehicular communication link. The results show that our proposal outperformed the isotropic antenna in terms of the communication range and response time, as well as other traditional machine learning approaches, such as genetic algorithms and mutation strategy-based particle swarm optimization.

Cooperative vehicular systems require that vehicles fuse sensor data and broadcast one-hop safety messages containing their kinematic information to enable vehicular applications based on incident detection. Several congestion control mechanisms have been proposed to mitigate channel congestion resulting from the continuous transmission of safety messages. This paper investigates the effect of message rate adaptation-based congestion control from a road safety perspective by evaluating the feasibility of prominent approaches, such as PULSAR, LIMERIC, reactive ETSI DCC, and SAE J2945/1, to support lane-changing maneuvers on multi-lane highways under varying conditions. Simulation results demonstrate that message size, vehicular density, losses at the physical layer, and observation time significantly influence the lane-changing application’s capability to detect unsafe maneuvers when congestion control is in action. Specific recommendations and guidelines for congestion control are provided to improve decision-making at the application level.

Automated vehicles (AVs) require proactive safety systems to safely operate in diverse driving scenarios. The complexity and challenges of high-risk driving maneuvers, such as complete overtaking, have led the research community to make significant efforts toward developing reliable operating modes for overtaking maneuver systems. These efforts aim to better understand the benefits and limitations of intrinsic operating modes in complete overtaking to take AVs to the next level. This paper delves into the challenges of complete overtaking maneuvers by developing and evaluating an overtaking system for AVs. Specifically, a vision and radar-based automated overtaking system is designed, which utilizes data from cameras and radars to track surrounding vehicles, evaluate overtaking feasibility based on overtaking time (OT) and time-to-collision (TTC), and decide the driving behavior of the ego vehicle. A modular architecture is developed encompassing perception, sensor fusion and tracking, collision detection and decision-making, overtaking planning, and motion control. Extensive simulations are conducted under varying conditions, utilizing a specialized framework consisting of RoadRunner, Automated Driving Toolbox, and MATLAB/Simulink, to assess the system’s ability to avoid potential collisions and enhance maneuver reliability. The findings highlight the significance of the operating mode in accident mitigation and contribute to the broader goal of achieving safe automated driving.

This paper investigates the use of unmanned aerial vehicles (UAVs) as flying base stations (BSs) to enhance fifth generation (5G) vehicular communications on highways, where traffic congestion and fluctuating user demand can challenge the capacity of terrestrial infrastructure. While UAV-assisted vehicular networking has attracted significant attention, many existing studies rely on simplified mobility, propagation, or communication models that limit the assessment of practical deployment performance. To address these limitations, we develop a realistic UAV-assisted vehicular networking framework that integrates microscopic traffic simulation through Simulation of Urban MObility (SUMO), network control via Traffic Control Interface (TraCI), and standard-compliant 5G communication modeling using MATLAB R2025b 5G Toolbox. The framework incorporates a 3rd Generation Partnership Project (3GPP) rural macro cell (RMa) highway scenario, detailed clustered delay line (CDL)-based channel characterization, and cross-layer communication procedures. Within this framework, we propose a low-complexity trajectory optimization strategy that adapts the UAV position in real time to maximize the average received signal to noise ratio (SNR) while respecting practical motion constraints. Simulation results demonstrate that adaptive UAV positioning enhances communication performance, achieving mean SNR gains of up to 2.04 dB, throughput improvement of up to 11.2%, and block error rate (BLER) reductions of up to 27.3%. These findings highlight the potential of UAV-assisted communications to enhance user-perceived quality of service (QoS) for bandwidth-demanding vehicular applications under realistic 5G highway operating conditions.

This paper investigates how cooperative perception can be leveraged for blind spot mitigation and to avoid potential vehicle collisions during lane-changing maneuvers in multi-lane highways. To achieve this goal, a simulation-based study of a lane-changing scenario relying on cooperative perception is comprehensively presented, where the obstructing vehicle uses vehicle-to-vehicle (V2V) communication to periodically transmit perception sensor data to the ego vehicle containing relevant state information of a vehicle approaching from the blind spot zone. For the ego vehicle, a cooperative-perception-based lane-changing system (CP-based LCh) is introduced, which uses sensor information and the received Collective Perception Messages (CPMs) to predict vehicle trajectories and assess in advance the reliability of the lane-changing maneuver. The lane-changing mechanism in the ego vehicle encompasses the following stages: most important object (MIO) identification, risk assessment and lane-changing decision, and driving behavior definition. The collision avoidance capability of the CP-based LCh system is investigated in comparison with a conventional blind spot detection (Conventional BSD) strategy, using a realistic simulation framework consisting of RoadRunner, Automated Driving Toolbox, and MATLAB/Simulink. The evaluation is conducted under different yaw angles, velocities, accelerations, transmission rates, safety margins, safety windows, probabilities of successful reception, and types of obstructing vehicles. Simulation results demonstrate that CP-based LCh consistently outperforms Conventional BSD in terms of Collision Prevention Rate (CPR) under the studied highway conditions, providing valuable insights into how cooperative perception can be leveraged to improve road safety.