Special Issue: Recent Advances in Intelligent Vehicular Networks and Communications

Over the last few decades, research on intelligent vehicular networks and communications has grown significantly due to the increasing demand for advanced vehicular technologies and the need for enhanced road safety, traffic management, and efficient transportation systems. As society advances and the need for smarter mobility solutions rises, intelligent vehicular networks and communications have emerged as dynamic research fields that introduce several challenges and opportunities for the development of innovative vehicular technologies [1].

Intelligent vehicular networks and communications are transforming the transportation sector by offering innovative services that range from autonomous driving and vehicle-to-everything (V2X) communication to advanced driver-assistance systems (ADAS) and smart traffic management solutions. In the current era of Industry 4.0, cutting-edge digital technologies and devices are extensively utilized for innovation and value creation in transportation, aiming to enhance road safety, efficiency, and overall user experience [2].

Artificial intelligence (AI) has garnered significant attention in the vehicular industry due to its potential for real-time decision-making, predictive maintenance, and optimizing traffic flow [3]. As data generation from vehicles and infrastructure increases, and computational resources become more powerful, AI technologies are becoming more accurate and efficient. Autonomous vehicles, equipped with AI and machine learning algorithms, are paving the way for self-driving cars, promising safer and more reliable transportation.

Advanced communication technologies, such as 5G and the Internet of Things (IoT), play a crucial role in enabling seamless V2X communication. These technologies facilitate real-time data exchange between vehicles, infrastructure, and pedestrians, contributing to the development of smart cities. Current research efforts on V2X technologies focus on enhancing scalability, interoperability, and security, to ensure reliable and secure communication networks [4,5,6].

As digital technologies continue to evolve and become more accessible, the integration of AI, IoT, and advanced communication systems in vehicular networks is driving the transformation of the transportation sector, paving the way for smarter, safer, and more efficient mobility solutions.

This Special Issue aimed to provide a platform for sharing state-of-the-art research and development on intelligent vehicular networks and communications. The content of this Special Issue focused on the architecture and implementations, communication and networking protocols, data security and privacy, and other enabling technologies for intelligent vehicular networks and communications.

We feel obliged to thank the scientific community, who responded with appreciable efforts, as a significant number of original research articles were submitted and considered for publication. Eleven (11) were finally accepted as full papers as a result of a careful blind peer review process with diligent editorial input from the Guest Editors. All accepted papers depict significant contributions and cover multidisciplinary application domains.

Zhang et al. (Contribution 1) authored the first contribution, which is titled “A Non-Intrusive Automated Testing System for Internet of Vehicles App Based on Deep Learning”. Researchers have been pioneering the idea of non-invasive testing of in-vehicle information screens of IoV applications. Through building a data set specifically for text detection and recognition on vehicle screens, they have presented a novel approach for this within the context automotive control screens, leveraging an advanced version of the FOTS framework. This approach significantly amplified the semantic potential of features by intertwining the detection and recognition of text within vehicle central-control screens. Additionally, this methodology incorporated a feature transformation module, bolstering its ability to handle features across multiple scales. In the end, both visual and quantitative analyses showed the proficiency of the method in fulfilling the text detection and recognition tasks within vehicular screens. This advancement holds profound implications for streamlining automated testing procedures within the burgeoning landscape of IoV applications.

The second contribution is titled “Packet Reordering in the Era of 6G: Techniques, Challenges, and Applications”, and it is authored by Lin et al. (Contribution 2). This review paper delved deeply into the intricate challenges posed by packet reordering in 6G-enabled vehicular networks, offering a detailed examination of their impact, underlying causes, and threats. Through a rigorous analysis, the paper uncovered the intricacies of packet reordering and the mechanisms behind its occurrence. To address the challenges associated with packet reordering, researchers introduced a range of innovative end-to-end methods and metrics tailored specifically for the 6 G context. Furthermore, they anticipated a growing trend toward integrating machine learning and data-driven optimization techniques to address packet reordering. These advanced methods, through leveraging the power of big data and artificial intelligence, can provide unprecedented insights into packet reordering behavior and enable more effective management strategies.

The third contribution is titled “Co-Simulation Platform with Hardware-in-the-Loop Using RTDS and EXata for Smart Grid” and is authored by Gong et al. (Contribution 3). In this paper, a new grid co-simulation platform was proposed to solve whole coupling network failure when it either breaks down due to component failure or attack in the smart grid system. This platform harnessed the strengths of RTDS for power simulations and EXata for communication modeling, linking them through a sophisticated protocol translation module that ensures seamless synchronization and format adaptation. Furthermore, it incorporated a versatile, programmable fault configuration interface (PFCI), empowering users to dynamically alter model parameters with ease. Additionally, the introduction of a real sub-network access interface (RSAI) facilitated direct interaction with physical grid elements or sub-nets within the SDVN realm, thereby enhancing the fidelity of simulations. To validate the efficacy of the platform, a series of rigorous tests were conducted. Several tests demonstrated that the platform has an obvious effect.

The fourth contribution was authored by St. Amour et al. (Contribution 4): “Data Rate Selection Strategies for Periodic Transmission of Safety Messages in VANET”. Due to the increasing number of vehicles, the limited bandwidth of the wireless channel used for vehicle-to-vehicle (V2V) communication can become congested, resulting in packets being dropped or delayed. To solve these problems, the aforementioned researchers proposed a decentralized congestion control algorithm. They found that BSMs transmitted with a lower bitrate have longer delivery times, but a lower SINR threshold ensured the accuracy of BSMs. However, BSMs transmitted with a higher bitrate may not reach the required SINR threshold, although they have a shorter sending time and reduce network congestion. Therefore, the congestion control algorithm was used to select a suitable bitrate value for each BSM according to the calculated CBR. The simulation results showed that the proposed algorithm was superior to the existing data rate-based algorithms in terms of packet reception and total channel load.

The fifth contribution is authored by Magsi et al. (Contribution 5), and its title is “A Machine Learning-Based Interest Flooding Attack Detection System in Vehicular Named Data Networking”. In this work, the authors proposed to use machine learning (ML) to detect an interest flooding attack (IFA) in vehicular named data networking (VNDN). NDN offers an effective network architecture for VANET to solve issues of the existing TCP/IPs network architecture. However, NDNs are extremely vulnerable to various attacks, and IFA is one of the most prevalent in VNDN. IFA consumes network resources through attacking intermediate nodes in NDN and sending non-existing interest packets. Therefore, these researchers used an ML binary classification system at RSUs to detect whether intermediate nodes were under attack. Various ML classifiers such as DT, K-nearest neighbor (KNN), random forest (RF), Gaussian naïve Bayes (GNB), and logistic regression (LR) were used for classification. The experimental results showed that the RF classifier achieves the highest accuracy (94%) in detecting IFA vehicles. After this, the researchers designed an attack prevention algorithm on Python, which made the intermediate node accept or reject the interest request according to the legality of the detected vehicles, ultimately preventing the IFA attack on VNDN.

The sixth contribution is titled “Multi-Frequency Channel Measurement and Characteristic Analysis in Forested Scenario for Emergency Rescue”, and it is authored by Guo et al. (Contribution 6). The authors of this study conducted a comprehensive study on wireless communication performance in forest scenarios, which was crucial for emergency rescue operations. The research involved a frequency channel measurement activity at 380 MHz, 640 MHz, and 1420 MHz in a virgin forest to accurately characterize wireless channels in such environments. The study began with the collection of measurement data, from which the average power delay profile (APDP) was derived, and multipath components (MPCs) were extracted. The authors then analyzed the root mean square (RMS) delay spread and path loss (PL) based on the MPCs, proposing a new path loss model tailored to forest conditions. This new model, known as the ABG model, considered frequency and distance factors, and it better captured the rapid loss characteristics observed in forest scenarios compared to existing models. Additionally, the authors calculated and modeled shadow fading (SF) as a normal distribution at the three frequency bands, highlighting the high instability and strong fluctuation in forested environments. The results, including detailed heatmaps and the fitted parameters of RMS delay spread and SF, provided reliable references for the evaluation and design of emergency communication systems in forest areas, addressing the unique challenges posed by these complex propagation environments.

The seventh contribution was titled “Hyperbolic-Embedding-Aided Geographic Routing in Intelligent Vehicular Networks”, and it was authored by Ying Pan and Na Lyu (Contribution 7). These authors introduced a hyperbolic-embedding-aided geographic routing strategy (HGR) designed to address routing voids in intelligent vehicular networks. Unlike traditional Euclidean geographic greedy routing strategies, the proposed HGR method employed a two-dimensional Poincaré hyperbolic disk for network topology embedding, enabling greedy forwarding based on the nodes’ hyperbolic coordinates. This innovative approach aimed to enhance routing efficiency and flexibility, which were essential for supporting diverse vehicle-to-everything (V2X) applications in dynamic, intelligent vehicular networks. Through a series of comparative simulation experiments, the effectiveness of the HGR strategy was validated, demonstrating a significant improvement in routing success rates with reduced routing path stretch, while incurring minimal additional routing computation time. This work contributed to the advancement of routing methodologies in vehicular networks, ensuring more reliable and efficient connectivity for various smart terminals, vehicles, roads, and users.

The eighth contribution was authored by Pang et al. (Contribution 8), titled “Research on the Evaluation and Prediction of V2I Channel Quality Levels in Urban Environments”. This study proposed a novel method for evaluating and forecasting the quality of vehicle-to-infrastructure (V2I) communication channels in urban settings using long-range (LoRa) technology. The approach introduced the concept of channel quality scoring, which offered a more precise description of channel quality compared to traditional packet reception rate (PRR) assessments. A channel quality assessment model, utilizing the gated recurrent unit (GRU) algorithm, used the current channel quality score of the vehicular terminal and the spatial channel parameters (SCP) of its location to classify channel quality levels with an accuracy of 97.5%. For prediction, the method focused on forecasting the channel quality score by calculating SCP for the vehicle’s subsequent temporal location, achieving spatial and temporal channel quality predictions through the Variational Mode Decomposition-Backoff-Bidirectional Long Short-Term Memory (VMD-BO-BiLSTM) algorithm, which reaches an R2 value of 0.9945. The study further employed a combined weighting-based approach to address the problem of channel quality assessment in mobile V2I scenarios. This method accounted for long-term and short-term variations in the channel, providing a scientific and sensitive measurement of the dynamically changing channel environment for vehicular terminals. By combining subjective and objective perspectives, this approach enhanced the accuracy of channel quality descriptions beyond what PRR alone can offer. The validation in three different link quality scenarios confirmed the method’s effectiveness. The findings indicated that this method significantly improved channel quality assessment and prediction, which had important implications for subsequent channel resource allocation in V2I scenarios.

The ninth contribution, titled “Performance Analysis of Internet of Vehicles Mesh Networks Based on Actual Switch Models”, was authored by Hu et al. (Contribution 9). The authors of this study proposed a network performance analysis model based on actual switches to address the problem that adopted simulation tools, such as OPNET, cannot achieve the dynamic link generation of IoV mesh networks. This model first constructed a typical IoV mesh network architecture and abstracted it into a mathematical model that describes how the link and topology change over time. Then, the task generation and forwarding model were proposed to obtain the actual traffic distribution of the network. This was achieved through user traffic modeling and the construction of node port routing tables to establish task scheduling, forwarding, and transmission strategies based on the actual switch model. Finally, a scientific network performance indicator system was constructed, which included the packet loss rate, task arrival rate, node load rate, and so on. Several experimental simulations were conducted to assess variations in multiple indicators across diverse vehicle densities and caching capacities. The simulation results demonstrated that, with rising task traffic and decreasing node caching capacity, the packet loss rate increases, and the task arrival rate decreases in the network. The proposed model can effectively evaluate the network performance across various traffic states and provide valuable insights for network construction and enhancement.

The tenth contribution was authored by Wang et al. (Contribution 10): “Vertical Switching Algorithm for Unmanned Aerial Vehicle in Power Grid Heterogeneous Communication Networks”. In this work, the authors investigated the vertical switching problem of wireless heterogeneous networks for unmanned aerial vehicles (UAVs) engaged in inspection tasks within power grid scenarios. The proposed model necessitated that a UAV conducting power grid inspections should plan its flight trajectory, avoid obstacles, and identify the optimal path to each inspection point. During the inspection process, the quality of communication services for the UAV was maintained. The UAV dynamically selected different networks for access at various locations, thereby addressing the dynamic network selection and vertical switching challenge. A method integrating trajectory planning and network selection was proposed, initially employing the A-star algorithm to determine suitable trajectories, followed by the evaluation and selection of networks using the Fuzzy Analytic Hierarchy Process (FAHP) to identify the most appropriate network. The numerical results demonstrated that this approach effectively met UAV inspection task requirements and reduced the number of network switches, thereby mitigating the issue of terminal vertical switches in power grid scenarios.

The eleventh contribution was authored by Naeem et al. (Contribution 11) and is titled “Road to Efficiency: V2V Enabled Intelligent Transportation System”. The authors reviewed the development of intelligent transportation systems, whose widespread application has improved mobility in traffic control and vehicle management while solving the traffic problems of urbanization. The study highlighted the great potential of intelligent transportation systems, which were subject to a comprehensive narrative review. The review focused on the basic communication system architectures of intelligent transportation systems, such as decentralized mesh networks, cloud-integrated hubs, edge computing-based architectures, blockchain-enabled networks, hybrid cellular networks, ad-hoc networks, and AI-driven dynamic networks. The main components and organizational frameworks of these communication architectures were compared and analyzed. At the same time, the contributions and limitations of different communication architecture applications were comprehensively analyzed. The inclusion criteria focused on articles published in English between 2015 and 2024. These authors concluded that there is a need for further research challenges and research directions, such as frequent topological changes, Scalability and Reliability, Autonomous and Cooperative Driving, etc., to accelerate the construction and deployment of V2V intelligent transportation systems worldwide.

The purpose and inspiration behind this Special Issue was to make a broad yet timely addition to the current body of literature. It is anticipated that the valuable methodologies featured in this Special Issue will be considered as beneficial and engaging, and will ultimately be acknowledged by both the industry and the scientific community. Our hope is that researchers will be motivated by the novel strategies presented, thereby advancing research across various multidisciplinary fields and encouraging further investigation in the area of intelligent vehicular networks and communications overall. Future efforts may include leveraging cutting-edge techniques and methodologies to enhance V2X communication, improve autonomous driving capabilities, and optimize resource allocation and task offloading in dynamic vehicular environments.