Zero-day attacks present a significant security threat to vehicular networks, exploiting vulnerabilities at both software and hardware levels within such systems that remain undiscovered. Mitigating these threats is essential to ensuring the safety and security of vehicular systems. Support Vector Machine (SVM) is a good candidate for anomaly detection of zero-day attacks within vehicular networks because it can handle highdimensional data and effectively distinguish between normal and abnormal patterns in complex and dynamic environments. A trained SVM on the normal operation data of in-vehicular network can identify flag deviations, thus making it effective in the detection of any previously unknown attack patterns, which is a common behaviour of zero-day attacks. In this paper, we introduce an anomaly detection method called “ZeroCAN” which models the behaviour of every single electronic control unit on the network with a separate SVM and a set of high-level features that capture the timing and data payload aspects of CANbus traffic. This approach achieves an anomaly detection rate of over $\mathbf{9 9 \%}$ and a false positive rate below $\mathbf{0. 0 1 \%}$ during normal operation in most cases.

Physical layer security is of paramount importance in vehicular cyber-physical systems, as it safeguards not only the privacy of sensitive data exchanged between vehicles and infrastructure but also ensures the integrity and reliability of the entire transportation network. Key generation plays a crucial role in establishing secure communication channels and facilitating the creation of unique cryptographic keys used for encryption, decryption, and authentication purposes. The secret key generation involves deriving secret bits by harnessing the inherent randomness present within the communication channels. The difficulty lies in precisely evaluating the randomness of the channel to achieve unanimous agreement on secure key generation within an unpredictable environment. In this line, we propose a combinatorial approach involving A2C and LSTM to decrease the key disagreement rate. A2C employs policy and value-based strategies to choose quantization levels predicated on the randomness of wireless channels and LSTM includes the partially observable radio channels and improves the environment. Based on our performance evaluation, the proposed A2C-LSTM method substantially accelerates the convergence rate by $\mathbf{5 0 - 6 0 \%}$ and reduces the Key Disagreement Rate (KDR) by $40 \%$.

The exploitation of radio channels’ inherent randomness for generating secret keys within a vehicular platoon offers a promising approach to securing communications in dynamic and unpredictable environments. The channel-based key generation leverages the fact that the physical characteristics of the radio channel, such as fading, shadowing, and multipath propagation, vary in a complex manner that makes it difficult for external adversaries to predict or replicate. A challenge lies in accurately assessing the channel’s randomness to ensure the generated keys are both secure and consistent across the platooning vehicles, especially in vehicular environments with high mobility and the ever-changing urban landscape. This paper proposes a novel channel-based key generation (DRL-KeyAgree) technique to enhance communication security within vehicular platoons through combinatorial deep reinforcement learning (DRL). DRL-KeyAgree addresses key disagreement among platooning vehicles by training advantage Actor-Critic (A2C), which integrates policy- and value-based strategies to dynamically select optimal quantization intervals adapting to the random wireless channels. Further incorporation of Long Short-Term Memory (LSTM) allows DRL-KeyAgree to capture the characteristics of partially observable radio channels, significantly enhancing the key agreement rate among vehicles. DRL-KeyAgree is rigorously evaluated using the standard National Institute of Standards and Technology (NIST) test suite.

In the mid-term future, vehicles will generate large amounts of data for both standalone usage (e.g., to recognize road features and external elements such as lanes, signs, and pedestrians) and cooperative usage (e.g., lane merging). However, processing the captured video and image data results comes with significant computational requirements (e.g., GPUs). Computer vision tasks, such as feature extraction, are unfeasible from a business perspective if performed directly in the User Equipment (UE), as automotive manufacturers are unwilling to increase the end-product’s costs. Thus, the logical solution is to collect and upload this data to be processed elsewhere. Nonetheless, processing the data as close to the vehicle is important due to latency constraints, thus calling for the use of Mobile Edge Computing (MEC). An additional benefit of this scenario, in which 5G connectivity enables data to be offloaded to the edge, is that the data from our car is not processed alone. Data from several sources, e.g., multiple vehicles and fixed cameras, can be offloaded to the edge node and processed together, enhancing its quality as more sources of data enhance the prediction output of machine-learning models. This demo showcases a video recording from a vehicle uploaded to an edge node via 5G software-defined-radio FPGA devices. There, a YOLO application to detect objects processes the video and communicates this information to the vehicle, ensuring QoS metrics even when the UE performs handover to a different cell or geographical area.

In wireless powered sensor networks (WPSN), data of ground sensors can be collected or relayed by an unmanned aerial vehicle (UAV) while the battery of the ground sensor can be charged via wireless power transfer. A key challenge of resource allocation in UAV-aided WPSN is to prevent battery drainage and buffer overflow of the ground sensors in the presence of highly dynamic lossy airborne channels which can result in packet reception errors. Moreover, state and action spaces of the resource allocation problem are large, which is hardly explored online. To address the challenges, a new data-driven deep reinforcement learning framework, DDRL-RA, is proposed to train flight resource allocation online so that the data packet loss is minimized. Due to time-varying airborne channels, DDRL- RA firstly leverages long short-term memory (LSTM) with precollected offline datasets for channel randomness predictions. Then, Deep Deterministic Policy Gradient (DDPG) is studied to control the flight trajectory of the UAV, and schedule the ground sensor to transmit data and harvest energy. To evaluate the performance of DDRL-RA, a UAV-ground sensor testbed is built, where real-world datasets of channel gains are collected. DDRL-RA is implemented on Tensorflow, and numerical results show that DDRL-RA achieves 19% lower packet loss than other learning-based frameworks.

Autonomous UAV cruising is gaining attention due to its flexible deployment in remote sensing, surveillance, and reconnaissance. A critical challenge in data collection with the autonomous UAV is the buffer overflows at the ground sensors and packet loss due to lossy airborne channels. Trajectory planning of the UAV is vital to alleviate buffer overflows as well as channel fading. In this work, we propose a Deep Deterministic Policy Gradient based Cruise Control (DDPG-CC) to reduce the overall packet loss through online training of headings and cruise velocity of the UAV, as well as the selection of the ground sensors for data collection. Preliminary performance evaluation demonstrates that DDPG-CC reduces the packet loss rate by under 5% when sufficient training is provided to the UAV.

Intersections are critical elements of urban traffic management and are identified as bottlenecks prone to traffic congestion and accidents. Intelligent intersection management plays a significant role in improving traffic efficiency and safety determining, among other metrics, the waiting time that vehicles incur when crossing an intersection. This work presents a preliminary analysis of the worst-case response time of intersection management protocols that handle mixed traffic with autonomous and human-driven vehicles. We deduce theoretical bounds for such time considered as the interval between the injection of a vehicle in the road system and its departure from the intersection, considering different intersection management protocols for mixed traffic, namely the Synchronous Intersection Management Protocol (SIMP) and several configurations of the conventional Round-Robin (RR) policy. Simulation results validate the analytical bounds partially. Ongoing work addresses the queue dynamics and its reliable detection by traffic simulators.

The unprecedented pervasiveness of IoT systems is pushing this technology into increasingly stringent domains. Such application scenarios become even more challenging due to the demand for encompassing the interplay between safety and security. The IEEE 802.15.4 DSME MAC behavior aims at addressing such systems by providing additional deterministic, synchronous multi-channel access support. However, despite the several improvements over the previous versions of the protocol, the standard lacks a complete solution to secure communications. In this front, we propose the integration of TAKS, an hybrid cryptography scheme, over a standard DSME network. In this paper, we describe the system architecture for integrating TAKS into DSME with minimum impact to the standard, and we venture into analysing the overhead of having such security solution over application delay and throughput. After a performance analysis, we learn that it is possible to achieve a minor impact of 1% to 14% on top of the expected network delay, depending on the platform used, while still guaranteeing strong security support over the DSME network.

Deterministic Synchronous Multichannel Extension (DSME) is a prominent MAC behavior first introduced in IEEE 802.15.4e that supports deterministic guarantees using its multisuperframe structure. DSME also facilitates techniques like multi-channel and CAP reduction that help to increase the number of available guaranteed timeslots in a network. However, no tuning of these functionalities in dynamic scenarios is supported in the standard. In this paper, we present an effective multisuperframe tuning technique that also helps to utilize CAP reduction in an effective manner improving flexibility and scalability, while guaranteeing bounded delay.

Wireless Sensor Networks have been enabling an ever increasing span of applications and usages in the industrial, domestic and commercial domains. Recent advancements in information and communication technologies have been fueling the increasing pervasiveness and ubiquity of this infrastructures, making them an obvious candidate to support the future Internet of Things. Among the prospective applications, however, there are those which present strict requirements in terms of timeliness and reliability, specially in the industrial domain. To address these, the IEEE 802.15.4 standard functionalities were recently enhanced by the IEEE 802.15.4e amendment. Ideas which are prominent in the industrial communication field such as frequency hopping, dedicated and shared timeslots and multichannel communication have been implemented in 802.15.4e. In this line, proposed MAC behaviors such as the Deterministic and Synchronous Multi-channel Extension (DSME) and Time Synchronous Channel Hopping (TSCH), are gaining a lot of attention. Nevertheless, to efficiently address the network demands in terms of latency, resources, and reliability, it is mandatory to carry out a thorough network planning. To achieve this, modeling the fundamental performance limits of such networks is of paramount importance to understand their behavior under the worst-case conditions and to make the appropriate design choices. Network Calculus is an established tool which can accurately compute the worst case bounds of a network. In this paper we provide an insight towards DSME and TSCH by modeling, using Network Calculus formalism, the delay bounds of these MAC behaviors. As a continuation of this work the end-to-end delay bounds will be derived for the rest of the MAC behaviors of IEEE 802.15.4e. Scheduling algorithms will be developed, analyzed and validated as a future work.