The rapid development and adaptation of advanced driver assistance systems (ADAS) and autonomous driving increases the burden of in-vehicle networks. In-vehicle networks are now required to provide a fast data rate and bounded delay for real-time operation, while conventional protocols such as controller area networks, local interconnected networks, and FlexRay begin to show limitations. Ethernet-based time-sensitive network (TSN) technology has been proposed as an alternative. TSN is a set of Ethernet standards being developed by the IEEE TSN task group, which aims to satisfy requirements such as real-time operation, stability, and low and bounded latency, and it can be used in automotive, industrial, and aerospace applications. This study introduces several standards for Ethernet traffic scheduling based on TSN technology and proposes a heuristic-based scheduling algorithm for Ethernet scheduling. In addition, three network configurations are simulated using OMNeT++ to show the applicability. The heuristic TSN scheduling algorithm is a straightforward and systematic procedure for practical network designers.
The automotive industry is experiencing a period of innovation, represented by the term CASE (connected, autonomous, shared, and electric). Among the innovative new technologies for automobiles, intelligent tire (iTire) collects road surface information through sensors installed inside a tire and informs the driver of the road conditions. iTire can promote safe driving. Various kinds of research on iTire is ongoing, and this paper proposes an algorithm to determine the road surface conditions while driving. Specifically, we have proposed a method for extracting the feature points of a frequency band, by converting acceleration data collected by sensors through fast Fourier transform (FFT) and determining road surface conditions via an artificial neural network. Lastly, the applicability of the algorithm was verified.
The accurate detection of tire tread wear plays an important role in preventing tire-related accidents. In previous studies, tire wear detection is performed by interpreting mathematical models and tire characteristics. However, this approach may not accurately reflect the real driving environment. In this study, we propose a tire tread wear detection system that utilizes machine learning to provide accurate results under real-road driving conditions. The proposed system comprises (1) an intelligent tire that samples the measured acceleration signals and processes them in a dataset, (2) a preprocessing component that extracts features from the collected data according to the degree of wear, and (3) a detection component that uses a deep neural network to classify the degree of wear. To implement the proposed system in a vehicle, we designed an acceleration-based intelligent tire that can transmit data over wireless networks. At speeds between 30 and 80 km/h, the proposed system was experimentally demonstrated to achieve an accuracy of 95.51% for detecting tire tread wear under real-road driving conditions. Moreover, this system uses only preprocessed acceleration signals and machine-learning algorithms, without requiring complex physical models and numerical analyses.
In this paper, we proposed a scheduling method to optimize a time-critical traffic using genetic algorithm in automotive ethernet networks. The proposed traffic scheduling method aims to generate a transmission schedule to minimize end-to-end delay, variation of reception time at the destination (jitter), and bandwidth utilization. These indicators should be simultaneously considered when generating a schedule since they are closely related to the performance of the networks. However, it is not easy to schedule the traffic considering these indicators together since they have different sizes and units. To solve this problem, the fitness function to evaluate performance of a chromosome has normalized indicators as parameters. A chromosome in the genetic algorithm consists of genes containing information (size, transmission period, and network delay) of messages to be scheduled. Finally, the applicability of the proposed schedule optimization method was examined by simulating an extended automotive Ethernet network for an autonomous vehicle.
Time-synchronization technology can provide a common notion of time among the participating nodes on a network. This is essential not only for protocol operation for time-critical services but also for the time stamp for information included in the message. Precise time information can be very crucial for such things as autonomous driving because there are various sensor measurements from multiple cameras, and radio detection and ranging (radar) and light detection and ranging (LiDAR) are used for perceiving the current situation via sensor fusion. A well-known synchronization method, IEEE 1588, denoted as the precision time protocol (PTP), can be used for various applications. For in-vehicle networks of autonomous cars, we have to consider that the network may be comprised of subnetworks based on different protocols such as controller area network (CAN) and Ethernet. However, implementing PTPs on such heterogeneous vehicle networks causes several problems. First, the PTP procedure must be modified to be implement on a CAN network. Second, to calculate the delay and offset for PTP, the processing delay that occurs during message conversion must be considered. In this paper, we propose a synchronization method for CAN–Ethernet networks to solve these problems. The performance of the proposed synchronization method is evaluated by experiments on a real CAN–Ethernet network.
This study investigates the optimization of traffic scheduling in autonomous vehicle networks using time-sensitive networking (TSN), a type of deterministic Ethernet. Ethernet has high bandwidth and compatibility to support various protocols, and its application range is expanding from office environments to smart factories, aerospace, and automobiles. TSN is a representative technology of deterministic Ethernet and is composed of various standards such as time synchronization, stream reservation, seamless redundancy, frame preemption, and scheduled traffic, which are sub-standards of IEEE 802.1 Ethernet established by the IEEE TSN task group. In order to ensure real-time transmission by minimizing end-to-end delay in a TSN network environment, it is necessary to schedule transmission timing in all links transmitting ST (Scheduled Traffic). This paper proposes network performance metrics and methods for applying machine learning (ML) techniques to optimize traffic scheduling. This study demonstrates that the traffic scheduling problem, which has NP-hard complexity, can be optimized using ML algorithms. The performance of each algorithm is compared and analyzed to identify the scheduling algorithm that best meets the network requirements. Reinforcement learning algorithms, specifically DQN (Deep Q Network) and A2C (Advantage Actor-Critic) were used, and normalized performance metrics (E2E delay, jitter, and guard band bandwidth usage) along with an evaluation function based on their weighted sum were proposed. The performance of each algorithm was evaluated using the topology of a real autonomous vehicle network, and their strengths and weaknesses were compared. The results confirm that artificial intelligence-based algorithms are effective for optimizing TSN traffic scheduling. This study suggests that further theoretical and practical research is needed to enhance the feasibility of applying deterministic Ethernet to autonomous vehicle networks, focusing on time synchronization and schedule optimization.