Autonomous driving with virtual rail – a wireless tracked and controlled self-driving vehicle system based on the integration of communication, sensing, and computing

Abstract

Safety and high efficiency are the two priorities of autonomous driving, which involves the integration of communication, sensing, and computing (ICSC). However, it is difficult to realize these priorities through the two solutions available at present, i.e., individual vehicles and networked vehicles. In this paper, enlightened by the compulsive-controlling strategies of railway transportation system, a promising solution called the autonomous driving with virtual rail (ADWVR) is proposed to improve safety and efficiency. In ADWVR, virtual rails, which are defined as moving trajectories, and the control strategies for vehicles are planned synergistically. With the support of enhanced ICSC, vehicles can be tracked and controlled along the virtual rail in real time through a cellular network. In this paper, the system architecture of ADWVR is presented, and the constitution and functions of each module are introduced. Moreover, the key technologies of ADWVR are discussed, including centralized control with a hierarchical architecture, ultra-reliable and low-latency wireless communications, and high-precision positioning. Five ADWVR implementations based on multiple positioning technologies are reviewed, where the deployment strategies and application scenarios are elaborated separately. The proposed ADWVR is a potential scheme for improving the safety and efficiency of autonomous driving and may provide guidance for future research.