What's in C-V2X (Cellular Vehicle-to-Everything) (6)?

C-V2X Integration Solutions: The 5G network-based PC5 C-V2X system integration solutions currently include the following categories:

• Converting traffic light control signals into C-V2X internal messages recognizable by RSU/OBU to implement SPAT applications. Autonomous vehicles are usually equipped with cameras and artificial intelligence to recognize traffic light information. However, recognition accuracy is easily affected by adverse weather or obstructions. This solution enhances robustness against any conditions that may hinder visual recognition.

• Utilizing artificial intelligence technology, which has demonstrated excellent performance in multiple fields, for VRUCW applications. Deep learning-based vulnerable road user detection and collision warning functions can be implemented through a PC5-based C-V2X system architecture.

• Integrating C-V2X into the autonomous driving system (ADS) to enhance safety. The ADS can monitor road conditions, detect potential problems, and take measures to avoid traffic accidents. The success of these projects will lay a solid foundation for the upcoming 5G NR-V2X.

I. Traffic Light Control System Integration: To implement SPAT applications locally, the system architecture shown in Figure 1 has been designed. The PC5-based C-V2X SPAT application has been successfully launched, where:

Figure 1. Traffic Light Control System Integration Architecture Diagram

• The system can directly collect traffic light information from the traffic light controller.
• The traffic light acquisition program is responsible for receiving roadside traffic light information; this includes traffic light phase, color, and remaining time, which are all sent to the roadside unit (RSU).
• The RSU reads this information and packages it into C-V2X protocol messages.
• The RSU broadcasts the C-V2X messages to the on-board unit (OBU) via the PC5 interface.
• The on-board unit (OBU) installed in the autonomous vehicle analyzes and filters this information, and then sends it to the autonomous driving system industrial PC (IPC) for deceleration or stop control.
• The user interface (UI) displays C-V2X technical information in an intuitive way.

II. VRUCW Application System Integration: The C-V2X VRUCW application based on PC5 is shown in Figure (2), where:

Figure 2. Schematic Diagram of VRUCW Integration System

• The VRUCW application can be considered a P2I2V service (Pedestrian-Infrastructure-Vehicle). IP cameras must be installed in the road area for line-of-sight (LOS) and non-line-of-sight (NLOS) monitoring.
• It uses an AI server equipped with a series of deep learning technologies (such as CNN (Convolutional Neural Network) and SSD (Single Shot Detector)). If any pedestrian passes through the camera's coverage area, the system will detect the object.
• The AI ​​server transmits the analysis results, including target recognition and motion prediction, to the Roadside Unit (RSU), which then broadcasts this information to all On-Board Units (OBUs) within its coverage area.
• The OBU is responsible for integrating vehicle information (such as speed, heading, and position) to determine if there is a collision risk. We use a target classification algorithm to determine the pedestrian's direction for subsequent calculation of the likelihood of a collision warning.
• Assuming there is a collision risk between the pedestrian and the vehicle, for example, if the distance between them is within 50 meters and the vehicle's speed exceeds 10 km/h, we trigger a collision warning through the algorithm.

III. Autonomous Driving System Integration: The integration of PC5-based C-V2X with the autonomous driving system is currently designed and implemented as shown in Figure (3), where:

Figure 3. Schematic Diagram of Autonomous Driving Integration System

• The Roadside Unit (RSU) receives information from the traffic light controller or the AI server. It then broadcasts this information within its coverage area using a predefined message format.
• The On-Board Unit (OBU) receives the broadcast messages through PC5-based C-V2X communication.
• The OBU connects to the industrial PC (IPC) of the autonomous driving system via the TCP/IP protocol. The OBU receives Global Navigation Satellite System (GNSS) and Controller Area Network (CAN) messages from the vehicle.
• The OBU uses advanced internal algorithms to determine if the situation is dangerous. It then sends corresponding warning messages to the autonomous driving system's IPC based on the situation.

At this point, the C-V2X technology is integrated into the autonomous driving system as expected.