Options
2025
Conference Paper
Title
Collective Perception for Autonomous Shuttles via 5G NR Sidelink
Abstract
Infrastructure-based perception significantly enhances autonomous driving by providing a broader and more comprehensive environmental understanding than vehiclemounted sensors alone. This paper presents a hybrid system architecture developed within the Shuttle2X project [1], which connects autonomous shuttles and intelligent infrastructure via Vehicle-to-Everything (V2X) technologies. The study focuses on the communication aspects of infrastructure-supported perception, contributing to the Collective Perception Message (CPM) standard and the implementation of 5 G NR sidelink communication to improve safety, efficiency, and reliability in autonomous driving. Due to the limited flexibility of fixed infrastructure setups, including sensors, communication units, and edge computing devices, a structured prototyping process is employed. It begins with a mobile infrastructure configuration, where the complete sensor suite, along with communication and edge computing units, remains portable for iterative testing and optimization. In the second phase, sensors are permanently mounted at key locations, while communication and processing units remain mobile to maintain adaptability during evaluation. The system also integrates a V2X-controlled digital traffic sign that dynamically adjusts signaling based on the current traffic situation. CPMs compliant with ETSI TS 103324 are generated from detected object lists and broadcast via a Keysight WaveBee RSU using 5G NR sidelink. On the vehicle side, a WaveBee Plus onboard unit receives and processes the CPMs, enriching the vehicle's situational awareness. The system enhances CPM v2.1.1, integrates 3GPP Release 16 -compliant sidelink functionality within the OpenAirInterface platform, and has undergone safety validation and RSSI-based sidelink reliability testing. These contributions advance cooperative perception and communication standards for intelligent transportation systems.
Author(s)