Experimental Insights into UDP-Based Video and Control Traffic over IEEE 802.11p ITS-G5

An IMRAD digest of testbed measurements on modulation, IP configuration, and queue management for latency-sensitive vehicular communication

Antonio Solida; Gaetano Orazio Cauchi; Salvatore Iandolo; Marco Savarese; Martin Klapez; Maurizio Casoni; Carlo Augusto Grazia· 2026· DOI 10.48550/arXiv.2606.13390

The core problem

Vehicular applications such as cooperative driving, teleoperation, and real-time perception depend on low-latency wireless communication. ITS-G5, based on IEEE 802.11p, is a key technology for direct vehicle-to-vehicle and vehicle-to-infrastructure communication. Despite its relevance, experimental studies of UDP-based traffic over IEEE 802.11p under realistic conditions remain limited. This paper addresses that gap with an experimental evaluation of UDP transmission over an IEEE 802.11p ITS-G5 testbed composed of Raspberry Pi-based onboard units and commercial roadside units. The study investigates the impact of different modulation and coding schemes (MCS), evaluates two network-layer configurations (IPv4 unicast and IPv6 multicast), and assesses the use of CAKE for active queue management. In addition to synthetic traffic generated with iPerf, the evaluation includes real-time video streaming using MPEG-TS over UDP to emulate latency-sensitive vehicular applications.

Innovation

The results show that the modulation scheme is the dominant factor influencing latency at low traffic loads. Higher-order modulations significantly reduce latency and variability. Under congested conditions, the choice of transmission mode and IP version becomes increasingly significant. IPv6 multicast exhibits greater delay dispersion than IPv4 unicast. Active queue management with CAKE does not seem to improve delay predictability. These findings are consistent across synthetic iPerf traffic and real-time MPEG-TS video streaming over UDP. In qualitative terms, the observed latency behavior can be summarized as: L = f (MCS, load, IP mode, AQM) where at low load the MCS term dominates, while at high load the load, IP mode, and AQM terms become more influential. The experiments indicate that higher-order MCS reduces both mean latency and its variance, whereas IPv6 multicast increases delay dispersion relative to IPv4 unicast. CAKE does not appear to reduce delay unpredictability in this testbed.

Vehicular applications such as cooperative driving, teleoperation, and real-time perception depend on low-latency wireless communication. ITS-G5, based on IEEE 802.11p, is a key technology for direct vehicle-to-vehicle and vehicle-to-infrastructure communication. Despite its relevance, experimental studies of UDP-based traffic over IEEE 802.11p under realistic conditions remain limited. This paper addresses that gap with an experimental evaluation of UDP transmission over an IEEE 802.11p ITS-G5 testbed composed of Raspberry Pi-based onboard units and commercial roadside units. The study investigates the impact of different modulation and coding schemes (MCS), evaluates two network-layer configurations (IPv4 unicast and IPv6 multicast), and assesses the use of CAKE for active queue management. In addition to synthetic traffic generated with iPerf, the evaluation includes real-time video streaming using MPEG-TS over UDP to emulate latency-sensitive vehicular applications.

The experimental platform is an IEEE 802.11p ITS-G5 testbed with Raspberry Pi-based onboard units and commercial roadside units. The authors vary the modulation and coding scheme (MCS) to study its effect on latency and variability. They compare two network-layer configurations: IPv4 unicast and IPv6 multicast. They also evaluate CAKE as an active queue management mechanism. Traffic generation includes synthetic UDP flows via iPerf and real-time video streaming using MPEG-TS over UDP to emulate latency-sensitive vehicular applications. The evaluation focuses on latency and delay dispersion under different traffic loads and configurations.

Why it matters

The experimental evidence suggests that for latency-sensitive vehicular services over ITS-G5, the modulation and coding scheme should be selected to minimize latency and variability, especially when traffic loads are low. As congestion increases, the network-layer configuration becomes more important: IPv4 unicast provides lower delay dispersion than IPv6 multicast in this testbed. The lack of improvement from CAKE active queue management indicates that delay predictability may be governed more by radio-level and IP-mode factors than by queue management in this setup. These practical insights can inform the configuration of ITS-G5 networks supporting cooperative driving, teleoperation, and real-time perception. The study is limited to the specific testbed and traffic patterns used, so generalization to other hardware, environments, and traffic mixes requires further experimental validation.

Who should read this

CS practitioners and researchers

Opening member content…