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

arXiv:2606.13390 · cs.NI · Submitted 2026-06-11 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Computer Science. The day in computer science: software, systems and theory on arXiv.

Tom: We're looking at the paper: "Experimental Insights into UDP-Based Video and Control Traffic over IEEE 802.11p ITS-G5".

Jane: Here's the gist of the paper. Experimental evaluation of UDP transmission over IEEE 802.11p using a real ITS-G5 testbed comprising Raspberry Pi–based OBUs and commercial RSUs shows that the modulation scheme is the main factor affecting delay under low traffic load Higher-order modulations significantly reduce latency and delay variability compared with BPSK As traffic load increases, the choice of transmission mode becomes more relevant, with IPv4 unicast providing more stable and predictable delay than IPv6 multicast The evaluation of CAKE queue management did not reveal measurable improvements in delay stability in the tested scenarios.

Tom: Experimental Setup and Testbed Configuration The study utilized a testbed composed of Raspberry Pi–based onboard units and commercial roadside units to evaluate UDP transmission over an IEEE 802.11p ITS-G5 environment The hardware setup involved the Raspberry Pi 5 as the primary platform, interfacing with a Mikrotik R11E-5HND wireless card operating in the 5.9 GHz ITS band Roadside Units were manufactured by Movyon Electronics and fully complied with European ITS-G5 (IEEE 802.11p) standards Video experiments involved capturing video at 1280×720 pixels at 30 frames per second, encapsulated in MPEG-TS over UDP Analysis of Traffic Performance The analysis examined the impact of different modulation and coding schemes (MCS), transmission modes, and network-layer configurations on latency performance Low traffic load (2 Mbps) showed that BPSK exhibits the highest delay, QPSK shows intermediate values, and higher-order modulations such as 16QAM achieve the lowest latency At 4 Mbps, delay oscillations became visible for BPSK, which is more pronounced in IPv6 multicast due to the lack of link-layer retransmission and reliability mechanisms Under high load (20 Mbps), 16QAM maintained significantly lower and more stable latency thanks to its higher data rate, while BPSK experienced abrupt truncation likely indicating increased packet loss Impact of Network Layer and Queue Management The comparison between IPv4 unicast and IPv6 multicast demonstrated that IPv4 unicast exhibited more stable and predictable latency than IPv6 multicast under congestion The results indicated that the CAKE does not introduce a measurable latency penalty, as observed results for both configurations appeared closely aligned Furthermore, the evaluation showed that queue management alone cannot compensate for the limitations introduced by lowrate modulation and multicast delivery The findings confirm that optimizing PHY-layer parameters and transmission strategies appears more effective than relying solely on active queue management mechanisms in ITS-G5 scenarios Future work will investigate more complex and dense vehicular scenarios, including heterogeneous traffic patterns and adaptive transmission strategies.

Jane: The paper concludes that the primary bottleneck lies in the physical layer and transmission characteristics rather than in buffer-induced queuing.

Paper summary: Tom: So we’re looking at this paper now, "Experimental Insights into UDP-Based Video and Control Traffic over IEEE 802 point 11p ITS-G5," and the main point here is that they finally put a real testbed together to see how UDP traffic actually behaves in this vehicular communication setup <ref:2606.13390#pg1,Experimental Insights into UDP-Based Video and Control Traffic over IEEE 802>. Jane Exactly, it’s about testing things like cooperative driving and teleoperation where low latency is everything, right? Tom Right. The paper claims that the modulation scheme, or MCS, is the biggest thing messing with delay under low traffic loads. Lu It seems they found that higher-order modulations significantly cut down both latency and how much it jumps around compared to using BPSK Meng Which is interesting because BPSK is usually the simplest thing you start with in communication systems.

Jane: And as the traffic load gets higher, things shift, and the choice of transmission mode starts becoming more important for stability. Tom That’s what they found too. They looked at IPv4 unicast versus IPv6 multicast under high load and said that IPv4 unicast gave them a more stable delay than IPv6 multicast Lu Which makes sense because UDP traffic, especially when you have multicast, doesn't have the same reliability mechanisms on the link layer to smooth things out Meng

Jane: But they also looked at active queue management using something called CAKE, and they didn't find any measurable improvement in delay stability in the scenarios they tested Tom That’s a bit disappointing because you expect some kind of buffer management technique to help with queuing issues. Lu They even showed that the queue management alone couldn't fix the problems caused by low-rate modulation or multicast delivery, which points back to the physical layer being key Meng

Jane: So, when you look at the whole picture of this paper, what does it actually mean for people who aren't deep in wireless networking? Tom It means that if you’re building a system relying on UDP for video or control signals in a vehicle, focusing on the physical layer parameters and your modulation strategy is way more impactful than just tuning the queue management software Lu It confirms that the bottleneck isn't usually the buffer itself, but how fast the signal can actually get through cleanly Meng

Jane: So, thinking about these authors Antonio Solida and Gaetano Orazio Cauchi and their team, what’s their main message when they wrap up this study on UDP transmission over IEEE 802 point 11p ITS-G5 <ref:2606.13390#pg1,over IEEE 802.11p ITS-G5>? Tom They are essentially saying that for real-time vehicular applications, you have to treat the modulation choice as a critical part of your design, not just a minor setting Lu It’s about making sure you pick something like 16QAM when traffic is low because it gives you that lower latency they measured Meng <ref:2606.13390#pg3>

Jane: And for the listener who isn't an engineer, what should they take away from this paper on UDP-Based Video and Control Traffic over IEEE 802 point 11p ITS-G5 <ref:2606.13390#pg1,UDP-Based Video and Control Traffic over IEEE 802.11p ITS-G5>? Tom The main thing is that if you’re dealing with real-time data in a connected vehicle environment, don't just assume the best software handles everything; you need to understand how the signal itself interacts with the channel Lu The paper shows that picking a better modulation scheme makes a real difference in how fast things move when traffic is light, and it stays stable when traffic gets heavy Meng

Jane: It seems like this work really zeroes in on the physical layer limitations, suggesting that software fixes for queuing aren't going to solve the core problem here Tom And that’s a big signal because it tells us where to focus our optimization efforts next Lu This sets up nicely for what we look at next, which is how this all ties back into the broader picture of vehicular communication systems.

Conclusion: Tom: So we’ve looked at how this paper tests UDP video and control traffic over IEEE eight hundred two point 11p in an ITS-G5 testbed, and now we're wrapping up with the big picture on what they found <ref:2606.13390#pg2>.

Jane: Right, so the main thing is that they didn't just look at one setup; they tested different things like video and control traffic across a real testbed with Raspberry Pi units and roadside units.

Lu: And what they really zero in on is that for UDP, the physical layer—the modulation—is way more important than just having good queue management software to handle the delays.

Meng: So even if you have the best buffer management system, it can't fix problems caused by how fast the signal itself is getting through.

Tom: Exactly. The paper shows that choosing a higher-order modulation, like sixteen QAM, really cuts down on both latency and how much that delay jumps around compared to using just BPSK when traffic is low.

Jane: And then as the traffic gets heavy, they found that IPv4 unicast gives you more stable delay than IPv6 multicast because of how the link layer handles things without any extra reliability bits.

Lu: The authors also checked something called CAKE queue management, and they found zero measurable improvement in delay stability across the scenarios they tested.

Meng: It means you can't just tune your software to fix every single queuing issue when you’re dealing with these low-rate physical layer limitations.

Tom: So what this paper really tells us is that optimizing the physical layer and transmission strategy is much more effective than relying only on active queue management mechanisms in ITS-G5 scenarios.

Jane: It confirms that the bottleneck isn't usually the buffer itself, but how fast the signal can actually get through cleanly.

Lu: The implication here is that for real-time vehicular applications, you need to focus heavily on picking the right modulation and transmission strategy from the start.

Tom: This suggests that future work should really look at more complex traffic patterns and adaptive transmission strategies in these dense vehicular environments.

Antonio Solida, Gaetano Orazio Cauchi, Salvatore Iandolo, Marco Savarese, Martin Klapez, Maurizio Casoni, Carlo Augusto Grazia

University of Modena and Reggio Emilia

cs.NI

Submitted: 2026-06-11

Updated: 2026-06-11

Comments: Accepted for publication at the V2X/NTN Workshop of the IEEE 2026 International Conference on Smart Applications, Communications and Networking (SmartNets 2026)

DOI: 10.1109/SmartNets69662.2026.11604774

Code: https://github.com/AlfaSierra92/DCC-over-802.11p

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 64/100

The gist: Experimental evaluation of UDP transmission over IEEE 802.11p using a real ITS-G5 testbed comprising Raspberry Pi–based OBUs and commercial RSUs shows that the modulation scheme is the main factor…

Key concepts

IEEE 802.11p ITS-G5
This is the specific wireless communication standard used for Intelligent Transportation Systems (ITS) in Europe. It defines how devices like vehicles and roadside units communicate wirelessly, allowing for reliable data exchange in smart road environments.
Modulation Scheme (MCS)
This refers to the way data is encoded onto a radio signal. Different schemes, like BPSK or 16QAM, determine how much information can be sent per transmission. The paper found that choosing a higher-order modulation helps reduce latency and delay variability.
IPv4 Unicast vs IPv6 Multicast
Unicast means sending data directly from one source to one specific destination, which the study found more stable for delay. Multicast sends the same data to many devices at once. The paper noted that multicast caused more delay oscillations compared to unicast under certain traffic conditions.
CAKE Queue Management
This is a system used to manage and control queues of data packets waiting to be transmitted. The research tested its effectiveness, but it concluded that CAKE did not provide any measurable improvement in delay stability in the tested scenarios.

Terminology

Summary

Experimental evaluation of UDP transmission over IEEE 802. p. 011p using a real ITS-G5 testbed comprising Raspberry Pi–based OBUs and commercial RSUs shows that the modulation scheme is the main factor affecting delay under low traffic load Higher-order modulations significantly reduce latency and delay variability compared with BPSK As traffic load increases, the choice of transmission mode becomes more relevant, with IPv4 unicast providing more stable and predictable delay than IPv6 multicast The evaluation of CAKE queue management did not reveal measurable improvements in delay stability in the tested scenarios. p. 5

Experimental Setup and Testbed Configuration

The study utilized a testbed composed of Raspberry Pi–based onboard units and commercial roadside units to evaluate UDP transmission over an IEEE 802.11p ITS-G5 environment The hardware setup involved the Raspberry Pi 5 as the primary platform, interfacing with a Mikrotik R11E-5HND wireless card operating in the 5. p. 19 GHz ITS band Roadside Units were manufactured by Movyon Electronics and fully complied with European ITS-G5 (IEEE 802. p. 211p) standards Video experiments involved capturing video at 1280×720 pixels at 30 frames per second, encapsulated in MPEG-TS over UDP

Analysis of Traffic Performance

The analysis examined the impact of different modulation and coding schemes (MCS), transmission modes, and network-layer configurations on latency performance

** Low traffic load (2 Mbps) showed that BPSK exhibits the highest delay, QPSK shows intermediate values, and higher-order modulations such as 16QAM achieve the lowest latency **

** At 4 Mbps, delay oscillations became visible for BPSK, which is more pronounced in IPv6 multicast due to the lack of link-layer retransmission and reliability mechanisms **

** Under high load (20 Mbps), 16QAM maintained significantly lower and more stable latency thanks to its higher data rate, while BPSK experienced abrupt truncation likely indicating increased packet loss **

Impact of Network Layer and Queue Management

The comparison between IPv4 unicast and IPv6 multicast demonstrated that IPv4 unicast exhibited more stable and predictable latency than IPv6 multicast under congestion The results indicated that the CAKE does not introduce a measurable latency penalty, as observed results for both configurations appeared closely aligned Furthermore, the evaluation showed that queue management alone cannot compensate for the limitations introduced by lowrate modulation and multicast delivery

Conclusion on Optimization Strategies

The findings confirm that optimizing PHY-layer parameters and transmission strategies appears more effective than relying solely on active queue management mechanisms in ITS-G5 scenarios Future work will investigate more complex and dense vehicular scenarios, including heterogeneous traffic patterns and adaptive transmission strategies. The paper concludes that the primary bottleneck lies in the physical layer and transmission characteristics rather than in buffer-induced queuing. p. 5

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve latency-sensitive vehicular applications such as cooperative driving, teleoperation, and real-time perception by implementing a system that dynamically selects higher-order modulation schemes like 16QAM when traffic loads are low to achieve the lowest latency, as Higher-order modulations significantly reduce latency and variability.

  2. Develop network protocols for congestion management that prioritize IPv4 unicast over IPv6 multicast configurations in congested conditions to ensure more stable and predictable delay, as IPv4 unicast exhibits more stable and predictable latency than IPv6 multicast.

  3. Enhance the performance of remote driving and teleoperation systems by integrating a CAKE active queue management algorithm into the network stack, although the paper notes that active queue management does not seem to improve delay predictability under tested conditions.

  4. Create video streaming subsystems capable of maintaining low and stable end-to-end latency for real-time video streams over IEEE 802.11p by utilizing higher-order modulation schemes, specifically 16QAM, which maintains the lowest and most stable delay values.

Related papers