What is it about?
Modern connected vehicles—from self-driving cars to smart taxis—rely heavily on streaming video, real-time navigation, and constant data sharing. This puts immense pressure on the network servers that deliver all that multimedia content. Traditionally, these servers are managed with rigid, static rules, leading to wasted electricity during quiet periods and overwhelmed, slow performance during traffic rushes. We developed an intelligent, self-correcting control system that acts like a smart thermostat for the entire data network. It continuously monitors how busy each server is, predicts upcoming traffic spikes before they happen, and automatically redistributes workloads or powers servers up and down to match the current demand. In our real-world simulations, this approach reduced energy consumption by up to 30%, ensured all servers shared the workload evenly, and successfully handled over 98% of all vehicle requests—even during the most extreme traffic peaks. The result is smoother video streaming, faster responses, and greener, more sustainable networks for the smart cities of tomorrow.
Featured Image
Photo by Soumik Ghosh on Unsplash
Why is it important?
What makes this work unique and timely? Today's connected vehicle networks are drowning in multimedia data—4K video streams, real-time navigation updates, and vehicle-to-vehicle safety communications—but the servers handling this traffic are managed using outdated, static rules. Most existing systems either leave servers running at full power even during quiet periods (wasting energy) or fail to scale up quickly enough during traffic surges (causing frustrating lag and buffering). Our work introduces a self-correcting "smart thermostat" for the entire multimedia server network. What makes it truly unique is the use of a PID controller—a proven engineering control mechanism—adapted for the first time to simultaneously manage both server load balancing and energy consumption in real time. Unlike traditional approaches that react to problems after they occur, our system continuously monitors server performance, predicts future demand using a lightweight machine-learning algorithm (NLMS), and proactively adjusts resources before bottlenecks form. This breakthrough is exceptionally timely for three reasons: The 5G/6G rollout is dramatically increasing the volume and speed of vehicular multimedia traffic, making static management approaches obsolete. Sustainability mandates from governments and corporations are demanding greener, more energy-efficient network infrastructures. Autonomous vehicle deployment requires ultra-reliable, low-latency multimedia delivery that legacy systems simply cannot guarantee. The difference it makes: Our framework delivers tangible, measurable improvements: up to 30% reduction in energy consumption, near-perfect load distribution across all servers, and the ability to successfully handle over 98% of vehicle requests—even during extreme traffic peaks (1,500+ simultaneous vehicles). This means streaming services run smoothly, navigation responds instantly, and network operators save significantly on electricity and cooling costs. Perhaps most importantly, this is not just a theoretical model—we validated our approach in a realistic testbed using actual servers, OpenFlow switches, and SUMO traffic simulation, demonstrating that it works in practice, not just in theory. This makes our framework a practical, deployable solution for the smart cities of tomorrow, directly addressing the urgent need for sustainable, high-performance multimedia delivery in next-generation vehicular networks.
Perspectives
As the author of this work, this research carries a deeply personal significance that goes far beyond technical achievement. My journey into this area began not in a laboratory, but during a long drive through a congested city, stuck in traffic while trying to stream a navigation update that kept buffering. I realized that while we talk endlessly about 5G speeds and autonomous vehicles, we rarely discuss the invisible backbone—the servers and network switches—that actually make these experiences possible. And even more rarely do we discuss how much energy these systems waste when they're poorly managed. What truly motivated me was a simple, almost embarrassing observation: I noticed that the servers running our multimedia networks were often running at full capacity even when few vehicles were on the road. It struck me as profoundly illogical—like keeping every light in a building blazing even when most rooms are empty. This seemed like such an obvious waste, yet it was the accepted norm in the industry. I couldn't shake the feeling that we could do better, and that better didn't require inventing entirely new technologies—it required applying proven control theory in a smart, adaptive way. The "eureka moment" came when I realized that a humble PID controller—the same technology that regulates temperature in my home thermostat—could be repurposed to manage server temperatures and loads simultaneously. This wasn't about flashy AI or complex algorithms; it was about elegant simplicity. Watching the simulations run for the first time, seeing the server loads balance themselves automatically and energy consumption drop by 30%, was genuinely exhilarating. It felt like watching a chaotic system finally find its rhythm. What I love most about this work is its direct, tangible impact. This isn't abstract theory—it's practical, deployable technology that can reduce electricity bills, lower carbon footprints, and make multimedia streaming smoother for millions of drivers. Every time I see a news report about smart cities or electric vehicle infrastructure, I feel a personal connection to this work, knowing that the underlying network management challenges are being addressed. If I could speak directly to readers, I would say this: Energy efficiency in networking isn't boring—it's urgent and it's solvable. We don't need to wait for 6G miracles or quantum computing breakthroughs. Sometimes the most powerful solutions come from adapting proven tools in clever ways. I hope this paper inspires other researchers and engineers to look at their own fields with fresh eyes, asking not "what new technology do we need?" but "what existing tools are we underutilizing?" Because sometimes, the answer is right in front of us, hiding in plain sight.
AhmadReza Montazerolghaem
University of Isfahan
Read the Original
This page is a summary of: Energy-Efficient Software-Defined 5G/6G Multimedia IoV: PID Controller-Based Approach, IEEE Transactions on Vehicular Technology, June 2026, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/tvt.2025.3649936.
You can read the full text:
Contributors
The following have contributed to this page
