CBET-EPSRC: TECAN - Telemetry-Enabled Carbon Aware Networking
Lead Research Organisation:
UNIVERSITY OF OXFORD
Department Name: Engineering Science
Abstract
Addressing climate change is of existential importance. Although there are many contributing factors, energy production and consumption has one of the most direct impacts on the environment. A significant portion of this can be attributed to computer networks, with some reports estimating that they consume 1.5x the energy of all data centres. However, in contrast to other large scale compute infrastructure, accounting for the carbon emissions of the network is extremely hard. A traffic flow can cross multiple administrative domains in different regions of the world, be processed in devices powered by different energy sources, and share the energy consumption with tens of thousands of other traffic flows.
The goal of this project is to quantify, and subsequently reduce, the carbon emissions of computer networks. Concretely, it proposes a comprehensive research agenda that will: (1) identify metrics that cover the end-to-end environmental cost of data transmission; (2) leverage programmable network elements to collect and report the metrics by developing novel carbon-intelligent network telemetry solutions; and (3) design novel, carbon-aware network routing algorithms that utilize the measurements. Overall, the project will enable accurate accounting of computer networks' carbon emissions and develop novel routing algorithms to minimize the carbon emissions of the Internet, combating climate change.
If successful, the project will help us meet the targets established by the UK government, which has committed to achieving 'net-zero' by 2050. Specifically, it will help communication providers to minimize their carbon emissions; enable policy makers to reliably check that companies meet government's net-zero goals; and allow consumers to choose communications and ICT providers based on their sustainability claims.
The goal of this project is to quantify, and subsequently reduce, the carbon emissions of computer networks. Concretely, it proposes a comprehensive research agenda that will: (1) identify metrics that cover the end-to-end environmental cost of data transmission; (2) leverage programmable network elements to collect and report the metrics by developing novel carbon-intelligent network telemetry solutions; and (3) design novel, carbon-aware network routing algorithms that utilize the measurements. Overall, the project will enable accurate accounting of computer networks' carbon emissions and develop novel routing algorithms to minimize the carbon emissions of the Internet, combating climate change.
If successful, the project will help us meet the targets established by the UK government, which has committed to achieving 'net-zero' by 2050. Specifically, it will help communication providers to minimize their carbon emissions; enable policy makers to reliably check that companies meet government's net-zero goals; and allow consumers to choose communications and ICT providers based on their sustainability claims.
Organisations
- UNIVERSITY OF OXFORD (Lead Research Organisation)
- Siemens AG (Collaboration)
- Technical University Berlin (Collaboration)
- Ori Industries 1 Ltd (Project Partner)
- Open Networking Foundation (Project Partner)
- Yale University (Project Partner)
- BT plc (Project Partner)
- Intel Corporation (UK) Ltd (Project Partner)
People |
ORCID iD |
| Noa Zilberman (Principal Investigator) |
Publications
El-Zahr S
(2025)
Carbon-Intelligent Content Scheduling in CDNs
El-Zahr S
(2025)
From Measurement to Emissions: Assessing the Carbon Footprint of Traffic Flows
in Proceedings of the ACM on Measurement and Analysis of Computing Systems
Wang Z
(2025)
NetFridgeS: Enabling Dynamic Frequency Scaling on Network Switches through Carbon-Aware Routing
in Proceedings of the ACM on Networking
Zilberman N
(2025)
The Small World Web of AI
| Description | IETF121 - Real Time Telemetry for Carbon Aware Networking |
| Geographic Reach | Multiple continents/international |
| Policy Influence Type | Contribution to new or improved professional practice |
| Impact | IEFT/IRTF activity just started |
| URL | https://datatracker.ietf.org/meeting/121/session/hotrfc |
| Description | IRTF SUSTAIN |
| Geographic Reach | Multiple continents/international |
| Policy Influence Type | Contribution to new or improved professional practice |
| Impact | First workgroup activity in IETF-121, with upcoming meeting in every IETF meeting (three times a year). |
| URL | https://www.irtf.org/sustain.html |
| Description | IRTF SUSTAIN - Carbon Aware Networking Challenges and Opportunities |
| Geographic Reach | Multiple continents/international |
| Policy Influence Type | Contribution to new or improved professional practice |
| URL | https://datatracker.ietf.org/meeting/123/materials/slides-123-sustain-carbon-aware-networking-challe... |
| Description | CHEDDAR: Communications Hub For Empowering Distributed ClouD Computing Applications And Research |
| Amount | £3,028,049 (GBP) |
| Funding ID | EP/X040518/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 05/2023 |
| End | 05/2026 |
| Description | GeCo: Siemens-TUB-Oxford Seed Funding |
| Organisation | Siemens AG |
| Country | Germany |
| Sector | Private |
| PI Contribution | Developing scalable, robust and sustainable generative communication networks for optimised AI deployment across industrial applications through advanced algorithms, programmable networks, and carbon-aware infrastructure. We led the sustainability and programmable networks part of the collaboration. |
| Collaborator Contribution | Developing scalable, robust and sustainable generative communication networks for optimised AI deployment across industrial applications through advanced algorithms, programmable networks, and carbon-aware infrastructure. Siemens led the use case development. TU Berlin led the algorithms and theory part of the collaboration. |
| Impact | Not yet. Paper under submission for publication. |
| Start Year | 2025 |
| Description | GeCo: Siemens-TUB-Oxford Seed Funding |
| Organisation | Technical University Berlin |
| Country | Germany |
| Sector | Academic/University |
| PI Contribution | Developing scalable, robust and sustainable generative communication networks for optimised AI deployment across industrial applications through advanced algorithms, programmable networks, and carbon-aware infrastructure. We led the sustainability and programmable networks part of the collaboration. |
| Collaborator Contribution | Developing scalable, robust and sustainable generative communication networks for optimised AI deployment across industrial applications through advanced algorithms, programmable networks, and carbon-aware infrastructure. Siemens led the use case development. TU Berlin led the algorithms and theory part of the collaboration. |
| Impact | Not yet. Paper under submission for publication. |
| Start Year | 2025 |
| Title | CICS - Carbon-Intelligent Content Scheduling |
| Description | The algorithm described in this application optimizes the delivery schedule of content to CDN sites. Prepositioning content is common in CDNs especially for static content like movies, series and software updates. This algorithm embeds the concept of carbon-intelligence in the scheduling of this content delivery. |
| Type Of Technology | Webtool/Application |
| Year Produced | 2025 |
| Open Source License? | Yes |
| Impact | Not yet. |
| URL | https://github.com/ox-computing/CICS |
| Title | Measurements2Emissions |
| Description | Carbon tracing at the flow level requires visibility in the network. The carbon cost of the flow should be traced at each device along the path of this flow. This is an application for flow-level carbon estimation. The project has two parts: (1) a power measurements part done on three types of switches and (2) a simulation part using the ns-3 simulator tool to quantify the scale of carbon per flow in a real network topology. |
| Type Of Technology | Webtool/Application |
| Year Produced | 2025 |
| Open Source License? | Yes |
| Impact | Currently being implemented on JOINER for wider adoption |
| URL | https://github.com/ox-computing/Measurements2Emissions |
| Title | SWW-AI |
| Description | SWW, or the Small World Web, reimagines the web by sending prompts instead of content. It uses generative AI to turn the prompts into content on the user's device. This repository contains the code developed to demonstrate SWW, and included in its publication: Noa Zilberman and Alex Jackson. "The Small World Web of AI". ACM HotNets 2025. https://doi.org/10.1145/3772356.3772420 |
| Type Of Technology | Webtool/Application |
| Year Produced | 2025 |
| Open Source License? | Yes |
| Impact | Not yet. |
| URL | https://github.com/ox-computing/SWW-AI |
| Description | Dagstuhl Seminar 24402 Organisation |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Dagstuhl Seminar 24402 - Greening Networking: Toward a Net Zero Internet What is the path towards a Net Zero Internet? Reducing carbon emissions to slow climate change is one of mankind's greatest challenges. Networking technology is key to that challenge - as a provider for solutions (such as enabling teleworking and reducing travel), but also as a contributor to greenhouse gas emissions, e.g. through significant power consumption. It is thus becoming important to make networks themselves "greener" and devise solutions that result in less carbon-intensity while continuing to meet increasing network traffic demands and service requirements. Complicating the matter, energy consumption relates to CO2 emissions in ways that vary a lot by network operators (through the use of "clean" energy). Many of today's improvements are driven by general advances in computing hardware as well as in transmission technology (antennas, lasers). While this is where the biggest opportunities for energy reduction may lie, it is important to extend questions of greenness to other aspects of carbon reduction, both temporal and spatial, and to other layers in the networking stack - to the data and control plane, to routing and traffic forwarding, to the ways in which networks are organized and deployed. For example, can network protocols be designed in ways that make them more carbon-efficient, e.g. by routing along a path that is primarily powered by renewable energy, so as to reduce CO2 emissions with proper path selection? Can control planes be designed in a way that allows to rapidly decommission resources that may currently not be needed without compromising other important goals such as network resilience or security? How much overhead is introduced by current architectures and protocols for cloud computing, media streaming, and content distribution and what is the energy conservation potential of alternative systems? What protocol advances could enable greener networking solutions? How can networks be optimized not just for QoS or utilization but for low environmental impact? The goal of this Dagstuhl Seminar was to bring together researchers of different perspectives with a vested interest on this topic, exchange ideas, and explore possibilities for synergy in their respective research directions. We wish to identify and prioritize the most impactful networking improvement to reduce carbon emission, define action items for a carbon-aware networking research agenda, and foster/facilitate research collaboration in order to reduce carbon emissions and to positively impact climate change. This seminar was organized into sessions pertaining to these themes, with presentations from leading researchers drawn from the invitees. Researchers in the six main topics - devices, protocols, metric, measurement, architecture, management - are invited to participate. The outcome of the seminar was to identify mechanisms to quantify (at the device, protocol, architecture levels) the carbon footprint of a network. This involves considering measurement tools, metrics, and management of the network. We defined a roadmap to deploy these tools into practical network settings and into standardization efforts. Additionally, we considered the key research items that would yield the most impact. |
| Year(s) Of Engagement Activity | 2024 |
| URL | https://www.dagstuhl.de/en/seminars/seminar-calendar/seminar-details/24402 |
| Description | Engineers without borders - Carbon Intelligent Computing: Saving the Planet Bit-by-Bit |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Undergraduate students |
| Results and Impact | A seminar for Engineers without Borders - Oxford branch. This student organisation aims to embed globally responsible engineering alongside what they learn in the lecture hall, connecting students across the university. About 15 students attended the seminar, presented by Prof Zilberman, which was followed by an hour long discussion of considerations, potential internships, ways to act and more. |
| Year(s) Of Engagement Activity | 2024 |
| URL | https://www.ewb-uk.org/inspire/chapters/ewb-oxford/ |