In short: A consortium led by Parallel Wireless, BT, and the Compound Semiconductor Centre has built modular Active Antenna Units under the Open Networks Ecosystem Competition that consume approximately 30 per cent less power than their conventional equivalents, removing one of the last credible objections to disaggregated radio networks.
Key Takeaways
- Open RAN does not have to mean higher energy bills — Project AURA's prototype antenna units achieved roughly 30% lower power consumption than the integrated equipment they replace
- Modular design enables targeted upgrades — operators can swap individual radio components without replacing the entire unit, extending hardware life and cutting embodied carbon
- The UK semiconductor supply chain is central to the result — the Compound Semiconductor Centre's GaN expertise drove the efficiency gains, keeping critical IP onshore
In a nutshell
The Energy Objection
For years, almost every boardroom conversation about Open RAN has included a version of the same question: "Doesn't all this disaggregation just use more power?" Traditional macro base stations are tightly integrated units where the vendor has optimised every thermal path and power rail — break that unit into modular, interoperable components and you might reasonably expect to lose some of that efficiency. This concern has given traditional equipment vendors a convenient talking point, and it has given operators genuine pause, because energy costs are quantifiable in a way that supply-chain diversification benefits are not.
Project AURA set out to answer the question with hardware, not slide decks.
What AURA Built
AURA was funded under DSIT's Open Networks Ecosystem Competition — the same programme that backed Aerix's ONE WORD project, the competition's largest funded effort at GBP 10 million. The consortium brought together Parallel Wireless, BT, and the Compound Semiconductor Centre in South Wales to design, fabricate, and test modular Active Antenna Units that conform to Open RAN interfaces whilst materially reducing power draw.
The key innovation sits in the power amplifier stage. AURA co-designed GaN power amplifiers with the digital front-end from the start, optimising envelope tracking and digital pre-distortion specifically for an O-RAN split-7.2x architecture. The Compound Semiconductor Centre's fabrication facility in Newport provided the GaN-on-SiC devices, keeping the entire semiconductor supply chain within the UK — a point whose importance we believe is easy to understate.
The result: antenna units consuming approximately 30 per cent less power than the conventional macro equipment they were benchmarked against. On a network of thousands of sites, that translates to tens of millions of pounds in annual electricity savings and a proportional cut in Scope 2 carbon emissions.
Modularity as a Sustainability Strategy
The power reduction is the headline, but modularity matters almost as much — and interestingly, it addresses a problem that rarely features in the Open RAN debate. In a conventional integrated AAU, if the PA technology moves on, the operator replaces the entire unit: housing, antenna elements, radome, mounting hardware, all into a skip. AURA's architecture allows the PA module to be swapped independently. When a more efficient amplifier generation arrives, the operator upgrades the module, not the unit. Over a 15-year site lifecycle, that approach could halve the total embodied carbon of the radio head. The trade-off is that modular designs introduce additional inter-module interfaces that must be maintained and tested, but we have seen that this engineering overhead is manageable relative to the sustainability benefit.
Why the UK Supply Chain Matters
The GaN-on-SiC devices at the heart of the efficiency gain were designed and fabricated in Wales. Open RAN's promise of vendor diversity is fundamentally only as real as the supply chain behind it — if every disaggregated radio unit depends on power amplifiers from a single overseas foundry, the supply-chain resilience argument rings hollow. AURA demonstrates that the UK has the semiconductor capability to anchor a critical layer of the Open RAN stack domestically.
BT's involvement signals that the operator community takes this seriously. The project tested the AAUs against BT's own network requirements, generating evidence that makes Open RAN a practical, economically rational choice. The energy objection was always the most dangerous criticism because it was the most quantifiable. Consequently, AURA has put a number on the answer — minus 30 per cent — and that number is difficult to argue with.
If you are planning a private 5G deployment and power consumption is a constraint, get in touch. You can read more about our work on open networks and the ONE WORD project on our about page.