Project
Energy Efficiency and Security in O-RAN
Putting sustainability at the heart of O-RAN through energy-saving mechanisms and digital twin simulations, enabling more efficient use of resources while reducing environmental impact.
Initial Approach
This project explores how telecommunications can evolve towards B5G/6G networks by addressing some of the industry’s most critical challenges, including cybersecurity, sustainability and scalability, while enabling interoperability and integration across multiple vendors.
With security as a core consideration, the project incorporates advanced frameworks to mitigate risks such as data poisoning, adversarial attacks and interface breaches. Real-time anomaly detection, secure-by-design AI/ML models and robust encryption mechanisms help protect the integrity of critical services.
In parallel, the project focuses on sustainability through energy-saving mechanisms and Digital Twin simulations, enabling more efficient resource management and reducing environmental impact.
By combining advanced technologies with close collaboration with industry leaders, the project develops future-ready solutions designed to meet the evolving demands of global connectivity.
Use case
The O-RAN Alliance places strong emphasis on security, with Working Group 11 (WG11) leading the development of specifications aimed at establishing a secure-by-design O-RAN architecture.
This work includes identifying potential security threats, assessing risks, defining security requirements and establishing specific protocols for different O-RAN components. WG11 also develops detailed testing specifications to ensure the robustness of these security measures.
One of the key outcomes of this work is the *O-RAN Security Threat Modeling and Risk Assessment* report, which identifies critical assets, defines threat models and establishes security principles alongside comprehensive risk assessments.
One such threat model is data poisoning of RICs. In this scenario, an operator may use ML models trained on manipulated data, leading to real-time decisions that could compromise network operations. Different security techniques have been developed to protect against data poisoning, including robust training algorithms, modified AI/ML architectures and proactive protection mechanisms.
However, while security is a priority, a fundamental question remains: how can O-RAN balance robust security measures with energy efficiency?
In this project, we analyse the relationship between techniques designed to protect against data poisoning and the energy cost associated with implementing these measures.
Innovation focus
Our innovation work on system-level security focuses on:
- Studying data poisoning attacks.
- Developing and assessing defence mechanisms against these attacks.
- Measuring the energy consumption associated with their use in AI-based systems.
Data poisoning, where malicious actors manipulate training data to compromise model integrity, represents a significant threat to modern networks. By analysing different defence techniques in depth, we aim to determine how effectively they protect against these attacks while also quantifying their energy requirements.
By assessing the trade-off between security robustness and energy efficiency, our work aims to identify strategies that can achieve the right balance between the two.
This approach not only strengthens the resilience of AI systems, but also helps make them more sustainable, generating knowledge that can support the deployment of secure, energy-efficient solutions in real-world environments.
Let’s talk
We look at every challenge in its real-world context to determine what’s worth building — and the right way to build it.