A new study from the cross-industry UK Spectrum Policy Forum (UKSPF), which was conducted by Smith Institute and Spectrivity, has recommended that the telecoms regulator, Ofcom, should adopt Artificial Intelligence (AI) technologies to help manage the nation’s finite wireless radio spectrum more efficiently.
The think-tank highlights how mobile networks, smart devices, and emerging technologies, like 6G based mobile broadband, are pushing current spectrum management strategies to their limits and that more innovation is needed to “prevent congestion” and maximise efficiency. “Traditional spectrum management methods often struggle to adapt to the dynamic needs of modern wireless networks, leading to inefficient use of valuable spectrum resources,” said the UKSPF.
According to the UKSPF, AI technologies could be used to “revolutionise spectrum management” by “optimising resource usage, reducing interference, and enhancing network performance“. This would come from leveraging data-driven decision-making and using AI to “dynamically assign spectrum based on real-time demand“. Some previous attempts to do the latter have had mixed success.
UKSPF – Key Recommendations
➤ AI for licensing applications: Large Language Models (LLMs) and predictive machine learning could be employed to streamline application reviews and assess potential interference risks, reducing manual workload and improving decision-making efficiency.
➤ AI for monitoring and compliance: AI-driven sensor placement and proactive monitoring could enable a shift from reactive to predictive compliance enforcement, ensuring better adherence to licensing terms.
➤ AI for spectrum sharing and interference management: Advanced machine learning models, including deep reinforcement learning, could enhance spectrum sharing strategies, minimising interference and optimising resource allocation.
➤ Synthetic data and international insights: AI could generate synthetic datasets based on international proxy data, addressing gaps in UK-specific information and improving predictive accuracy.
➤ Simulations and digital twin development: The study advocates for the creation of realistic RF environment simulations and, in the long term, a comprehensive digital twin to test and refine new policies and technologies before real-world implementation.
Matthew Evans, Director of Markets and Chief Operating Officer at techUK, said:
“The UK has the opportunity to drive a new wave of spectrum management. By embracing AI, regulators can improve flexibility, ensure efficient spectrum utilisation, and support the continued expansion of wireless technologies that power our digital economy.
This report gives clear guidelines to regulators and industry on how they can use AI-driven solutions to drive efficiency for the telecommunications sector.”
At the time of writing, the full report hadn’t yet been published (it should be by the time you read this), although we’d agree with the broad thrust that there is potential here for AI to help enhance UK spectrum management and regulatory oversight. At the very least this is something that Ofcom should explore and trial, before deciding upon whether the pros outweigh the cons.