Teaching the Network to Think: Smarter Fibre for the Future of Connectivity

How artificial intelligence is improving spectrum utilisation and boosting performance in next-gen optical systems


Lead Institution:

  • University of Bristol (Smart Internet Lab / HASC)
Project Partners:
  • Chalmers University of Technology, Sweden
  • Universidad Carlos III de Madrid, Spain
Supported by:
Challenge:
  • Traditional optical network algorithms are inefficient for managing multi-band spectrum.
Solution:
  • A deep reinforcement learning (DRL)-based framework that dynamically provisions services based on real-time network conditions and quality of transmission (QoT).
Impact:
  • Reduced service blocking by 35%–85%; improved spectral efficiency and scalability for 5G/6G readiness.

THE CHALLENGE: Today’s networks are under growing pressure

Today’s networks are under growing pressure. As emerging technologies like 5G, 6G, AR/VR, IoT, cloud services, and AI continue to evolve, they’re driving an explosion in data demand, placing unprecedented strain on existing fibre infrastructure. And this demand is only going to increase.

Traditional optical networks mostly operate within the C-band, but multi-band (MB) elastic optical networks (EONs) offer a powerful alternative. By unlocking additional spectral regions such as the L and S bands, MB-EONs expand the usable spectrum from around 5 THz to as much as 20 THz. So far, so good.

The challenge? Managing this expanded spectrum is becoming increasingly complex. Conventional algorithms simply aren’t up to the task. The result is a high rate of service blocking, where connection requests are denied or dropped, leading to poor user experience and significant inefficiencies in resource use.

THE SOLUTION: A new approach to spectrum management

A new approach to spectrum management is required. The team wanted to establish new ways of addressing four main issues:

  • Service blocking
  • Complex spectrum management
  • Quality of transmission (QoT)
  • Scalability

The team turned to AI and deep reinforcement learning (DRL), which is the same novel technology used to power things like autonomous driving algorithms and robots learning to walk, jump, or perform a backflip.

Reinforcement Learning (RL) is a type of machine learning where an (AI) agent learns to make increasingly better decisions by interacting with its environment. It gets rewarded for correct calculations and optimisations and penalties for bad decisions.

Deep learning (neural networks) allows the agent to handle more complex, high-dimensional environments.

KEY TECHNICAL INNOVATIONS: Applying DRL to network management

Applying DRL to network management, the team developed an innovative framework designed to do a number of things.

Learn and continually improve

The new framework learns by interacting with simulated networks, rewarding itself for successful provisioning. Inputs to the DRL include both route-level and band-level features, giving it a rich view of the whole network and current conditions. This allows it to adapt and make holistically informed decisions.

The DRL accelerates its own learning and optimises for long-term throughput (not just short-term wins). The innovative reward function enables it to learn effective strategies way beyond what traditional heuristics can achieve.

The system trains itself to get better over time instead of needing constant human fine-tuning.

Handle complex spectrum management

DRL handles the complexity of multiple bands and modulation formats that traditional algorithms simply cannot. It can dynamically allocate spectrum across MB-EONs, making more efficient use of the available spectrum.

Provide sufficient signal quality

The built-in QoT assessment model ensures that allocated resources provide sufficient signal quality, avoiding wasted or unusable connections. It does this by profiling the real performance of modulation formats per channel, so the algorithm makes physics-aware decisions. This means fewer wasted resources and ultimately more reliable connections overall.

Network adaptivity

AI adapts and scales with growing demand, unlike static rules.

THE RESULTS

  • Service blocking reduced by 35% – 85% compared with existing heuristic methods: Making networks significantly more efficient and capable of handling surging traffic.
  • Improved resource utilisation: Allocating spectrum more efficiently and with higher QoT guarantees.
  • Scalability: System adapts to increased traffic loads and network complexity.
  • Operational efficiency: Operators can serve more users and higher bitrates without physically laying more fibres (which is costly, disruptive and not always viable – especially in built-up areas). Boosting capacity in this way offers a more effective and sustainable way to scale.

This method avoids the need for costly, high-emission fibre deployment, aligning with the UK’s sustainability goals and net-zero strategy by maximising existing infrastructure efficiency. This innovation could benefit many real-word applications including 5G and 6G backhaul, high throughput networks and even data centre interconnects for hyper-scalers such as Google and Meta. Without solving these issues, users face slower speeds, dropped connections, and spiralling infrastructure costs. For users, this innovation means improved services such as consistently reliable streaming and improved mobile performance.

TIME TO DEPLOYMENT

  • Short term: Algorithm validated in simulation; lab prototypes are feasible now
  • Medium term (3–5 years): Field trials for 5G backhaul expected as MB optical systems evolve
  • Long term (5–10 years): Key enabler for 6G core network infrastructure and widespread MB-EON adoption

This is not something operators can roll out tomorrow, but it is an important stepping stone towards preparing optical infrastructure for 6G-scale traffic. By laying the groundwork now, operators will be ready for a future where 6G-scale traffic is not longer a challenge, but a significant opportunity.


“Proud to see how AI can reshape optical networking for the next generation of scalable, resilient infrastructures.” Shuangyi Yan (Associate Professor & Programme Director at University of Bristol)


TEAM & PAPERS

University of Bristol – Smart Internet Lab

  • Yiran Teng
  • Haiyuan Li
  • Shuangyi Yan
  • Dimitra Simeonidou

Chalmers University of Technology, Sweden

  • Carlos Natalino
  • Paolo Monti

Universidad Carlos III de Madrid, Spain

  • Farhad Arpanaei
  • Alfonso Sánchez-Macián

Supporting Projects

  • ECO-eNET – EU-SNS funded initiative for energy-efficient confluent networks
  • ALLEGRO EU – Horizon Europe project for ultra-low energy secure networks

Paper: DRL-Assisted QoT-Aware Service Provisioning in Multi-Band Elastic Optical Networks – https://ieeexplore.ieee.org/abstract/document/11131684

Download the case study: https://allspectrumhub.org/wp-content/uploads/2025/09/Case-Study-DRL-Assisted-QoT-Aware-Service-Provisioning_01-0925.pdf 


Want to stay involved? Keep an eye on our channels for updates, opportunities, and news from across the HASC community.

Sign up for our newsletter.

Spectrum Management: Challenges and Innovations in Future Connectivity

Spectrum is the invisible infrastructure of our connected world. Every time we send a message, stream a video, or connect a device, we are relying on carefully managed slices of spectrum – the defined radio and optical frequencies through which wireless digital information flows. But this space is becoming increasingly crowded as our demand for mobile data grows year on year. By 2030, Ofcom has estimated that UK mobile data demand may be 7.5 to 52 times the level of 2021, and as high as 19 to 540 times by 2035. Managing the spectrum has therefore become both a highly complex and increasingly urgent challenge.

What Is Spectrum Management?

Although the electromagnetic spectrum is theoretically continuous, only certain ranges are practical for communications. Very low frequencies cannot carry high data rates, while extremely high frequencies – such as X-rays or gamma rays – are unsafe or impractical. This means that the usable portion (from 300 kHz to 300 GHz) is a finite resource.

Spectrum management is the coordinated allocation and management of usable frequencies to ensure wireless communication services can operate efficiently and without interfering with each other. Regulators (such as Ofcom for the UK) assign specific frequency bands for designated purposes – such as mobile phone networks, public safety functions, and satellite services. Telecommunication providers bid for frequency band licenses, then build infrastructure to deliver services. 

Why Spectrum Management Matters

Communications infrastructure is now as essential to our societies as electricity or water. In 2023, the UK Government described spectrum as a ‘critical national asset’, with ‘more efficient and intelligent use of spectrum’ a priority for continued growth across many sectors.

But just as demand for mobile data is rising exponentially, available spectrum in the most practical frequency range is becoming scarce. Usage is also becoming highly concentrated; for instance, most (around 80%) of wireless traffic now occurs indoors, carried over Wi-Fi or indoor mobile coverage. This places disproportionate pressure on lower frequency bands that propagate more easily through walls and other obstacles.

Meanwhile, emerging technologies, such as the future 6G, are pushing into higher frequency bands, including terahertz, which are technically more challenging and typically demand line-of-sight transmission. As more frequency bands come on board, this adds to the increasing complexity of the network, already convoluted by many legacy systems.

Energy consumption is also an increasing concern. If the UK is to achieve its Net Zero targets, reducing the energy footprint of data delivery is a key priority – however, improving energy-efficiency does not always align with achieving best spectral efficiency.

Meeting these challenges will require coordinated innovation that integrates technological advances with industry leadership and a supportive policy framework. HASC works at the intersection of these areas, forging towards a vision of seamless, energy-efficient spectrum management for tomorrow’s communication networks.

Addressing Congestion Through New Spectrum Access

Through our Connectivity work stream, led by University College London, HASC is investigating novel technologies that can use new regions of the spectrum, including terahertz and optical wireless domains. The optical spectrum is particularly attractive; unlike radio frequencies, much of the optical spectrum is not subject to regulatory licensing, and offers around 3,000 times more bandwidth. To tap into this opportunity, HASC researchers at Cambridge are pioneering Li-Fi: a high-speed wireless technology that uses light instead of radio waves. Experimental systems have achieved ultra-fast data transmission speeds of 100 Gbps; future work will build scalable systems that can support next-generation applications, and reduce pressure on the congested radio spectrum.

University of Cambridge demonstrates Wavelength Division Multiplexing Li-Fi at 100 Gbps

In the wired domain, a promising innovation is hollow-core fibres, where light travels through a hollow (air-filled) core instead of solid glass. These have the potential to transmit data with lower loss and latency, enabling higher speeds and energy efficiency in high-capacity settings such as data centres. Meanwhile HASC researchers at the University of Oxford are exploring how to link wired and wireless communication networks using light, by developing virtual fibres to provide high-speed wireless connectivity indoors. Data carried over fibres via light is then directed by a base station in the ceiling to travel wirelessly through free-space to a mobile terminal and back into a wired connection.

Improvements can also still be made in the radio frequency domain. Reassigning bandwidth in underutilised frequencies and decommissioning legacy networks, such as 2G, both offer ‘low-hanging fruit’ to free up capacity. A key focus for HASC, meanwhile, is exploring how we can ‘engineer’ the propagation environment using intelligent surfaces to overcome the natural travel limitations of higher-frequency signals, which could free up pressure on lower frequencies. This could include, for instance, engineered panels that guide radio signals more effectively through complex environments.

Managing Complexity with Adaptive and Multi-Band Systems

HASC’s Adaptivity programme, led by the University of Bristol, is addressing how we can efficiently manage the increasing heterogeneity of communications networks, with fibre, Wi-Fi, mobile, and optical systems operating in parallel. This work includes developing algorithms and AI-enabled control frameworks that can intelligently assess the most effective connectivity mode in real time, taking into account factors such as congestion, bandwidth availability, interference, and energy consumption.

A core element of this work is multi-band communication, where different portions of the spectrum (terahertz, radio frequency, and optical) are used simultaneously to enhance throughput and resilience. For instance, during emergencies or natural disasters, adaptive networks could quickly shift traffic from damaged wired infrastructure to wireless alternatives, maintaining critical communications when they are needed most.

This approach already shows promise in optical systems. Recent HASC research into multi-band optical transmission has demonstrated that by using extra spectrum beyond the traditional C-band (such as L- and S-bands) fibre networks could quadruple their capacity. Managing this expanded spectrum is technically challenging, but by applying deep reinforcement learning (a branch of AI that learns by trial and error) the research team reduced service blocking by up to 85% compared with conventional techniques, while keeping decision times practical.

Quantum and Spectrum Management

Communications systems are becoming increasingly complex and ensuring these remain secure is an ongoing challenge. Integrating new quantum technologies within the spectrum used by current systems, however, could provide enhanced security benefits. Through the Security challenge, led by the University of Cambridge, HASC researchers are investigating quantum key distribution (QKD) for both wired and wireless communication systems. QKDs use quantum mechanics to generate cryptographic keys that immediately alter their state in response to attempts to intercept or observe them. This makes them theoretically immune to eavesdropping, since this would alert the network to the presence of the breach.

Improved Energy Efficiency in Data Transport

Many of these technologies that HASC is exploring have the added potential benefit of reducing energy needs besides improving connectivity. Intelligent reflecting surfaces, for instance, reduce wasted transmitted power from poorly-propagating signals, whilst AI-enabled control frameworks could automatically select the most energy-efficient communication mode for a given situation. New types of optical fibre, meanwhile, could transmit much higher volumes of data, improving energy usage. In combination, these advances could enable future networks to deliver vastly more data without a proportional increase in energy use.

Looking to The Future

Ultimately, the goal of seamless, efficient spectrum management is inextricable from HASC’s vision of combining wired and wireless internet technologies into a single resource in order to deliver intelligent ‘all-spectrum’ end-to-end connectivity. Only then can we design future-proof networks that can intelligently adapt to meet the current and ever-evolving data demands of modern societies. By integrating advances across fibre, wireless, AI, and quantum, HASC is ensuring the UK remains at the forefront of research-led innovation to manage this unseen, yet essential asset.