ÌÇÐÄVlog

Scientific questions & interdisciplinary approaches

AI-RAN optimisation: How can we integrate advanced AI/ML architectures—including Foundation Models, Generative AI, and Neurosymbolic AI—directly into the Radio Access Network (AI-RAN) to enable real-time, self-evolving resource management?

Goal-Oriented semantic communication: How can task-oriented semantic communication (SemComs) be engineered to drastically compress transmission bandwidth while maintaining strict context and message fidelity across multi-user environments?

Decentralised & edge intelligence: How can we orchestrate decentralised optimisation models—such as Split Learning, Swarm Learning, and peer-to-peer Federated Learning—to coordinate resource allocation at the edge without sacrificing security or signalling budget?

Network digital twins: How can real-time, data-driven virtual mirrors be deployed alongside physical architectures to safely simulate, predict, and implement risk-free configuration policies?

Metasurface wave-domain processing (RIS/SIM): How can multi-layer Stacked Intelligent Metasurfaces (SIM) and Beyond-Diagonal RIS (BD-RIS) be mathematically modelled and optimised to perform wave-domain analog precoding, near-field holographic beamforming, and interference-free integrated sensing and communication (ISAC) directly in the electromagnetic wave regime?

Obstacles & target environments

Training & Latency Bottlenecks: Massive data requirements and lengthy training intervals typical of deep neural networks, which conflict with real-time, mission-critical 6G latency boundaries.

Heterogeneity & Scale Complexity: Extreme architectural scale, multi-domain network heterogeneity, and the non-convex optimization complexity of mapping real-time system variables.

Imperfect network realities: System vulnerability to imperfect channel state information (ICSI), non-linear impairments, and processing limits inherent to resource-constrained edge hardware.

Metasurface & electromagnetic hardware constraints: High RF chain power dissipation, near-field signal blockage vulnerabilities, inter-element mutual coupling, and real-time phase-shift configuration constraints.

Target Environments: Highly dynamic Open-RAN (O-RAN) ecosystems, intelligent edge-cloud cooperative frameworks, near-field multi-user cellular hotspots, and green, low-cost/low-RF-chain base station deployments.

Expected outputs

Self-Configuring AI-RAN frameworks: Predictive radio access frameworks capable of forecasting traffic patterns, minimising over-provisioning, and dynamically allocating spectrum on-the-fly.

Task-oriented semantic protocol layers: Intelligent, symmetric encoder-decoder networks (e.g., combining transformers and lightweight architectures) that dynamically switch modes based on statistical channel conditions to balance performance and complexity.

Privacy-preserving decentralised toolkits: Robust, hardware-agnostic machine learning modules tailored for real-time edge processing and distributed load balancing.

Autonomous digital twin mirrors: End-to-end multi-domain simulation platforms providing actionable scripts for self-optimisation, self-healing, and self-configuration of physical network nodes.

Programmable wave-domain optimisation suites: High-efficiency wave-based analog precoders and alternating optimisation routines that maximise sum rates, generate precise sensing beams, and map outage performance for multi-sector BD-RIS/SIM architectures while significantly slashing digital power demands.

Scientific questions & interdisciplinary approaches

Joint communication & positioning (ISAC): How can joint communication and positioning techniques be optimised at the waveform and receiver levels for high-dynamics LEO satellite systems?

AI-driven mobility & handovers: How can machine learning engines be used to make intelligent, data-driven handover decisions between satellite and terrestrial networks without disrupting service?

Semantic & cooperative communications: How can semantic communication (SemComs) be coupled with multi-satellite cooperative massive MIMO to dramatically compress transmission data rates while preserving context and message fidelity?

3GPP standards convergence: How do we orchestrate radio resource management and network orchestration to seamlessly unify Terrestrial Networks (TN) and Non-Terrestrial Networks (NTN) into a standardised access continuum?

Obstacles & target environments

High-velocity kinematics: Severe Doppler shifts, fast-changing time-varying propagation conditions, and unequal propagation latencies native to high-speed LEO/NGSO systems.

Hardware & architectural constraints: Local oscillator phase noise diffusion, the need for tight multi-satellite synchronisation, and computational complexity over limited onboard satellite processing power.

Spectrum coexistence: Coexistence challenges between GNSS and 5G signals, along with severe inter-satellite and multi-user interference.

Target environments: Integrated Space-Terrestrial Networks (STNs), direct-to-device (D2D) cellular systems, multi-orbit (GSO/NGSO) satellite mega-constellations, and massive IoT verticals (e.g., logistics and smart farming).

Expected outputs

Hardware-agnostic ML handover engines: Modular software frameworks and software-defined validation platforms that elevate 5G-NTN handover success rates above 95% while drastically minimising signalling overhead.

Robust physical layer tracking filters: Advanced signal processing and tracking algorithms capable of operating under low Signal-to-Disturbance-plus-Noise Ratios (SDNR) and imperfect channel state information (ICSI).

3GPP-compliant solutions: Dynamic, traffic-aware satellite network management protocols fully aligned with 3GPP standards (Releases 17–19) to cost-effectively scale up capacity.

Next-Gen waveforms & equipment: Multi-band and multi-orbit antenna architectures for user equipment (UE), integrated with joint-sensing communication waveforms.

Scientific questions & interdisciplinary approaches

Network scaling & architectural integration: How can Quantum Key Distribution (QKD) be seamlessly hybridised across terrestrial fibre-optic backbones and satellite segments to build a scalable, cross-border European Quantum Communication Infrastructure (EuroQCI)?

Cross-layer design optimisation: How do we optimise security and efficiency across all layers—from low-level physical optics and signal processing to high-level network routing and cryptography?

Classical-to-Quantum Signal Paradigms: How can classical wireless communication principles (e.g., MIMO, spatial multiplexing, and diversity techniques) be repurposed to mitigate crosstalk and stabilise continuous-variable (CV) states?

Theoretical & variational informatics: How can we leverage variational signal processing and mathematical modelling to estimate unknown quantum channels and establish bounds for multi-user entanglement distribution?

Obstacles & target environments

Physical layer degradation: Stochastic channel fading (e.g., atmospheric log-normal distributions), signal attenuation over long distances, optical path crosstalk, and high-rate, low-latency sync constraints.

Security threats: The impending compromise of traditional cryptography by quantum computing, alongside physical-layer hardware vulnerabilities such as single-photon detector time-shift attacks.

Target environments: Critical multi-user networks requiring future-proof security, including FinTech infrastructures, national data centres, lossy quantum topologies, and satellite-to-ground optical communication links.

Expected outputs

Functional infrastructure & testbeds: Operational national quantum networks (such as LUQCIA and Lux4QCI) deployed as real-world testbeds with integrated, proprietary Key Management Systems (KMS).

Validated signal frameworks: Mathematical frameworks and wavepacket engineering strategies proving that spatial-mode diversity optimises secret key rates under harsh fading conditions.

Algorithmic solutions: Rapidly converging variational algorithms for channel estimation and traffic-aware, cost-optimised routing protocols designed to maximise entanglement success across lossy topologies.

Standardised security blueprints: Validated multi-orbit/multi-band architectural layouts engineered to safeguard public and private telecommunications infrastructure.