UKI-FNI Pilot Project Phase 3 – Summaries

This pilot will deliver a practical Zero‑Trust framework that keeps 5G Open RAN network slices compliant with SLAs even when the RAN extends over non‑terrestrial networks (NTN). We will develop and adapt novel lightweight device/edge attestation and post‑quantum keying (ML‑KEM) secure the control loop end‑to‑end and thereby enable an Open RAN compliant framework for Non Terrestrial Networks. We will evaluate on the University of Surrey’s NTN digital‑twin environment and replicate results on the IIT Hyderabad 5G testbed, producing open datasets and a short memo for the O‑RAN Software Community. The outcome is a containerised MVP that shows how Zero‑Trust policies can safely automate slice admission and remediation under satellite‑style delays.

Objectives:

1. Architecture & threat model: Finalise a Zero‑Trust control loop for Open RAN + NTN (A1/KPM/RC/O1).
2.  xApp (near‑RT RIC): Slice‑admission controller that evaluates KPM against per‑slice SLAs and enforces RC Style‑2/Action‑6 policies.
3. rApp (Non‑RT RIC): A1 policy distribution and lightweight learning for long‑term SLA behaviour.
4. PQC integration: Prototype ML‑KEM (Kyber) keying; provide CPU‑only fallback.
5.  Evaluation: Surrey NTN digital‑twin scenarios → IITH 5G testbed replication; release curated datasets and an O‑RAN memo.
6.  MVP deliverable: Containerised demo showcasing secure, automated slice admission under satellite conditions.

This pilot project develops a novel RIS-aided Integrated Sensing and Communication (ISAC) framework within an Open RAN architecture to enable secure, intelligent, and adaptive wireless networks for future 6G systems. By integrating reconfigurable intelligent surfaces with ISAC functionalities, the project enables real-time environmental awareness and dynamic control of electromagnetic propagation. A key innovation is the development of RIS-aware xApps that operate within the O-RAN ecosystem to jointly optimise communication performance, sensing accuracy, and physical-layer security. The system is capable of detecting anomalies such as eavesdropping and unauthorized localization while enhancing signal quality and coverage, particularly in non-line-of-sight scenarios. The outcome is a deployable, scalable, and secure-by-design platform that supports next-generation cyber-physical systems and future network infrastructures.

Project Objectives

• Develop a RIS-assisted ISAC framework aligned with Open RAN principles
• Design and implement RIS-aware xApps for real-time sensing and control
• Enable dynamic beamforming and environment shaping using RIS
• Achieve secure-by-design wireless communication, including detection of eavesdropping and anomalies
• Improve localization accuracy and situational awareness using ISAC
• Validate the system through a real-world O-RAN testbed with USRP and RIS integration
• Demonstrate low-latency, scalable, and programmable wireless environments for future 6G applications
• Figure 1 shows the capability of RIS in directing the beams
• Figure 2 shows the capability of ISAC in integrating both Communication and sensing using ISAC xApp for security and environmental awareness.

“The vision for 6G involves massive device density, ultra-reliability, and AI-native network-compute convergence, to support scalability and resilience. However, as 6G networks become more complex and interconnected, reliability and resilience challenges increase. The existing solutions to detect and respond to faults, and recover from failures are mostly reactive, requiring human intervention or centralized control. The ability to autonomously detect and mitigate faults, attacks, and outages, and the corresponding resilience issues remain limited. This project aims to design and validate a new cognitive plane powered by AI agents and agentic workflows within a fully distributed multi-agent system in future 6G network fabric, that embed state-of-the-art multimodal large language models (MLLMs) and reinforcement learning (RL) for autonomous fault mitigation, self-healing/optimization, and zero-touch failure recovery. The ultimate goal is to cut down the network response time to faults, attacks, and failures embedding autonomous AI agents at various layers in the network fabric. These agents will reason over the federated network data bases and interact via a standards-aligned Agent-Based Interface (ABI) and a Model Context Protocol (MCP) to achieve resilient, intent-based orchestration of 6G network service functions. Our approach embeds resilience as a core design principle, enabling rapid fault isolation, self-healing, and adaptive recovery from cyber-attacks or large-scale outages.”

-Collaborators: “University of Surrey (UoS), UK, and Indian Institute of Technology (IIT), Delhi”

-Objectives: “1) Design of coordinator and domain specific AI agents in the emulator environment, 2) Integration in UoS testbed, validation and Optimisation.”

Traditional cellular architectures impose fundamental limitations on 5G/6G performance, where the rigid cell-based control plane creates inefficiencies in mobility management (e.g., handover failures) and limits the potential of user-centric massive multiple input mmultip output (MIMO) systems. Although 5G introduced beam management, it remains tethered to specific Cell-IDs, restricting global coordination. This project (BEAM-RAN) studies a paradigm shift to a fully beam-centric architecture, where fixed cells are eliminated and the active beam concept is explored. The active beams are the dynamic logical entities defined by frequency, spatial location, and directional filtering, managed globally to serve users. The project brings together a strong UK-India consortium of six academic partners and four industry partners coming equally from both countries with complementary expertise spanning digital twins, 5G/6G systems, beamforming, AI/ML, O-RAN, quantum technologies, and experimental testbeds, creating a cohesive team well-positioned to deliver impact. The UK academic partners bring strength in theory, twin modelling, RAN architecture and testbed infrastructure. Indian academic partners bring strength in physical layer design, security, quantum applications, and system-level AI/ML. Industry partners from both sides ensure relevance, standard alignment, and rapid prototyping and implementation.

Project objectives:

(O1) To define the logical entity model for “active beams” (defined by frequency, spatial location, and filter) and specify the necessary changes to the O-RAN architecture.

(O2) To explore the development of a Digital Twin prototype capable of modelling the new beam-centric user plane and a framework for integrating quantum technologies into the control plane.

(O3) To study a 6G use-case and to prepare a roadmap for the bilateral UK-India physical testbed.

Open RAN is poised to play a central role in future 6G networks, with early adoption emerging in private 5G deployments. However, the very features that make Open RAN appealing—its software-defined, multi-vendor architecture—also contribute to increased operational expenditure (OPEX) due to the complexity of integration, configuration, and ongoing management. This project addresses these challenges through a collaboration between the Networked Systems Group at the University of Edinburgh and the IOS-MCN group at the Indian Institute of Science (IISc). It focuses on enabling efficient, scalable, and sustainable Open RAN by leveraging machine learning for automated configuration and compute resource optimization. Validation will be conducted on two complementary testbeds: the IOS-MCN lab at IISc and a campus-scale outdoor private 5G Open RAN deployment at the University of Edinburgh.

Project objectives

This project aims to reduce the OPEX of Open RAN-based private 5G networks through two complementary means: first through automated RAN configuration and second through performance-constrained resource optimization, by leveraging machine learning based approaches for both.

As quantum computing advances, classical cryptographic algorithms currently used in 5G/6G networks—such as those in the Transport Layer Security (TLS) protocol—become vulnerable to “Harvest Now, Decrypt Later” attacks. The project will integrate NIST-approved quantum-secure algorithms suite into an O-RAN testbed at the University of Surrey. By conducting realistic benchmarking, the research will quantify the tax or performance overhead the PQC introduces to network latency, jitter, and bandwidth, providing a foundational roadmap for secure-by-design 6G ecosystems.

The work is structured to move PQC evaluation from theoretical simulation to real-world infrastructure validation. The objectives of the project are:

• Integration: Developing and testing an open-source hybrid TLS 1.3 library integrated with O-RAN Alliance compliant E2 Node emulators and xApps.
• KPI Benchmarking: Identifying and analyzing Key Performance Indicators (KPIs) like handshake latency and stress/load across various security levels.
• Scalability Testing: Conducting realistic performance measurements on the Surrey testbeds, specifically focusing on massive machine-type communications (mMTC) where millions of devices connect simultaneously.
• Knowledge Sharing: Demonstrating the impact of hybrid PQC TLS 1.3 to the UK-India community through joint workshops and peer-reviewed publications.

Advisory & Industrial Support: Strategic guidance is provided by an advisory board at QUB (including Prof. Maire O’Neill) and the CSIT Industrial Advisory Board to ensure the research aligns with global industry standards.