Machine Learning-Based, Networking and Computing Infrastructure Resource Management of 5G and Beyond Intelligent Networks
Overview
The rapid expansion of megacities and the rise of smart city applications have led to unprecedented growth in data traffic, placing immense demands on urban digital infrastructure. Current networks rely on diverse wireless, optical, and computing technologies owned by various providers, including small-scale operators. As 5G becomes more widespread, expected to serve up to 65% of the population, the challenge lies in integrating edge computing, virtualised Radio Access Network (vRAN), and Open RAN (O-RAN) technologies into a unified, low-latency, secure system. This integration is complicated by the isolated development of these technologies and the need for architectures that scale flexibly across complex urban environments.
The MARSAL project addressed these challenges by introducing a novel paradigm of elastic virtual infrastructures that transparently integrated advanced radio access, networking, management, and security technologies. It focused on three core pillars: network design, virtual infrastructure, and network security. MARSAL advanced cell-free networking and mmWave integration aligned with O-RAN standards, while enhancing Mobile Edge Computing and network slicing through layered analytics. Furthermore, it developed machine learning-based mechanisms to ensure data privacy and security in multi-tenant environments. Overall, MARSAL delivered an end-to-end framework for efficient, secure data transfer, processing, and storage tailored to the needs of future smart megacities.
Objectives
Design and demonstrate scalable, distributed cell-free massive-MIMO networks supporting massive Access Points deployments.
Design and implement a cell-free virtual Radio Access Network for B5G, aligned with the O-RAN Alliance architecture.
Architect a disaggregated, Software-Defined-Networking control plane towards Fixed-Mobile Convergence.
Deploy an Elastic Edge Computing paradigm with Cloud-Native technologies.
Ensure policy-driven security, privacy and trust in multi-tenant infrastructures.
Deliver a Self-driven infrastructure with pervasive, Machine-Learning-driven control.
Implement and achieve proof-of-concept of the MARSAL solutions.
Promote dissemination, standardisation and exploitation of MARSAL.
eBOS Role in the Project
eBOS led the dissemination and communication activities of the MARSAL project, implementing a comprehensive plan that included publications in leading journals and conferences, the organisation of stakeholder workshops with testbed demonstrations, and participation in targeted expos and exhibitions to showcase proof-of-concept prototypes. eBOS also coordinated outreach efforts through a dedicated website, social media, newsletters, and printed materials, while issuing press releases and business-oriented articles to maximise visibility and highlight the commercial potential of MARSAL technologies, in collaboration with all project partners.
eBOS also led the implementation and integration of MARSAL’s comprehensive security framework, with a strong focus on protecting multi-tenant infrastructures. This involved enhancing confidentiality protocols for EP message passing, integrating a blockchain-based Smart Contract platform to manage resource ownership and network slice permissions, and deploying a policy-driven NFS Gateway to enforce privacy, trust, and probabilistic data integrity. Additionally, eBOS led the integration of the Threat Detection and Threat Analysis engines, enabling system-wide threat correlation and enhancing data-plane security through the isolation and micro-segmentation of malicious flows.
Consortium
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101017171.