In recent years, the Internet of Things (IoT) revolutionized network computing by offering a multitude of different applications and services made available to everyday users. The volume of connected devices is expected to grow exponentially in the following years, with projections estimating billions of devices – averaging six to seven devices per person. This challenge creates in turn a demand for efficient resource allocation that would effectively handle the physical resources of the base stations of network cells, either macro cells or small cells. If we consider that future networks will not discard current network infrastructures, but will attempt to build upon the existing ones, then the challenge of taking full advantage of current network physical resources in order to serve as many network users as possible is looming large.

Based on the above challenges, the main objective of the ERA5G-Beyond project is to study, design, develop and evaluate technologies and techniques towards efficient resource allocation in 5G and Beyond Networks. Our goal consists of two different phases:

                 i.  In the first phase, we will examine the possibilities of current promising technologies, such as Multi-User Multiple-Input Multiple-Output (MU-MIMO) and Downlink and Uplink Decoupling (DUDe). We will investigate possible efficiency shortcomings or loopholes that limit their performance and we will propose new mechanisms/algorithms or improve the existing ones targeting at achieving efficient allocation of physical resources in heterogeneous networks (HetNets).

                 ii. In the second phase, we will apply techniques inspired by Machine Learning (ML) and Game Theory (GT) on the proposed mechanisms/algorithms in order to add prediction models and user fairness, which will in turn refine/optimize the performance of MU-MIMO and DUDe in terms of resource allocation, throughput, spectral efficiency and energy consumption.

Our research team, led by the Principal Investigator, has contributed to the research on mobile networks and on the abovementioned technologies providing a great number of publications in peer-reviewed scientific journals and conferences. The consortium is completed by experts in ML and GT who have significant experience in data mining, analysis of dynamic systems, ML-based problem solving, strategic aspects of decision making, price of anarchy analysis, coordination mechanisms, and more. As a result, we will build upon the knowledge and experience we have obtained over the past years.