PhD defence of Mobeen Mahmood – Enhancing Coverage and Capacity in Massive MIMO Wireless Communications Systems
Enhancing Coverage and Capacity in Massive MIMO Wireless Communications Systems
Abstract
The ITU’s 2030 (6G) framework focuses on enhancing terrestrial networks by integrating aerial communications for peak data rate, ubiquitous coverage, and sensing. In this regard, massive multiple-input multiple-output (mMIMO) is considered as a key enabling technology for large-scale deployments in beyond 5G mobile networks. By utilizing a large number of antennas at base station (BS), mMIMO enables state-of-the-art hybrid beamforming (HBF) and MIMO techniques that are powerful tools for improving end-user experience and capacity in both uplink and downlink. The two-stage HBF architecture is considered a promising solution in mMIMO systems to provide high data rates with much-reduced hardware complexity/cost. We aim to address a set of objectives, including antenna array configurations, beamforming optimization, interference suppression, sum-rate maximization, improved energy efficiency, high self-interference suppression (SIS) quality, and unmanned aerial vehicle (UAV) deployment to enhance both coverage and capacity in terrestrials and UAVs-assisted terrestrial networks.
First, we study different array structures, which can be used at BS to support both aerial and ground users. We consider half-duplex communications and investigate how the system performance can vary based on (i) users’ angular location, and (ii) number of users. In this regard, we design HBF schemes based on users’ angular locations for both full-resolution and low-resolution hardware components to reduce multi-user (MU) interference.
Secondly, we explore full-duplex (FD) communications. We aim to mitigate strong self-interference (SI) and maximize the total achievable rate based on over-the-air (OTA) measurements of the SI channel measured in an anechoic chamber for a sub-6 GHz frequency band. By using perturbation-based HBF for SI suppression and by exploiting spatial degrees-of-freedom (DoF) due to the use of large antenna arrays, our objective is to bring the SI level down to the noise floor, thus avoiding the use of costly/complex analog cancellation circuits commonly used in FD circuitry. We propose different HBF solutions,
which can suppress SI up to 80 dB by optimizing (i) variable gain-controllers, (ii) perturbing the directed beams, and (iii) selecting the best Tx/Rx antennas.
Finally, we consider a UAV-assisted terrestrial system to enhance coverage and capacity by using UAVs as relays between BS and users. The UAV's mobility and flexibility add degrees of freedom, mitigating signal fading and attenuation, and extending coverage to obscured users. We explore a joint HBF approach that optimizes UAV location and power allocation, sequentially designing HBF stages for the BS and UAV. A deep learning-based solution is developed, using offline training and online prediction to maximize achievable rates while reducing runtime by 99%. We further extend this analysis to multiple UAV systems for broader coverage and capacity in dynamic environments.