PhD defence of Yuanzhe Gong – Metamaterial-Based Large-Scale Array for Half-Duplex/Full-Duplex mMIMO

Abstract

Massive multiple-input multiple-output (mMIMO) technology has emerged as a key enabler for future wireless systems, providing substantial gains in capacity and coverage through large-scale antenna arrays. With strategically designed beamforming algorithms, these arrays can precisely direct radiation toward intended users while effectively suppressing interference, thereby boosting spectral efficiency and signal quality. Metamaterial-based structures with tunable electromagnetic (EM) properties present promising avenues for further improving large-scale array performance by facilitating miniaturized, high-isolation antenna elements and enabling novel interference mitigation strategies. Moreover, hybrid beamforming (HBF) schemes that integrate radio-frequency (RF) beamformers with digital baseband precoders/combiners using fewer RF chains effectively reduce both hardware and computational complexity while maintaining performance through the joint design of the two stages. This thesis investigates and develops metamaterial-based large-scale antenna arrays alongside beamforming optimization techniques for both half-duplex (HD) multi-user mMIMO (MU-mMIMO) and full-duplex (FD) mMIMO systems. Extensive full-wave simulations and experimental measurements evaluate the performance of various prototype arrays, considering factors such as array aperture size, sub-array placement, and surrounding environments.

In the first part, various nulling control beamforming (NCB) optimization algorithms are proposed, leveraging both deterministic and nature-inspired approaches while accounting for realistic EM characteristics such as heterogeneous element radiation patterns. These algorithms target far-field, angular-separation-based interference scenarios and near-field, location-differential cases involving spherical wavefronts in extra-large arrays. Full-wave simulations illustrate how beam radiation patterns transition from near-field to far-field propagation with increasing distance. Through simulations and experimental measurements conducted in an anechoic chamber with the proposed array prototypes, the NCB techniques effectively create deep, accurately aligned nulls that reduce residual multi-user interference (MUI) to noise levels while maintaining the desired signal gain.

In the second part, the thesis explores FD mMIMO communications, where concurrent transmission and reception further enhance spectral efficiency. Metamaterial-based large-scale arrays offer additional degrees of freedom for suppressing self-interference (SI), a critical challenge in FD operations. A multi-stage transmit-receive isolation framework is proposed, encompassing: (1) antenna isolation techniques, (2) joint uplink/downlink beamforming optimization, and (3) integration of metamaterial absorber structures. Experimental results demonstrate the effectiveness of these methods in collectively mitigating SI, preventing receiver front-end saturation and nonlinear distortion. Through experimental measurements, the study examines the correlation between beam radiation patterns and beam-level mutual coupling, as well as the effectiveness of various isolation techniques, offering valuable and practical insights into the development of beam-level isolation strategies. By combining these isolation measures with self-interference cancellation techniques, residual SI is reduced to the noise floor, thereby ensuring reliable FD mMIMO communication performance.