sst |
CRLB-Based Antenna Selection for Angle of Arrival Estimation
The new generation of wireless systems aims to enable smarter and more sustainable connectivity. Growing demands for data and digitalization require technologies that make more efficient use of available resources, raising the question of how to design sustainable communication networks. Cellular systems, where base stations serve many devices, are central to this challenge, as they must meet strict energy and environmental constraints.
This research work investigates energy efficiency and resource allocation in Multiple-Input Multiple-Output (MIMO) systems, focusing on the localization phase, where users’ Angles of Arrival (AoAs) are estimated. The Cramér–Rao Lower Bound (CRLB) for AoA estimation in massive MIMO with linear antenna arrays is analyzed under dense multipath channel conditions. It is shown that the CRLB converges to a deterministic value regardless of the multipath distribution. Based on this result, a novel formulation of localization energy efficiency is introduced and optimized via antenna selection.
Then, the analysis is then extended to planar antenna arrays. A real-time antenna selection strategy is developed to minimize the CRLB of azimuth AoA in planar antenna arrays. The proposed method uses a greedy algorithm whose optimal starting point is obtained. The study indicates that the proposed strategy achieves near-optimal performance while remaining computationally efficient. The numerical results demonstrate that with the proposed antenna selection strategies, high percentages of the achievable CRLB performance can be obtained using only a small fraction of the available antennas, hence improving the efficiency of resource utilization.
Jury members :
Prof. Luc Vandendorpe (UCLouvain)(Promoteur)
Prof. David Bol (UCLouvain) (Président)
Prof. Jérôme Louveaux (UCLouvain) (Secrétaire)
Prof. Sofie Pollin (KU Louvain)
Prof. Marc Moonen (KU Louvain)
Pay attention : the public defense of Masoud ARASH will also take place in the form of a videoconference
