Doctor of Philosophy (Ph.D.)
Degree Granting Department
Hüseyin Arslan, Ph.D.
Gokhan Mumcu, Ph.D.
Ismail Uysal, Ph.D.
Nasir Ghani, Ph.D.
Alessio Gaspar, Ph.D.
Ali Imran, Ph.D.
Beam Selection, Lens Antenna Subarrays, Millimeter Wave, Precoding
The ever-growing demand for higher data rates and greater data capacity at lower cost has led the mobile cellular industry to investigate new physical layer techniques and possible utilization of unused spectrums at higher frequencies for next-generation cellular networks. Thus, exploitation of the millimeter-wave (mmWave) spectrum and non-orthogonal multiple-access (NOMA) have been envisioned as the most promising enablers in meeting capacity demand. Due to the smaller wavelengths offered in mmWave frequencies, it is possible to deploy many antennas into a relatively smaller physical space in mmWave frequencies. This property leads to a promising integration between mmWave and massive multiple-input multiple-output (M-MIMO) architecture to surmount the severe free-space pathloss thanks to the high directional beamforming gain. MmWave M-MIMO also offers significantly improved spectral efficiency by allowing simultaneous transmission of multiple data streams and utilizing the abundant and large bandwidth. However, conventional digital precoding causes excessive power consumption and hardware cost when directly adopted for mmWave M-MIMO since each antenna element necessitates its own radio-frequency (RF) chain. This problem can be addressed by beamspace MIMO (B-MIMO) because it can reduce the required RF chain by taking advantage of inherent sparsity in mmWave channels and applying a proper beam selection. On the other hand, NOMA enhances spectral efficiency by multiplexing multiple users' signals in the power domain using the same time and frequency resources, where the detection of multiple users' signals is performed by successive interference cancellation (SIC).
The research has concentrated on the beam selection problem, precoding design in B-MIMO, and spectral/energy efficiency enhancement in mmWave M-MIMO and NOMA. Specifically, the dissertation addresses the following:
First, we investigate the complexity reduction of the existing beam selection algorithms with incremental QR precoder (I-QR-P) and decremental QR precoder (D-QR-P). The proposed two-stage and three-stage algorithms reduce the complexity of D-QR-P and I-QR-P, respectively. Both aim to lower complexity by decreasing the candidate beam size by eliminating the beams with no contribution to any user and using matrix perturbation theory to update QR decompositions.
Second, we propose a hybrid precoding algorithm for the lens antenna subarray (LAS)-MIMO architecture in mmWave to control the LAS design efficiently. The precoding problem is formulated as a sparse reconstruction problem due to the inherent sparsity of the mmWave channel. The proposed algorithm is an iterative process developed jointly using artificial bee colony (ABC) optimization with orthogonal matching pursuit (OMP) algorithms. In each iteration, the algorithm randomly selects the switches for each lens using ABC and then uses OMP to approximate optimal unconstrained precoders.
Third, we investigate the spectral efficiency and energy efficiency tradeoff in downlink NOMA with the consideration of the quality of service (QoS) requirements. The non-convex multi-objective optimization problems are solved using population-based multi-objective evolutionary algorithms (MOEAs).
Finally, we propose an algorithm for the user-cell association problem in M-MIMO ultra-dense heterogeneous networks (UDHNs)., where the spectral and energy efficiency tradeoff is addressed. To this end, we formulate a convex multi-objective optimization problem and convert it into a single objective optimization problem where a priority is assigned for the spectral efficiency and energy efficiency with a weighting factor. The problem aims to maximize the weighted sum of spectral efficiency and energy efficiency. As a solution, Lagrange duality analysis is performed, and a distributed game theoretical user-cell association (GTUCA) algorithm is developed, considering the fairness among users.
Scholar Commons Citation
Cetinkaya, Sinasi, "On the Performance Enhancement of Beamspace MIMO and Non-orthogonal Multiple Access for Future Cellular Networks" (2023). USF Tampa Graduate Theses and Dissertations.