Towards low complexity matching theory for uplink wireless communication systems

Al Ayidh, Abdulrahman Saeed (2022) Towards low complexity matching theory for uplink wireless communication systems. PhD thesis, University of Glasgow.

Abstract

Millimetre wave (mm-Wave) technology is considered a promising direction to achieve the high quality of services (QoSs) because it can provide high bandwidth, achieving a higher transmission rate due to its immunity to interference. However, there are several limitations to utilizing mm-Wave technology, such as more extraordinary precision hardware is manufactured at a higher cost because the size of its components is small. Consequently, mm-Wave technology is rarely applicable for long-distance applications due to its narrow beams width. Therefore, using cell-free massive multiple input multiple output (MIMO) with mm-Wave technology can solve these issues because this architecture of massive MIMO has better system performance, in terms of high achievable rate, high coverage, and handover-free, than conventional architectures, such as massive MIMO systems’ co-located and distributed (small cells). This technology necessitates a significant amount of power because each distributed access point (AP) has several antennas. Each AP has a few radio frequency (RF) chains in hybrid beamforming. Therefore more APs mean a large number of total RF chains in the cell-free network, which increases power consumption. To solve this problem, deactivating some antennas or RF chains at each AP can be utilized. However, the size of the cell-free network yields these two options as computationally demanding. On the other hand, a large number of users in the cell-free network causes pilot contamination issue due to the small length of the uplink training phase. This issue has been solved in the literature based on two options: pilot assignment and pilot power control. Still, these two solutions are complex due to the cell-free network size.

Motivated by what was mentioned previously, this thesis proposes a novel technique with low computational complexity based on matching theory for antenna selection, RF chains activation, pilot assignment and pilot power control. The first part of this thesis provides an overview of matching theory and the conventional massive MIMO systems. Then, an overview of the cell-free massive MIMO systems and the related works of the signal processing techniques of the cell-free mm-Wave massive MIMO systems to maximize energy efficiency (EE), are provided. Based on the limitations of these techniques, the second part of this thesis presents a hybrid beamforming architecture with constant phase shifters (CPSs) for the distributed uplink cell-free mm-Wave massive MIMO systems based on exploiting antenna selection to reduce power consumption. The proposed scheme uses a matching technique to obtain the number of selected antennas which can contribute more to the desired signal power than the interference power for each RF chain at each AP. Therefore, the third part of this thesis solves the issue of the huge complexity of activating RF chains by presenting a low-complexity matching approach to activate a set of RF chains based on the Hungarian method to maximize the total EE in the centralized uplink of the cell-free mm-Wave massive MIMO systems when it is proposed hybrid beamforming with fully connected phase shifters network.

The pilot contamination issue has been discussed in the last part of this thesis by utilizing matching theory in pilot assignment and pilot power control design for the uplink of cell-free massive MIMO systems to maximize SE. Firstly, an assignment optimization problem has been formulated to find the best possible pilot sequences to be inserted into a genetic algorithm (GA). Therefore, the GA will find the optimal solution. After that, a minimum-weighted assignment problem has been formulated regarding the power control design to assign pilot power control coefficients to the quality of the estimated channel. Then, the Hungarian method is utilized to solve this problem. The simulation results of the proposed matching theory for the mentioned issues reveal that the proposed matching approach is more energy-efficient and has lower computational complexity than state-of-the-art schemes for antenna selection and RF chain activation. In addition, the proposed matching schemes outperform the state-of-the-art techniques concerning the pilot assignment and the pilot power control design. This means that network scalability can be guaranteed with low computational complexity.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from King Khalid University.
• Subjects: T Technology > T Technology (General)
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Supervisor's Name: Imran, Professor Muhammad Ali and Sambo, Dr. Yusuf
• Date of Award: 2022
• Depositing User: Theses Team
• Unique ID: glathesis:2022-83171
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 07 Oct 2022 14:12
• Last Modified: 07 Oct 2022 14:14
• Thesis DOI: 10.5525/gla.thesis.83171
• URI: https://theses.gla.ac.uk/id/eprint/83171

Related URLs

• Article DOI
• Article DOI
• Article DOI
• Article DOI