Author: Mahmoud Sallam

Publisher:

Published: 2022

Total Pages: 0

ISBN-13:

"Massive MIMO (mMIMO), using a very large number of antenna elements to effectively serve multiple users (MU) simultaneously, has shown great promise to enhance the system capacity in 5G and beyond. Hybrid beamforming/precoding is one of the most popular mMIMO architectures, aiming to reduce the complexity by using a beamforming stage to bring the very large number of antenna elements to a much smaller number of RF chains, and consequently a small baseband (BB) precoder, sufficient for MU communications. The RF and BB precoders are designed to minimize the interference among users as much as possible.However, when users are too close, it may not be possible to serve them using different beams. This leads to higher interference among users and a significant decrease in capacity, especially when using high transmit power.In this work, we present three approaches for power allocation (PA) in HBF-mMIMO system to maximize the sum-rate of the users based on the effective channel obtained after beamforming/precoding.We start by simplifying the problem with a negligible-interference assumption, which leads to the water-filling (wf) strategy. A parameter called "channel indicator" is proposed accordingly, which can indicate the effects of interference on the channel quality and the success of the WF solution in a given scenario by a single value. By ignoring interference, WF solution tends to allocate more power to users having larger main-channel gains.WF solution is simple and acceptable when interference is low. However, when interference becomes significantly high, WF solution is far from the optimum PA. We continue by introducing a direct-solution (DS) approach for solving the PA optimization problem in presence of interference. The relationship between the channel gains and the sum-rate function is studied in detail, especially for the two-user and three-user cases. We focus on studying the impacts of the direct (main) channel and the levels of interference users caused and received by users on the sum-rate function.A user ranking algorithm based on the channel indicator is proposed, saving time and computational power for the DS.The Lagrangian solution is also derived for the general case as an alternative PA approach, providing a general generating equation. DS tends to reduce power allocated to users causing more interference.DS can offer the optimum PA solution at the costs of high complexity, especially for a large number of users. For this, we consider the iterative water-filling (IWF) as an extension of WF that takes the interference into account.The behavior of the IWF is studied in detail, by investigating the effect of the channel gains and transmit power on its trajectory and convergence. A modified IWF algorithm is proposed based on the user ranking algorithm to improve the performance of the IWF in high interference scenarios.IWF has the tendency to reduce powers allocated to users receiving more interference"--