Massive MIMO for Future Wireless Networks: Energy Efficiency, Interference Management and Symbiotic IoT Networks

Abstract

Due to the dramatic increase in wireless data traffic and the energy consumption of wireless networks, spectral- and energy-efficient wireless networks are imperative. Using multiple-input multiple-output (MIMO) transceiver structures and cell den- sification through small cell (SC) deployment increases both spectral efficiency and energy efficiency significantly and meets future network requirements, but also brings new challenges. These are severe interference, limited fronthaul capacity, computa- tional complexity, cost, and power consumption. Promisingly, radio frequency (RF) energy harvesting, that helps technologies such as Internet of things (IoT) to further reduce the power consumption of devices while providing the desired quality of service (QoS), can benefit from MIMO systems in an overlaying network. This thesis designs high spectrum and energy-efficient cellular networks via three main objectives: 1) design and performance evaluation of an energy-efficient network by integrating MIMO and SC deployments with well-designed interference management and resource allocation methods; 2) design and performance evaluation of computa- tionally efficient precoding algorithms for co-located and cell-free (CF) massive MIMO (mMIMO) systems; 3) design and performance evaluation of energy-harvesting IoT networks underlying and symbiotic with massive MIMO cellular networks. First, we focus on maximizing the energy efficiency of a MIMO-enabled heterogeneous cloud radio access network (H-CRAN) as a candidate architecture for beyond 5G cellular systems. A joint radio resource block allocation and antenna selection algorithm is proposed for the SCs, and a single RF chain structure is considered for the mMIMO macro base station (BS). Moreover, while coordinating transmissions between cells subject to user-centric clustering, an energy-efficient beamforming design, and power allocation optimization problem is formulated and its solution is proposed. Second, we address the implementation complexity of matrix inversion associated with precoding in mMIMO systems. We investigate the convergence of different iterative matrix inversion methods in the presence of small-scale fading, large-scale fading, and spatial correlation and compare their performance and complexity. Third, by considering the coexistence of CF mMIMO and symbiotic backscatter communication and deriving the upper bound for signal-to-interference-plus-noise ratios (SINRs) and also the average harvested power, we provide a novel insight toward efficient implementation of massive machine-type communications (mMTC) use case of 5G and beyond cellular networks.