Spectral efficiency is one of the most important measures of the performance of wireless communication systems owing to scarcity and cost of the radio spectrum. The increase in spectral efficiency provides higher data rates to the user, lower network cost to the operator, coverage extension and higher service reliability as well. Intercell interference due to frequency reuse is one of the key impairments in wireless systems. Multiple-input multiple-output (MIMO) technique has been developed to enhance the desired signal power (and hence mitigating the effects of intercell interference) and with employing simple linear signal processing technique, can strongly mitigate the interference resulting from co-channel users. This technique is mainly used to achieve spatial diversity for boosting the communication link reliability by combating fading, and providing spatial multiplexing to increase data rates without extra bandwidth by exploiting multipath. Distributed antenna system (DAS) has recently gained substantial interest due to its ability to reduce transmitted power thereby lowering the out-of-cell interference effects, maximise the coverage and improve the spectral efficiency. The combination of MIMO techniques with DAS, so-called distributed MIMO (D-MIMO) systems, is now being exploited and largely succeeded to fulfil the services of the fourth generation (4G) wireless systems. Very recently, one of crucial significance approach to reducing the radiated power and improving spectral efficiency to cope with fifth generation (5G) wireless systems is the use of large-scale MIMO (also referred to as massive MIMO) technology, which utilizes a large number of antennas, i.e., tens to hundreds, typically at the base station (BS) side. Presently, in the light of the rapid evolution of wireless systems into 5G, the integration of wireless power transfer (WPT) with newly wireless systems has seen a great deal of attention as a potential solution for powering energy-constrained wireless systems, especially with shortening communication links by emerging new technologies, e.g., D-MIMO and massive MIMO. This thesis is devoted to investigating and comparing the performance of three different MIMO systems. More specifically, the thesis focuses on analysing the spectral efficiency of a comprehensive model of self-powered MU-MIMO systems employing linear ZF technique at the BS for both perfect and imperfect channel state information (CSI) cases. The results demonstrate the impact of practical channel impairments, e.g., spatial correlation, shadowing and co-channel interference (CCI), and system parameters, e.g., the number of BS and user antennas, signal to noise ratio (SNR) and channel estimation error, on the spectral efficiency of the system. Besides, from a spectral efficiency perspective, a proposed model of a combination of MIMO and massive MIMO technologies with DAS in the presence of linear receivers at the processing unit (PU) is considered and compared to a centralised MIMO (CMIMO) system. The obtained results provide a wide range of insights into the effects of system parameters on the spectral efficiency and reveal that the proposed distributed MIMO system outperforms the C-MIMO system. In the context of wireless powered MIMO systems, this work investigates the performance of a power beacon (PB)-assisted wireless powered C-MIMO system, including one multi-antenna BS and a number of single-antenna users powered by randomly deployed PBs in the presence of ZF receiver at the BS. Also, two modes for radiation from the PBs are assumed and compared, one is the beamforming radiation mode (BRM), and the other is the isotropic radiation mode (IRM).