In this thesis, we investigate the energy efficiency (EE) problem in multiple-input multiple-output (MIMO) systems. First, different from conventional energy-efficient optimization approaches in point-to-point MIMO spatial multiplexing schemes that require iterative numerical algorithms, we derive an optimal solution in a closed form with less computational complexity, which provides a valuable insight upon the relation between the optimum EE and system parameters such as circuit power and channel conditions. Since perfect CSI at the transmitter needs additional radio resources, we also consider equal power loading scenario under perfect CSI at the receiver and limited feedback information at the transmitter as a means to save energy. To determine the best energy-efficient transmit antennas subset, we suggest a novel antenna selection algorithm with good performance and low complexity that enables us to achieve higher EE performance. Moreover, this thesis addresses energy efficiency optimization problems for downlink multiuser (MU) MIMO systems with block-diagonal (BD) transceivers, subject to a joint power and rate constraints for each user.
