PRECODING ALGORITHMS FOR MILLIMETER WAVE MIMO SYSTEMS

Abstract

The high demand for bandwidth has recently attracted a lot of interest in under- utilized millimeter wave (mm-wave) frequency spectrum. Mm-wave multiple-input- multiple-output (MIMO) systems can play a vital role in increasing system capacity and hence can address the strict requirements set by International Telecommunica- tion Union (ITU) for 5G systems. Hybrid precoder is preferred in mm-wave MIMO systems over conventional precoding schemes as it reduces the cost and power con- sumption of the system by decreasing number of signal mixers and analog to digital converters. However, the optimal design of such hybrid structure has not been fully understood. In this thesis, hybrid precoding algorithms are proposed and compared with analog and digital precoding techniques. The results will then be compared to a new class of alternating minimization (AltMin) algorithms which are introduced recently which can achieve a near optimal solution for hybrid precoder. Insights from these algorithms for fully connected structures are drawn and presented. Finally a low complexity algorithm used in the practical structure will be investigated and a modi ed version of the algorithm using least square method is proposed. The results of the proposed algorithm will show a reduction in complexity and improvement in performance of low complexity AltMin algorithm.