Performance Evaluation of Low Complexity Massive MIMO Techniques for SC-FDE Schemes

Abstract

Massive-MIMO technology has emerged as a means to achieve 5G's ambitious goals; mainly to obtain higher capacities and excellent performances without requiring the use of more spectrum. In this thesis, focused on the uplink direction, we make a study of performance of low complexity equalization techniques as well as we also approach the impact of the non-linear elements located on the receivers of a system of this type. For that purpose, we consider a multi-user uplink scenario through the Single Carrier with Frequency Domain Equalization (SC-FDE) scheme. This seems to be the most appropriate due to the low energy consumption that it implies, as well as being less favorable to the detrimental effects of high envelope fluctuations, that is, by have a low Peak to Average Power Ratio (PAPR) comparing to other similar modulations, such as the Orthogonal Frequency Division Multiplexing (OFDM). Due to the greater number of antennas and consequent implementation complexity, the equalization processes for Massive- MIMO schemes are aspects that should be simplified, that is, they should avoid the inversion of matrices, contrary to common 4G, with the Zero Forcing (ZF) and Minimum Mean Square Error (MMSE) techniques. To this end, we use low-complexity techniques, such as the Equal Gain Combining (EGC) and the Maximum Ratio Combining (MRC). Since these algorithms are not sufficiently capable of removing the entire Inter-Symbol Interference (ISI) and Inter-User Interference (IUI), we combine them with iterative techniques, namely with the Iterative Block with Decision Feedback Equalizer (IB-DFE) to completely remove the residual ISI and IUI. We also take into account the hardware used in the receivers, since the effects of non-linear distortion can impact negatively the performance of the system. It is expected a strong performance degradation associated to the high quantization noise levels when implementing low-resolution Analog to Digital Converters (ADCs). However, despite these elements with these configurations become harmful to the performance of the majority of the systems, they are considered a desirable solution for Massive-MIMO scenarios, because they make their implementation cheaper and more energy efficient. In this way, we made a study of the impact in the performance by the low-resolution ADCs. In this thesis we suggest that it is possible to bypass these negative effects by implementing a number of receiving antennas far superior to the number of transmitting antennas.