Investigation of 5G Performance for Millimeter Wave Transmission using Efficient Beamforming Channel Estimation and NOMA Non Orthogonal Multiple Access Techniques

Abstract

The worldwide shortage of microwave bandwidth has become a key issue that may limit the development of wireless networks required to support emerging applications such as massive Machine Type Communication (mMTC), enhanced Mobile Broadband (eMBB), and Ultra-Reliable Low Latency Communications (URLLC), etc. For this reason, the unused Millimetre Wave (mm-Wave) frequency band of 30 GHz to 300 GHz has received a lot of attention as a potential medium for 5G wireless networks. However, the communication at this frequency band suffers from higher propagation losses, reflection losses and penetration losses due to its smaller wavelength. newlineTo compensate for these losses, mm-Wave systems will require massive antenna arrays with beamforming to provide high directive gains. Owing to the smaller wavelength of mm-Wave bands, more antennas can be packed within the same physical dimension at both transmitter and receiver ends. However, conventional MIMO systems with a specific RF chain connected to each antenna are unfeasible for the mm-Wave band due to excessively high costs and massive energy consumption of high-resolution analog-to-digital converters, digital-to-analog converters, and power amplifiers. To reduce hardware cost, complexity, and power consumption, hybrid MIMO architectures have been proposed for use in mm-Wave 5G systems. This research work considers an investigation of 5G system performance for mm-Wave transmission using efficient channel estimation, beamforming, and NOMA techniques. newlineThe first section of the thesis focuses on the development of an optimal channel estimation newlinesolution for the sparse nature of mm-Wave channels. A Compressed-Sensing (CS) newline newlinealgorithm is proposed to make it easier to estimate the beamspace channel matrix from a sparse channel model. Experimental results achieved through MATLAB reveals that the proposed algorithm shows improved normalized mean square error over other benchmark methods at different signal to noise ratio level and requires less computation time. newlineThe second part of the