Recursive Least Squares based Kalman Hybrid Precoding for Multi-User Millimeter Wave Massive MIMO Systems

Abstract

Massive MIMO and mmWave are critical enabling technologies in the 5G cellular network. mmWave has a higher frequency range, which corresponds to a very short wavelength, resulting in high path loss and blockage in communication systems. To address these challenges, Integrating mmWave and massive MIMO can improve communication performance and RLS Kalman Filter-based Hybrid beamforming is the best technique for minimizing bit error rates, efficiently utilizing energy, and improve spectral efficiency. This thesis compares the performance of various precoding techniques, including Zero Forcing (ZF), Minimum Mean Square Error (MMSE), Kalman, Normalized Least Mean Squares (NLMS) Kalman, Minimum Square Error Fully digital (MSE Fully digital), and Proposed Recursive Least Squares (RLS) Kalman Hybrid Precoding, using performance metric parameters such as bit error rate, energy efficiency, and spectral efficiency. For the simulation purpose Matlab R2018a is used. Simulation outcomes demonstrate that the Proposed Recursive Least Squares (RLS) Kalman Hybrid Precoding achieved the best energy efficiency, spectral efficiency performance, and bit error rate with increased transmitter and receiver antennae and fewer users. The RLS Kalman Hybrid Precoding improved the spectral efficiency by almost 4.105 bps/Hz at 20dB with ten channel paths compared to the Zero Forcing (ZF) Hybrid Precoding and almost 3.758 bps/Hz, 3.159 bps/Hz, 2.107 bps/Hz, and 0.57 bps/Hz for the MMSE, Kalman, NLMS Kalman, and MSE Fully Digital Hybrid Precoding techniques, respectively, under the same conditions.