Channel Estimation using Parameter Analysis for massive MIMO Systems using Deep Learning

mm-Wave and massive MIMO systems have emerged as feasible technologies for strengthening wireless communications performance by deploying numerous antennas at both the transmitter and receiving ends. Optimizing the performance of massive MIMO systems mandates reliable assessment of the channel characteristics. Conventional methods are not suitable for massive MIMO systems because of its complexity and dynamic nature. In this paper, analysis of important parameters for channel estimation in massive MIMO systems, such as antenna orientation and channel magnitude response for OFDM and non-OFDM systems. This paper introduces the Deep massive MIMO generator tool, and it trains the deep learning models based on the input data to estimate the channel parameters more accurately. For analysis, orientation of Base Station antenna done with different angles [(0,0, -45), (0,0, -60),) (0,0, -120)], and BS orientation is based on the signal strength and blockage are considered. And also observed the channel matrix magnitude with different OFDM parameters. The tremendous advantages of parameter analysis by using Deep mMIMO generator is expected to be implemented in intelligent reflecting surfaces (IRS), and blockage prediction in massive MIMO systems.