A Novel Method for 5 Generation Multiple-Input Multiple Output Orthogonal Frequency Division Multiplexing Channel Estimation using Cauchy Evading Golden Tortoise Adaptive-Bi Directional-Long Short-Term Memory

Abstract

Multiple-Input Multiple Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM), which could provide a large number of antennas, is one among the key technologies of 5G. For recovering the transmitted symbols at the receiver, a key role is played by the Channel Estimation (CE) in MIMO-OFDM. But it is indisputable recently. Hence, a novel method for 5G MIMO-OFDM CE using Cauchy Evading Golden Tortoise Adaptive-Bi Directional-Long Short-Term Memory (CEGTA-Bi-LSTM) is proposed. Primarily, input like text, video, and audio are given for the encoding process at the transmitter; in addition, it is encoded using the Gaussian Distribution-Low Density Parity Check (GD-LDPC) algorithm. Next, the encoded signal is given to the rate matching. Then, by employing Combined Bits-Quadrature Phase Shift Keying (CB-QPSK), the signal is modulated. Next, the output is inputted into the pulse shaping, which is conducted by deploying Truncated Linear-Raised Cosine Filter (TL-RCF). The pulse-shaped signal is fed into the pilot insertion phase where the reference signal is inserted. Afterward, the signal is inputted into the E-IFFT phase. Next, the output from the E-IFFT passes through the channel with AWGN noise; also, it is given to the receiver phase in which the reverse process of the transmitter occurs. Then, from the receiver output, the channel is estimated. As per the experimentation, the proposed system depicted superior performance when analogized to the prevailing methodologies.