Performance Evaluation of Training ResNet Receiver Under Mobility for 5G NR Wireless Communication Systems

This paper investigates the bit error rate (BER) performances of a 5G NR-based uplink wireless communication system having a Residual Network-based (ResNet) receiver. The performances are analyzed in the clustered delay line (CDL) channel model under various training scenarios based on the speed of a transmitter. Computer simulation results confirm the robustness of a ResNet Receiver model trained under the highest transmitter speed scenario to recover the signals from a transmitter with various speeds. Two different channel encoders are also presented to demonstrate the consistency of the robustness.

I. Introduction

The fifth generation of wireless technology (5G) promises faster speeds, lower latency, and more reliable connections than its predecessors. This technology is currently being rolled out in many countries worldwide and is envisioned to have a major impact on industries such as telecommunications, health, and entertainment.