Experimental Open-Source-Based Evaluation Platform for Highly Spectral-Efficient 5G With Simplified UTW-OFDM

Abstract

Fifth-generation mobile communication (5G) systems are increasingly being deployed in both commercial and private wireless networks to meet the growing demand for high-speed, reliable connectivity. While 5G systems adopt orthogonal frequency-division multiplexing (OFDM) for its high data rates and spectral efficiency, OFDM is known to generate large out-of-band emissions (OOBE), which must be suppressed to maximize spectrum usage. In this study, to tackle this issue, we propose and develop an experimental 5G full-stack evaluation platform that implements a waveform-shaping function for OFDM signals. The platform utilizes software-defined radio and open-source 5G software compliant with third-generation Partnership Project standards. We implement a simplified universal time-domain windowed OFDM as an application of the waveform shaping. This is a waveform shaping technique that can strongly suppress OOBE by applying a long time-domain window to the conventional cyclic prefix OFDM symbol. The transmission performance of the proposed platform was evaluated using a complete 5G system, which includes a 5G base station, user equipment, and a 5G core network. The effectiveness of the proposed platform is verified through link-level computer simulations. The results demonstrate that the block error rate characteristics exhibited a signal-to-noise power ratio difference of less than 1 dB between the platform and simulations and achieved an OOBE suppression of up to 24 dB at a bandwidth of 40 MHz. Furthermore, connectivity with a commercial 5G device demonstrated the feasibility of achieving OOBE suppression of 22 dB at a bandwidth of 100 MHz with a tolerable decrease of 18% in user throughput.