Fairness-Constrained Throughput Optimization for Coexistence of WiFi and Duty-Cycle 5G NR in the Unlicensed Spectrum

In this paper, we consider a WiFi and Duty-Cycle-based 5G NR-U coexisted network in the unlicensed spectrum, and study how to maximize the total network effective throughput under two widely-adopted notions of fairness, throughput fairness and 3GPP fairness, by adaptively tuning the system parameters. By properly characterizing the inter-network collisions between WiFi and 5G NR-U, explicit expressions of the maximum network effective throughput, the corresponding optimal duty-cycle ratio of 5G NR-U network and the optimal initial backoff window size of WiFi nodes are obtained, and verified by simulation results. Comparison between the performance limits with two fairness constraints shows that a higher maximum network throughput can be achieved under 3GPP fairness when the ratio of the number of WiFi nodes and the number of 5G NR-U nodes does not exceed a certain threshold. The analysis shows that DC-based 5G NR-U network can fairly coexist with WiFi while maximizing the throughput performance of the coexisting network through meticulous system parameter tuning, which sheds crucial insights for practical network design for unlicensed spectrum sharing.