QoS-Aware User Selection and Resource Assignment for Coexistence of NR-U and Wi-Fi-Enabled IoT Networks

Amidst overwhelming demands for spectrum in the licensed wireless networks caused by the explosive breakthrough of the Internet of Things (IoT), the unlicensed band has been a tremendous resource in enhancing massive connectivity. This has also accelerated the creation of cutting-edge radio access technologies (RATs) such as 5G new unlicensed (NR-U) to support IoT networks. However, these efforts have been frustrated by several performance challenges including increased interference which has led to the deterioration in the quality-of-service (QoS). Moreover, wireless engineers are tasked to develop RATs whose specifications comply with several inter-RAT coexistence regulations. In this paper, we propose a joint IoT user selection and resource assignment algorithm to improve the performance of the 5G NR-U-enabled IoT system under quality-of-service (QoS) constraints with regulated interference to the Wi-Fi-enabled IoT system. Specifically, we formulate a mixed-integer non-linear programming problem (MINLP) and decompose it into two subproblems including user selection and power allocation. We adopt the difference of concave functions (DC) approach to solve the resulting optimization problems. We also exploit low-complexity algorithms based on matching theory for IoT user selection and dual-decomposition for power allocation. Especially, we derive a closed form of the power allocation expression using the Lagrangian method. Through various simulations, we demonstrate that the proposed algorithm enables multi-user diversity and significantly improves the spectral efficiency compared to the conventional scheme.