Maximizing Downlink User Connection Density in NOMA-aided NB-IoT Networks Through a Graph Matching Approach

• Published in: 2022 IEEE 96th Vehicular Technology Conference (VTC2022-Fall) pp. 1 - 7
• Main Authors: Mishra, Shashwat, Salaun, Lou, Gorce, Jean-Marie, Chen, Chung Shue
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2022
• Subjects: B5G; Bipartite graph; Bipartite Graph Matching; Downlink; Frequency diversity; Industrial IoT; Machine-to-Machine Communication; Massive Connectivity; NOMA; Performance evaluation; Quality of service; Vehicular and wireless technologies
• ISSN: 2577-2465
• DOI: 10.1109/VTC2022-Fall57202.2022.10012847

We develop a framework for maximizing the number of transmitted packets for devices in a Narrowband Internet of Things (NB-IoT) network using non-orthogonal multiple access (NOMA) in the downlink. The base station (BS) chooses one of the multiple available physical resource blocks (PRBs) that are well separated in frequency for a device, giving them the advantage of exploiting frequency diversity. The scheduling strategy focuses on the two-fold problem involving efficient device clustering and optimum power allocation. This problem is a mixed-integer non-convex problem. We propose a bipartite graph matching approach, termed minimum weight full matching with pruning (MWFMP), to address the problem over multiple PRBs and solve it under the quality-of-service (QoS), allowable PRB, power budget, and interference constraints. Additionally, we provide a comparison with a greedy heuristic, the multi-PRB stratified device allocation (MPSDA), where we extend our previous work for a single PRB connectivity problem. Furthermore, we compare our algorithms to orthogonal multiple access (OMA) scheduling, which is prevalent in legacy LTE networks. We show that our algorithms steadily outperform the connectivity performance offered by OMA.