UAV-Assisted Cooperative & Cognitive NOMA: Deployment, Clustering, and Resource Allocation

TL;DR

This paper proposes a UAV-assisted cooperative and cognitive NOMA framework that optimizes user clustering, channel assignment, and resource allocation to enhance spectrum efficiency and fairness in next-generation wireless networks.

Contribution

It introduces a novel joint deployment and resource allocation method using closed-form solutions for UAV-assisted CCR-NOMA, improving efficiency and fairness.

Findings

Achieves 100% accuracy in user clustering and channel assignment.

Delivers optimal resource allocation with less energy and time than geometric programming.

Provides closed-form expressions for coverage probability considering practical impairments.

Abstract

Cooperative and cognitive non-orthogonal multiple access (CCR-NOMA) has been recognized as a promising technique to overcome issues of spectrum scarcity and support massive connectivity envisioned in next-generation wireless networks. In this paper, we investigate the deployment of an unmanned aerial vehicle (UAV) as a relay that fairly serves a large number of secondary users in a hot-spot region. The UAV deployment algorithm must jointly account for user clustering, channel assignment, and resource allocation sub-problems. We propose a solution methodology that obtains user clustering and channel assignment based on the optimal resource allocations for a given UAV location. To this end, we derive closed-form optimal power and time allocations and show it delivers optimal max-min fair throughput by consuming less energy and time than geometric programming. Based on optimal resource allocation, the optimal coverage probability is also provided in closed-form, which takes channel estimation errors, hardware impairments, and primary network interference into account. The optimal coverage probabilities are used by the proposed max-min fair user clustering and channel assignment approaches. The results show that the proposed method achieves 100% accuracy in more than five orders of magnitude less time than the optimal benchmark.