Performance of Transmit SNR Weighted K-Means Clustering in NOMA-Based FANETS

Abstract

Flying ad hoc networks (FANETs), consisting of selforganizing unmanned aerial vehicles (UAVs), offer infrastructureless connectivity in a wide range of mission critical applications. In this paper, a non-orthogonal multiple access (NOMA) based FANET is considered in which cooperative UAVs act as decodeand-forward relays to serve the user equipments (UEs) in their coverage area. The partitioning of UEs into clusters around their nearest UAV and the UAV locations are determined using transmit signal-to-noise ratio (SNR) weighted K-means clustering. The weighting accounts for the constraint that UAVs in practical deployments may operated with different transmit powers due to, e.g., targeted coverage range, available energy budget, actual or intended mission duration. Simulation results for the outage probability and sum rate of the NOMA-based FANET with transmit SNR weighted K-means clustering for Nakagami-m fading are provided for a variety of system parameters.