Approaching K-Means for Multiantenna UAV Positioning in Combination With a Max-SIC-Min-Rate Framework to Enable Aerial IoT Networks

In long-range wireless communication networks, the fading channels described in channel state information are strongly related to distance and the path loss exponent and represent a major challenge in delivering the performance required to support emerging applications. Conveniently, multiple antennas and cooperative relays are efficient solutions that can combat fading channels, thereby improving networking capacity and transmission reliability. This study investigated the use of multi-antenna unmanned aerial vehicle (UAV)s as aerial Internet of Things (IoT) relays and employed their direct line-of-sight benefits to assist IoT wireless networks. To improve the outage probability, system throughput, and energy efficiency (EE), we first considered a combination of transmit antenna selection at the transmitter and the selection combining technique at the receiver to determine the best channel from the pre-coding channel matrix. Using a practical model in a three-dimensional earth environment in combination with the K-means algorithm, we then investigated optimal UAV placement to obtain optimal channel state information for the non-orthogonal multiple access (NOMA) -IoT device cluster globally, thereby ensuring the quality of service for the IoT devices. We introduced a max-successive interference cancellation-min-rate framework for non-ordered NOMA devices, thus deriving theoretical expressions in novel closed forms for two independent scenarios: (
$i$
) Rayleigh and (
$ii$
) Nakagami-
$m$
fading channels. By optimizing the UAV placement, the investigated results applied to the UAV scheme delivered better performance in a NOMA-IoT network than in a terrestrial relay (TR) scheme. Finally, the study examines a variety of models and presents algorithms for Monte Carlo simulations to verify the theoretical results.