Table 5.

Phases of UAV-based network models (continuation).

Phase	Approaches	Strategies/Models	Advantages/Findings
Novel: >2016	Rural Panorama Addressing	Energy consumption optimization aims to improve aerial node missions and connectivity in the countryside by using a graph-based structure [8], besides an optimized model called RURALPLAN [79,80].	▪The multi-period graph approach derives into Genetic Algorithms. It guarantees the coverage and efficient management of UAV consumed energy. ▪RURALPLAN can reduce energy consumption by up to 60%. ▪The deployment of UAV-based networks can adopt a short-distance LOS, decreasing installation costs. ▪By considering a set of optical fiber links to support the backhaul network, the capital and operation expenditures can be compensated, simplifying the stated model.
		Analysis of joined-architecture networks, mixing UAVs and GEO/LEO satellites, to increase radius coverage and state the usability of aerial nodes to assist fixed-infrastructure networks in the countryside [81,82].	▪The use of aerial nodes, acting as relays, can cover vast rural extensions, addressing further mobile network generations—such as 5G—to implement steady-links IoT devices. ▪Bearing in mind the optimizing cellular networks aim in the countryside, heritage functionalities of LTE can achieve prominent coverage radius in the sub-1 GHz bands, raising RF propagation. ▪Since Non-Terrestrial Networks may be an integral part of 5G infrastructure, UAVs become the bedrock of a mixed-architecture network, especially in collecting data in massive MTC types of application.
		LTE networks can provide coverage by UAV nodes in rural areas, chiefly to boost the Command and Control downlink channel, despite the raised interference due to height dependency [83,84].	▪The dependency of the large-scale path loss on the drone’s height may be challenging for achieving significant growth in coverage level, boosting the aerial-node’s perceived interference level. ▪Applying network diversity, it is possible to improve the network coverage level and its reliability, since SINR would be better than the achieved −6 dB index under the full-load assumption. ▪The interference conditions—because drastically changed UAV height— will determine channel characterization to assess wireless remote controls for the aerial nodes.
		Boosting aerial coverage of rural area network deployment to clear limitations by interference mitigation techniques [85].	▪Interference canceling and antenna beam selection are strategies to improve overall—aerial and terrestrial— system performance. ▪The abovementioned schemes will gain a 30% of throughput and achieve a 99% reliability increase. ▪Downlink and Uplink radio interference trigger poor performance within aerial traffic.
		A Non-Orthogonal Multiple Access (NOMA) layout for UAV-assisted networks to provide emergency services in rural areas [86].	▪The proposal carry out the performance of terrestrial users enhancement, resulting in a by-device that is consumed in energy minimizing. ▪The proposed user-centric strategy follows stochastic geometry approaches for terrestrial users—placed into Voronoi cells—served by UAVs, achieving the location model of both nodes and UEs. ▪In the case of the NOMA-assisted multi-UAV framework, the analysis of coverage probability can aim to properly set up the network’s power allocation factors and targeted rates.
	Cellular Network Advance	Optimization of the UAV-mounted base stations (MBSs) placement, setting forth a Geometric Disk Cover (GDC) algorithmic solution, which coats with all ground terminals (GTs) in an inward spiral manner [87].	▪The correct deployment of MBSs can cover a set of k nodes with a minimum number of disks of a given circular surface with radius r. ▪The computational complexity may be significantly reduced when coverage starts from the perimeter of the area boundary.
		The Path Loss (PL) Characterization for urban, suburban, and rural environments enhances the access technologies for low-altitude aerial networks, considering UAV height effects on the channel [88,89].	▪By introducing a Correction Factor (CF), which relies on the UAV altitude, the large-scale fading and the PL of the A2G channel will be accurately characterized. ▪In urban contexts, PL increases with horizontal distance. In the case of rural zones, PL is irrelevant to UAV heights, albeit it approximates to free-space propagation models at heights around 100 m. ▪UAV-based networks face a large amount of neighboring interference due to the down-tilted antenna pattern of cellular networks. Moreover, the coverage behavior will be affected beneath this scheme.
		Improvement of coverage and capacity for future 5G configurations of aerial networks beneath two algorithmic approaches, entropy-based network formation [90] and latency-minimal 3D cell association scheme [91].	▪By correctly selecting the UAV controller and then performing network bargaining, the aerial base station could top off a more remarkable improvement on its throughput, SINR per UE capacity in the order of 6.3% and minimal delays and error rates. ▪With the increase in simultaneous requests within the next-generation heterogeneous wireless network, entropy approaches appear to be suitable for overcoming UAV allocation and Macro Base Station decision problems. ▪Lifting 3D configuration for aerial cellular networks, a yield of reducing up to 46% in the average total latency would enhance spectral efficiency.
		Optimal design of aerial nodes trajectory in cellular-enabled UAV communication with Ground-BS (GBS) subject to quality-of-connectivity constraints about the link GBS-UAV [92].	▪The optimization problem converges in a non-convex approach to find high-quality approximate trajectory solutions. ▪Channel’s delay-sensitive rates and SNR requirements restrict the target communication performance. ▪UAV’s mission completion time may guarantee an efficient method for checking the strategy’s feasibility.
		Cooperation of small and mini drones can further enhance the performance of the coverage area of FANETs—even other aerial-kind networks—by establishing a hierarchical structure of efficient collaboration of drones [93,94].	▪In the case of ultra-dense networks, the approach efficiently broadens the common issues such as sparse and low-quality coverage and the non-steady aerial links. ▪The rapidly unfolding of UAV carries out in the non-dependency of geographical constraints and implies system performance lifting by establishing LOS communication links in most scenarios. ▪Among other advantages—at the top of cooperative distributed UAV networks— are the distributed gateway-selection algorithms used and stability-control regimes.