AI-Based UAV Networks | Encyclopedia MDPI

AI-Based UAV Networks: History

Please note this is an old version of this entry, which may differ significantly from the current revision.

Subjects: Computer Science, Artificial Intelligence

Contributor:

To enhance the overall performance of the UAV networks and to address some specific problems, new features in the network are being designed as autonomous features. This approach not only provides optimum solutions for the targeted problems but also supports the dynamic properties of a UAV network.

artificial intelligence
autonomous UAV
AI-based UAV networks

1. Resource Management and Network Planning

Optimal management and planning of network resources are very crucial for the success of any network. This is particularly true when it comes to networks where human intervention is minimal. For UAVs, this feature is being focused on by many researchers, and various challenges are being identified and addressed. Table 2 provides a summary of related surveys on UAV network resource management and planning.

Table 2. Summary of related survey: UAV network resource management and planning.

Scope	Autonomous Features	Computational Intelligence	Channel Modeling	Interference Management	Security and Safety	Reference
Cooperative path planning, collisions	Path planning	√	Non-convex modeling	-	-	[31]
Cooperative path planning, resource allocation	Path planning	√	-	-	-	[32]
Cooperative UAVs, trajectory conflicts	Conflict detection and resolution	√	-	-	-	[33]
Channel modeling		√	Shadowing channel	√	-	[35]
Interference management	Path planning	√	Rician distribution	√	-	[36]
Interference-aware path planning	Path planning	√	Free-space path loss with 6 GHz	√	-	[36]
Interference management	Spatial configuration	√	Customized	√	-	[38]
Interference management	Transmission power, trajectory planning	√	Large-scale path loss	√	-	[40]
Interference management	Transmission power	√	Free-space path loss	√	-	[41]
Resource management	Energy consumption, transmission power	√	Various	√	-	[42]
Collision free navigation	Trajectory planning	√	-	-	-	[43]
Risk-aware path planning	Path planning	√	-	-	-	[44]
Mobility challenges		-	-	-	√	[45]
Resource management	User association	√	Free-space path loss	√	-	[46]
Delay-aware throughput maximization	Trajectory planning	√	Free-space path loss	-	-	[47]
UAV placement	Energy efficiency and optimization	√	Path loss outdoor/indoor penetration	-	-	[48]
Collision free navigation	Trajectory planning	√	-	√	-	[49]
Swarm-based UAV	Path planning	√	-	-	-	[50]
Physical layer security	security and cooperation	√	Free-space path loss	√	√	[51]
Secure UAV communication	Cooperative scheduling	√	Free-space path loss	-	√	[52]
Physical layer security	Cooperative trajectory and optimization	√	Free-space path loss	√	√	[53]
Physical layer security	Cooperative resource allocation	√	Free-space path loss	-	√	[54]
Quality of Experience (QoE)	Cooperative resource allocation	√	LOS and Non-LOS	-	-	[55]
Secure UAV communication	UAV defense	√	-	-	√	[56]
Software-defined radio (SRD)	Localization of unwelcomed UAVs	√	-	√	√	[57]

In [42], the autonomous distribution of resource management for UAV networks is being discussed. The authors have used the concept of game theory and highlighted various game models that may be used for optimal resource management among UAVs. In specific, five-game theory models were discussed, including coalition, potential, graphical, mean-field, and Stackelberg. Each model is explained based on its goals, design of utility function, and strategies, which implies their application areas.

Reference [43] focused on real-time planning for UAV’s path under dynamic conditions. A discrete algorithm, along with a probabilistic graph, is being used to achieve a path with no collisions. Another similar study to explore path planning for UAVs is conducted in [44]. In this study, the information from both static and dynamic paths is utilized to provide a path. A step-by-step adaptive path planning technique is proposed to achieve optimal results. Another interesting idea is that UAVs are considered additional users in the 5G cellular networks [45]. The study highlights various challenges that are encountered by service providers to facilitate new users.

2. Multiple Access and Routing Protocols

Multiple access and routing protocols are another challenging domain for UAV networks. As UAVs are supposed to be co-located with other networks, the natural choice for multiple access for UAVs was initially space division multiple access techniques [58,59,60]. However, as time passes, many researchers have started exploring other techniques that prove to be more efficient in the case of UAV networks. Most of these include orthogonal, non-orthogonal, or rate-splitting techniques. Table 3 provides a summary of related surveys on UAV multiple access and routing protocols.

In orthogonal multiple access techniques, the interference is reduced by ensuring that the simultaneous communications are orthogonal to one another. Multiple research studies [46,61] have been conducted to achieve this goal. Some propose to use time division multiple access (TDMA) and frequency division multiple access (FDMA) in their system models. Another group of researchers focused on using code division multiple access (CDMA) [62,63,64]. On the other hand, orthogonal frequency division multiple access (OFDMA) is explored by some researchers [47,65,66], while some other research studies are conducted exploring space division multiple access (SDMA) [67,68,69,70,71,72].

To overcome the above-stated limitation of orthogonal multiple access techniques, much work is being performed using non-orthogonal techniques, mainly using joint optimization of the base stations (BSs), including power [73,74,75], trajectory [76,77,78], and placement [79,80]. Although non-orthogonal multiple access techniques can perform better by covering many users with non-orthogonal yet correlated channels, however, there is a case when its performance can degrade exponentially, i.e., when the number of antennas is greater than the number of users scheduled. To avoid such circumstances, another effective technique is explored.

Rate-splitting multiple access techniques for UAVs seem to overcome the problems of both orthogonal and non-orthogonal multiple access.

3. Power Control and Energy Efficiency

For drones and unmanned aerial vehicles (UAVs), power control and energy efficiency are other very critical realms. As these network nodes are not on the ground, an uninterrupted supply of power and efficient utilization of energy become very crucial. Many research studies have addressed this area and proposed several solutions, some targeting multiple power sources, including battery [87,88], hydrogen fuel [89,90], solar [91,92], and hybrid [93,94], that can be utilized by the UAVs. While others focus on how efficiently energy can be consumed. Energy efficiency affects all the operations of a UAV. Many studies have been conducted to explore UAV placement for optimal energy efficiency [48,95].

Another aspect explored in this domain is path planning. Various approaches are being used to explore this area, including sample-based [49,96], space search [50], and biological searching techniques [97,98,99].

Table 3. Summary of related survey: UAV multiple access and routing protocols.

Scope	Autonomous Features	Computational Intelligence	Channel Modeling	Interference Management	Security and Safety	Reference
Channel access	Cyclic multiple channel access	√	Free-space path loss with LOS	-	-	[61]
MAC protocol	Energy consumption, Packet-error-rate (PER)	√	Free-space path loss with LOS	-	-	[62,63]
Performance evaluation of MAC	PER	√	Rician fading	-	-	[64]
Trajectory optimization	Trajectory planning	√	Free-space path loss with LOS, correlated Rician fading, Rayleigh fading, Rician K-factor	√	-	[67]
mm wave UAV cellular network	Beam forming	√	Quasi-static, Rayleigh fading	√	-	[68]
Channel access with time-modulated array (TDM)	Beam forming, performace	√	Free-space path loss with LOS	√	-	[70]
Trajectory optimization	Trajectory planning, power control	√	Additive white Gaussian noise (AWGN)	√	-	[74]
MAC protocol	Power optimization	√	Rician fading	√	-	[75]
MAC protocol	Throughput optimization	√	Free-space path loss	-	-	[76]
MAC protocol	Power optimization	√	Free-space path loss with LOS	√	-	[77]
MAC protocol	Trajectory planning, resource management	√	LOS and Non-LOS (NLOS)	√	-	[78]
MAC protocol	Power optimization	√	LOS, NLOS	√	-	[79]
MAC protocol	Power and placement optimization	√	Additive white Gaussian noise (AWGN)	√	-	[80]
Rate-splitting	Beam forming	√	Additive white Gaussian noise (AWGN)	√	-	[82]
Rate-splitting	Spectral efficiency	√	Additive white Gaussian noise (AWGN)	√	-	[86]

This entry is adapted from the peer-reviewed paper 10.3390/drones7050322

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.