OPEN ACCESS | Review

Multi-UAV networks for disaster monitoring: challenges

and opportunities from a network perspective
Indu Chandran and Kizheppatt Vipin
Department of Electrical and Electronics Engineering, BITS Pilani – K.K. Birla Goa Campus, India

Corresponding author: Indu Chandran (email: [email protected])

Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Abstract
Disasters, whether natural or man-made, demand rapid and comprehensive responses. Unmanned aerial vehicles (UAVs),
or drones, have become essential in disaster scenarios, serving as crucial communication relays in areas with compromised
infrastructure. They establish temporary networks, aiding coordination among emergency responders and facilitating timely
assistance to survivors. Recent advancements in sensing technology have transformed emergency response by combining the
collaborative power of these networks with real-time data processing. However, challenges remain to consider these networks
for disaster monitoring applications, particularly in deployment strategies, data processing, routing, and security. Extensive
research is crucial to refine ad hoc networking solutions, enhancing the agility and effectiveness of these systems. This ar-
ticle explores various aspects, including network architecture, formation strategies, communication protocols, and security
concerns in multi-UAV networks for disaster monitoring. It also examines the potential of enabling technologies like edge com-
puting and artificial intelligence to bolster network performance and security. Further, the article provides a detailed overview
of the key challenges and open issues, outlining various research prospects in the evolving field of multi-UAV networks for
disaster response.
Key words: multi-UAV network, architecture, mobility, optimization, disaster, security

1. Introduction Despite notable advancements in wireless communication

In recent decades, the world has faced a growing threat technology, addressing communication challenges during
from natural disasters, with their impact on human life and disaster relief activities remains an ongoing concern. The lit-
infrastructure becoming increasingly pronounced. A study erature in disaster research underscores significant limita-
by the Centre for Research on the Epidemiology of Disasters tions in executing first response operations, especially when
(CRED) and the United Nations Office for Disaster Risk Re- the terrestrial communication networks are partially dam-
duction (UNDRR) reveals a concerning trend, indicating that aged or completely destroyed. In search-and-rescue (SAR) mis-
between 2000 and 2021, natural disasters caused 1.23 million sions, the need for real-time and reliable communication is
deaths and incurred a substantial economic cost of US$2.98 important, as first responders must coordinate their actions
trillion (Cred and UNDRR Centre for Research on the Epi- and collaborate with other teams. Micheletto et al. (2018) ad-
demiology of Disasters (CRED) 2021). Despite technological vocate for the use of flying ad hoc networks to offer commu-
advancements improving disaster management efficiency, a nication support in disaster scenarios, with unmanned aerial
slight increase in the number of deaths compared to the pre- vehicles (UAVs) serving as communication gateways among
vious two decades is attributed to the rising frequency and first responders across various locations in the affected area.
severity of these events, particularly due to the effects of An equally interesting topic that is often compared with
global warming. The COVID-19 pandemic has underscored UAV networks is the Internet of Things (IoT) (Aggarwal and
the urgency for innovative disaster prevention and mitiga- Kumar 2020; Boursianis et al. 2022). Though both of these
tion methods. Simultaneously, as disasters like the cyclone networks work on a similar paradigm of collaborative wire-
in Indonesia, the Tornado in Kentucky, landslides in China, less networking, they differ in terms of fundamental prin-
the typhoon in the Philippines, and the flash floods in India ciples and applications (McEnroe et al. 2022). IoT is charac-
witnessed in the years 2020–2021 continue to wreak havoc, terized by a wide-ranging network of diverse devices, includ-
the need for effective post-disaster communication systems ing sensors, actuators, and everyday objects, interconnected
becomes evident. The golden time within the first hours after through existing communication infrastructures. Its primary
a disaster is crucial for saving lives, necessitating the devel- goal is to facilitate data exchange for automation and im-
opment of reliable and quickly deployable emergency com- proved efficiency in various domains, such as smart homes,
munication networks. healthcare, and industrial processes. In contrast, ad hoc UAV

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 1

 Canadian Science Publishing

Fig. 1. UAV applications in disaster management.

Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Fig. 2. Network scenario in a disaster.

networks focus on the dynamic formation of temporary com- The development of collaborative UAV networks also poses
munication links among UAVs, typically in scenarios where challenges, necessitating advanced communication and con-
traditional infrastructure is lacking or impractical. Ad hoc trol systems, distributed algorithms, and addressing safety
UAV networks prioritize mobility, flexibility, and rapid de- concerns. The control system must adeptly manage the in-
ployment, making them suitable for applications like disas- tricacies of multi-UAV operations, optimizing the trajectories
ter response, surveillance, or military operations. While both while ensuring safe and efficient mission execution. An ad-
technologies leverage wireless communication, their specific ditional challenge lies in developing distributed algorithms
architectures, scalability, and intended use cases distinguish that enable UAV collaboration while preserving autonomy ca-
them significantly. Its applications range from surveillance, pable of navigating diverse circumstances such as node fail-
medical aid delivery, SAR, and providing relayed communica- ures, communication failures, hostile attacks, and changing
tions to many more (Luo et al. 2019). Figure 1 outlines various environmental conditions. Thus, careful consideration of all
applications of UAVs during disasters. The dynamic and flexi- the network aspects is required to ensure efficiency, relia-
ble nature of these vehicles and the non-requirement of a pre- bility, and security. From a network perspective, the article
existing network infrastructure allow them to quickly adapt delves into multiple facets of multi-UAV networks for disas-
to changing environmental conditions. They can be deployed ter monitoring and surveillance, encompassing recent devel-
rapidly to areas that are difficult to reach by the responders, opments in communication protocols, network architecture,
thus playing a key role in remote or disaster-stricken loca- topology, path optimization, fault tolerance, routing, and se-
tions. This is particularly valuable for coordinating rescue curity.
and relief efforts as well as facilitating information exchange
among affected communities. This integration offers a dy-
namic and efficient solution to bridge gaps in communica-
1.1. State-of-the-art surveys and our
tion infrastructure, particularly in the aftermath of natural contributions
disasters or emergencies. Figure 2 illustrates a network sce- In recent years, there has been a surge in interest in de-
nario during disasters. velopment of multi-UAV networks, particularly focusing on

2 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Table 1. A comparison chart of existing surveys.

Deployment/ Mobility Communication Channel Coverage and
Reference architecture model protocols modelling Routing connectivity Security
Deepak et al. (2019)       
Shakhatreh et al. (2019)       
Jahir et al. (2019)       
Luo et al. (2019)       
Hentati and Fourati (2020)       
Garnica-Peña and Alcántara-Ayala       
(2021)
Matracia et al. (2022)       
Javaid et al. (2023)       
Our survey       
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

cutting-edge technologies like UAV swarm-based edge com- cuss the different aspects of network architecture for disas-
puting and machine learning (ML) techniques. While it is true ter surveillance, including topology, communication archi-
that there exist several survey papers on UAV networks, it is tecture, and formation strategies.
essential to recognize that the landscape of UAV applications
is constantly evolving, and emerging technologies demand
2.1. Topology
a fresh perspective and in-depth analysis. The uniqueness of
The topology of a network can have a significant im-
the article lies in its comprehensive exploration of network-
pact on its performance, efficiency, scalability, and reliabil-
centric aspects related to multi-UAV systems, particularly in
ity (Bekmezci et al. 2013). There are various types of network
the context of disaster monitoring. To the best of our knowl-
topologies related to single UAV and multi-UAV systems that
edge, there is no survey that has extensively covered all the
well suit various applications, each with its own advantages
network aspects from a disaster perspective. Table 1 outlines
and drawbacks.
the available surveys and evaluates each of them across vari-
ous network aspects.
The major contributions of this work are as follows:
2.1.1. Single UAV systems
Single UAV systems refer to the deployment and operation
r Provides valuable insights into network-centric aspects as-
of UAVs as independent entities, distinct from collaborative
sociated with multi-UAV systems. By exploring different or multi-UAV systems. It is important to note that while the
communication protocols, formation control techniques, system is not limited to a single UAV node, there could be
network architectures, trajectory optimization schemes, multiple UAV nodes working concurrently to deliver services.
data management and routing, and security aspects, the In this configuration, each UAV operates as an individual unit
work presents an overview of the conventional methodolo- under the direct control of a ground station (GCS), as denoted
gies that have evolved to help mitigate disasters, thereby in Fig. 3a. The range of flight for each UAV depends upon the
ensuring speedy recovery efforts to save human lives. communication range of the GCS, which serves as a central
r Emphasizes the paramount importance of security in the
command hub for vehicle navigation and operation. This in-
context of multi-UAV networks used for disaster monitor- dividualized approach allows for precise control and manage-
ing. The paper delves into the potential vulnerabilities that ment of each UAV, making it particularly suitable for tasks
these networks may face and explores solutions to over- where autonomy and simplicity take precedence over collab-
come the threats that can compromise their reliability and orative efforts.
functionality. From the perspective of a disaster recovery network, sin-
r Makes a significant contribution by identifying and ana-
gle UAV systems play a role in tracking victims and vehicle
lyzing the challenges inherent in deploying multi-UAV net- movements. However, a notable challenge remains in the
works to pave the way for unlocking the full potential of limited capability of a single UAV node to execute tasks ef-
multi-UAV networks in disaster applications. fectively. Some researchers are addressing this concern by
optimizing sensor functionality, hardware, and software to
enhance the capabilities of single-UAV nodes. A single UAV
2. System architecture system proves advantageous in observing multiple objects,
Disaster surveillance requires an efficient network archi- facilitating rapid object localization (Bekmezci et al. 2013).
tecture that can facilitate the coordination of UAVs with other Furthermore, it aids in investigating the precise extent of
agents in the network. A well-designed network architecture an operational area that has been disrupted post-disaster.
can enable effective communication and information shar- Tasks such as navigation, monitoring, and control can be ef-
ing, which are critical for situational awareness and decision- ficiently performed using a single UAV system, eliminating
making in disaster scenarios. In this article, we will dis- the need for additional resources to accomplish these tasks.

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 3

 Canadian Science Publishing

Fig. 3. UAV communication architectures: (a) single UAV system, (b) multi-UAV ad hoc network, (c) multi-group ad hoc network,
and (d) multi-layer ad hoc network.
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Nasr et al. (2019) suggested using UAVs for wireless commu- erating multiple rays intersecting with a horizontal plane
nication in areas without terrestrial infrastructure, with a to form a specific region. The steepest descent method with
focus on public safety, especially for rescuing shipwrecked the Armijo searching algorithm is used to estimate the tar-
individuals. Drones detect and locate victims through emer- get height, with the criterion of minimizing the area. The
gency signals from safety jackets, employing Received Sig- method does not require prior target height information,
nal Strength Indicator (RSSI) measurements for localization. making it versatile across different conditions.
The mobility of UAVs is utilized to create virtual anchors, en-
hancing victim localization. The proposed method involves
2.1.2. Multi-UAV systems
launching UAVs from a naval base to patrol the disaster
area and scanning for emergency signals from safety jack- During disasters, the limitations of single UAV systems
ets equipped with transmitters. A strategy to track ground- become apparent. The challenges of maintaining consistent
moving targets in complex indoor and outdoor environments communication links with the ground station can arise due
with UAVs based on computer vision is presented in Chen et to the dynamic and unpredictable nature of disaster environ-
al. (2017). An embedded camera is provided on the UAV plat- ments. Additionally, scalability poses a significant concern,
form to provide a real-time video stream to the onboard com- as a single UAV may be insufficient to cover large and com-
puter, where the target recognition and tracking algorithms plex disaster areas effectively. To address these challenges,
are implemented. A probabilistic estimate for the monitored researchers have turned to the concept of multi-UAV sys-
area by a single UAV tracking multiple objects is calculated in tems (Micheletto et al. 2018). A comprehensive review of var-
Albert and Imsland (2017). A constant velocity model is con- ious applications involving multiple UAV systems that have
sidered to establish a performance bound for position esti- been developed in recent years is presented in Skorobogatov
mate errors, determining the period of visit when monitoring et al. (2020). In a multi-UAV architecture, UAVs operate col-
multiple objects. Li et al. (2016) introduced a fast target local- laboratively, forming a network where they can communi-
ization method for a single UAV, addressing issues such as cate and coordinate with the other members of the net-
accuracy, real-time requirements, and cost constraints in ex- work. This shift from a centralized approach to a distributed
isting methods. The proposed approach involves multi-point and interconnected network allows UAVs to share informa-
observation of the target along an expected trajectory, gen- tion, make collective decisions, and adapt their strategies

4 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

based on real-time data. The advantage of multi-UAV systems works organize UAVs in a layered structure, facilitating effi-
lies in their ability to enhance redundancy, reliability, and cient communication and coordination. The network is di-
scalability during disaster response efforts (Micheletto et al. vided into tiers, each with specific functionalities. The higher
2018). By distributing tasks among multiple UAVs, the sys- tiers typically consist of UAVs with advanced computational
tem becomes more resilient to individual failures and dis- capabilities, acting as coordinators and relaying information
ruptions. Furthermore, the coordination and communication to the lower tiers. This hierarchical arrangement enhances
capabilities among peer UAVs within the network empower scalability and adaptability in dynamic environments. At the
these vehicles to make informed decisions collectively. This top tier, command-and-control UAVs oversee the entire net-
collaborative decision-making process enhances situational work, managing task distribution and communication pro-
awareness, enabling a more efficient and adaptive response tocols. Intermediate tiers may include relay UAVs responsi-
to the dynamic conditions prevalent in disaster-stricken ble for data transmission and coordination within their re-
regions. spective clusters. The lower tiers consist of task-specific UAVs,
In the context of multi-UAV systems, various network ensuring a distributed approach to mission execution. The
topologies are considered to facilitate communication and study presented by Celtek et al. (2019) introduced a solution
coordination among the UAVs, as shown in Figs. 3b, 3c, and to the limited communication range in drone applications by
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

3d. Three prominent topologies that are often explored in proposing a hierarchical tree topology-based wireless drone
the literature are multi-star, mesh, and hierarchical mesh network. The network comprises three main components:
(Esrafilian et al. 2020). In a multi-star architecture, each UAV control center, master drone, and slave drones. The tree
forms a local star connection, and one node from each star model emphasizes the effectiveness of a well-organized drone
extends its connection to the ground station. This architec- swarm in completing applications in a shorter timeframe. A
ture has extended scalability when compared to single UAV similar approach is addressed in Chen et al. (2021) on the for-
systems; however, scalability is limited by the number of con- mation control problem in fixed-wing UAV swarms through
nections that the ground station can handle. If any of these the establishment of a group-based hierarchical architecture.
central nodes fail, it can disrupt the communication and co- The UAVs are organized into non-overlapping groups, each
ordination within that particular group of UAVs. With mul- with a selected group leader. The group leaders coordinate
tiple star connections, the communication overhead also in- path coverage to manage the mission process among dif-
creases, as each UAV needs to maintain connections not only ferent groups, while followers track their direct leaders for
with its local central node but also with the ground station. inner-group coordination.
This architecture is not generally preferred in disaster mis- Clustered mesh networks organize UAVs into tightly knit
sions as the fixed structure of star connections may limit the groups or clusters, promoting collaboration within each
flexibility to adapt to changing mission requirements or dy- cluster and efficient inter-cluster communication (Uddin et
namic environmental conditions. al. 2018). Each cluster operates semi-autonomously, with a
In a mesh topology, every node is interconnected, allow- designated cluster head coordinating the activity. This ap-
ing information to hop through intermediate nodes to reach proach minimizes the need for direct communication be-
its destination. The self-forming and self-healing nature of tween all UAVs, reducing complexity and resource consump-
this network allows for the automatic establishment of con- tion. Within a cluster, UAVs can share information rapidly,
nections between neighbouring UAVs, contributing to a sys- facilitating real-time decision-making. The cluster head man-
tem that is both scalable and capable of reconfiguring it- ages intra-cluster communication, ensuring that tasks are dis-
self as UAVs enter or leave the network. In Portmann and tributed effectively among its members. Inter-cluster com-
Pirzada (2011), the applicability of wireless mesh network munication is achieved through designated gateway UAVs
(WMN) is assessed in the realms of public safety, disaster re- that relay information between clusters, optimizing the use
covery, and crisis management communication. The analy- of available bandwidth.
sis aims to determine the extent to which WMN technol- Clustered mesh networks are well-suited for applications
ogy aligns with the unique communication needs in sce- requiring localized coordination, such as search and rescue
narios related to public safety, disaster recovery, and cri- missions or monitoring specific geographic areas. They of-
sis management. Dey and Ray (2017) built an ad hoc mesh fer a balance between decentralized decision-making within
network of UAVs tailored for disaster management and re- clusters and coordinated efforts across the entire network.
mote sensing applications. The network established con- A fully autonomous and adaptive disaster recovery network
nections among multiple UAVs, enhancing the coverage of based on a traditional cell network structure is presented in
the observable area. The approach is based on process pat- Bupe et al. (2015) with 7-cell clusters arranged hexagonally,
terns, which define context-dependent behaviours of UAVs utilizing MAVLink for communication. The algorithm estab-
in various situations. Ganesh et al. (2021) introduced the lishes a hierarchical structure by designating higher-ranked
ubiquity network (UbiQNet) architecture, which leverages UAVs as super nodes, and centrally managing UAV cells. Zobel
drones to establish a mesh network for communication dur- et al. (2019) introduced strategies to enhance the efficiency
ing emergency situations. It was primarily aimed at allow- of inter-cluster flights for data ferry UAVs, seamlessly in-
ing victims to relay their situation and location to responders tegrating various optimization functions to accommodate
efficiently. scenarios with multiple objectives. A comprehensive frame-
The mesh topology can be broadly classified as hierarchi- work for leveraging mini-UAVs in disaster monitoring, em-
cal and clustered (Ueyama et al. 2014). Hierarchical mesh net- phasizing the benefits of a distributed network structure, is

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 5

 Canadian Science Publishing

Table 2. A brief analysis of the literature on UAV deployment.

Reference Description
Zhao et al. (2018) Presented two UAV deployment algorithms: a centralized one for known ground user positions and a distributed
motion control algorithm for on-demand coverage without global information
Zhang and Duan (2017) Developed an optimal deployment algorithm for emergency UAV deployment, minimizing delay in covering a
target area by dispatching diverse UAVs from a central location
Malandrino et al. (2019) Investigated using UAVs for wireless coverage in emergencies, and solved an optimization problem to maximize
user throughput and ensure fairness across disaster-affected areas
Busnel et al. (2019) Proposed a distribution algorithm for autonomous target discovery and self-organization of UAVs, ensuring
connectivity within a multi-hop aerial wireless network
Wang et al. (2019) Developed an adaptive UAV scheme using the majority rule for sector selection without real-time user tracking.
Optimizes UAV displacement to enhance throughput and transmission probability, accounting for user density
Panda et al. (2019) Aimed to create and deploy a cost-effective, user-friendly emergency communication network, supporting on-site
surveillance to ensure robustness, with connection management
Hydher et al. (2020) Investigated optimal placement of UAVs as aerial base stations for enhanced network connectivity, spectral
efficiency, and maximum quality of service (QoS) requirements
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Jin et al. (2020) Examined emergency response needs and utilized regional disaster susceptibility, traffic inconvenience, and terrain
complexity for recommendations on UAV and payload deployment in different regions
Masroor et al. (2021) Addressed UAV deployment for disaster management in wireless networks, optimizing UAV placement for
minimum distance, cost, and quantity through an integer linear optimization problem (ILP)
Lin et al. (2022) Proposed an adaptive UAV deployment scheme aiming to optimize the deployment location of UAVs for enhanced
coverage of ground nodes while minimizing energy consumption

presented in Joshi et al. (2020). The architecture facilitates covered. However, these methods either used probability
exploration of large and disjoint terrains through the forma- functions or neglected communication conditions in deploy-
tion of multiple clusters. To ensure isolation between clusters ment.
and optimize network energy, UAVs employ adaptive power Liu et al. (2020) utilized deep reinforcement learning (RL)
communication techniques (Ramesh 2014). for UAV deployment, emphasizing long-term communication
coverage. Nevertheless, this approach assumed a fixed UAV al-
titude and communication range, overlooking the impact of
2.2. Deployment obstacles in actual communication scenarios. Existing work
While UAVs offer various benefits in emergency scenar- often generates line of sight (LoS) and non-line of sight be-
ios, the challenging task of deploying them effectively for tween UAVs and GNs randomly based on probability func-
optimal coverage remains a significant issue. Al-Hourani et tions, neglecting real-world scenarios where specific infor-
al. (2014) introduced an analytical approach for optimiz- mation about the entire area, such as GN distribution and
ing the UAV altitude, aiming to maximize radio coverage building characteristics, is not directly accessible to the UAV.
for ground nodes (GNs). Building upon this, Mozaffari et In Lin et al. (2022), the objectives and constraints are re-
al. (2016) proposed an efficient deployment strategy utiliz- defined to consider real-world complexities. The approach
ing circle packing theory to ascertain the optimal UAV lo- optimized information collection for energy-efficient com-
cations based on the number of UAVs. Delving into the cov- munication at GNs by determining ideal UAV hovering loca-
erage problem, Alzenad et al. (2017) broke down UAV de- tions. To conserve energy given the limited computing power,
ployment into horizontal and vertical dimensions. The hor- the strategy minimizes broadcasts to the GNs. The UAVs are
izontal deployment was modelled as the smallest enclosing guided to optimal positions through an adaptive deployment
circle problem and a circle placement problem. In Dong scheme, considering environmental factors. Metaheuristic-
et al. (2018), a thorough examination of UAV communica- based multi-objective optimization algorithms like MOPSO,
tion characteristics and collaborative coverage led to the NSGA-II, SPEA2, and PESA-II are employed in Gupta and
derivation of an optimal deployment density function to Varma (2021) to find optimal UAV placements, balancing con-
achieve maximum coverage of GNs. The approach predom- flicting network objectives such as target coverage, Quality
inantly relied on average path loss for UAV location deter- of Service, and energy consumption in the post-disaster sce-
mination, aiming for maximum wireless coverage range or nario. Hydher et al. (2020) introduced a simplified approach
node count. In Zhao et al. (2018), a centralized deployment for optimizing UAV positions and assigning user equipment
algorithm and a distributed motion control algorithm for (UE) in an aerial base station (ABS) network with the objective
node coverage are presented. Lyu et al. (2017) introduced of maximizing total spectral efficiency while ensuring mini-
a novel placement algorithm, deploying the mobile base mal quality of service (QoS). Table 2 lists all the recent studies
station of the UAV in a spiral manner until all GNs were on UAV deployment.

6 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

2.3. Mobility model decision criterion alongside area coverage and network con-
The performance of protocols in ad hoc networks is signifi- nectivity, contributing to efficient energy management and
cantly influenced by node mobility (Bekmezci et al. 2013). The mission success based on neighbour-informed movements.
node movements are depicted using mobility models, which Azmi et al. (2021) explored existing research on UAV mobil-
can closely reflect real-life scenarios in the designated con- ity models, network technologies, and performance, with the
text. Analysis of mobility models for UAV networks in disaster primary goal of identifying the most effective mobility model
scenarios can be approached by considering two subclasses: for search and rescue missions.
those based on UAV mobility and those based on survivor mo-
bility. 2.3.2. Mobility model based on survivor/rescuer
mobility
Many existing mobility models, such as the random mo-
2.3.1. Mobility model based on UAV mobility
bility model, RWP model, random walk model, and random
While various mobility models, such as the random mo- direction model, exhibit unrealistic patterns in the context
bility model, random waypoint (RWP) model, random walk of disaster operations (Sahingoz 2014). In SAR scenarios, res-
model, and random direction model, have been proposed for cue teams do not move randomly; their movements are in-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

mobile ad hoc networks (MANETs), their suitability for aerial fluenced by obstacles like walls, debris, trees, and various
networks is limited for several reasons (Bani and Alhuda other environmental factors. So, the existing mobility mod-
2016). In the case of the random mobility models, the distri- els based on randomness do not hold good. In Aschenbruck
bution and movement patterns of nodes throughout the en- et al. (2007), a disaster area (DA) mobility model for disaster
tire simulation area do not correspond to the characteristics scenarios is discussed, which divides simulation areas into
of aerial networks. This discrepancy is particularly evident in sub-regions (e.g., incident site, hospital zone) with manual
scenarios involving aerial networks deployed for disaster re- node assignments. Despite efforts to mimic real scenarios,
sponse, where the unique dynamics and spatial constraints the model relies on the RWP mobility model for rescue agent
of such situations are not accurately captured by these con- movement, especially in the disaster site sub-area. Pomportes
ventional models. Therefore, alternative mobility models tai- et al. (2011) proposed a composite mobility (CoM) model for
lored to the specific challenges of aerial networks in disaster disasters, combining RPGM and Levy-walk models for group
scenarios are required. mobility, with obstacle avoidance based on a geographic map
A semi-random circular movement (SRCM) model was de- and the Dijkstra algorithm. However, the CoM model re-
signed in Wang et al. (2010) for UAVs to gather informa- lies on an accurate map and may pose challenges in disas-
tion while hovering at specific locations. The model is for- ter scenarios with modified or non-existent infrastructures
mulated for curved movement scenarios, with preliminary (Conceição and Curado 2013). A human behaviour for disaster
results demonstrating uniform node distributions and ro- areas (HBDA) mobility model is discussed in Conceição and
bust performance. The pheromone repel model (Kuiper and Curado (2013), designed to emulate realistic node movements
Nadjm-Tehrani 2006), derived from the three-way random in search operations for evaluating mobile wireless network
mobility model (Xie et al. 2018), guides UAV movements based performance in disaster scenarios. The HBDA prioritizes area
on a fixed turn radius and a probability distribution influ- coverage and minimizes search time, utilizing a force vec-
enced by a map of recently visited positions. This model pro- tor system to balance proximity and distance to neighbour
vides good coverage, but has connectivity issues (Kuiper and nodes. A three-dimensional mobility model designed for dy-
Nadjm-Tehrani 2006). Sanchez-Garcia et al. (2016) introduced namic, uncertain environments is discussed in Wang et al.
a self-deployment algorithm for aerial ad hoc networks in (2018b) to enhance emergency rescue missions by addressing
disaster scenarios, utilizing the Jaccard dissimilarity metric challenges posed by dynamic, distributed, dense, and irreg-
to guide UAV movements. The mobility model aimed to es- ular obstacles in rescue areas. Mahiddin et al. (2021) present
tablish a flexible communication infrastructure for disaster a review of existing mobility models for studying rescue en-
victims. Though the above models proved effective in cer- tity movements in disaster scenarios. The primary goal of this
tain aspects, they had certain flaws as they were available in work is to identify an ideal mobility model that realistically
two dimensions. In response to this, a 3D extension of the captures the movements of rescue entities in disaster scenar-
SRCM mobility model, termed 3DSRCM, is introduced in Mi ios.
and Dai (2021). The model incorporates a novel pheromone
track selection mechanism to enhance scanning coverage
and employs an orbit switch method for smoother trajectory 2.4. Network requirements
transitions. Additional features include a track lock mecha-
2.4.1. Rapid response
nism and highly uniform randomness to prevent potential
UAV collisions. A novel distributed mobility model for au- In disaster recovery networks, the foremost requirement
tonomous UAV fleets engaged in area exploration missions is a rapid response. The ability to swiftly deploy and es-
is presented in Messous et al. (2016). Unlike existing mod- tablish communication channels while considering the mo-
els, it uniquely integrates the remaining energy level as a bility of nodes is critical during emergencies. A quick

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 7

 Canadian Science Publishing

response ensures that timely and efficient coordination ity to handle the sessions generated by both the victims and
among emergency responders can take place, minimizing the the disaster relief crew members. Thus, the multi-UAV net-
impact of the disaster and potentially saving lives. work solution should have the capacity to support this traffic
scenario.

2.4.2. Network lifetime

2.4.7. Ease of use and equipment cost
In the aftermath of a disaster, when conventional com-
munication infrastructures may be compromised, the dis- In the context of disaster recovery networks, two critical
aster recovery network becomes a lifeline for emergency factors are ease of use and equipment cost. User-friendly fea-
response operations. During the rescue mission, the net- tures contribute to the ease of use, enabling response teams
work’s ability to deliver uninterrupted services is critical for to quickly and effectively establish communication infras-
coordinating efforts, disseminating information, and ensur- tructure. Affordable network equipment allows for scalability
ing effective communication among response teams. Beyond and broader accessibility, ensuring that even organizations
the completion of the rescue mission, the network remains with limited budgets can deploy effective communication
instrumental in supporting ongoing recovery operations, systems during crises. Balancing cost-effectiveness with per-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

providing a vital communication backbone until the restora- formance is essential to making disaster recovery networks
tion of regular infrastructure. accessible, sustainable, and impactful in their support for
emergency response efforts.
2.4.3. Interoperability
Interoperability among UAVs is a crucial aspect in disaster 2.4.8. Outdoor and indoor scenarios
recovery and emergency response operations. It refers to the A disaster-resilient network necessitates the capability to
ability of different UAVs, often from diverse manufacturers operate seamlessly in both indoor and outdoor environ-
or models, to communicate, collaborate, and share informa- ments. Disasters can strike in various settings, ranging from
tion seamlessly. In disaster scenarios, various UAVs may be de- urban areas and buildings to remote outdoor locations. The
ployed for tasks such as aerial reconnaissance, surveillance, adaptability of the network to function effectively in indoor
or search and rescue. Ensuring interoperability among UAVs spaces, such as shelters, facilities, or structures, is essen-
is essential for effective coordination and resource optimiza- tial for maintaining communication during evacuations or
tion. It allows different UAVs to share real-time data, coordi- within emergency response centres. Simultaneously, the abil-
nate their movements, and collectively contribute to a com- ity to operate in outdoor environments is critical for address-
prehensive understanding of the disaster area. ing disasters that occur in open spaces or areas with limited
infrastructure. Outdoor functionality is particularly crucial
2.4.4. Network coverage for search and rescue missions, surveillance, and the coordi-
nation of response efforts in the affected regions.
When a disaster occurs, the communication infrastructure
may be partitioned into a number of disjoint areas. In this
case, a disaster recovery system should be such that it can be 2.4.9. High precision for localization and search
quickly used to interconnect the different regions of disaster. operation
If no part of the pre-existing infrastructure is available after
Effective subject or survivor localization enhances the over-
the disaster, then it should be possible to deploy a solution
all efficiency of response teams by enabling them to navigate
that can cover the disaster area with one network or a cluster
complex and dynamic disaster environments. It allows for
of networks that can be interconnected to permit communi-
targeted and expedited deployment of resources to specific
cation across the affected area.
locations, minimizing response time and increasing the like-
lihood of successful outcomes.
2.4.5. Support for heterogeneous traffic types
The ability of the ad hoc network to support voice, data, 2.4.10. Scalability
and video applications is a major concern. Some proposed so-
In disaster missions, scenarios may arise where the net-
lutions are voice-only solutions, while others are data and/or
work experiences fluctuations in the number of UAVs, either
video-only solutions. A desirable feature of the disaster-
with the addition of more UAVs to the network or the depar-
resilient ad hoc network is its ability to support different traf-
ture of UAVs from the network. These dynamic changes can
fic types.
be influenced by evolving mission requirements, the need
for additional surveillance or rescue capabilities, or the com-
2.4.6. Network capacity
pletion of specific tasks by individual UAVs. Network adapt-
Consider a situation where some or all the victims in a dis- ability is crucial to accommodate these variations, ensuring
aster area have devices with which they can communicate seamless integration and disengagement of UAVs without
with the outside world, but the infrastructure is damaged by compromising overall communication and operational effi-
the disaster. The ad hoc network must have sufficient capac- ciency.

8 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

2.5. Critical insights and gaps in existing assuming uniform user distribution have been discussed, an
studies efficient deployment strategy tailored for disasters is lacking
In disaster scenarios, robust system architectures are es- in the literature. Existing studies often treat UAVs as inde-
sential for UAV networks to enable efficient coordination pendent entities without considering backhaul connectivity.
and rapid decision-making. Relevant UAV architectures to Moreover, they often assume fixed user locations, which is
match the scale, type, and specific challenges of each dis- not the case in disasters, where rescue teams and victims
aster are crucial. Factors like flexibility, adaptability, robust- may relocate to safer areas. Therefore, object detection and
ness, and operation in resource-limited environments must tracking have become vital. While various mobility models
be considered when designing UAV architectures for disaster based on randomness have been explored, ground mobility
response and monitoring. For localized incidents or smaller- during disasters cannot be approximated as random. Further-
scale disasters, simpler UAV architectures may suffice, with more, existing deployment models overlook victims stuck in-
single UAV systems or small groups performing tasks such as side buildings, necessitating the incorporation of non-line-of-
damage assessment, search and rescue, or supply delivery to sight communication paths into the optimization problem.
isolated areas. However, for widespread disasters covering A generalized mobility model adaptive to all types of disas-
large or multiple areas, more complex architectures are typi- ters is lacking in current studies, necessitating further explo-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

cally needed. Multi-UAV systems, swarm-based approaches, or ration. While the studies by Aschenbruck et al. (2007) and
networked UAV fleets operating in coordination with ground- Pomportes et al. (2011) consider adaptive mobility schemes
based and satellite systems can be explored to provide com- for rescue teams and users, there is a need to incorporate ob-
prehensive coverage and efficient aid delivery. Despite their stacle avoidance without relying on accurate maps, as such
theoretical efficiency, practical design challenges and envi- maps may not be available in disaster scenarios. Models based
ronmental limitations require further investigation. A ma- on human behaviour have also been explored, but predicting
jor challenge is the heterogeneity of UAVs, where theoreti- real-time human behaviour is not practically feasible as it de-
cal analysis often assumes homogeneous sets of UAVs, but in pends on the type of disaster as well.
practice, differences in batteries, payload capacity, and en-
durance impact architecture selection. While heterogeneous 3. Communication aspects
network studies are recently gaining attention, their practi-
cality remains largely unexplored. Terrain also plays a cru- In the aftermath of a disaster, establishing swift and reli-
cial role in communication coverage, with obstacles such as able communication networks is paramount for effective co-
mountains or buildings potentially blocking signals between ordination in relief efforts. UAVs play a crucial role by form-
ground bases and UAVs, particularly in urban areas. Node ing ad hoc networks through various wireless technologies.
failures can disrupt network topology, necessitating topol-
ogy updates when a UAV fails or is introduced. While UAV 3.1. Wireless technologies
coordination enhances system reliability by accommodating While the literature explored multiple wireless links such
topology updates, it also places additional workload on ac- as Bluetooth (Khan et al. 2019) and zigBee (Acosta-Coll et al.
tive UAVs due to missions previously assigned to failed agents. 2021) to provide resilience in wireless sensor networks, the
To maintain system efficiency without overloading network choice narrows down to considerations of cost, communica-
agents, controlled redundancy is crucial. Task offloading to tion range, compatibility, and regulatory compliance for a
peer UAVs or high-altitude platforms (HAPs) is considered UAV network. Cellular communication solutions such as 2G,
for complex computations. In disjoint missions, clustering is 3G, 4G, and others offer expanded coverage, yet their effec-
considered beneficial, where one UAV acts as the cluster head tiveness hinges on the presence of base stations (BSs), mak-
and others as cluster members. However, all these mecha- ing them unsuitable for emergency communication systems
nisms involve developing robust localization algorithms re- in scenarios lacking infrastructure (Avanzato and Beritelli
silient to environmental uncertainties. Localization and map- 2019). Furthermore, they entail a substantial drain on device
ping methods such as SLAM, GNSS, and computer vision face energy, contributing to escalated operational costs (Mekki
difficulties in disaster environments due to factors like de- et al. 2019). Technologies such as Wi-Fi and WiMAX stand
bris, smoke, and low visibility. Challenges such as poor light- out, with Wi-Fi (IEEE 802.11) gaining prominence due to its
ing and occlusions complicate feature extraction and local- ubiquity, cost-effectiveness, and the widespread use of smart-
ization accuracy (LA) in computer-vision methods. The ef- phones equipped with Wi-Fi capabilities among survivors. Al-
fectiveness of these systems in disaster scenarios remains though WiMAX is superior in coverage, its dependence on
underexplored. specific infrastructure and receiver requirements can limit
Another critical aspect of architectural design is resource its utility in the dynamic and often unpredictable context of
allocation. While path planning and trajectory optimization disaster response. Harrington et al. (2020) outline the devel-
have been extensively studied in the literature, research fo- opment of a multiple drone network within a Wi-Fi environ-
cusing on maximizing the utilization of available resources ment, highlighting the ability to coordinate and control the
by deploying UAVs in optimal numbers is still limited. Addi- drones autonomously through the exchange of flight com-
tionally, efficient 3D placement of UAVs to ensure maximum mands. However, latency in transmitting commands could
coverage of the area and the victims is in its early stages of ex- potentially disrupt the coordination and synchronization of
ploration. The altitude of the UAV operation also impacts the flight plans within a network of multiple drones. Hayat et
payload it can carry. Although heuristic-based optimizations al. (2015) investigated the application of IEEE 802.11n and

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 9

 Canadian Science Publishing

802.11ac standards in aerial WiFi networks. Indoor and out- potential use cases and effective operational ranges of the sys-
door performance tests, including multi-sender networks in tem, particularly in scenarios where distant UAVs could uti-
infrastructure and ad hoc modes, are conducted to simulate lize nearby flying UAVs as relays to broadcast their telemetry
real-world UAV scenarios. Though implementation in small data to the LoRaWAN ground network. Utilizing IEEE 802.11s-
UAVs showed promising data rates and throughput indoors, it based mesh networking within UAV swarms, as proposed in
faced challenges outdoors due to low signal strength, result- Morgenthaler et al. (2012), offers direct visibility between
ing in decreased throughput as the UAV moved away from nodes, particularly at high altitudes, enabling Wi-Fi signals
the base station. For it to be considered for disaster appli- to propagate freely without physical obstacles. However, ele-
cations, additional research is required to enhance reliabil- vated flight altitudes mitigate interference from urban Wi-Fi
ity and throughput, particularly considering the high mobil- networks, potentially expanding operational ranges.
ity of UAVs. Gu et al. (2015) focused on the integration of To harness the advantages of communication protocols
WiFi signals into airborne networks for the swift deployment and mitigate their respective limitations for the imple-
of WiFi infrastructures, particularly in scenarios lacking ex- mentation of UAV swarms, a mesh strategy is proposed
isting communication infrastructure, such as disasters. A in Davoli et al. (2021). This involves integrating LoRa- and
notable aspect of the study involves the utilization of direc- IEEE 802.11s-based communication patterns through oppor-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

tional antennae for long-range WiFi signal transmission, aim- tunistic switching, handling, and management mechanisms.
ing to boost signal strength and minimize interference with This approach aims to address various use cases involving
other WiFi channels. The research conducted field tests to as- UAV swarms while striking a balance between operational
sess how distance impacts WiFi signal throughput, shedding range and available bandwidth. By combining different wire-
light on the practical considerations of employing directional less communication networks— —utilizing a LoRa-based layer
antennae in UAV-based communication systems. for long-distance and payload-constrained communications,
Various research studies have explored the potential adop- and an IEEE 802.11s-based layer for mid-range and uncon-
tion of LoRa-based protocols, such as LoRaWAN, for UAV com- strained payload applications— —a coexistence framework is
munications, showcasing diverse applications, implementa- established. Operating on separate layers, a smart switching
tions, and use cases with long-range capabilities. In Stellin mechanism is essential for selecting the most appropriate
et al. (2020), LoRaWAN is utilized where multiple UAVs func- network based on real-time considerations such as task re-
tion as aerial LoRaWAN gateways. This configuration provides quirements and network quality conditions. This proposed
network coverage for ground-based LoRaWAN end nodes, switching mechanism is designed to optimize performance
leveraging LoRaWAN solely for ground-to-UAV applications in diverse application scenarios by considering the capabili-
in emergency scenarios, while the primary network relies ties and constraints of each available communication proto-
on Wi-Fi, assuming LOS links. A similar methodology is out- col in terms of operational range and admitted payload.
lined in Sharma et al. (2018), where multiple UAVs collaborate
to establish a sophisticated LoRaWAN-centric urban surveil-
lance system, ensuring efficient and targeted network cov- 3.2. Communication link design
erage. However, across these referenced works, LoRa-based Unmanned aerial systems employ two distinct wireless
communications predominantly serve ground-to-UAV inter- communication links (Li et al. 2020). The first is the air-to-
actions, connecting on-field nodes with the LoRaWAN net- ground (A2G) and ground-to-air (G2A) link, establishing con-
work. This approach significantly extends network coverage nectivity between the UAV and a GN. The second is the air-to-
for specific use cases in designated operational regions but air (A2A) link, facilitating communication between multiple
is not suitable for critical missions in remote areas. Another UAVs engaged in collaborative flight tasks. Design considera-
approach to UAV-enabled flying LoRaWAN networking is de- tions for both links involve tailored specifications to accom-
tailed in Saraereh et al. (2020), with a specific focus on disas- modate particular communication scenarios, incorporating
ter management applications. This strategy combines flying factors such as communication connectivity, flight trajecto-
LoRaWAN GWs with Wi-Fi communications from a UAV-to- ries, and the probability of successful completion of flight
ground access point (AP) and LoRaWAN end nodes. The ob- tasks.
jective is to accompany emergency operators, collecting data A mobility-aware multi-UAV placement strategy for estab-
and positions. Even in this context, LoRa-based communi- lishing a disaster-resilient communication network is dis-
cations play a vital role in gathering data from ground-end cussed in Peer et al. (2020). The strategy formulates an op-
nodes, ensuring a reliable network for emergency applica- timization problem aiming to maximize the coverage of
tions while tracking the movement of ground devices. An ground users by UAVs while adhering to the constraint of
alternative scenario involving UAV-to-ground LoRaWAN com- UAV flight time. Notably, the study focuses on emergency
munications is outlined in Rahmadhani et al. (2018). The au- first responders as the ground users, modelling their mobility
thors detailed the utilization of LoRaWAN for communica- within the disaster-affected region. The connectivity restora-
tion between flying UAVs and the existing LoRaWAN ground tion strategy outlined in Kurt et al. (2021) involves moving
network. This application focuses on transmitting crucial specific nodes based on a connected dominating set heuris-
telemetry data, including GNSS coordinates (latitude, longi- tic, designed to optimize connectivity in dynamic and chal-
tude, and altitude), drone speed, and heading direction. How- lenging environments. The strategy focused on maintaining
ever, the work has not explored the use of LoRa for communi- a core for the wireless ad hoc network, ensuring that every
cation among multiple flying UAVs. This oversight limits the node remains reachable to every other node. By identifying

10 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

and moving specific nodes strategically, the approach aims icant performance degradation compared to alternative ap-
to minimize movement while maximizing the restoration of proaches.
connectivity, a critical factor in disaster response where time Matolak and Fiebig (2019) offer an overview of the wire-
and resources are often constrained. Li et al. (2020) proposed less channel for UAVs. A comparative analysis with various
an optimized UAV-aided data collection design for emergency other channel types is conducted, and existing measure-
scenarios, prioritizing mission completion time. It strategi- ments and models are examined. Cui et al. (2020) conducted
cally addresses trajectory, altitude, velocity, and data link A2G channel measurements at various frequencies (1, 4, 12,
optimization challenges with ground users, employing algo- and 24 GHz) for UAV-based wireless communications. Utiliz-
rithms tailored to minimize mission time effectively. How- ing the 3rd Generation Partnership Project channel model, it
ever, addressing the optimization problem involving continu- extracted crucial path loss coefficients and introduced a novel
ous trajectory variables may present a challenge due to poten- autocorrelation model for shadow fading. Statistical analysis
tial computational complexity and scalability issues. More- reveals the log-logistic distribution as the best fit for small-
over, discretizing the trajectory for optimization could poten- scale fading. Second-order statistical characteristics, such as
tially lead to errors and reliance on the number of discretized level crossing rate (LCR) and average fade duration (AFD), are
points, which may influence the overall quality of the solu- employed for a comprehensive understanding of the fading
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

tion. Do-Duy et al. (2021) addressed the joint optimization of behaviour. A detailed survey on UAV communication chan-
real-time deployment and resource allocation for UAV-aided nel modelling is presented in Yan et al. (2019). The paper
relay systems in emergency scenarios such as disaster relief. addressed the research gaps by conducting a comprehen-
It introduced a rapid K-means-based user clustering model sive analysis of A2G, and A2A channel measurements and
and optimal power and time allocation, utilizing UAVs as fly- modelling in the context of UAV and aeronautical commu-
ing base stations for immediate network recovery and ongo- nications across diverse scenarios. It offers design guidelines
ing connectivity maintenance during and after disasters. The for UAV communication link budget management, consider-
work has only analyzed a straightforward network configu- ing link losses and channel fading effects. Additionally, the
ration consisting of a source node and a single UAV-assisted work analyzed the benefits of receive/transmit diversity gain
relay node. This simplification limits the complexity of the and spatial multiplexing gain achieved through multiple-
network model in practical, harsh environments. antenna-assisted UAV communications. Jiang et al. (2020) pro-
In contrast to the exploration of A2G/G2A links, studies posed a 3D MIMO channel model for A2G communications in
focusing on A2A links concentrate on facilitating collabora- UAV environments, considering the dynamic motion of both
tive flight tasks involving multiple UAVs. Given that A2A can the UAV transmitter and ground receiver. They introduced
be modelled as line-of-sight propagation, its link quality is an angular estimation algorithm for real-time departure and
influenced by the mobility of multiple UAVs. Consequently, arrival angles and also explored time-varying spatial cross-
the research direction shifts from enhancing A2A link qual- correlation functions and temporal auto-correlation func-
ity to coordinating the positions of multiple UAVs for col- tions based on different moving directions and velocities. The
laboration. Due to the dynamic mobility of UAVs, these col- effect of Doppler power spectral densities and power delay
laborative positions constantly change, necessitating the de- profiles on the channel model is also studied.
velopment of routing protocols, a specific aspect covered in
Section 5.
3.4. Trajectory planning and localization
In a disaster scenario where time is of the essence, effi-
3.3. Channel modelling cient trajectory planning and localization can provide real-
In UAV-to-ground communications, the wireless environ- time mapping of the disaster area, thereby helping respon-
ment is marked by high mobility and LoS propagation con- ders to act swiftly. By planning trajectories, responders can
ditions (Matolak and Sun 2015; Li et al. 2019). However, the cover a larger area and increase the chances of locating sur-
assumption of LoS is often not met due to the presence of sub- vivors or assessing the extent of damage in a more organized
stantial obstacles (Khuwaja et al. 2018), leading to shadow- manner. This information is vital for decision-making, allow-
ing or large-scale fading. This phenomenon results in unpre- ing responders to adapt their strategies based on the current
dictable variations in mean envelope levels. Another distinct situation and ensuring a coordinated and effective disaster
aspect observed in mobile communication environments is response.
double-scattering (DSc) propagation. This type of fading oc- An intriguing facet of employing UAVs as network in-
curs when either the transmitter, the receiver, or significant frastructure involves crafting optimal trajectories, consid-
scatterers in their vicinity are in motion (Salo et al. 2000; ering factors such as collision avoidance, meeting terres-
Andersen and Kovacs 2002). A new channel model that con- trial user demand, addressing energy constraints, and man-
siders both mobility and shadowing effects is presented in aging flight duration. This has spurred researchers to ex-
Bithas et al. (2020). The study analyzed a UAV-based commu- plore optimal path planning within UAV-assisted wireless
nication system operating in a shadowed DSc channel. The networks (Jiang and Swindlehurst 2012; Di Franco and But-
model is versatile, representing various fading/shadowing tazzo 2015; Fadlullah et al. 2016; Wang et al. 2018a). In Wang
conditions through easily evaluated mathematical functions. et al. (2018a), the authors delved into the joint optimiza-
Additionally, a low-complexity UAV selection policy is pro- tion of a UAV path and transmit power, aiming to maxi-
posed, reducing signal processing complexity without signif- mize the least average throughput within a specified time

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 11

 Canadian Science Publishing

frame. They proposed a suboptimal algorithm for the as- convergence time and LA with lower computational costs. It
sociated non-convex optimization, incorporating trajectory also outperforms typical routing protocols in packet delivery,
and power budget constraints. Meanwhile, Jiang and Swindle- delay, and overhead, while consuming less energy and pro-
hurst (2012) focused on a multi-antenna UAV base station, longing network lifetime.
examining optimal trajectories to maximize the sum rate of
uplink communications. Di Franco and Buttazzo (2015) intro-
duced an energy-aware UAV trajectory planning strategy for 3.5. Critical insights and gaps in existing
photogrammetric sensing of a designated area. studies
While Jiang and Swindlehurst (2012), Di Franco and But- A robust channel model is essential for effective link design
tazzo (2015), and Wang et al. (2018a) concentrated on a soli- and evaluation in multi-UAV networks, especially in disaster
tary UAV, Fadlullah et al. (2016) considered a swarm of UAVs scenarios where reliable communication is vital. While UAV-
to ensure maximum network coverage and connect discon- to-ground links share similarities with satellite-to-ground
nected terrestrial heterogeneous networks. In response, re- communications, they also possess unique aspects specific
searchers introduced a centralized dynamic path planning to UAV operations that must be carefully considered. Fac-
algorithm deployed at the control station, aimed at improv- tors such as obstacles, ground station height, and UAV alti-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

ing network latency and throughput. Wu et al. (2018) delved tude significantly influence line-of-sight probability and sig-
into the joint optimization of UAV trajectory, user schedul- nal propagation, thus impacting path loss and fading char-
ing, and association to maximize the minimum downlink acteristics. Multipath components, which vary with UAV al-
throughput for terrestrial users. A survey on 3D UAV place- titude, contribute to the severity of fading and delay dis-
ment and trajectory optimization is presented in Lakew et al. tortion. Additionally, the relative difference in velocities be-
(2020), also exploring the existing challenges and research tween communicating entities can induce a Doppler shift,
issues. Demiane et al. (2020) focused on using UAVs in disas- which varies based on UAV speed and altitude. Antenna pat-
ter scenarios with compromised communication infrastruc- terns, influenced by UAV material and structure, play a crit-
ture to accurately locate potential survivors. The approach ical role in signal reception and transmission. Thus, chan-
involves collecting RSSI data from mobile devices in differ- nel modelling must encompass diverse environmental con-
ent cells of varying importance. The study formulates and ditions, including terrain, obstacles, and reflective surfaces,
solves two subproblems: identifying strategic waypoints for to accurately represent real-world scenarios. Future studies
UAV positioning and constructing an optimal UAV trajectory. should explore a wide range of UAV-to-ground communica-
However, challenges surrounding the reliability of wireless tion scenarios, considering various velocities, antenna con-
signals in disaster zones, ensuring robust communication be- figurations, and time-varying or time-invariant channel char-
tween the UAV and the server, and accurately locating victims acteristics. Hybrid deterministic/stochastic models may pro-
in rapidly changing and unpredictable conditions pose signif- vide more accurate representations of UAV-to-ground chan-
icant hurdles. Numerous techniques have been suggested for nels, incorporating multipath clustering and spatial consis-
solving the localization problem in UAV networks, with the tency tracking.
majority relying on distance measurement methods such as The type of traffic flow within a network, whether real-
trilateration and bilateration (Lee and Scholtz 2002; Gezici et time, periodic, or delay-tolerant, significantly influences net-
al. 2005). However, these methods often face challenges like work design. For missions like search and rescue, coordi-
flip ambiguity, introducing substantial measurement errors nation and data sharing are crucial. Centralized decision-
due to environmental noise affecting the accuracy of distance making includes the exchange of location and heading direc-
measurements between nodes (Zhang et al. 2012). The high tion, while online decisions require bidirectional communi-
mobility of UAVs in network routing protocols necessitates cation. Decentralized missions require the exchange of times-
enhanced location accuracy at frequent intervals. UAV nodes, tamps and maps. Distance between units is another impor-
characterized by limited transmission range, dynamic links, tant factor affecting communication links. Cellular technolo-
and constrained battery power, require special consideration gies such as LTE, WiMAX, and UMTS are suitable for longer
for developing energy-efficient routing protocols. The local- distances, but they rely on fixed infrastructure, which in dis-
ization method in the routing protocol serves the purpose of asters may not be available. Shorter distances often rely on
swiftly determining the precise location of any node. Existing ZigBee and WiFi due to their suitability and availability. Ex-
localization methods vary in terms of LA, computation cost, isting research underscores the superiority of 802.11 mesh
and error rate, with categorizations based on characteristics networking over cellular for small networks. However, as net-
like static versus mobile nodes, anchor-based versus anchor- works grow, WiFi emerges as a viable option due to its scala-
free, sparse versus dense network, indoor versus outdoor, bility. Yet, for very large networks, WiFi alone may not suf-
and range-based versus range-free (Guo et al. 2019). Range- fice, prompting exploration of alternative technologies. In
based localization methods, known for higher location accu- critical scenarios like disaster missions, where real-time data
racy and lower error rates, are preferred in a broader range transmission is vital, the feasibility of using WiFi for large-
of applications compared to range-free localization meth- scale aerial coordination within the latency bounds needs
ods. A swarm intelligence-based localization (SIL) and cluster- to be examined. Additionally, longer distances impact UAV
ing (SIC) scheme for emergency communications in UAV net- energy consumption, prompting consideration of cellular
works is presented in Arafat and Moh (2019a). The 3D SIL al- handover-like schemes. Studies across IEEE 802.11 standards
gorithm, using particle swarm optimization (PSO), improves consistently demonstrate WiFi’s ability to support various

12 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Fig. 4. (a) UAV-swarm assisted edge computing. (b) Edge computing-assisted UAV swarm.
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

applications with fewer multi-hops. Nonetheless, unresolved mote servers. These servers process the information, conduct
issues persist for modern UAV applications relying on WiFi. complex analyses, and deliver valuable outputs back to the
Emerging alternatives enable unidirectional communication, UAVs or disaster response teams. By offloading data process-
but bidirectional support for diverse data types remains lim- ing tasks to remote servers, it reduces the computational bur-
ited. LoRaWANs, with their standalone and hybrid schemes, den on the UAVs themselves. As a result, UAVs can conserve
are being debated regarding their suitability for UAV commu- onboard resources, such as processing power and battery life,
nication. which are critical for prolonged and efficient mission execu-
tion.
Among the recent UAV swarm-enabled edge computing ap-
4. Data processing and decision-making plications, a prominent focus of research has been on uti-
Effective decision-making is critical in all stages of disaster lizing UAV swarms as Mobile Edge Computing (MEC) servers
management, influencing the success of rescue missions and (Zhang et al. 2020a; Abrar et al. 2021; Zhan et al. 2021). In this
related events (Alsamhi et al. 2022). Handling the complexity paradigm, specific UAVs within the swarm are equipped with
of big data analysis during disasters calls for computational computing capabilities, effectively acting as MEC servers.
intelligence and real-time algorithms. These technologies en- These UAVs can process data locally or offload computation
able swift decision-making, analyze diverse data structures, to other UAVs within the swarm. This concept, referred to as
extract relevant information, and present insights through UAV swarm-assisted edge computing, as shown in Fig. 4a, en-
various strategies (Donratanapat et al. 2020). As UAVs are of- ables decentralized data processing, reducing reliance on dis-
ten limited in size and payload capacity, on-board data pro- tant central servers and minimizing communication delays.
cessing must strike a balance between processing capabili- It is particularly advantageous for applications requiring real-
ties and energy efficiency to ensure optimal flight endurance time data analysis and decision-making. Another paradigm
and mission success. With the rise of multi-UAV swarm sys- involves UAV swarms as users seek computing services from
tems, data processing takes on a whole new level of complex- nearby servers due to their inherent limitations in size and
ity. In swarm-based architectures, UAVs collaborate and co- computational capabilities (Zhou et al. 2020), as shown in
ordinate their actions to accomplish shared objectives. This Fig. 4b. In this scenario, UAVs constantly interact with edge
requires seamless communication and distributed decision- computing resources, leveraging optimization methods to
making among the swarm members. Each UAV in the swarm handle switch-overs. Swarm optimization is also resistant to
may possess different data processing capabilities, necessitat- noise and uncertainty in the optimization problem, making
ing intelligent task allocation and load balancing to optimize it appropriate for real-world applications in uncertain envi-
the overall performance of the swarm. During disaster situa- ronments. Furthermore, because they are population-based
tions, efficient data processing becomes vital to derive action- methods, these algorithms can explore several solutions in
able insights from the collected data promptly. One approach parallel, providing a good balance between exploration and
to process data during disaster situations is by leveraging re- exploitation of the search space. This feature increases the

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 13

 Canadian Science Publishing

likelihood of finding the global optimum rather than the lo- cessing is a viable solution, particularly crucial in such sce-
cal optimum, which is very useful when dealing with nonlin- narios, facilitating tasks like structural damage assessment
ear and multi-modal situations. Some of the widely used op- and hazard identification. In large-scale disasters like floods
timization algorithms include PSO, ant colony optimization or wildfires, remote sensing technologies enable UAVs to
(ACO), bee colony optimization (BCO), firefly algorithm (FA), gather valuable data over vast areas, aiding in rapid dam-
gray wolf optimization (GWO), and glow-worm swarm algo- age assessments and response coordination. Techniques like
rithm (GSO) (Vanitha and Padma 2014). A comparison of all data fusion and compression optimize data transmission, re-
the swarm optimization algorithms across various parame- ducing latency. In smaller to medium-scale disasters, pho-
ters like computational complexity, convergence speed, mer- togrammetry creates detailed 3D maps, offering insights into
its, and limitations is presented in Table 3. structural integrity and terrain changes. Furthermore, ML al-
Innovative approaches incorporating information technol- gorithms enhance real-time decision-making across all dis-
ogy, artificial intelligence, ICT tools, and ML are instrumen- aster scenarios, trained on extensive datasets for tasks like
tal in enhancing all stages of disaster response (Chamola et object detection and survivor identification. However, con-
al. 2021; Hernandez et al. 2022; McEnroe et al. 2022). The ap- cerns arise regarding predictions beyond available data and
plication of AI techniques in disaster management encom- the evolving nature of hazards, challenging the reliance on
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

passes various ML and DL methods. ML methods, including AI for resource deployment in disaster management. Despite
SVM, NB, DT, RF, LR, and KNN, are employed. DL methods these challenges, developing powerful and cost-effective AI-
involve diverse artificial neural network architectures like based tools remains an emerging area of research, aiming to
convolutional neural networks (CNNs), MLP, recurrent neu- improve analysis accuracy and speed.
ral networks (RNNs), LSTM, transformers, and GANs (Yu et al. In scenarios where UAVs face resource limitations, such as
2018; Arinta and Andi 2019; Sun et al. 2020). ML and DL fa- storage and processing power, data offloading to external sys-
cilitate the utilization of extensive and intricate datasets to tems becomes necessary to conserve onboard resources and
develop predictive systems and aid in disaster response and extend mission duration. Various data offloading schemes,
recovery. These techniques leverage the manipulation of di- including transmission to cloud servers or ground stations,
verse data types from multiple sources to identify patterns, have been discussed in the literature. While this may intro-
offering valuable insights otherwise challenging to discern. duce latency compared to onboard processing, it proves ef-
Munawar et al. (2021) provide an overview of recent progress fective in disaster missions. However, for latency-critical mis-
in flood management, particularly in the post-disaster phase, sions like search and rescue, UAV-based edge computing tech-
emphasizing developments in image processing, artificial in- niques offer a more efficient solution. By integrating UAVs
telligence (AI), and integrated approaches. The authors have with high onboard processing capabilities, while lower end
also focused on reviewing the history of flood events and re- UAVs forward the data for analysis, bandwidth usage and la-
sponses in Munawar et al. (2022), emphasizing the utiliza- tency are minimized, ensuring efficient data handling in con-
tion of AI techniques for flood risk mitigation. It proposed gested or unreliable communication networks. Edge comput-
an AI/ML-based early flood warning system for aged care fa- ing devices enable UAVs to leverage their onboard processing
cilities in the Hawkesbury-Nepean region, incorporating UAV capabilities to solve complex data tasks without relying on ex-
and path planning for timely disaster response and evacua- ternal infrastructure, supporting response efforts in remote
tion. Linardos et al. (2022) aim to offer an overview of research or inaccessible areas. Though HAPs are considered for data of-
studies conducted since 2017, focusing on the application of floading, practical implementations are in their early stages.
ML and deep learning (DL) methods in disaster management. Challenges like intelligent HAP-UAV channel modelling and
Specific areas of interest include disaster and hazard predic- security need extensive research. Balancing energy efficiency
tion, risk and vulnerability assessment, disaster detection, and security, along with automatic switch-over schemes dur-
early warning systems, disaster monitoring, damage assess- ing node failures, is crucial for uninterrupted data flow in
ment, post-disaster response, and relevant case studies. Ad- real-time missions. Integrating these advancements into the
ditionally, the work analyzed recently developed ML and DL network infrastructure can significantly enhance disaster
applications in the field of disaster management. A detailed mitigation efforts.
survey of MI/AI schemes for disaster management is detailed
in Sun et al. (2020), providing thorough insights about the
role of these techniques at different phases of the disaster.
5. Data routing schemes
The work has also identified potential challenges for these When the distance between a UAV and a ground station
techniques to be explored by the research community. surpasses the communication range, an alternative UAV can
act as a relay to maintain connectivity between them. How-
ever, the challenges of an unstable wireless link, frequent
4.1. Critical insights and gaps in existing topology changes, and the high mobility of UAVs make tra-
studies ditional MANET routing protocols impractical for UAV net-
UAVs rely heavily on their computational capabilities for works. Asadpour et al. (2014) observed the shortcomings in a
data processing and decision-making. However, not all pro- majority of the MANET routing protocols, notably prolonged
posed solutions are applicable in disaster scenarios due to convergence time and high routing overhead, hindering their
their reliance on factors like network connectivity, band- adaptability to the dynamic nature of aerial networks. High-
width, and infrastructure. Implementing onboard image pro- lighting the specific characteristics of microaerial networks,

14 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Table 3. Comparison of swarm optimization algorithms.

Particle swarm Ant colony optimization Bee colony optimization Gray wolf optimization Glowworm swarm
Algorithm optimization (PSO) (ACO) (BCO) Firefly algorithm (FA) (GWO) optimization (GSO)
Inspired from Flocking behaviour of Foraging behaviour of Foraging behaviour of Flashing pattern of Pack of grey wolves Bio-luminescence
birds ant species honey bees fireflies chasing their prey behaviour of glowworm
Parameters Current velocity, Pheromone update Velocity and position Attractiveness of firefly Position update of Luciferin update
personal best, global based on source omega wolves
best
Computational O(NT) O(NS) O(NS) O(2NS) O(NS) O(NS)
complexity
Memory No Yes No No Yes No
Search techniques Mutation and selection Mutation and selection Mutation and selection Mutation Mutation and selection Mutation
Convergence speed Fast Slow Slow Fast Slow Slow
Scalability Poor Good Poor Good Good Good
Merits Good diversity Suitable to solve mix Suitable to solve Good exploration, can Strong versatility, Suitable for solving
variable problems, have high-dimensional easily escape from local reduces operational multi-dimensional
feedback system constrained problems minima time for optimization problems
high-dimensional with equality and
problems inequality constraints
Limitations Gets trapped easily in Uncertain convergence Unable to provide Not good at Prone to local Poor accuracy, easily
local minima for time complete optimal exploitation, have low stagnation, limited trapped in local minima
high-dimensional solution diversity exploration, less
problems accurate

Canadian Science Publishing

15
 Canadian Science Publishing

particularly the availability of GPS data, the work proposed UAV swarm life cycle. Harounabadi et al. (2015) introduced
a better approach to mobile ad hoc networking (BATMAN), a trajectory-aware geographical (TAG) routing for cognitive
emphasizing the potential of geographical routing. Never- radio ad hoc networks featuring UAV nodes. TAG utilizes tra-
theless, the paradigm required further refinement to opti- jectory information from UAVs and ensures that a UAV is not
mize its functionality. An optimized way-points (OPWP) al- chosen as the next hop if it is expected to fly inside a primary
gorithm based on message ferry route design is discussed user (PU) region or in close proximity to it. This approach is
in Tariq et al. (2006). This algorithm is designed for sparse designed to safeguard real-time packets from potential de-
MANETs, offering efficient performance without necessitat- lays caused by PU activity. Shumeye Lakew et al. (2020) have
ing real-time collaboration between nodes and the ferry. reviewed UAV classification metrics and deployment issues,
The OPWP ferry route consists of strategically selected way- serving as a basis for classifying FANET communication ar-
points and associated waiting times, tailored according to chitecture. It proposed a new taxonomy for routing protocols
the node mobility model. Delay-tolerant networks (DTN) al- in FANETs, providing thorough discussions and comparative
low nodes to store messages until forwarding is possible, studies.
a mechanism known as store–carry–forward (SCF). This en- A location-aided delay-tolerant routing (LADTR) protocol
hances routing robustness in the face of disruptions. DTN is for UAV networks, specifically designed for post-disaster oper-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

primarily designed for sparse networks with infrequent con- ations, is presented in Arafat and Moh (2018). The protocol in-
tact opportunities, making it less efficient in well-connected corporates location-awareness and utilizes an SCF technique.
networks. Many DTN routing schemes resort to packet repli- Notably, ferrying UAVs are introduced for efficient SCF, mark-
cation to reduce delivery delays and increase delivery prob- ing the first instance of their use for routing in UAV networks.
ability. However, this replication introduces substantial stor- This innovation aims to enhance connection paths between
age and bandwidth overhead, potentially diminishing perfor- searching UAVs and ground stations, thereby reducing end-to-
mance in well-connected networks. end delays and improving packet delivery ratios. A dynamic
A hybrid MANET-DTN routing approach designed to en- priority packet scheduling scheme designed for maintaining
hance the performance of a MANET routing protocol is dis- high QoS in post-disaster UAV-assisted MANETs is discussed in
cussed in Raffelsberger and Hellwagner (2013) by incorpo- Gao et al. (2021). The scheme considers not only the current
rating local packet buffers. This protocol stores data pack- packet delay but also anticipates the impact of future trans-
ets in the absence of an end-to-end path and sends them missions on priority assignment. The Gauss–Markov Mobil-
when a route is identified. Despite its benefits, this scheme ity Model is employed to capture the dynamic characteristics
may experience significant delays due to the absence of a of nodes. Additionally, the scheme integrates factors such as
mechanism for advancing data delivery when no end-to- node movement, topology instability, and time-varying chan-
end path is available. Aung et al. (2017) introduced a data- nel quality into the priority assignment process. Table 4 sum-
delivery solution for opportunistic networks, consisting of marizes routing protocols tailored for UAV networks in dis-
two main algorithms: store–carry–cooperative forward rout- aster environments, and the state-of-the-art schemes are out-
ing and information epidemic control. The data forward- lined in Table 5.
ing algorithm involves proactive monitoring by nodes, uti-
lizing direct and two-hop cooperative forwarding opportuni-
ties, and adaptively switching between cooperative and reac-
tive SCF routing. Additionally, an information epidemics con- 5.1. Critical insights and gaps in existing
trol algorithm is proposed for earlier control signal distribu- studies
tion and a faster recovery rate, with the effectiveness studied Designing a routing protocol for reliable communication
using the susceptible-infected-recovered model. In a disaster in UAV networks poses significant challenges due to high
scenario, a forwarding-based protocol like a geographic rout- mobility, dynamic topology, and uneven UAV distribution.
ing protocol may be more suitable (Huda et al. 2012; Fajardo In missions like search and rescue, minimizing latency and
et al. 2014). Location-aware message delivery (LMD) is intro- ensuring high data transmission rates are crucial. While
duced in Huda et al. (2012) as a communication solution for achieving zero delay is impractical, efforts focus on mini-
short messages among individuals in disaster-stricken areas. mizing latency within certain limits. Proactive routing pro-
Emphasizing power conservation and message delivery ratio tocols, which require frequent table updates, may not be
as crucial design goals, the authors presented LMD as a rout- ideal for highly dynamic environments. In contrast, reactive
ing protocol capable of exchanging messages without net- protocols like BATMAN offer efficiency, however, with poten-
work infrastructure or a continuous end-to-end path. The sys- tial limitations in packet size. To address frequent link dis-
tem performs well beyond certain node density levels, mak- connections, the store-carry-forward technique is proposed
ing it a preferable option over energy-consuming multi-copy in literature by having nodes carry packets until finding a
strategies for communication in disaster-stricken areas. To suitable neighbour; however, this approach, particularly in
address challenges posed by limited UAV battery capacity and sparse networks, can introduce delays. Greedy forwarding-
unpredictable network connectivity in disaster sites, Yang et based schemes select the next node based on the minimum
al. (2019) investigated an energy-efficient multi-hop data rout- distance, but they may fail to find closer nodes, leading to lo-
ing algorithm for UAV-aided medical assistance. Prioritizing cal minimum problems. Multi-hop AODV-based protocols are
quality-of-service, the research focuses on minimizing energy noted for their adaptability to harsh conditions with minimal
consumption, considering transmission rate, time delay, and overhead and have been extensively investigated.

16 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Table 4. Routing protocols for a multi-UAV network in a disaster environment.

Classification Routing protocol
Reactive r Ad hoc on-demand distance vector (AODV)
r Location-aided routing (LAR)
r Temporally-ordered routing algorithm (TORA)
r Associativity-based routing (ABR)

Opportunistic r Opportunistic routing protocol (ORP)

r Multi-copy opportunistic routing (MCOR)
r Cooperative communication-based opportunistic routing (CCOR)
r Geographic and opportunistic routing (GEAR)
r Spray and wait (SnW)

Ad hoc r Centralized routing

r Decentralized routing
r Collaborative routing
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Delay-tolerant networks (DTNs) r Epidemic routing

r Spray and wait routing
r PRoPHET routing
r MaxProp routing
r Message ferry routing

Wireless mesh network (WMN) r Hybrid wireless mesh protocol (HWMP)

r Better approach to mobile adhoc networking (BATMAN)
r Optimized link state routing (OLSR)
r Hybrid wireless mesh protocol v2 (HWMPv2)
r Multi-gateway routing (MGR)
r Proactive routing protocol for mesh networks (BMF)
r Distance routing effect algorithm for mobility (DREAM)

Table 5. A brief analysis of the literature on routing schemes.

Reference Description
Arafat and Moh (2018) Introduced the location-aided delay-tolerant routing (LADTR) protocol for post-disaster UAV networks, employing
ferrying UAVs to enhance store-carry-forward technique, improving connection paths, reducing delays, and
increasing packet delivery ratio
Fu et al. (2022) Introduced a UAV Routing System (UAVRS) to determine optimal routes for inspecting damages and monitoring
transmission lines and roads in real time for distribution networks
Faiz et al. (2024) Proposed a framework optimizing a two-echelon vehicle routing problem using ground vehicles to transport
drones for delivering medical aid to trapped populations
Mohammed Ahmed et al. Evaluated the efficiency of routing protocols (AODV, DSR, OLSR, and ZRP) in disaster scenarios for UAVs
(2021) communication mesh, addressing challenges to enhance network performance
Yin et al. (2017) Introduced a greedy and position-assisted routing protocol for highly mobile UAVs, mitigating transmission delay
challenges
Arafat and Moh (2019b) Reviewed UAV network routing protocols, categorizing them and conducting a qualitative comparison based on
features, characteristics, and performance
Yang et al. (2019) Examined challenges in UAV-aided disaster management and proposed an energy-efficient routing algorithm
ensuring quality-of-service, while addressing issues such as limited UAV battery capacity and unstable network
connectivity
Khan et al. (2018) Focused on topology-based routing protocols, reviewing their features, pros, and cons. Evaluates selected protocols
through simulation analyses for end-to-end delay, throughput, and network load
Bousbaa et al. (2020) Explored fleet routing in FANETs and proposed a geo-cast protocol mitigating UAV mobility challenges, enhancing
average delay, packet delivery, and throughput

Position-based routing schemes offer an alternative ap- to address this issue, but practical testing remains limited.
proach for disaster missions, where every node shares its Moreover, these schemes rely on GPS locations to determine
position with others in the network. However, in pure the next node, leading to drawbacks in global accuracy and
geographic-based routing schemes, the periodic transmission security, such as susceptibility to GPS spoofing attacks in
of beacons to update positions results in significant network highly mobile networks. From our analysis, it is evident that
overhead. Various alternative protocols have been studied there is no single protocol suitable for all types of disasters.

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 17

 Canadian Science Publishing

Instead, a combination of protocols tailored to specific disas- ity. From the comprehensive survey on security and safety
ter scenarios may be considered. For instance, proactive pro- considerations presented in Altawy and Youssef (2017) for
tocols may be suitable for smaller networks with limited mo- civilian applications, the potential attacks associated with a
bility, while position-based protocols are preferred for larger UAV network in disaster response scenarios are identified and
networks with high mobility. Position-based routing offers listed in Fig. 5. The primary security challenge in communi-
advantages such as reduced overhead and scalability, as the cation protocols for UAVs lies in securing data transmitted
next hop is chosen based on positions. However, the complex- over vulnerable connections like WiFi. UAVs typically trans-
ity of position-based protocols is higher compared to proac- mit data to ground stations via wireless links that are sus-
tive ones, as beacons need to be transmitted frequently. De- ceptible to attacks. To safeguard against interception, encryp-
spite the multitude of protocols discussed in the literature, tion schemes such as the advanced encryption standard are
many focus on theoretical improvements rather than practi- commonly used. However, its efficiency in real-time applica-
cal application, overlooking the importance of aligning with tions, particularly with high data transfer rates, poses chal-
specific mission requirements. Furthermore, protocols that lenges (Heron 2009). Another significant security concern is
meet security requirements are often underexplored, high- potential attacks on UAVs, aiming to seize control or disrupt
lighting the ongoing need for research in routing protocols communication with the ground control station (GCS). Attack
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

for critical missions with dynamic network topologies. methods include jamming, spoofing, and false data injection
AI-based routing is a growing trend, leveraging ML algo- attacks. A comprehensive literature review on UAV security
rithms for optimal route selection based on network per- is presented in Shafique et al. (2021), examining vulnerabili-
ceptions. However, frequent topology changes challenge pre- ties in existing protocols and proposing potential solutions. It
dictions, prompting research into node-based position pre- analyzes threats, including WiFi insecurity, jamming attacks,
dictions to reduce reliance on network topology. Further and fuzzing attacks, outlining how these vulnerabilities can
research is needed to fully harness the potential of AI- be exploited. The study also emphasizes risks in packet for-
based solutions, integrating advanced ML algorithms for warding and routing protocols within UAVs, highlighting po-
path planning, routing, and resource allocation. Computa- tential threats to security.
tional complexity and power consumption must be consid- In the context of disaster response and management, the
ered, especially for battery-powered UAVs. Connectivity is- spread of false or manipulated information can have seri-
sues due to limited UAV range and battery drain necessi- ous consequences. It can lead to misinformation and confu-
tate energy-based predictions and lightweight protocols. In- sion and hinder the effectiveness of emergency response ef-
tegrating online path planning algorithms can minimize forts. To protect data transmission, various symmetric cryp-
connectivity issues, particularly in static mission areas like tographic and steganographic methods have been proposed
search and rescue or wildfire monitoring. Sophisticated AI in the literature (Naji et al. 2009; Vegh and Miclea 2014).
approaches that can predict nodes, environments, and con- However, securely distributing and managing the keys be-
nectivity quality can address network challenges comprehen- comes a complex task in symmetric schemes, particularly
sively. Lightweight protocols incorporating energy efficiency in large-scale systems, making it susceptible to various at-
and connectivity optimization can further enhance UAV net- tacks, including brute-force attacks. This paved the way for
work performance. asymmetric cryptographic schemes, where key exchange was
made possible through the Diffie–Hellman key exchange al-
gorithm (Boneh no date). This algorithm establishes a ses-
6. Security challenges and approaches sion key between communicating entities, ensuring secure
Given the inherent broadcast vulnerability in wireless com- data transmission once the session key is in place. In Wang
munications, ensuring security is paramount for current et al. (2023), an innovative, secure, and energy-efficient data
and future wireless network designs. The attention in re- sharing scheme for urban drone rescue networks is pre-
search has increasingly turned to physical layer security, dis- sented. The approach employs a lightweight, infrastructure-
tinct from traditional key-based cryptographic techniques free blockchain framework to ensure data security and trace
(Mukherjee et al. 2014). Unlike upper-layer methods, physical misbehaviour at disaster sites. In Wesson et al. (2014), a data
layer security safeguards wireless data transmissions with- authentication protocol employing an asymmetric key algo-
out relying on secret keys and intricate algorithms, render- rithm technique is proposed to verify the authenticity of data
ing it more suitable for large-scale decentralized networks received by the UAV, distinguishing between communication
(Liu et al. 2017). Recent research explores the potential of from the authentic ground station and potential eavesdrop-
full-duplex techniques at the source or legitimate destina- pers. In Sahingoz (2013), a public-key exchange protocol en-
tion, aiming to double spectral efficiency. To identify the ables sensor nodes to authenticate each other before com-
research gaps in the above methods, a secure transmission munication. Two nodes act as communicating parties, ex-
scheme is designed for UAV wiretap channels in the pres- changing encrypted messages with public and private keys.
ence of a full-duplex active eavesdropper in Liu et al. (2017). In Valentin-Alexandru et al. (2019), a trust-based protocol is
Two major contributions of the work include a derived com- proposed where sensors are assigned trust values. The UAV
pact expression for the hybrid outage probability, consider- combines these values and avoids communication with sen-
ing short-distance LoS links in UAV-aided communication sys- sors having negative trust values. In Yoon et al. (2017), an
tems, and the determination of an optimal power allocation authentication protocol for UAV security is proposed. The
factor at the source to minimize the hybrid outage probabil- UAV is authenticated based on comparing received data with

18 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Fig. 5. Security attacks on the UAV network in a disaster scenario.

Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

maintained indexes, allowing take-off for successful authen- While blockchain technology offers advantages such as data
tication and disconnecting communication for unsuccessful confidentiality and entity validation, making it suitable for
attempts. Challenges of this scheme include high bandwidth, trust-sensitive applications, certain limitations were identi-
cost, and processing time for large data. A detailed review of fied (Shafique et al. 2021). In larger networks, the blockchain
various authentication schemes is presented in Zhou et al. may grow to a size that makes it impractical for agents
(2020), Shafique et al. (2021), and Pandey et al. (2022). to maintain a complete record. Additionally, the time re-
To address the challenges posed by jamming GPS signals, quired to process a new block could be deemed inefficient,
adopting alternative navigation methods becomes essential. particularly in UAVs that typically operate with an average
Wu and Johnson (2010) proposed a vision and inertial navi- flight autonomy of approximately 25 min. In response to
gation system, enabling autonomous navigation for drones the identified challenges, a UAV network identity authenti-
in situations where GPS signals are unavailable. Existing cation scheme is presented in Li et al. (2019) that relies on
research on vision-based navigation is comprehensively re- the elliptic curve ECC algorithm (Li 2002). The system uti-
viewed in Balamurugan et al. (2017), and it is concluded that lized ECC digital certificates as proof of identity for autho-
a visual odometry-based approach is more efficient in terms rized access. Authentication of drone identity is achieved
of memory and computational power. The authors also pro- through the application of the elliptic curve-based ECDSA
posed a modular multi-sensor data fusion technique designed signature algorithm (Johnson et al. 2001), chosen for its effi-
for UAV navigation in GPS-denied environments. Similar to ciency in terms of computation and resource consumption.
GPS signals, the attackers can jam or spoof the position and Subsequently, the ECDH exchange algorithm (Haakegaard
velocity broadcasts of the and Lang 2015) is employed to generate a session key for
UAVs, potentially causing collisions or redirecting the secure UAV communication. To address the issues identified
drone to a desired location to physically capture it. To miti- with existing ECC-based protocols, such as inflexibility and
gate this challenge, De Melo et al. (2021) introduced a system backward security issues, a lightweight authentication pro-
designed to validate the identity and location of UAVs. It com- tocol over elliptic curve is proposed in Zhang et al. (2023).
bines a public-key authentication method with a movement It ensures backward secrecy of session keys, provides flex-
plausibility check for groups of UAVs. This method periodi- ibility, and exhibits minimal time cost compared to other
cally assesses the credibility of neighbouring UAV locations, authentication methods, enhancing overall security. Table 6
providing enhanced security by detecting intruders deviating outlines the recent studies on security aspects of UAV net-
from expected trajectories. works.
Another significant security concern in UAV networks is
intrusion attacks. A concise survey of state-of-the-art intru-
sion detection systems (IDS) in the context of networked 6.1. Critical insights and gaps in existing
UAV environments is presented in Choudhary et al. (2018). studies
The classification is based on various factors, including in- In the realm of highly dynamic UAV networks, tradi-
formation gathering sources, deployment strategies, detec- tional security methods relying on asymmetric approaches
tion methods, detection states, IDS acknowledgment, and in- face challenges due to the absence of a central authority
trusion types. Various authentication schemes are studied in for issuing digital signatures and managing key storage. As
the literature to overcome the challenge of intrusion attacks. a result, researchers have delved into alternative avenues
The concepts of blockchain for authentication mechanisms such as distributed certificate-based techniques, pre-key dis-
were explored in Jensen et al. (2019) and Lv et al. (2021). tribution algorithms, and blockchain technology. Recently,

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 19

 Canadian Science Publishing

Table 6. A brief analysis of the literature on security approaches.

Reference Description
Rodday et al. (2016) Demonstrated security vulnerabilities, potential man-in-the-middle attacks, and proposed countermeasures to
enhance security and resilience
He et al. (2016) Underscored the crucial need for communication security in UAVs, highlighting potential vulnerabilities and
presenting a low-cost implementation of GPS spoofing and WiFi attacks
Abdallah et al. (2019) Proposed a secure disaster surveillance UAV system considering energy constraints and limited computation,
utilizing ring-learning with errors for confidentiality and data redundancy for accuracy in a two-tier cluster
network
Sun et al. (2019) Investigated physical layer security issues in UAV wireless communications, addressing eavesdropping
vulnerabilities and proposing techniques like trajectory design and resource allocation, alongside applications of
advanced technologies for improved security and spectral efficiency
Li et al. (2019) Proposed a lightweight identity authentication method based on elliptic curve cryptography, ensuring two-way
identity authentication and key consistency verification and providing a more secure and efficient solution
compared to traditional methods
Wang et al. (2019) Addressed the power allocation strategy for a UAV swarm-enabled aerial network to enhance physical layer security
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

against eavesdropping
Ch et al. (2020) Introduced a blockchain technology solution to enhance security and privacy for UAVs used in aerial surveying,
employing elliptic curve cryptography and SHA for data integrity
Alladi et al. (2020b) Introduced a lightweight mutual authentication scheme based on physical unclonable functions (PUFs) for securing
communication between UAVs and ground station, addressing vulnerabilities, and providing formal security
analysis
Abro et al. (2022) Examined evolving UAV applications, focusing on detection, classification, and tracking advancements while
addressing security and privacy concerns through control signal jamming and proposing strategies to enhance
UAV security and privacy
Iqbal (2021) Explored for intelligent defense mechanisms within UAV operating systems, examining existing security issues,
proposing solutions, and discussing research challenges for secure UAV operating systems
Tsao et al. (2022) Surveyed security and privacy issues in UAVs, focusing on FANETs. Categorizes threats and defense mechanisms,
analyzes UAV routing protocols, and discusses research challenges and future directions
Pandey et al. (2022) Conducted a comprehensive survey on security issues in UAV-aided networks, addressing cyber threats, privacy
concerns, and mitigation techniques, while integrating key wireless communication technologies and emerging
topics like machine learning and blockchain
Asghar Khan et al. (2022) Addressed security and privacy concerns in UAV-enabled intelligent transportation systems by proposing a
privacy-preserving authentication scheme combining hyperelliptic curve cryptography, digital signature (Asghar
Khan et al. 2022), and hash function

hardware-driven solutions like physically unclonable func- for battery-powered UAVs. Cloud offloading schemes mitigate
tions (PUFs) have garnered attention for their potential to this but add complexity and cost, particularly concerning
enhance security for mobile UAVs. Despite their theoretical data security in disaster scenarios. Software-defined network-
exploration, practical implementation remains limited in ex- ing introduces delays and single points of failure, making it
isting studies. Security considerations within UAV networks less suitable for low-latency missions. Distributed and multi-
vary depending on factors like network size and deployment controller approaches offer alternatives, while lightweight
environment. Solutions tailored to specific applications are encryption schemes and optimized cryptographic protocols
necessary since generic approaches like data encryption can- can enhance scalability and efficiency. Blockchain-enabled se-
not address diverse threats such as GPS spoofing or jamming curity techniques show potential in managing tasks and se-
attacks. Addressing each attack separately is imperative, ne- curing swarm UAV networks, yet they suffer from computa-
cessitating measures like alternative localization methods tional delays, a major concern in resource-constrained UAVs.
and secure handover mechanisms. While UAVs offer line-of- Fog computing techniques have garnered attention for their
sight communication, this also exposes vulnerabilities, with applicability in low-latency missions, offering the capabil-
data confidentiality at risk due to interception by rogue UAVs. ity to handle large data volumes while providing quality of
Different mission types face distinct security challenges; for service, scalability, adaptivity, reduced platform dependency,
instance, disaster surveillance missions may encounter eaves- and low latency. This makes them well-suited for missions
dropping or man-in-the-middle attacks, while search and res- like disaster mitigation, as they offer security against threats
cue operations are susceptible to GPS spoofing and denial such as GPS spoofing, hijacking, eavesdropping, and denial of
of service. Incorporating security into routing protocols and service attacks. Despite these advantages, fog computing has
exploring physical layer security techniques are avenues for not been fully explored and has limitations, including a lack
mitigating these risks. of task sharing and inter-layer resourcing. Task sharing can
ML approaches, including CNNs and RNNs, offer promise distribute the workload evenly, enhancing network lifetime,
for detecting anomalies and intrusions. However, their adop- while inter-layer resourcing facilitates interaction and data
tion introduces computational complexity, posing challenges offloading among heterogeneous devices. Addressing these

20 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Fig. 6. Literature survey chart.

Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

challenges could enhance the suitability of fog computing for ber of research publications in the domain of UAV networks
UAV networks. for disaster applications over the last five years (2016–2023).
Privacy concerns, alongside security, are paramount in UAV
systems, necessitating measures to prevent unauthorized ac-
cess and mitigate privacy leakage. However, detecting and 7.1. Cyber-physical security
identifying malicious UAVs is a critical challenge, with exist- UAV networks face significant security issues, such as
ing literature lacking comprehensive studies on algorithms wire-tapping, malicious attacks, jamming, and control sig-
for detecting spying actions in specific areas of interest. nal forgery. These vulnerabilities may compromise mission-
While watermark-based schemes have been proposed for de- critical data and UAV operation security. While security has
tection (Nassi et al. 2019), their applicability is limited to been a focal point in recent research, there is a need for fur-
UAVs within the range of Wi-Fi first-person view. In scenarios ther exploration into the implementation and hardware re-
involving multiple UAVs, distinguishing between legitimate source requirements of existing schemes. The impact on pay-
and malicious ones becomes increasingly challenging, under- load capacity and computational resources can significantly
scoring the need for extensive research in security fields. Ad- affect UAV battery life and mission accomplishment.
ditionally, ensuring security during UAV flight missions re- The major attacks that predominate in a disaster environ-
quires forensic analysis, but collecting real-time data from ment are eavesdropping, signal jamming, intrusion attacks,
each UAV and conducting thorough testing to identify se- and physical capture attacks. Eavesdropping, to a greater ex-
curity breaches poses logistical challenges. Addressing these tent, can be mitigated through lightweight cryptographic
challenges necessitates future research efforts aimed at de- techniques, but they are yet to be tested and analyzed on
veloping robust detection algorithms and enhancing forensic UAV swarms. To enhance intrusion detection in computer
analysis capabilities within UAV systems. networks with high bandwidth and data traffic, researchers
have turned to ML and DL algorithms. However, the effective-
ness of these schemes needs to be thoroughly investigated
in terms of training and precision. To mitigate physical cap-
7. Future research directions ture attacks, PUFs are gaining popularity and will definitely
The preceding discussions clearly show that UAVs have the receive major attention in the coming years. PUFs are de-
possibility of significantly increasing the efficiency of disas- vices that generate responses based on intrinsic variations
ter management operations by providing critical situational (McGrath et al. 2019). A striking feature that is explored for
awareness and delivering relief and supplies to affected ar- physical security is that, any attempt to tamper with the PUF,
eas. Despite the benefits provided by UAVs and technology such as micro-probing, renders it useless. However, existing
advancements, the research on UAV-swarm networks is still research on these devices is confined to theoretical simu-
in its infant stage, and there are many open issues that need lations, with insufficient consideration given to the impact
to be addressed and investigated. The difficulties range from of temperature variations and other environmental factors
regulatory issues to technical constraints, and they can have (Alladi et al. 2020a; Garcia-Bosque et al. 2020). PUF can also
a considerable impact on the effectiveness of UAV-based dis- add uniqueness and randomness to key generation and will
aster mitigation efforts. In this context, it is critical to eval- definitely receive major attention in the coming years. The
uate and comprehend the issues connected with UAV-based ability of these devices to mitigate other attacks is also to be
disaster mitigation to devise effective solutions and exploit explored. Thus, continuous research and collaboration are re-
the full potential of these technologies. Figure 6 represents quired to keep up with increasing security concerns in the
the current research trend, deduced from the average num- field of UAVs.

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 21

 Canadian Science Publishing

Furthermore, future research directions should investigate curity protocols tailored to the resource constraints of UAVs
countermeasure approaches to mitigate jamming attacks in would further enhance the applicability of this approach in
UAV relay schemes. The RL approach (Zhang et al. 2020b) real-world scenarios.
is a promising avenue, leveraging RL and transfer learning
to optimize relay signal power against jamming, even with-
7.3. Joint computation, communication, and
out prior knowledge of network topology and signal mod-
els. Exploring the scalability, robustness, and real-world ap- control
plicability of such countermeasure techniques would be in- Integration of space, air, and ground communications is
strumental in enhancing the security of drone communi- a current development in aerial networks. The joint com-
cation networks. Some of the other research perspectives putation, communication, and control challenge in UAVs in-
may include the development of AI-driven security solu- volves optimizing and coordinating these three aspects for
tions capable of adapting to evolving threats, exploring less- efficient and reliable UAV operations. UAVs often face limi-
computationally intensive blockchain technology for secure tations in on-board computational resources, reliable com-
data sharing among UAVs, and addressing privacy concerns munication links, and precise control systems, which adds
through privacy-preserving algorithms. Additionally, further latency in mission execution. The challenge arises due to the
interconnected nature of these elements, where computation
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

research to strengthen physical security measures for UAVs

could mitigate the risks associated with unauthorized access affects communication and control, and vice versa. Overcom-
and tampering. ing this challenge requires a holistic approach, incorporating
advanced algorithms for computation, communication pro-
tocols, and control systems. Technologies such as edge com-
7.2. Routing issues puting, MIMO communication, adaptive control, and intelli-
UAVs typically operate in low-density environments with gent decision-making play a crucial role in addressing this
high mobility, leading to frequent changes in network topol- challenge and unlocking the full potential of UAVs in disas-
ogy and disconnections among communication nodes. This ter scenarios. This integration has the potential to create a
instability adversely affects routing efficiency and perfor- seamless and pervasive connectivity environment for the effi-
mance, making the design of routing protocols challeng- cient communication and exchange of data between various
ing. It is obvious that traditional routing protocols designed platforms, such as satellites, HAPs, UAVs, and ground com-
for MANETs and VANETs are not sufficient in UAV networks munication systems. However, innovative solutions are re-
due to the high mobility and frequent changes in topol- quired to overcome the inherent challenges of network het-
ogy. While existing schemes in the literature have individu- erogeneity in agent handovers and ensure the reliability of
ally succeeded in incorporating parameters like hop count, these integrated networks. Future research should delve into
link quality, congestion, and energy efficiency, a standard- developing adaptive algorithms that dynamically distribute
ized routing scheme that integrates all these factors to en- tasks and responsibilities among heterogeneous UAVs, taking
sure low overhead and a high packet delivery rate during into account their unique performance metrics. This involves
disaster scenarios is yet to be established. The existing ap- designing intelligent decision-making frameworks capable
proaches, such as SCF, indeed led to the nodes not requir- of optimizing the utilization of resources across a diverse
ing to maintain continuous routing table for the routes, but set of UAVs in a network. Researchers can explore the de-
it has added overhead to the transmission process. Further, sign of communication protocols that account for the diverse
achieving a high packet delivery ratio during disasters with communication capabilities of UAVs. This involves develop-
routing protocols may still result in mission failures due to ing adaptive routing algorithms that consider the bandwidth
adverse effects from malicious replays and false signalling. requirements of different UAVs and optimizing data transfer
Additionally, many state-of-the-art protocols are simulation- strategies based on the communication constraints present
based, and the mobility rates considered may not always align in the heterogeneous network. Various techniques for sen-
with real-time scenarios. Testing and evaluating these proto- sor fusion that integrate data from diverse sensors on UAVs
cols in real-time disaster-like scenarios are essential for stan- to enhance the overall perception and information gathering
dardization. Therefore, to fully harness the UAV network po- capabilities of the network are to be extensively explored in
tential, routing requires substantial attention. Factors such as the coming years to ensure accurate decision-making in het-
inter-node synchronization, heterogeneity, energy efficiency, erogeneous environments.
collision avoidance, decentralized control, communication The analysis of synchronization, resource allocation, and
range management, and real-time decision-making are also scheduling in drone networks highlights the importance
to be further investigated in the routing process. A notable of optimizing connectivity, especially in densely populated
and pioneering work involves the integration of security mea- swarm configurations. As drones play a crucial role in re-
sures into routing protocols to ensure the robust protection laying units to enhance networking solutions, future re-
of data traffic. Existing works (Yadav et al. 2017; Patil et al. search should delve into advanced synchronization mecha-
2020) still need to be deeply investigated, and future direc- nisms to achieve optimal resource allocation and schedul-
tions in this line of research could explore advanced cryp- ing strategies. The throughput optimization for both A2A
tographic techniques, anomaly detection mechanisms, and and A2G links poses challenges, particularly in the con-
intrusion prevention strategies within routing protocols. Ad- text of Internet of Drones (IoD) systems where various
ditionally, the development of lightweight and efficient se- metrics need consideration. Addressing the computational

22 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

complexity of solving optimization problems related to and time-variant characteristics. Conventional determinis-
throughput maximization is crucial. Additionally, there is a tic and stochastic models are inadequate for characterizing
need to explore energy-efficient solutions, given the trade-off these channels, which depend on factors such as UAV alti-
between maximizing throughput and increased energy con- tude, type, elevation angle, and propagation environment.
sumption, especially in battery-powered drones. Further, the UAV network is highly dynamic. As UAVs move
through the airspace, their relative motion with respect to
7.4. Energy-efficient coordination the ground stations and other nodes causes a frequency shift
Coordinating a swarm of UAVs for cooperative mission in the signals transmitted and received. The Doppler effect
planning is a complex and challenging task. The problem can lead to fluctuations in signal frequencies, which further
becomes NP-hard as the size of the swarm and mission ob- complicates the channel characteristics and requires care-
jectives increase, leading to exponential computational re- ful consideration in channel modelling and communication
quirements. The UAV network is also affected by environ- system design. Developing an accurate and generic channel
mental changes, making real-time coordination difficult. Fur- model for UAV-to-ground communications requires compre-
thermore, the energy-constrained nature of UAVs raises con- hensive simulations and measurements in various environ-
cerns about energy-efficient coordination. Addressing these ments. Also, the lack of a standardized communication chan-
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

challenges requires decentralized approaches, hierarchical nel for UAV authentication and authorization poses chal-
control, distributed optimization, learning-based methods, lenges for FANETs. The initialization phase for UAVs could
energy-aware routing, task allocation strategies, and dynamic be streamlined with the implementation of a standardized
resource management. Resolving these issues will enable ef- communication channel. Moreover, as the number of UAVs in
fective deployment of UAV swarms in various applications, the network grows, so does the possibility of inter-UAV inter-
making continued research essential for advancement in this ference. This can result in decreased network performance
field. Therefore, future research should concentrate on devel- and dependability. Researchers are investigating new inter-
oping energy-aware algorithms that consider the diverse en- ference management strategies, like beam-forming and in-
ergy constraints of UAVs. This involves designing optimiza- terference cancellation, to help alleviate the consequences of
tion algorithms for task scheduling, path planning, and col- interference in multi-UAV networks.
laborative decision-making that account for variations in en-
ergy availability among heterogeneous UAVs, ensuring effi- 7.7. Deployment
cient use of resources throughout missions. One of the most critical challenges is designing the optimal
3D placement and flight path, as these factors significantly
impact the performance of UAV-assisted wireless communi-
7.5. Trajectory optimization cations. The optimal altitude of UAV base stations and their
Optimizing flight trajectories is challenging due to prac- flight trajectories depend heavily on environmental condi-
tical constraints on UAV networks deployed in disaster sce- tions, application scenarios, and the number of UAVs em-
narios. To reduce communication delay, UAVs need to move ployed. While existing studies on UAV base station placement
close to ground users while maintaining interconnections have shown promising results, further research is needed to
with neighbouring UAVs. An optimal trajectory ensures end- optimize 3D UAV deployment, taking into account the unique
to-end link connections and sufficient coverage of the target features of UAVs. Many existing works on computing optimal
area. To achieve this, a dynamic trajectory control method is UAV positions assume static UAV base stations, which is un-
required, considering communication range, adaptive plan- realistic in most cases. Therefore, designing a dynamic UAV
ning, collaborative decision-making, coverage, energy effi- base station deployment strategy for UAV-assisted wireless
ciency, real-time optimization, and safety. While some stud- networks, considering UAV mobility and other constraints,
ies have addressed energy-efficient optimization, they of- presents an interesting research problem. Interdisciplinary
ten rely on generic energy consumption models, which may collaboration and continuous innovation are necessary to
not be the most suitable fit for disaster scenarios. Disaster- overcome these hurdles and fully harness the potential of
stricken areas may have a more complex and dynamic envi- UAVs in diverse applications.
ronment with debris, damaged structures, and unpredictable
terrain. Navigating through such areas requires advanced
sensing and processing capabilities to detect and avoid obsta- 8. Conclusions
cles. All these factors adversely affect the energy consump- Unmanned aerial networks have proven remarkably effec-
tion rates of the nodes. While advancements in trajectory tive in critical missions, reaching hard-to-access areas to pro-
planning for UAV swarms and edge computing have proved vide essential assistance. While substantial research has been
effective individually, the integration of both areas, known dedicated to this field, the absence of standardized schemes
as UAV swarm-enabled edge computing, is still relatively new on communication, task scheduling, data processing, and tra-
and remains in its early stages of exploration. jectory optimization still remains a challenge in this domain.
The vast scope of UAVs requires focused research in specific
7.6. Channel modelling areas to address the intricate challenges unique to their di-
UAV-to-ground communication channels are more com- verse applications. The authors of the article have identi-
plex and susceptible to blockage than traditional ground fied the prevailing research trends, emphasizing the priori-
communication channels due to their distinctive 3D space tization of research in data processing capabilities and the

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 23

 Canadian Science Publishing

security of both devices and data in the network. This arti- Aggarwal, S., and Kumar, N. 2020. Path planning techniques for un-
cle presents a substantial body of research, examining liter- manned aerial vehicles: a review, solutions, and challenges. Com-
put. Commun. 149(July 2019): 270–299. doi:10.1016/j.comcom.2019.
ature across various network aspects, including deployment,
10.014.
trajectory optimization, routing, and security. It also outlines Albert, A., and Imsland, L. 2017. Performance bounds for tracking mul-
research areas set to be extensively explored in the coming tiple objects using a single UAV. In 2017 International Conference on
years. Each facet of these networks presents abundant op- Unmanned Aircraft Systems, ICUAS 2017. pp. 1539–1546. doi:10.1109/
ICUAS.2017.7991440.
portunities for research exploration, particularly in mission-
Al-Hourani, A., Kandeepan, S., and Lardner, S. 2014. Optimal LAP altitude
critical scenarios such as disasters. for maximum coverage. IEEE Wireless Commun. Lett. 3(6): 569–572.
doi:10.1109/LWC.2014.2342736.
Alladi, T., Chamola, V., Naren, and Kumar, N. 2020a. PARTH: a two-stage
Article information lightweight mutual authentication protocol for UAV surveillance net-
works. Comput. Commun. 160(March): 81–90. doi:10.1016/j.comcom.
History dates 2020.05.025.
Received: 26 September 2023 Alladi, T., Naren, Bansal, G., Chamola, V., and Guizani, M. 2020b. SecAu-
thUAV: a novel authentication scheme for UAV-ground station and
Accepted: 20 February 2024 UAV–UAV communication. IEEE Trans. Veh. Technol. 69(12): 15068–
Accepted manuscript online: 27 February 2024 15077. doi:10.1109/TVT.2020.3033060.
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Version of record online: 2 April 2024 Alsamhi, S.H., Shvetsov, A.V., Kumar, S., Shvetsova, S.V., Alhartomi, M.A.,
Hawbani, A., et al. 2022. UAV computing-assisted Search and Rescue
Mission framework for disaster and harsh environment mitigation.
Copyright Drones, 6(7): 1–21. doi:10.3390/drones6070154.
© 2024 The Author(s). This work is licensed under a Creative Altawy, R., and Youssef, A.M. 2017. Security, privacy, and safety aspects
Commons Attribution 4.0 International License (CC BY 4.0), of civilian drones: a survey. ACM Trans. Cyber-Phys. Syst. 1(2): 1–25.
doi:10.1145/3001836.
which permits unrestricted use, distribution, and reproduc- Alzenad, M., El-Keyi, A., Lagum, F., and Yanikomeroglu, H. 2017. 3-D
tion in any medium, provided the original author(s) and placement of an unmanned aerial vehicle base station (UAV-BS) for
source are credited. energy-efficient maximal coverage. IEEE Wireless Commun. Lett. 6(4):
434–437. doi:10.1109/LWC.2017.2700840.
Andersen, J.B., and Kovacs, I.Z. 2002. Power distributions revisited. In Pro-
Data availability ceedings of COST273 3rd Management Committee Meeting [Preprint].
This manuscript does not report data. Arafat, M.Y., and Moh, S. 2018. Location-aided delay tolerant routing pro-
tocol in UAV networks for post-disaster operation. IEEE Access, 6:
59891–59906. doi:10.1109/ACCESS.2018.2875739.
Author information Arafat, M.Y., and Moh, S. 2019a. Localization and clustering based on
swarm intelligence in UAV networks for emergency communica-
Author ORCIDs tions. IEEE Internet Things J. 6(5): 8958–8976. doi:10.1109/JIOT.2019.
2925567.
Indu Chandran https://orcid.org/0000-0002-1374-9737 Arafat, M.Y., and Moh, S. 2019b. Routing protocols for unmanned aerial
vehicle networks: a survey. IEEE Access, 7: 99694–99720. doi:10.1109/
Author contributions ACCESS.2019.2930813.
Arinta, R.R., and Andi, E.W.R. 2019. Natural disaster application on
Writing – original draft: IC big data and machine learning: a review. In 2019 4th International
Writing – review & editing: KV Conference on Information Technology, Information Systems and
Electrical Engineering, ICITISEE 2019. Vol. 6: 249–254. doi:10.1109/
Competing interests ICITISEE48480.2019.9003984.
Asadpour, M., Egli, S., Hummel, K.A., and Giustiniano, D. 2014. Routing
The authors declare there are no competing interests. in a fleet of micro aerial vehicles: first experimental insights. In AIR-
BORNE 2014 – Proceedings of the 2014 ACM Workshop on Airborne
Funding information Networks and Communications, Co-located with MobiHoc 2014. pp.
No funding was received to assist with the preparation of this 9–10. doi:10.1145/2636582.2636832.
Aschenbruck, N., Padilla, E.G., Gerharz, M., Frank, M., and Martini, P.
manuscript. 2007. Modelling mobility in disaster area scenarios. In MSWiM’07:
Proceedings of the Tenth ACM Symposium on Modeling, Analysis,
References and Simulation of Wireless and Mobile Systems. pp. 4–12. doi:10.
1145/1298126.1298131.
Abdallah, A., Ali, M., Mišić, J., and Mišić, V. 2019. Efficient security scheme Asghar Khan, M., Ullah, I., Alkhalifah, A., Ur Rehman, S., Ali Shah,
for disaster surveillance UAV communication networks. Information, J., Uddin, I., et al. 2022. A provable and privacy-preserving
10(2): 43. doi:10.3390/INFO10020043. authentication scheme for UAV-enabled intelligent transportation
Abrar, M., Ajmal, U., Almohaimeed, Z.M., Gui, X., Akram, R., and Mas- systems. IEEE Trans. Ind. Inf. 18(5): 3416–3425. doi:10.1109/TII.2021.
roor, R. 2021. Energy efficient UAV-enabled mobile edge computing 3101651.
for IoT devices: a review. IEEE Access, 9: 127779–127798. doi:10.1109/ Aung, C.Y., Ho, I.W.-H., and Chong, P.H.J. 2017. Store–carry–cooperative
ACCESS.2021.3112104. forward routing with information epidemics control for data delivery
Abro, G., Zulkifli, S., Masood, R., Asirvadam, V., and Laouiti, A. 2022. in opportunistic networks. IEEE Access, 5: 6608–6625. doi:10.1109/
Comprehensive review of UAV detection, security, and communica- ACCESS.2017.2690341.
tion advancements to prevent threats. Drones, 6(10). doi:10.3390/ Avanzato, R., and Beritelli, F. 2019. An innovative technique for
drones6100284. identification of missing persons in natural disaster based on
Acosta-Coll, M., Solano-Escorcia, A., Ortega-Gonzalez, L., and Zamora- drone-femtocell systems. Sensors (Switzerland), 19(20). doi:10.3390/
Musa, R. 2021. Forecasting and communication key elements for s19204547.
low-cost fluvial flooding early warning system in urban areas. Int. Azmi, I.N., Yussoff, Y.M., Kassim, M., and Tahir, N.M. 2021. A mini-review
J. Electr. Comput. Eng. 11(5): 4143–4156. doi:10.11591/ijece.v11i5. of flying ad hoc networks mobility model for disaster areas. Int. Trans.
pp4143-4156. J. Eng. 12(10): 1–12. doi:10.14456/ITJEMAST.2021.191.

24 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Balamurugan, G., Valarmathi, J., and Naidu, V.P.S. 2017. Survey on UAV Davoli, L., Pagliari, E., and Ferrari, G. 2021. Hybrid lora-ieee 802.11 s
navigation in GPS denied environments. In International Conference opportunistic mesh networking for flexible UAV swarming. Drones,
on Signal Processing, Communication, Power and Embedded System, 5(2). doi:10.3390/drones5020026.
SCOPES 2016 – Proceedings. pp. 198–204. doi:10.1109/SCOPES.2016. Deepak, G.C., Ladas, A., Sambo, Y.A., Pervaiz, H., Politis, C., and Imran,
7955787. M.A. 2019. An overview of post-disaster emergency communication
Bani, M., and Alhuda”, “N. 2016. Flying ad-hoc networks: routing pro- systems in the future networks. IEEE Wireless Commun. 26(6): 132–
tocols, mobility models, issues. Int. J. Adv. Comput. Sci. Appl. 7(6): 139. doi:10.1109/MWC.2019.1800467.
162–169. doi:10.14569/ijacsa.2016.070621. De Melo, C.F.E., Dapper E Silva, T., Boeira, F., Stocchero, J.M., Vinel, A.,
Bekmezci, I., Sahingoz, O.K., and Temel, Ş. 2013. Flying ad-hoc networks Asplund, M., and De Freitas, E.P. 2021. UAVouch: a secure identity and
(FANETs): a survey. Ad Hoc Netw. 11(3): 1254–1270. doi:10.1016/J. location validation scheme for UAV-networks. IEEE Access, 9: 82930–
ADHOC.2012.12.004. 82946. doi:10.1109/ACCESS.2021.3087084.
Bithas, P.S., Nikolaidis, V., Kanatas, A.G., and Karagiannidis, G.K. 2020. Demiane, F., Sharafeddine, S., and Farhat, O. 2020. An optimized UAV tra-
UAV-to-ground communications: channel modeling and UAV selec- jectory planning for localization in disaster scenarios. Comput. Net-
tion. IEEE Trans. Commun. 68(8): 5135–5144. doi:10.1109/TCOMM. works, 179(June): 107378. doi:10.1016/j.comnet.2020.107378.
2020.2992040. Dey, S., and Ray, S. 2017. Ad-hoc networked UAVs as aerial mesh net-
Boneh, D. no date. The decision Diflie-Hellman problem. work for disaster management application and remote sensing: an
Boursianis, A.D., Papadopoulou, M.S., Diamantoulakis, P., Liopa- approach. pp. 301–304.
Tsakalidi, A., Barouchas, P., Salahas, G., et al. 2022. Internet of Things Di Franco, C., and Buttazzo, G. 2015. Energy-aware coverage path plan-
(IoT) and agricultural unmanned aerial vehicles (UAVs) in smart ning of UAVs. In Proceedings - 2015 IEEE International Conference
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

farming: a comprehensive review. Internet of Things (Netherlands), on Autonomous Robot Systems and Competitions, ICARSC 2015. pp.
18: 100187. doi:10.1016/j.iot.2020.100187. 111–117. doi:10.1109/ICARSC.2015.17.
Bousbaa, F.Z., Kerrache, C.A., Mahi, Z., Tahari, AElK, Lagraa, N., and Do-Duy, T., Nguyen, L.D., Duong, T.Q., Khosravirad, S.R., and Claussen,
Yagoubi, M.B. 2020. GeoUAVs: a new geocast routing protocol for fleet H. 2021. Joint optimisation of real-time deployment and re-
of UAVs. Comput. Commun. 149(July 2019): 259–269. doi:10.1016/j. source allocation for UAV-aided disaster emergency communications.
comcom.2019.10.026. IEEE J. Sel. Areas Commun. 39(11): 3411–3424. doi:10.1109/JSAC.2021.
Bupe, P., Haddad, R., and Rios-Gutierrez, F. 2015. Relief and emergency 3088662.
communication network based on an autonomous decentralized UAV Dong, C., Xie, J., Dai, H., Wu, Q., Qin, Z., and Feng, Z. 2018. Optimal de-
clustering network. In Conference Proceedings— —IEEE SOUTHEAST- ployment density for maximum coverage of drone small cells. China
CON, 2015 June. pp. 1–8. doi:10.1109/SECON.2015.7133027. Commun. 15(5): 25–40. doi:10.1109/CC.2018.8387984.
Busnel, Y., Caillouet, C., and Coudert, D. 2019. Self-organized disaster Donratanapat, N., Samadi, S., Vidal, J.M., and Sadeghi Tabas, S. 2020. A na-
management system by distributed deployment of connected UAVs tional scale big data analytics pipeline to assess the potential impacts
In 6th International Conference on Information and Communica- of flooding on critical infrastructures and communities. Environ.
tion Technologies for Disaster Management, ICT-DM 2019. [Preprint] : Modell. Software, 133: 104828. doi:10.1016/j.envsoft.2020.104828.
doi:10.1109/ICT-DM47966.2019.9032964. Esrafilian, O., Gangula, R., and Gesbert, D. 2020. Autonomous UAV-aided
Celtek, S.A., Durdu, A., and Kurnaz, E. 2019. Design and simulation of the mesh wireless networks. In IEEE INFOCOM 2020 - IEEE Conference
hierarchical tree topology based wireless drone networks. In 2018 In- on Computer Communications Workshops, INFOCOM WKSHPS 2020.
ternational Conference on Artificial Intelligence and Data Processing, pp. 634–640. doi:10.1109/INFOCOMWKSHPS50562.2020.9162753.
IDAP 2018. pp. 1–5. doi:10.1109/IDAP.2018.8620755. Fadlullah, Z.Md., Takaishi, D., Nishiyama, H., Kato, N., and Miura, R. 2016.
Ch, R., Srivastava, G., Reddy Gadekallu, T., Maddikunta, P.K.R., and Bhat- A dynamic trajectory control algorithm for improving the communi-
tacharya, S. 2020. Security and privacy of UAV data using blockchain cation throughput and delay in UAV-aided networks. IEEE Network,
technology. Journal of Information Security and Applications, 55(Oc- 30(1): 100–105. doi:10.1109/MNET.2016.7389838.
tober): 102670. doi:10.1016/j.jisa.2020.102670. Faiz, T.I., Vogiatzis, C., and Noor-E-Alam, Md 2024. Computational ap-
Chamola, V., Hassija, V., Gupta, S., Goyal, A., Guizani, M., and Sikdar, B. proaches for solving two-echelon vehicle and UAV routing problems
2021. Disaster and pandemic management using machine learning: a for post-disaster humanitarian operations. Expert Syst. Appl. 237(PB):
survey. IEEE Internet Things J. 8(21): 16047–16071. doi:10.1109/JIOT. 121473. doi:10.1016/j.eswa.2023.121473.
2020.3044966. Fajardo, J.T.B., Yasumoto, K., Shibata, N., Sun, W., and Ito, M. 2014.
Chen, H., Wang, X., Shen, L., and Cong, Y. 2021. Formation flight of fixed- Disaster information collection with opportunistic communication
wing UAV swarms: a group-based hierarchical approach. Chin. J. Aero- and message aggregation. J. Inf. Process. 22(2): 106–117. doi:10.2197/
naut. 34(2): 504–515. doi:10.1016/j.cja.2020.03.006. ipsjjip.22.106.
Chen, S., Guo, S., and Li, Y. 2017. Real-time tracking a ground moving tar- Fu, J., Nunez, A., and De Schutter, B. 2022. Real-time UAV routing strategy
get in complex indoor and outdoor environments with UAV. In 2016 for monitoring and inspection for postdisaster restoration of distri-
IEEE International Conference on Information and Automation, IEEE bution networks. IEEE Trans. Ind. Inf. 18(4): 2582–2592. doi:10.1109/
ICIA 2016. pp. 362–367. doi:10.1109/ICInfA.2016.7831851. TII.2021.3098506.
Choudhary, G., Sharma, V., You, I., Yim, K., Chen, I.R., and Cho, J.H. Ganesh, S., Gopalasamy, V., and Sai Shibu, N.B. 2021. Architecture for
2018. Intrusion detection systems for networked unmanned aerial drone assisted emergency ad-hoc network for disaster rescue opera-
vehicles: a survey. In 2018 14th International Wireless Communica- tions. In 2021 International Conference on COMmunication Systems
tions and Mobile Computing Conference, IWCMC 2018. pp. 560–565. and NETworkS, COMSNETS 2021, vol. 2061. pp. 44–49. doi:10.1109/
doi:10.1109/IWCMC.2018.8450305. COMSNETS51098.2021.9352814.
Conceição, L., and Curado, M. 2013. Modelling mobility based on hu- Gao, M., Zhang, B., and Wang, L. 2021. A dynamic priority packet schedul-
man behaviour in disaster areas. In Lecture Notes in Computer Sci- ing scheme for post-disaster UAV-assisted mobile ad hoc network In
ence (including subseries Lecture Notes in Artificial Intelligence and IEEE Wireless Communications and Networking Conference, WCNC,
Lecture Notes in Bioinformatics), 7889 LNCS. pp. 56–69. doi:10.1007/ 2021 March. doi:10.1109/WCNC49053.2021.9417537.
978-3-642-38401-1_5. Garcia-Bosque, M., Senorans, G.D., Azqueta, C.S., and Celma, S. 2020.
Cred and UNDRR Centre for Research on the Epidemiology of Introduction to physically unclonable fuctions: properties and ap-
Disasters (CRED), I.Ucl. 2021. Disaster year in review 2020 plications. In ECCTD 2020 – 24th IEEE European Conference on
global trends and perspectives. Cred, 62: 2020–2021. Available Circuit Theory and Design [Preprint]. doi:10.1109/ECCTD49232.2020.
from https://www.cred.be/index.php?q=work-with-us%0Ahttps: 9218404.
//cred.be/sites/default/files/CredCrunch62.pdf. Garnica-Peña, R.J., and Alcántara-Ayala, I. 2021. The use of UAVs for land-
Cui, Z., Briso-Rodriguez, C., Guan, Ke, Zhong, Z., and Quitin, F. 2020. slide disaster risk research and disaster risk management: a literature
Multi-frequency air-to-ground channel measurements and analysis review. J. Mount. Sci. 18(2): 482–498. doi:10.1007/s11629-020-6467-7.
for UAV communication systems. IEEE Access, 8: 110565–110574. doi: Gezici, S., Tian, Zhi, Giannakis, G.B., Kobayashi, H., Molisch, A.F., Poor,
10.1109/ACCESS.2020.2999659. H.V., and Sahinoglu, Z. 2005. Localization via ultra-wideband radios:

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 25

 Canadian Science Publishing

a look at positioning aspects of future sensor networks. IEEE Signal Jin, W., Yang, J., Fang, Y., and Feng, W. 2020. Research on application
Process. Mag. 22(4): 70–84. doi:10.1109/MSP.2005.1458289. and deployment of UAV in emergency response. In ICEIEC 2020 -
Gu, Y., Zhou, M., Fu, S., and Wan, Y. 2015. Airborne WiFi networks Proceedings of 2020 IEEE 10th International Conference on Elec-
through directional antennae: an experimental study. In 2015 IEEE tronics Information and Emergency Communication. pp. 277–280.
Wireless Communications and Networking Conference, WCNC 2015. doi:10.1109/ICEIEC49280.2020.9152338.
pp. 1314–1319. doi:10.1109/WCNC.2015.7127659. Johnson, D., Menezes, A., and Vanstone, S. 2001. The elliptic curve digi-
Guo, Q., Zhang, Y., Lloret, J., Kantarci, B., and Seah, W.K.G. 2019. A tal signature algorithm (ECDSA). Int. J. Inf. Secur. 1(1): 36–63. doi:10.
localization method avoiding flip ambiguities for micro-UAVs with 1007/s102070100002.
bounded distance measurement errors. IEEE Trans. Mobile Comput. Joshi, A., Dhongdi, S., Kumar, S., and Anupama, K.R. 2020. Simulation
18(8): 1718–1730. doi:10.1109/TMC.2018.2865462. of multi-UAV ad-hoc network for disaster monitoring applications.
Gupta, M., and Varma, S. 2021. Optimal placement of UAVs of an aerial International Conference on Information Networking, January 2020.
mesh network in an emergency situation. J. Ambient Intell. Human. pp. 690–695. doi:10.1109/ICOIN48656.2020.9016543.
Comput. 12(1): 343–358. doi:10.1007/s12652-020-01976-2. Khan, M.A., Khan, I.U., Safi, A., and Quershi, I.M. 2018. Dynamic rout-
Haakegaard, R., and Lang, J. 2015. The elliptic curve Diffie–Hellman ing in flying ad-hoc networks using topology-based routing protocols.
(ECDH). p. 4 Available from http://koclab.cs.ucsb.edu/teaching/ec Drones, 2(3): 1–15. doi:10.3390/drones2030027.
c/project/2015Projects/Haakegaard+Lang.pdf [accesssed 10 February Khan, M., Qureshi, I., and Khanzada, F. 2019. A hybrid communication
2020]. scheme for efficient and low-cost deployment of future flying
Harounabadi, M., et al. 2015. TAG: trajectory aware geographical routing ad-hoc network (FANET). Drones, 3(1): 16. doi:10.3390/
in cognitive radio ad hoc networks with UAV nodes. Lect. Notes Inst. DRONES3010016.
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

Comput. Sci. Soc.-Inf. Telecommun. Eng. 155: 111–122. doi:10.1007/ Khuwaja, A.A., Chen, Y., Zhao, N., Alouini, M.-S., and Dobbins, P.
978-3-319-25067-0_9. 2018. A survey of channel modeling for uav communications. IEEE
Harrington, P., Ng, W.P., and Binns, R. 2020. Autonomous drone con- Commun. Surv. Tut. 20(4): 2804–2821. doi:10.1109/COMST.2018.
trol within a Wi-Fi network. In 2020 12th International Symposium 2856587.
on Communication Systems, Networks and Digital Signal Processing, Kuiper, E., and Nadjm-Tehrani, S. 2006. Mobility models for UAV group
CSNDSP 2020. [Preprint]. doi:10.1109/CSNDSP49049.2020.9249585. reconnaissance applications. In Second International Conference on
Hayat, S., Yanmaz, E., and Bettstetter, C. 2015. Experimental analysis Wireless and Mobile Communications, ICWMC 2006. pp. 2–8. doi:10.
of multipoint-to-point UAV communications with IEEE 802.11n and 1109/ICWMC.2006.63.
802.11ac. In IEEE International Symposium on Personal, Indoor and Kurt, A., Saputro, N., Akkaya, K., and Uluagac, A.S. 2021. Distributed con-
Mobile Radio Communications, PIMRC, December 2015. pp. 1991– nectivity maintenance in swarm of drones during post-disaster trans-
1996. doi:10.1109/PIMRC.2015.7343625. portation applications. IEEE Trans. Intell. Transport. Syst. 22(9): 6061–
He, D., Chan, S., and Guizani, M. 2016. Communication security of 6073. doi:10.1109/TITS.2021.3066843.
unmanned aerial vehicles. IEEE Wireless Commun. 24(4): 134–139. Lakew, D.S., Masood, A., and Cho, S. 2020. 3D UAV placement and tra-
doi:10.1109/MWC.2016.1600073WC. jectory optimization in UAV assisted wireless networks. In Interna-
Hentati, A.I., and Fourati, L.C. 2020. Comprehensive survey of UAVs com- tional Conference on Information Networking, January 2020. pp. 80–
munication networks. Comput. Stand. Interfaces, 72(June): 103451. 82. doi:10.1109/ICOIN48656.2020.9016553.
doi:10.1016/j.csi.2020.103451. Lee, J.-Y., and Scholtz, R.A. 2002. Ranging in a dense multipath environ-
Hernandez, D., Cano, J.-C., Silla, F., Calafate, C.T., and Cecilia, J.M. 2022. ment using an UWB radio link. IEEE J. Sel. Areas Commun. 20(9):
AI-enabled autonomous drones for fast climate change crisis assess- 1677–1683. doi:10.1109/JSAC.2002.805060.
ment. IEEE Internet Things J. 9(10): 7286–7297. doi:10.1109/JIOT.2021. Li, B., Fei, Z., and Zhang, Y. 2019. UAV communications for 5G and be-
3098379. yond: recent advances and future trends. IEEE Internet Things J. 6(2):
Heron, S. 2009. Advanced encryption Standard (AES). Network Secur. 2241–2263. doi:10.1109/JIOT.2018.2887086.
2009(12): 8–12. doi:10.1016/S1353-4858(10)70006-4. Li, J., Zhao, H., Wang, H., Gu, F., Wei, J., Yin, H., and Ren, B. 2020. Joint
Huda, M.N., Yasmeen, F., Yamada, S., and Sonehara, N. 2012. An ap- optimization on trajectory, altitude, velocity, and link scheduling for
proach for short message resilience in disaster-stricken areas. In In- minimum mission time in UAV-aided data collection. IEEE Internet
ternational Conference on Information Networking. pp. 120–125. Things J. 7(2): 1464–1475. doi:10.1109/JIOT.2019.2955732.
doi:10.1109/ICOIN.2012.6164362. Li, T., Jianfeng, M., Pengbin, F., Yue, M., Xindi, M., Jiawei, Z, et al. 2019.
Hydher, H., Jayakody, D.N.K., Hemachandra, K.T., and Samarasinghe, T. Lightweight security authentication mechanism towards UAV net-
2020. Intelligent UAV deployment for a disaster-resilient wireless net- works. In Proceedings - 2019 International Conference on Network-
work. Sensors (Switzerland), 20(21): 1–18. doi:10.3390/s20216140. ing and Network Applications, NaNA 2019. pp. 379–384. doi:10.1109/
Iqbal, S. 2021. A study on UAV operating system security and future re- NaNA.2019.00072.
search challenges. In 2021 IEEE 11th Annual Computing and Com- Li, V.C. 2002. Advances in ECC research. American Concrete Institute, ACI
munication Workshop and Conference, CCWC 2021. pp. 759–765. Special Publication, SP-206. pp. 373–400. doi:10.14359/12264.
doi:10.1109/CCWC51732.2021.9376151. Li, Y., QingYang, C., Sheng, X., and Xi, H.Z. 2016. A fast target localization
Jahir, Y., Atiquzzaman, M., Refai, H., Paranjothi, A., and Lopresti, P.G. method with multi-point observation for a single UAV. In Proceedings
2019. Routing protocols and architecture for disaster area network: a of the 28th Chinese Control and Decision Conference, CCDC 2016. pp.
survey. Ad Hoc Networks, 82: 1–14. doi:10.1016/j.adhoc.2018.08.005. 5389–5394. doi:10.1109/CCDC.2016.7531961.
Javaid, S., Saeed, N., Qadir, Z., Fahim, H., He, B., Song, H., and Bilal, M. Lin, Na, Liu, Y., Zhao, L., Wu, D.O., and Wang, Y. 2022. An adaptive UAV
2023. Communication and control in collaborative UAVs: recent ad- deployment scheme for emergency networking. IEEE Trans. Wireless
vances and future trends. IEEE Trans. Intell. Transp. Syst. 24(6): 5719– Commun. 21(4): 2383–2398. doi:10.1109/TWC.2021.3111991.
5739. doi:10.1109/TITS.2023.3248841. Linardos, V., Drakaki, M., Tzionas, P., and Karnavas, Y. 2022. Machine
Jensen, I.J., Selvaraj, D.F., and Ranganathan, P. 2019. Blockchain tech- learning in disaster management: recent developments in meth-
nology for networked swarms of unmanned aerial vehicles (UAVs). ods and applications. Mach. Learn. Knowl. Extract. 4(2): 446–473.
In 20th IEEE International Symposium on A World of Wireless, Mobile doi:10.3390/make4020020.
and Multimedia Networks, WoWMoM 2019. [Preprint]. doi:10.1109/ Liu, C., Quek, T.Q.S., and Lee, J. 2017. Secure UAV communication in
WoWMoM.2019.8793027. the presence of active eavesdropper. In 2017 9th International Con-
Jiang, F., and Swindlehurst, A.L. 2012. Optimization of UAV heading for ference on Wireless Communications and Signal Processing, WCSP
the ground-to-air uplink. IEEE J. Sel. Areas Commun. 30(5): 993–1005. 2017 - Proceedings, January 2017. pp. 1–6. doi:10.1109/WCSP.2017.
doi:10.1109/JSAC.2012.120614. 8171198.
Jiang, H., Zhang, Z., Wang, C.-X., Zhang, J., Dang, J., Wu, L., and Zhang, Liu, C.H., Ma, X., Gao, X., and Tang, J. 2020. Distributed energy-efficient
H. 2020. A novel 3D UAV channel model for A2G communication en- multi-UAV navigation for long-term communication coverage by
vironments using AoD and AoA estimation algorithms. IEEE Trans. deep reinforcement learning. IEEE Trans. Mobile Comput. 19(6):
Commun. 68(11): 7232–7246. doi:10.1109/TCOMM.2020.3011716. 1274–1285. doi:10.1109/TMC.2019.2908171.

26 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079

 Canadian Science Publishing

Luo, C., Miao, W., Ullah, H., McClean, S., Parr, G., and Min, G. 2019. Un- Munawar, H.S., Mojtahedi, M., Hammad, A.W.A., Ostwald, M.J.,
manned aerial vehicles for disaster management. pp. 83–107. doi:10. and Waller, S.T. 2022. An AI/ML-based strategy for disaster re-
1007/978-981-13-0992-2_7. sponse and evacuation of victims in aged care facilities in
Lv, Z., Qiao, L., Hossain, M.S., and Choi, B.J. 2021. Analysis of using the Hawkesbury-Nepean Valley: a perspective. Buildings, 12(1).
blockchain to protect the privacy of drone big data. IEEE Network, doi:10.3390/buildings12010080.
35(1): 44–49. doi:10.1109/MNET.011.2000154. Naji, A.W., Zaidan, A.A., Zaidan, B.B., Hameed, S.A., and Khalifa, O.O.
Lyu, J., Zeng, Y., Zhang, R., and Lim, T.J. 2017. Placement optimization of 2009. Novel approach for secure cover file of hidden data in the un-
UAV-mounted mobile base stations. IEEE Commun. Lett. 21(3): 604– used area within EXE file using computation between cryptography
607. doi:10.1109/LCOMM.2016.2633248. and steganography. J. Comput. Sci. 9(5): 294–300.
Mahiddin, N.A., Affandi, F.F.M., and Mohamad, Z. 2021. A review on mo- Nasr, I., Chekir, M., and Besbes, H. 2019. Shipwrecked victims localiza-
bility models in disaster area scenario. Int. J. Adv. Technol. Eng. Ex- tion and tracking using UAVs. In 2019 15th International Wireless
plor. 8(80): 848–873. doi:10.19101/IJATEE.2021.874084. Communications and Mobile Computing Conference, IWCMC 2019.
Malandrino, F., Chiasserini, C.-F., Casetti, C., Chiaraviglio, L., and pp. 1344–1348. doi:10.1109/IWCMC.2019.8766534.
Senacheribbe, A. 2019. Planning UAV activities for efficient user cov- Nassi, B., Ben-Netanel, R., Shamir, A., and Elovici, Y. 2019.
erage in disaster areas. Ad Hoc Networks, 89: 177–185. doi:10.1016/j. ‘Drones’ cryptanalysis— —smashing cryptography with a flicker.
adhoc.2019.04.001. In Proceedings— —IEEE Symposium on Security and Privacy, May 2019.
Masroor, R., Naeem, M., and Ejaz, W. 2021. Efficient deployment of UAVs pp. 1397–1414. doi:10.1109/SP.2019.00051.
for disaster management: a multi-criterion optimization approach. Panda, K.G., Das, S., Sen, D., and Arif, W. 2019. Design and deployment of
Comput. Commun. 177: 185–194. doi:10.1016/J.COMCOM.2021.07. UAV-aided post-disaster emergency network. IEEE Access, 7: 102985–
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

006. 102999. doi:10.1109/ACCESS.2019.2931539.

Matolak, D.W., and S., Ruoyu, 2015. Unmanned aircraft systems: air- Pandey, G.K., Gurjar, D.S., Nguyen, HaH, and Yadav, S. 2022. Secu-
ground channel characterization for future applications. IEEE Veh. rity threats and mitigation techniques in UAV communications: a
Technol. Mag. 10(2): 79–85. doi:10.1109/MVT.2015.2411191. comprehensive survey. IEEE Access, 10(September): 112858–112897.
Matolak, D.W., and Fiebig, U.C. 2019. UAV channel models: review and doi:10.1109/ACCESS.2022.3215975.
future research. In 13th European Conference on Antennas and Prop- Patil, P.A., Deshpande, R.S., and Mane, P.B. 2020. Trust and opportunity
agation, EuCAP 2019. pp. 1–5. based routing framework in wireless sensor network using hybrid
Matracia, M., Saeed, N., Kishk, M.A., and Alouini, M.-S. 2022. Post- optimization algorithm. Wireless Pers. Commun. 115(1): 415–437.
disaster communications: enabling technologies, architectures, and doi:10.1007/S11277-020-07579-6/FIGURES/7.
open challenges. IEEE Open J. Commun. Soc. 3(June): 1177–1205. Peer, M., Bohara, V.A., and Srivastava, A. 2020. Multi-UAV place-
doi:10.1109/OJCOMS.2022.3192040. ment strategy for disaster-resilient communication network In
Mcenroe, P., Wang, S., and Liyanage, M. 2022. A survey on the conver- IEEE vehicular technology conference, November 2020. doi:10.1109/
gence of edge computing and AI for UAVs: opportunities and chal- VTC2020-Fall49728.2020.9348687.
lenges. IEEE Internet Things J. 9(17): 15435–15459. doi:10.1109/JIOT. Pomportes, S., Tomasik, J., and Vèque, V. 2011. A composite mobility
2022.3176400. model for ad hoc networks in disaster areas. REV J. Electron. Com-
Mcgrath, T., Bagci, I.E., Wang, Z.M., Roedig, U., and Young, R.J. 2019. A mun. 1(1): 62–68. doi:10.21553/rev-jec.10.
PUF taxonomy. Appl. Phys. Rev. 6(1). doi:10.1063/1.5079407. Portmann, M., and Pirzada, A.A. 2011. Wireless mesh networks for public
Mekki, K., Bajic, E., Chaxel, F., and Meyer, F. 2019. A comparative study safety and crisis management applications. IEEE Eng. Manage. Rev.
of LPWAN technologies for large-scale IoT deployment. ICT Express, 39(4): 114–122. doi:10.1109/EMR.2011.6093893.
5(1): 1–7. doi:10.1016/j.icte.2017.12.005. Raffelsberger, C., and Hellwagner, H. 2013. A hybrid MANET-DTN routing
Messous, M.A., Senouci, S.M., and Sedjelmaci, H. 2016. Network connec- scheme for emergency response scenarios. In 2013 IEEE International
tivity and area coverage for UAV fleet mobility model with energy Conference on Pervasive Computing and Communications Work-
constraint. In IEEE Wireless Communications and Networking Con- shops, PerCom Workshops 2013. pp. 505–510. doi:10.1109/PerComW.
ference, WCNC, September 2016. pp. 1–6. doi:10.1109/WCNC.2016. 2013.6529549.
7565125. Rahmadhani, A,R., Isswandhana, R., Giovani, A., and Syah, R.A. 2018. Lo-
Mi, J., and Dai, Z. 2021. A 3D smooth mobility model based on semi- RaWAN as secondary telemetry communication system for drone de-
random circular movement for FANETs. In 2021 7th International livery. In Proceedings – 2018 IEEE International Conference on In-
Conference on Computer and Communications, ICCC 2021. pp. 954– ternet of Things and Intelligence System, IOTAIS 2018. pp. 116–122.
959. doi:10.1109/ICCC54389.2021.9674424. doi:10.1109/IOTAIS.2018.8600892.
Micheletto, M., Petrucci, V., Santos, R., Orozco, J., Mosse, D., Ochoa, S., Ramesh, M.V. 2014. Design, development, and deployment of a wireless
and Meseguer, R. 2018. Flying real-time network to coordinate disas- sensor network for detection of landslides. Ad Hoc Networks, 13(Part
ter relief activities in urban areas. Sensors (Switzerland), 18(5): 1–20. A): 2–18. doi:10.1016/J.ADHOC.2012.09.002.
doi:10.3390/s18051662. Rodday, N.M., De Schmidt, R.O., and Pras, A. 2016. Exploring security vul-
Mohammed Ahmed, S.B., Hussain, S.A., Abdul Latiff, L., Ahmad, N., and nerabilities of unmanned aerial vehicles. In Proceedings of the NOMS
Mohd Sam, S. 2021. Performance evaluation of FANET routing proto- 2016 – 2016 IEEE/IFIP Network Operations and Management Sympo-
cols in disaster scenarios. In Proceedings of the 2021 IEEE Symposium sium (NOMS). pp. 993–994. doi:10.1109/NOMS.2016.7502939.
on Future Telecommunication Technologies, SOFTT 2021. pp. 46–51. Sahingoz, O.K. 2013. Multi-level dynamic key management for scalable
doi:10.1109/SOFTT54252.2021.9673152. wireless sensor networks with UAV. Lect. Notes Electric. Eng. 214
Morgenthaler, S., Braun, T., Zhao, Z., Staub, T., and Anwander, M. 2012. LNEE, pp. 11–19. doi:10.1007/978-94-007-5857-5_2.
UAVNet: a mobile wireless mesh network using unmanned aerial ve- Sahingoz, O.K. 2014. Networking models in flying ad-hoc networks
hicles. In 2012 IEEE Globecom Workshops, GC Workshops 2012. pp. (FANETs): concepts and challenges. J. Intell. Robot. Syst.: Theory Appl.
1603–1608. doi:10.1109/GLOCOMW.2012.6477825. 74(1–2): 513–527. doi:10.1007/s10846-013-9959-7.
Mozaffari, M., Saad, W., Bennis, M., and Debbah, M. 2016. Efficient de- Salo, J., El-Sallabi, H.M., and Vainikainen, P. 2000. Statistical analysis of
ployment of multiple unmanned aerial vehicles for optimal wireless the multiple scattering radio channel. Comput. Sci. Commun. Dictio-
coverage. IEEE Communications Letters, 20(8): 1647–1650. Available nary, 54(11): 1401–1401. doi:10.1007/1-4020-0613-6_15358.
at: doi:10.1109/LCOMM.2016.2578312. Sanchez-Garcia, J., García-Campos, J.M., Toral, S.L., Reina, D.G., and Bar-
Mukherjee, A., Fakoorian, S.A.A., Huang, J., and Swindlehurst, A.L. 2014. rero, F. 2016. A self organising aerial ad hoc network mobility model
Principles of physical layer security in multiuser wireless networks: a for disaster scenarios. In Proceedings - 2015 International Conference
survey. IEEE Commun. Surv. Tut. 16(3): 1550–1573. doi:10.1109/SURV. on Developments in eSystems Engineering, DeSE 2015. pp. 35–40.
2014.012314.00178. doi:10.1109/DeSE.2015.12.
Munawar, H.S., Hammad, A.W.A., Waller, S.T., Thaheem, M.J., and Saraereh, O.A., Alsaraira, A., Khan, I., and Uthansakul, P. 2020. Perfor-
Shrestha, A. 2021. An integrated approach for post-disaster flood mance evaluation of UAV-enabled LoRa networks for disaster man-
management via the use of cutting-edge technologies and UAVs: a agement applications. Sensors (Switzerland), 20(8): 1–18. doi:10.3390/
review. Sustainability (Switzerland), 13(14). doi:10.3390/su13147925. s20082396.

Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079 27

 Canadian Science Publishing

Shafique, A., Mehmood, A., and Elhadef, M. 2021. Survey of security pro- Wesson, K.D., Humphreys, T.E., and Evans, B.L. 2014. Can cryptography
tocols and vulnerabilities in unmanned aerial vehicles. IEEE Access, secure next generation air traffic surveillance. IEEE Secur. Privacy,
9: 46927–46948. doi:10.1109/ACCESS.2021.3066778. 1–8.
Shakhatreh, H., Sawalmeh, A.H., Al-Fuqaha, A., Dou, Z., Almaita, E., Wu, A.D., and Johnson, E.N. 2010. Autonomous flight in GPS-denied en-
Khalil, I., et al. 2019. Unmanned aerial vehicles (UAVs): a survey on vironments using monocular vision and inertial sensors. In AIAA In-
civil applications and key research challenges. IEEE Access, 7: 48572– fotech at Aerospace 2010. pp. 1–19. doi:10.2514/6.2010-3510.
48634. doi:10.1109/ACCESS.2019.2909530. Wu, Q., Zeng, Y., and Zhang, R. 2018. Joint trajectory and communication
Sharma, V., You, I., Pau, G., Collotta, M., Lim, J., and Kim, J. design for multi-UAV enabled wireless networks. IEEE Trans. Wireless
2018. LoRaWAN-based energy-efficient surveillance by drones for Commun. 17(3): 2109–2121. doi:10.1109/TWC.2017.2789293.
intelligent transportation systems. Energies, 11(3). doi:10.3390/ Xie, J., Wan, Y., Wang, B., Fu, S., Lu, K., and Kim, J.H. 2018. A compre-
en11030573. hensive 3-dimensional random mobility modeling framework for air-
Shumeye Lakew, D., Sa’ad, U., Dao, N.-N., Na, W., and Cho, S. 2020. Rout- borne networks. IEEE Access, 6: 22849–22862. doi:10.1109/ACCESS.
ing in flying ad hoc networks: a comprehensive survey. IEEE Com- 2018.2819600.
mun. Surv. Tut. 22(2): 1071–1120. doi:10.1109/COMST.2020.2982452. Yadav, S., Trivedi, M.C., Singh, V.K., and Kolhe, M.L. 2017. Securing AODV
Skorobogatov, G., Barrado, C., and Salamí, E. 2020. Multiple UAV routing protocol against black hole attack in MANET using outlier de-
systems: a survey. Unmanned Syst. 8(2): 149–169. doi:10.1142/ tection scheme. In 2017 4th IEEE Uttar Pradesh Section International
S2301385020500090. Conference on Electrical, Computer and Electronics, UPCON 2017,
Stellin, M., Sabino, S., and Grilo, A. 2020. LoRaWAN networking in mobile January 2018. pp. 1–4. doi:10.1109/UPCON.2017.8251012.
scenarios using a WiFi mesh of UAV gateways. Electronics (Switzer- Yan, C., Fu, L., Zhang, J., and Wang, J. 2019. A comprehensive survey
Drone Syst. Appl. Downloaded from cdnsciencepub.com by 158.62.23.212 on 02/06/25

land), 9(4): 1–20. doi:10.3390/electronics9040630. on UAV communication channel modeling. IEEE Access, 7: 107769–
Sun, W., Bocchini, P., and Davison, B.D. 2020. Applications of artificial 107792. doi:10.1109/ACCESS.2019.2933173.
intelligence for disaster management, natural hazards. Springer, The Yang, Y., Qiu, X., Li, S., Wang, J., Chen, W., Hung, P.C.K., and Zheng,
Netherlands. doi:10.1007/s11069-020-04124-3. Z. 2019. Energy-efficient data routing in cooperative UAV swarms
Sun, X., Ng, D.W.K., Ding, Z., Xu, Y., and Zhong, Z. 2019. Physical layer for medical assistance after a disaster. Chaos, 29(6). doi:10.1063/1.
security in UAV systems: challenges and opportunities. IEEE Wireless 5092740.
Commun. 26(5): 40–47. doi:10.1109/MWC.001.1900028. Yin, C., Xiao, Z., Cao, X., Xi, X., Yang, P., and Wu, D. 2017. Enhanced rout-
Tariq, M.M.B., Ammar, M., and Zegura, E. 2006. Message ferry route de- ing protocol for fast flying UAV network. In 2016 IEEE International
sign for sparse ad hoc networks with mobile nodes. In Proceedings Conference on Communication Systems, ICCS 2016. [Preprint].doi:10.
of the International Symposium on Mobile Ad Hoc Networking and 1109/ICCS.2016.7833587.
Computing (MobiHoc). pp. 37–48. doi:10.1145/1132905.1132910. Yoon, K., Park, D., Yim, Y., Kim, K., Yang, S.K., and Robinson, M. 2017.
Tsao, K.-Y., Girdler, T., and Vassilakis, V.G. 2022. A survey of cyber secu- Security authentication system using encrypted channel on UAV net-
rity threats and solutions for UAV communications and flying ad-hoc work. In Proceedings - 2017 1st IEEE International Conference on
networks. Ad Hoc Networks, 133: 102894. doi:10.1016/J.ADHOC.2022. Robotic Computing, IRC 2017. pp. 393–398. doi:10.1109/IRC.2017.56.
102894. Yu, M., Yang, C., and Li, Y. 2018. Big data in natural disaster man-
Uddin, M.A., Mansour, A., Jeune, DLe, Ayaz, M., and Aggoune, El-HM 2018. agement: a review. Geosciences (Switzerland), 8(5). doi:10.3390/
UAV-assisted dynamic clustering of wireless sensor networks for crop geosciences8050165.
health monitoring. Sensors, 18(2): 555. doi:10.3390/S18020555. Zhan, C., Hu, H., Liu, Z., Wang, Z., and Mao, S. 2021. Multi-UAV-
Ueyama, Jo, Freitas, H., Faical, B.S., Filho, G.P.R., Fini, P., Pessin, G., et al. enabled mobile-edge computing for time-constrained IoT applica-
2014. Exploiting the use of unmanned aerial vehicles to provide re- tions. IEEE Internet Things J. 8(20): 15553–15567. doi:10.1109/JIOT.
silience in wireless sensor networks. IEEE Commun. Mag. 52(12): 81– 2021.3073208.
87. doi:10.1109/MCOM.2014.6979956. Zhang, Q., Chen, J., Ji, L., Feng, Z., Han, Z., and Chen, Z. 2020a. Response
Valentin-Alexandru, V., Ion, B., and Victor-Valeriu, P. 2019. Energy effi- delay optimization in Mobile edge computing enabled UAV swarm.
cient trust-based security mechanism for wireless sensors and un- IEEE Trans. Veh. Technol., 69(3): 3280–3295. doi:10.1109/TVT.2020.
manned aerial vehicles. In Proceedings of the 11th International Con- 2964821.
ference on Electronics, Computers and Artificial Intelligence, ECAI Zhang, S., Liu, Y., Han, Z., and Yang, Z. 2023. A lightweight authentication
2019. pp. 5–10. doi:10.1109/ECAI46879.2019.9041986. protocol for UAVs based on ECC scheme. Drones, 7(5). doi:10.3390/
Vanitha, S., and Padma, T. 2014. A survey on swarm intelligence algo- drones7050315.
rithms. Int. J. Comput. Sci. Mobile Comput. 3(5): 994–998. Zhang, X., and Duan, L. 2017. Optimization of emergency UAV deploy-
Vegh, L., and Miclea, L. 2014. A new approach towards increased security ment for providing wireless coverage. In 2017 IEEE Global Commu-
in cyber-physical systems. In International Conference on Systems, nications Conference, GLOBECOM 2017 – Proceedings, January 2018.
Signals, and Image Processing. pp. 175–178. pp. 1–6. doi:10.1109/GLOCOM.2017.8254659.
Wang, H., Ren, G., Chen, J., Ding, G., and Yang, Y. 2018a. Unmanned Zhang, Y., Chen, Y., and Liu, Y. 2012. Towards unique and anchor-free
aerial vehicle-aided communications: joint transmit power and tra- localization for wireless sensor networks. Wireless Pers. Commun.
jectory optimization. IEEE Wireless Commun. Lett. 7(4): 522–525. 63(1): 261–278. doi:10.1007/s11277-011-0337-0.
doi:10.1109/LWC.2018.2792435. Zhang, Y., Mou, Z., Gao, F., Jiang, J., Ding, R., and Han, Z. 2020b. UAV-
Wang, N., et al. 2018b. UAV 3D mobility model oriented to dynamic enabled secure communications by multi-agent deep reinforcement
and uncertain environment, Lecture Notes in Computer Science (in- learning. IEEE Trans. Veh. Technol. 69(10): 11599–11611. doi:10.1109/
cluding subseries Lecture Notes in Artificial Intelligence and Lecture TVT.2020.3014788.
Notes in Bioinformatics). Springer International Publishing. doi:10. Zhao, H., Wang, H., Wu, W., and Wei, J. 2018. Deployment algorithms
1007/978-3-030-05057-3_48. for UAV airborne networks toward on-demand coverage. IEEE J. Sel.
Wang, W., Guan, X., Wang, B., and Wang, Y. 2010. A novel mobility Areas Commun. 36(9): 2015–2031. doi:10.1109/JSAC.2018.2864376.
model based on semi-random circular movement in mobile ad hoc Zhou, J., Yang, J., and Lu, Lu 2020. Research on multi-UAV networks in
networks. Inf. Sci. 180(3): 399–413. doi:10.1016/j.ins.2009.10.001. disaster emergency communication. IOP Conf. Ser.: Mater. Sci. Eng.,
Wang, X., Feng, W., Chen, Y., and Ge, N. 2019. UAV swarm-enabled aerial 719(1). doi:10.1088/1757-899X/719/1/012054.
CoMP: a physical layer security perspective. IEEE Access, 7: 120901– Zhou, Y., Rao, B., and Wang, W. 2020. UAV swarm intelligence: recent
120916. doi:10.1109/ACCESS.2019.2936680. advances and future trends. IEEE Access, 8: 183856–183878. doi:10.
Wang, Y., Su, Z., Xu, Q., Li, R., Luan, T.H., and Wang, P. 2023. A secure 1109/ACCESS.2020.3028865.
and intelligent data sharing scheme for UAV-assisted disaster rescue. Zobel, J., Lieser, P., Drescher, B., Freisleben, B., and Steinmetz, R. 2019.
IEEE/ACM Trans. Network. 1–17. doi:10.1109/TNET.2022.3226458. Optimizing inter-cluster flights of post-disaster communication sup-
Wang, Z., Duan, L., and Zhang, R. 2019. Adaptive deployment for port UAVs. In Proceedings— —Conference on Local Computer Net-
UAV-aided communication networks. IEEE Trans. Wireless Commun. works, LCN, October 2019. pp. 364–371. doi:10.1109/LCN44214.2019.
18(9): 4531–4543. doi:10.1109/TWC.2019.2926279. 8990801.

28 Drone Syst. Appl. 12: 1–28 (2024) | dx.doi.org/10.1139/dsa-2023-0079