10REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
RED DE DRONES EN MOVIMIENTO BASADA EN SDN
PARA LA TRANSMISIÓN DE VIDEO EN TIEMPO REAL
PARA VIGILANCIA ÁREA
Dipartimento di Ingegneria Informatica, Modellistica, Elettronica e Sistemistica
Università della Calabria
Rende CS, Italy
RESUMEN
Este trabajo presenta y evalúa una red de
drones basada en redes definidas por software
(Software Defined Networking, SDN) para la
transmisión de video en tiempo real orientada
a la vigilancia aérea. La arquitectura utiliza un
backbone cableado de puntos de acceso (AP)
gestionados por el controlador Ryu, empleando
el Protocolo de Árbol de Expansión (Spanning
Tree Protocol, STP) para evitar bucles, mientras
que los drones actúan como nodos inalámbricos
que transmiten video en tiempo real hacia una
estación base que simula el centro de control. La
simulación integra CoppeliaSim y Mininet-WiFi
mediante un servidor socket, y el streaming de
video se genera con VLC. la escalabilidad se
estudia incrementando el número de drones de
tres a siete donde se analizan las métricas como:
el throughput efectivo pasa de 2,75 a 7,21 Mbit/s,
el ancho de banda medio se mantiene entre 6,93
y 7,99 Mbit/s, el jitter permanece por debajo de 1
ms y el tiempo de ida y vuelta (Round-Trip Time,
RTT) varía de 8,53 a 8,99 ms, mientras que la
pérdida de paquetes aumenta de 21,34 % a 24,43
%. Al comparar distintos exponentes del modelo
de propagación para tres drones, el RTT crece de
12,57 ms (exponente 2) a 18,53 ms (exponente
4), mientras que el throughput se mantiene
alrededor de 2,75–2,76 Mbit/s y la pérdida de
paquetes entre 31 % y 32 %. En conjunto, la
arquitectura escala adecuadamente hasta cinco
drones y presenta una congestión moderada con
siete. Como trabajo futuro se propone extender la
Mobile SDN-Based Drone Network for Real-Time Video
Streaming in Aerial Surveillance
ABSTRACT
This paper presents and evaluates a drone
network based on Software-Defined Networking
(SDN) for real-time video transmission aimed at
aerial surveillance. The architecture uses a wired
backbone of access points (APs) managed by the
Ryu controller, employing the Spanning Tree
Protocol (STP) to prevent loops, while the drones
act as wireless nodes that transmit real-time video
to a base station simulating the control center.
The simulation integrates CoppeliaSim and
Mininet-WiFi through a socket server, and video
streaming is generated using VLC. Scalability
is studied by increasing the number of drones
from three to seven, analyzing metrics such
as: effective throughput, which increases from
Fecha de Recepción: 13/08/2025. Fecha de Aceptación: 02/12/2025. Fecha de Publicación: 28/01/2026
REVISTA PERSPECTIVAS
VOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
arquitectura a redes con múltiples controladores
SDN y estudiar protocolos de enrutamiento
específicos para drones en la transmisión de
video, incorporando el análisis de la Calidad de
Experiencia (Quality of Experience, QoE), la
Calidad de Servicio (Quality of Service, QoS) y
el consumo energético.
Palabras Clave: Redes Definidas por Software,
Vehículo Aéreo No Tripulado, Transmisión de
video en tiempo real, Vigilancia aérea, Protocolo
de Árbol de Expansión, Throughput, Jitter,
Pérdida de paquetes, Handover.
DOI: https://doi.org/10.47187/perspectivas.8.1.249
Anthonny Flores flrnhn98e242z605v@studenti.unical.it
Sebastian Ruiz rzjmsb97r14z605w@studenti.unical.it
iD
iD
11REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
I. Introduction
Aerial surveillance has advanced considerably
due to recent progress in Unmanned Aerial
Vehicles (UAVs), whose versatility has enabled a
wide range of military and civilian applications.
Current literature highlights that UAV systems
are increasingly used in domains such as law
enforcement, border monitoring, and emergency
response due to their ability to operate in areas that
are inaccessible or unsafe for ground personnel
[1]. In situations involving natural disasters or
hazardous environments, UAVs provide a practical
means of acquiring timely information while
avoiding the risks associated with direct human
intervention [2].
An Aerial Surveillance System is usually a
remotely piloted or pre-programmed flying
device that can transmit data in real time
back to control centers to make up an aerial
surveillance system. Since they can move beyond
road infrastructure, these UAVs have several
2.75 to 7.21 Mbit/s; average bandwidth, which
remains between 6.93 and 7.99 Mbit/s; jitter,
which stays below 1 ms; and round-trip time
(RTT), which ranges from 8.53 to 8.99 ms, while
packet loss increases from 21.34% to 24.43%.
When comparing different propagation model
exponents for three drones, RTT increases from
12.57 ms (exponent 2) to 18.53 ms (exponent
4), while throughput remains around 2.75–2.76
Mbit/s and packet loss between 31% and 32%.
Overall, the architecture scales adequately up to
five drones and shows moderate congestion with
seven. As future work, it is proposed to extend
the architecture to networks with multiple SDN
controllers and to study drone-specific routing
protocols for video transmission, incorporating
the analysis of Quality of Experience (QoE),
Quality of Service (QoS), and energy
consumption.
Keywords: Software Defined Networking,
Unmanned Aerial Vehicle, Real-Time Video
Streaming, Aerial Surveillance, Spanning Tree
Protocol, Throughput, Jitter, Packet Loss,
Handover.
advantages over terrestrial vehicles, such as faster
operating speeds and greater mobility [2][3]. Even
though UAV technology is becoming more and
more significant, maintaining proper security is
still a major worry. Conventional surveillance
techniques, like depending only on security
guards, frequently have trouble keeping an eye on
large urban areas. When compared to traditional
Closed-Circuit Television (CCTV) systems, UAV
systems greatly improve surveillance capabilities
by overcoming these constraints by offering
extensive aerial coverage [4].
The use of multiple aerial vehicles instead of
a single drone base offers both economic and
operational advantages. Deploying several small
drones connected through a communication
system is more cost-effective than relying on
one large drone, while also providing wider
coverage and faster task completion. A major
benefit of multi-drone systems is their ability to
preserve mission continuity even if individual
units fail, thanks to cooperative communication.
Multiple UAVs can collaborate to form an
Unmanned Aerial Vehicle Network (UAVNet),
which is inherently more robust and capable of
covering larger areas of interest. Such networks
are designed for scalability, allowing additional
drones to be incorporated as operational demands
evolve. Furthermore, the communication links
between drones support the formation of aerial
relay networks capable of distributing information
across extensive geographic regions [5],[6].
Various reviews on UAV networks show that
systems consisting of multiple aerial vehicles
are particularly suitable for civil monitoring and
surveillance applications, as they allow for greater
coverage, increased robustness against failures, and
greater flexibility in mission design compared to
single-drone deployments. In these proposals, UAVs
are explicitly treated as network nodes that exchange
information and cooperate with each other, giving
rise to UAV networks or Flying Ad Hoc Networks
(FANETs) capable of providing connectivity and
aerial relay functions over large areas [7], [8].
Using Software-Defined Networking (SDN) in
UAV networks is a new way to handle a lot of data
I. Introduction
Aerial surveillance has advanced considerably
due to recent progress in Unmanned Aerial
Vehicles (UAVs), whose versatility has enabled a
wide range of military and civilian applications.
Current literature highlights that UAV systems
are increasingly used in domains such as law
enforcement, border monitoring, and emergency
response due to their ability to operate in areas that
are inaccessible or unsafe for ground personnel
[1]. In situations involving natural disasters or
hazardous environments, UAVs provide a practical
means of acquiring timely information while
avoiding the risks associated with direct human
intervention [2].
An Aerial Surveillance System is usually a
remotely piloted or pre-programmed flying
device that can transmit data in real time
back to control centers to make up an aerial
surveillance system. Since they can move beyond
road infrastructure, these UAVs have several
2.75 to 7.21 Mbit/s; average bandwidth, which
remains between 6.93 and 7.99 Mbit/s; jitter,
which stays below 1 ms; and round-trip time
(RTT), which ranges from 8.53 to 8.99 ms, while
packet loss increases from 21.34% to 24.43%.
When comparing different propagation model
exponents for three drones, RTT increases from
12.57 ms (exponent 2) to 18.53 ms (exponent
4), while throughput remains around 2.75–2.76
Mbit/s and packet loss between 31% and 32%.
Overall, the architecture scales adequately up to
five drones and shows moderate congestion with
seven. As future work, it is proposed to extend
the architecture to networks with multiple SDN
controllers and to study drone-specific routing
protocols for video transmission, incorporating
the analysis of Quality of Experience (QoE),
Quality of Service (QoS), and energy
consumption.
Keywords: Software Defined Networking,
Unmanned Aerial Vehicle, Real-Time Video
Streaming, Aerial Surveillance, Spanning Tree
Protocol, Throughput, Jitter, Packet Loss,
Handover.
advantages over terrestrial vehicles, such as faster
operating speeds and greater mobility [2][3]. Even
though UAV technology is becoming more and
more significant, maintaining proper security is
still a major worry. Conventional surveillance
techniques, like depending only on security
guards, frequently have trouble keeping an eye on
large urban areas. When compared to traditional
Closed-Circuit Television (CCTV) systems, UAV
systems greatly improve surveillance capabilities
by overcoming these constraints by offering
extensive aerial coverage [4].
The use of multiple aerial vehicles instead of
a single drone base offers both economic and
operational advantages. Deploying several small
drones connected through a communication
system is more cost-effective than relying on
one large drone, while also providing wider
coverage and faster task completion. A major
benefit of multi-drone systems is their ability to
preserve mission continuity even if individual
units fail, thanks to cooperative communication.
Multiple UAVs can collaborate to form an
Unmanned Aerial Vehicle Network (UAVNet),
which is inherently more robust and capable of
covering larger areas of interest. Such networks
are designed for scalability, allowing additional
drones to be incorporated as operational demands
evolve. Furthermore, the communication links
between drones support the formation of aerial
relay networks capable of distributing information
across extensive geographic regions [5],[6].
Various reviews on UAV networks show that
systems consisting of multiple aerial vehicles
are particularly suitable for civil monitoring and
surveillance applications, as they allow for greater
coverage, increased robustness against failures, and
greater flexibility in mission design compared to
single-drone deployments. In these proposals, UAVs
are explicitly treated as network nodes that exchange
information and cooperate with each other, giving
rise to UAV networks or Flying Ad Hoc Networks
(FANETs) capable of providing connectivity and
aerial relay functions over large areas [7], [8].
Using Software-Defined Networking (SDN) in
UAV networks is a new way to handle a lot of data
12REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
collection quickly and easily. By separating the
network's control and data planes, SDN technology
makes data management processes more flexible
and gives users more control. The control plane is
in charge of network signaling, route calculation,
system management, and configuration. It is
the part of the network that decides how it will
behave. On the other hand, the data plane’s job is
to send packets to their next destinations. So, you
can think of the network as a distributed structure
that connects a lot of independent devices. SDN
architecture separates the control functions from
the hardware infrastructure, putting all control
operations in a single programmable controller.
This centralized control unit lets network admins
change and improve how the network works in
real time, quickly adapting to new situations with
a specific controller app [7],[9],[10].
In wired SDN systems with a static infrastructure,
programmability refers to the ability of the
control plane to modify data paths as needed,
while the data plane implements these decisions
by forwarding packets through the assigned
interfaces. In contrast, when SDN is employed in
UAV networks (UAVNets), programmability also
involves managing the movement of the UAVs to
prevent collisions or to enhance the performance
of the applications, selecting or updating routing
paths, and adjusting transmission parameters such
as data rate or transmission power in response to
performance or energy constraints, among several
other functions [5].
From a communications perspective, UAV
networks are characterized by high mobility,
predominantly line-of-sight links, and a rapidly
and frequently changing topology. Recent tutorials
on UAV-assisted communications highlight that
this three-dimensional and dynamic nature of
the channel requires flexible network control
mechanisms capable of adapting to variations
in the radio environment and connectivity [11].
In this context, the introduction of a logically
centralized and programmable control plane, such
as that provided by SDN, is a natural choice for
reacting to topology changes and adjusting routes
and link configurations in near real time.
One of the major difficulties in these environments
is the occurrence of broadcast storms. In traditional
networks that contain loops, Spanning Tree
Protocol is typically employed to create a loop-free
logical topology and thereby prevent uncontrolled
broadcast propagation. In contrast, within an SDN
environment, the controller leverages its global
view of the topology to compute a spanning tree
in a centralized manner to mitigate the same issue
[12].
In this paper, we propose a simple SDN-based
drone network design for Aerial Surveillance.
The goal is to provide an architecture that is
easy to understand and configure while ensuring
reliable backhaul connectivity for real-time
video transmission. Because of its simplicity, the
network can be deployed rapidly in unforeseen
or urgent situations requiring immediate aerial
monitoring. The design incorporates the STP to
prevent loops in the backbone and maintain stable
packet forwarding. In addition, the simulation
integrates Mininet-WiFi with CoppeliaSim,
enabling a combined evaluation of network
behavior and drone mobility. Recent literature on
experimentation with UAV swarms emphasizes
that, in many cases, a single simulation platform
is not sufficient to accurately model both robotic
behavior and network aspects. In particular, it has
been shown that rigorous experimental design for
multi-UAV systems often requires the combination
of specialized tools, such as robotic simulators
and network emulators, in order to jointly capture
mobility, sensory perception, and communication
performance [13]. This view supports the use of
the hybrid environment adopted in this work,
where CoppeliaSim handles the kinematics of
the drones and Mininet-WiFi emulates the SDN-
based wireless network.
The proposed design enables uninterrupted video
streaming over strategic areas, such as regions
with elevated crime rates. Key performance
metrics are analyzed to evaluate system behavior.
The following sections describe the methodology
and present the results that characterize the
performance and scalability of the network
architecture.
collection quickly and easily. By separating the
network's control and data planes, SDN technology
makes data management processes more flexible
and gives users more control. The control plane is
in charge of network signaling, route calculation,
system management, and configuration. It is
the part of the network that decides how it will
behave. On the other hand, the data plane’s job is
to send packets to their next destinations. So, you
can think of the network as a distributed structure
that connects a lot of independent devices. SDN
architecture separates the control functions from
the hardware infrastructure, putting all control
operations in a single programmable controller.
This centralized control unit lets network admins
change and improve how the network works in
real time, quickly adapting to new situations with
a specific controller app [7],[9],[10].
In wired SDN systems with a static infrastructure,
programmability refers to the ability of the
control plane to modify data paths as needed,
while the data plane implements these decisions
by forwarding packets through the assigned
interfaces. In contrast, when SDN is employed in
UAV networks (UAVNets), programmability also
involves managing the movement of the UAVs to
prevent collisions or to enhance the performance
of the applications, selecting or updating routing
paths, and adjusting transmission parameters such
as data rate or transmission power in response to
performance or energy constraints, among several
other functions [5].
From a communications perspective, UAV
networks are characterized by high mobility,
predominantly line-of-sight links, and a rapidly
and frequently changing topology. Recent tutorials
on UAV-assisted communications highlight that
this three-dimensional and dynamic nature of
the channel requires flexible network control
mechanisms capable of adapting to variations
in the radio environment and connectivity [11].
In this context, the introduction of a logically
centralized and programmable control plane, such
as that provided by SDN, is a natural choice for
reacting to topology changes and adjusting routes
and link configurations in near real time.
One of the major difficulties in these environments
is the occurrence of broadcast storms. In traditional
networks that contain loops, Spanning Tree
Protocol is typically employed to create a loop-free
logical topology and thereby prevent uncontrolled
broadcast propagation. In contrast, within an SDN
environment, the controller leverages its global
view of the topology to compute a spanning tree
in a centralized manner to mitigate the same issue
[12].
In this paper, we propose a simple SDN-based
drone network design for Aerial Surveillance.
The goal is to provide an architecture that is
easy to understand and configure while ensuring
reliable backhaul connectivity for real-time
video transmission. Because of its simplicity, the
network can be deployed rapidly in unforeseen
or urgent situations requiring immediate aerial
monitoring. The design incorporates the STP to
prevent loops in the backbone and maintain stable
packet forwarding. In addition, the simulation
integrates Mininet-WiFi with CoppeliaSim,
enabling a combined evaluation of network
behavior and drone mobility. Recent literature on
experimentation with UAV swarms emphasizes
that, in many cases, a single simulation platform
is not sufficient to accurately model both robotic
behavior and network aspects. In particular, it has
been shown that rigorous experimental design for
multi-UAV systems often requires the combination
of specialized tools, such as robotic simulators
and network emulators, in order to jointly capture
mobility, sensory perception, and communication
performance [13]. This view supports the use of
the hybrid environment adopted in this work,
where CoppeliaSim handles the kinematics of
the drones and Mininet-WiFi emulates the SDN-
based wireless network.
The proposed design enables uninterrupted video
streaming over strategic areas, such as regions
with elevated crime rates. Key performance
metrics are analyzed to evaluate system behavior.
The following sections describe the methodology
and present the results that characterize the
performance and scalability of the network
architecture.
13REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
II. Background
A. Drones
Drones, also called unmanned aerial vehicles
(UAVs), are aircraft that fly without a human pilot
on board. They have become very popular due
to their mobility, flexibility, and adjustable flight
heights. These characteristics allow drones to be
used in many areas, such as military operations,
surveillance, telecommunications, medical supply
deliveries, and rescue missions. In wireless
communication systems, drones can act as aerial
base stations. Additionally, they can serve as
mobile devices, connecting directly to cellular
networks for tasks like live video streaming or
delivering packages. Another way to classify
drones is by their design. Fixed wing: drones like
small airplanes fly at high speeds, cannot hover,
but have longer flight times. Rotary wing: drones
such as quadcopters that can hover in place but
typically have shorter flight durations due to
higher energy consumption [13].
Compared with ground vehicles and fixed
infrastructure, aerial drones benefit from a
substantially higher probability of establishing line-
of-sight (LoS) links due to their elevated position
and reduced obstruction, a behavior extensively
characterized in UAV propagation models [14].
B. Software-Defined Networks (SDN)
Software-defined is a network architecture where
network control is decoupled from forwarding and
is directly programmable. So, SDN is defined by
two characteristics, namely decoupling of control
and data planes, and programmability on the
control plane [15].
SDN separates the routing and forwarding
decisions of networking elements (e.g., routers,
switches, and access points) from the data plane.
Network administration and management become
simple because the control plane only deals
with the information related to logical network
topology, the routing of traffic, and so on. In
contrast, the data plane orchestrates the network
traffic according to the established configuration
in the control plane [16].
C. Ryu Controller
The controller is one of the most relevant elements
in Software Defined Networking (SDN), and it
is responsible for managing and programming
different network applications. There are many
controllers with different programming languages,
and the protocol versions they support. They are
also designed for different environments, like data
centers or cloud computing. The RYU controller
is an open-source option developed in Python. It
supports versions of the OpenFlow (OF) protocol
[17].
Fig. 1 illustrates the RYU controller’s main
components, which help in developing network
applications and managing networks. Some of
its tools include OFconfig, used to set up the
OpenFlow protocol; the Open Virtual Switch
Database (OVSDB) library, which manages switch
settings and allows users to create, edit, or delete
flow table rules; and the NETConfig library,
which applies configurations to devices across the
network [18].
Fig. 1. Architecture of RYU controller. Source: Adapted from [18]
D. Mininet WiFi
Mininet is a tool used for emulating networks.
Allows create virtual hosts, switches, links,
and controllers, all within one machine. This is
possible due to container-based virtualization,
which allows the system to behave like a real
network. It’s a cost-effective and reliable option
for building and testing applications that use
OpenFlow. With Mininet, there’s no need to set
up physical hardware to try out different network
II. Background
A. Drones
Drones, also called unmanned aerial vehicles
(UAVs), are aircraft that fly without a human pilot
on board. They have become very popular due
to their mobility, flexibility, and adjustable flight
heights. These characteristics allow drones to be
used in many areas, such as military operations,
surveillance, telecommunications, medical supply
deliveries, and rescue missions. In wireless
communication systems, drones can act as aerial
base stations. Additionally, they can serve as
mobile devices, connecting directly to cellular
networks for tasks like live video streaming or
delivering packages. Another way to classify
drones is by their design. Fixed wing: drones like
small airplanes fly at high speeds, cannot hover,
but have longer flight times. Rotary wing: drones
such as quadcopters that can hover in place but
typically have shorter flight durations due to
higher energy consumption [13].
Compared with ground vehicles and fixed
infrastructure, aerial drones benefit from a
substantially higher probability of establishing line-
of-sight (LoS) links due to their elevated position
and reduced obstruction, a behavior extensively
characterized in UAV propagation models [14].
B. Software-Defined Networks (SDN)
Software-defined is a network architecture where
network control is decoupled from forwarding and
is directly programmable. So, SDN is defined by
two characteristics, namely decoupling of control
and data planes, and programmability on the
control plane [15].
SDN separates the routing and forwarding
decisions of networking elements (e.g., routers,
switches, and access points) from the data plane.
Network administration and management become
simple because the control plane only deals
with the information related to logical network
topology, the routing of traffic, and so on. In
contrast, the data plane orchestrates the network
traffic according to the established configuration
in the control plane [16].
C. Ryu Controller
The controller is one of the most relevant elements
in Software Defined Networking (SDN), and it
is responsible for managing and programming
different network applications. There are many
controllers with different programming languages,
and the protocol versions they support. They are
also designed for different environments, like data
centers or cloud computing. The RYU controller
is an open-source option developed in Python. It
supports versions of the OpenFlow (OF) protocol
[17].
Fig. 1 illustrates the RYU controller’s main
components, which help in developing network
applications and managing networks. Some of
its tools include OFconfig, used to set up the
OpenFlow protocol; the Open Virtual Switch
Database (OVSDB) library, which manages switch
settings and allows users to create, edit, or delete
flow table rules; and the NETConfig library,
which applies configurations to devices across the
network [18].
Fig. 1. Architecture of RYU controller. Source: Adapted from [18]
D. Mininet WiFi
Mininet is a tool used for emulating networks.
Allows create virtual hosts, switches, links,
and controllers, all within one machine. This is
possible due to container-based virtualization,
which allows the system to behave like a real
network. It’s a cost-effective and reliable option
for building and testing applications that use
OpenFlow. With Mininet, there’s no need to set
up physical hardware to try out different network
14REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
setups, since it can build custom and complex
topologies virtually. It also comes with a simple
Python interface that makes it easy to design and
test networks. [19].
Mininet-WiFi adds wireless functionality by
creating virtual Wi-Fi stations (STAs) and access
points (APs) that use the mac80211/SoftMAC
driver. The driver stack is found in most current
Linux wireless cards. Since the majority of Linux
wireless drivers rely on mac80211/SoftMAC,
Mininet-WiFi can access almost all the features
of real Wi-Fi adapters and gives users very fine,
low-level control over each wireless packet [20].
E. Coppelia Sim
Coppelia Sim is used for algorithm development,
factory automation simulations, fast prototyping
and verification, robotics-related education, and
remote monitoring. Coppelia Sim is based on
a distributed control architecture; each object/
model can be individually controlled via a remote
API client (Python, Lua, Java, MATLAB, Octave,
C, C++, Rust) [21].
F. Propagation Model
Propagation path models represent a set of algorithms
and mathematical equations that are used for signal
strength estimation in a particular terrain profile.
Propagation models can be classified into three types
of models. Empirical models, Deterministic models,
and Statistical models. Empirical models use a set
of equations obtained from the results of several
measurements. Deterministic Models use reflection
& diffraction laws, which govern electromagnetic
signal propagation. Statistical models model the
terrain profile as a series of random variables and
depend on probability analysis to predict path loss.
These models need the least information about the
terrain profile and are the least accurate. There are
three different area types, namely, Rural, Suburban,
and Urban. Rural Area [22].
G. Handover Effect
Handover or handoff is the procedure by which
a mobile node transfers its wireless link from
one cell to another, reassigning radio resources
such as frequency, time slot, spreading code, or
a combination thereof without interrupting the
ongoing communication. The transfer is typically
triggered when the device crosses the coverage
boundary of a cell or when the signal quality drops
below a set threshold, thereby preserving session
continuity and maintaining quality of service in
mobile systems [23].
H. Spanning Tree Protocol
Spanning Tree Protocol (STP) is a protocol that
operates in the data link layer (Layer 2) of the OSI
model. STP allows for defining a loop-free topology
by preventing broadcast storms that occur when there
are loops. This protocol operates by exchanging
messages between switches to determine a root
bridge, which is the central point of the spanning
tree. Then, the switches calculate the shortest path to
the root bridge and disable any redundant links that
could create loops. When a link that is part of the
active links is disabled, the protocol searches for an
alternative link in the network [24].
The root bridge is identified by a unique Bridge
ID, which consists of two parts: a configurable
priority value and the bridge’s MAC address.
In STP, bridges exchange Bridge Protocol Data
Units (BPDUS) to evaluate and share information
about the configuration of bridges and ports, which
determines whether ports should be forwarded or
blocked. Therefore, STP defines three types of
roles that ports can take.
To establish these roles, the following
considerations are as follows:
• The switch with the lowest ID is designated as
the root bridge. This switch sets all its ports as
designated ports.
• The port on each switch with the lowest cost
to the root bridge will be determined as the
root port, and the remaining ports will be
configured as designated ports. If the cost is
the same, the designated port is chosen based
on the lowest port ID.
setups, since it can build custom and complex
topologies virtually. It also comes with a simple
Python interface that makes it easy to design and
test networks. [19].
Mininet-WiFi adds wireless functionality by
creating virtual Wi-Fi stations (STAs) and access
points (APs) that use the mac80211/SoftMAC
driver. The driver stack is found in most current
Linux wireless cards. Since the majority of Linux
wireless drivers rely on mac80211/SoftMAC,
Mininet-WiFi can access almost all the features
of real Wi-Fi adapters and gives users very fine,
low-level control over each wireless packet [20].
E. Coppelia Sim
Coppelia Sim is used for algorithm development,
factory automation simulations, fast prototyping
and verification, robotics-related education, and
remote monitoring. Coppelia Sim is based on
a distributed control architecture; each object/
model can be individually controlled via a remote
API client (Python, Lua, Java, MATLAB, Octave,
C, C++, Rust) [21].
F. Propagation Model
Propagation path models represent a set of algorithms
and mathematical equations that are used for signal
strength estimation in a particular terrain profile.
Propagation models can be classified into three types
of models. Empirical models, Deterministic models,
and Statistical models. Empirical models use a set
of equations obtained from the results of several
measurements. Deterministic Models use reflection
& diffraction laws, which govern electromagnetic
signal propagation. Statistical models model the
terrain profile as a series of random variables and
depend on probability analysis to predict path loss.
These models need the least information about the
terrain profile and are the least accurate. There are
three different area types, namely, Rural, Suburban,
and Urban. Rural Area [22].
G. Handover Effect
Handover or handoff is the procedure by which
a mobile node transfers its wireless link from
one cell to another, reassigning radio resources
such as frequency, time slot, spreading code, or
a combination thereof without interrupting the
ongoing communication. The transfer is typically
triggered when the device crosses the coverage
boundary of a cell or when the signal quality drops
below a set threshold, thereby preserving session
continuity and maintaining quality of service in
mobile systems [23].
H. Spanning Tree Protocol
Spanning Tree Protocol (STP) is a protocol that
operates in the data link layer (Layer 2) of the OSI
model. STP allows for defining a loop-free topology
by preventing broadcast storms that occur when there
are loops. This protocol operates by exchanging
messages between switches to determine a root
bridge, which is the central point of the spanning
tree. Then, the switches calculate the shortest path to
the root bridge and disable any redundant links that
could create loops. When a link that is part of the
active links is disabled, the protocol searches for an
alternative link in the network [24].
The root bridge is identified by a unique Bridge
ID, which consists of two parts: a configurable
priority value and the bridge’s MAC address.
In STP, bridges exchange Bridge Protocol Data
Units (BPDUS) to evaluate and share information
about the configuration of bridges and ports, which
determines whether ports should be forwarded or
blocked. Therefore, STP defines three types of
roles that ports can take.
To establish these roles, the following
considerations are as follows:
• The switch with the lowest ID is designated as
the root bridge. This switch sets all its ports as
designated ports.
• The port on each switch with the lowest cost
to the root bridge will be determined as the
root port, and the remaining ports will be
configured as designated ports. If the cost is
the same, the designated port is chosen based
on the lowest port ID.
15REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
Fig.2 shows the different roles of each port. The
Root Port receives BPDU packets originating from
the Root Bridge. The Designated Port forwards
BPDU packets received from the Root Bridge
to other ports, and Non-designated Ports block
frame transmission to prevent network loops.
Besides, STP protocol defines four main states for
each port:
• Blocking: This port is blocked to prevent
loops in the network.
• Listening: The port processes BPDUs and
waits for new information that could cause it
to return to the blocking state.
• Forwarding: Operates normally by
forwarding and receiving frames.
• Disabled: This port neither forwards
frames nor participates in the spanning tree
configuration
Although STP is a protocol used in Ethernet
networks, it can also be used in modern networks
such as Software-Defined Networks (SDN).
However, it has certain limitations due to its
lack of suitability for these types of networks. In
large-scale networks, this protocol presents some
limitations in loop prevention [24].
• Designated ports will be set as forwarding
ports, while the other ports will be blocked
ports.
A. Equipment and Materials
1) Hardware: The hardware components used to
simulate this study include the following:
• DELL Inspiron 15 laptop with 16 GB of RAM
and a 512 GB SSD, Intel Core i5 7th 2.5GHz
2) Software: The Software and operating system
used in this project are as follows:
• Ubuntu 20.04.6
• Coppelia Sim
• Mininet Wi-Fi
• Ryu Controller
• VLC media player
• Wireshark
B. Network Topology
Fig.3 shows the SDN-based drone network
topology, consisting of a wired backbone with
four strategically placed Access Points (APs),
forming a robust infrastructure that covers the
area of interest for aerial surveillance. Within
this coverage area, three drones were deployed to
analyze the network metrics. The drones used in
the simulation are quadcopters that feature four
vertical rotors which provide lift and control. An
example of such a drone is the Parrot AR.Drone.
Fig. 2. STP Port roles
Fig. 3. Topology of SDN-based drone
III. Methodology
Fig.2 shows the different roles of each port. The
Root Port receives BPDU packets originating from
the Root Bridge. The Designated Port forwards
BPDU packets received from the Root Bridge
to other ports, and Non-designated Ports block
frame transmission to prevent network loops.
Besides, STP protocol defines four main states for
each port:
• Blocking: This port is blocked to prevent
loops in the network.
• Listening: The port processes BPDUs and
waits for new information that could cause it
to return to the blocking state.
• Forwarding: Operates normally by
forwarding and receiving frames.
• Disabled: This port neither forwards
frames nor participates in the spanning tree
configuration
Although STP is a protocol used in Ethernet
networks, it can also be used in modern networks
such as Software-Defined Networks (SDN).
However, it has certain limitations due to its
lack of suitability for these types of networks. In
large-scale networks, this protocol presents some
limitations in loop prevention [24].
• Designated ports will be set as forwarding
ports, while the other ports will be blocked
ports.
A. Equipment and Materials
1) Hardware: The hardware components used to
simulate this study include the following:
• DELL Inspiron 15 laptop with 16 GB of RAM
and a 512 GB SSD, Intel Core i5 7th 2.5GHz
2) Software: The Software and operating system
used in this project are as follows:
• Ubuntu 20.04.6
• Coppelia Sim
• Mininet Wi-Fi
• Ryu Controller
• VLC media player
• Wireshark
B. Network Topology
Fig.3 shows the SDN-based drone network
topology, consisting of a wired backbone with
four strategically placed Access Points (APs),
forming a robust infrastructure that covers the
area of interest for aerial surveillance. Within
this coverage area, three drones were deployed to
analyze the network metrics. The drones used in
the simulation are quadcopters that feature four
vertical rotors which provide lift and control. An
example of such a drone is the Parrot AR.Drone.
Fig. 2. STP Port roles
Fig. 3. Topology of SDN-based drone
III. Methodology
16REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
Each drone operates as a wireless AP with a
coverage of 20 meters, that continuously sends
live video data through the AP backbone network
that has a coverage area of 50 meters by each
AP. In this case both the AP and drone use WIFI
for communication allowing handover when the
drone is far from the AP. Therefore, if the drone is
far and the coverage is insufficient, it changes the
AP and continues sending the information.
The APs create a coordinated network, centrally
managed by an SDN controller (Ryu), which can
be used to optimize, route and improve network
traffic. Therefore, the controller allows maintains
communication links and adjusts the data route
in real time, ensuring efficient use of network
resources to maintain video data quality. The video
data transmitted by the drones is sent through the
backbone network to the base station, which is
represented by an AP that has two interfaces, one
wireless and one wired, allowing communication
with the server and other APs.
C. Performance Metrics
This project uses Coppelia Sim in conjunction with
Mininet-WiFi and the Ryu controller to simulate
the behavior of drones navigating the surveillance
area, as shown in Fig. 4. This simulation does not
consider aspects such as obstacles or wind, which
facilitates the evaluation of network performance
under optimal operating conditions. The metrics
are analyzed at the base station, which is where
the data arrives, to understand the behavior of the
network.
The metrics evaluated are throughput, packet loss,
jitter, round-trip time (RTT), and bandwidth.
D. Drone Movement and Positioning
For the implementation of the topology, specific
parameters were established in each drone for its
movement and the camera coverage. The camera
used in the simulation is configured with a FOV
(Field of View) of 60 degrees vertically, meaning
it has a conical aperture of 60 degrees with a field
of view from top to bottom. This FOV is the same
for all drones.
The movement of each drone is done diagonally
up and down, considering movements from the left
and right, until it forms a square. Each drone motor
is configured to move 100 steps, considering a
displacement value of 0.005. Therefore, the drone
travels 0.5 meters along the X and Y axes. The
total displacement of the configured drone will be
10 meters on each side. Therefore, the maximum
coverage area of each drone is 100 square meters
as shown in Fig.5.
Related work on disaster management with
multi-UAV systems and FANET networks use
predefined trajectories and geometric coverage
models to study how flight paths and field of
view influence the monitored area and network
connectivity. These studies show that trajectory
design and coordination between multiple UAVs
have a direct impact on both coverage quality and
wireless link stability [6], [25]. This justifies the
use of simple movement patterns, such as square or
systematic sweep trajectories, to analyze the effect
of mobility on network coverage and performance
in a controlled manner.
Fig. 4. Coppelia Sim simulation of the drone behavior. Fig. 5. Scanning area of the drone.
Each drone operates as a wireless AP with a
coverage of 20 meters, that continuously sends
live video data through the AP backbone network
that has a coverage area of 50 meters by each
AP. In this case both the AP and drone use WIFI
for communication allowing handover when the
drone is far from the AP. Therefore, if the drone is
far and the coverage is insufficient, it changes the
AP and continues sending the information.
The APs create a coordinated network, centrally
managed by an SDN controller (Ryu), which can
be used to optimize, route and improve network
traffic. Therefore, the controller allows maintains
communication links and adjusts the data route
in real time, ensuring efficient use of network
resources to maintain video data quality. The video
data transmitted by the drones is sent through the
backbone network to the base station, which is
represented by an AP that has two interfaces, one
wireless and one wired, allowing communication
with the server and other APs.
C. Performance Metrics
This project uses Coppelia Sim in conjunction with
Mininet-WiFi and the Ryu controller to simulate
the behavior of drones navigating the surveillance
area, as shown in Fig. 4. This simulation does not
consider aspects such as obstacles or wind, which
facilitates the evaluation of network performance
under optimal operating conditions. The metrics
are analyzed at the base station, which is where
the data arrives, to understand the behavior of the
network.
The metrics evaluated are throughput, packet loss,
jitter, round-trip time (RTT), and bandwidth.
D. Drone Movement and Positioning
For the implementation of the topology, specific
parameters were established in each drone for its
movement and the camera coverage. The camera
used in the simulation is configured with a FOV
(Field of View) of 60 degrees vertically, meaning
it has a conical aperture of 60 degrees with a field
of view from top to bottom. This FOV is the same
for all drones.
The movement of each drone is done diagonally
up and down, considering movements from the left
and right, until it forms a square. Each drone motor
is configured to move 100 steps, considering a
displacement value of 0.005. Therefore, the drone
travels 0.5 meters along the X and Y axes. The
total displacement of the configured drone will be
10 meters on each side. Therefore, the maximum
coverage area of each drone is 100 square meters
as shown in Fig.5.
Related work on disaster management with
multi-UAV systems and FANET networks use
predefined trajectories and geometric coverage
models to study how flight paths and field of
view influence the monitored area and network
connectivity. These studies show that trajectory
design and coordination between multiple UAVs
have a direct impact on both coverage quality and
wireless link stability [6], [25]. This justifies the
use of simple movement patterns, such as square or
systematic sweep trajectories, to analyze the effect
of mobility on network coverage and performance
in a controlled manner.
Fig. 4. Coppelia Sim simulation of the drone behavior. Fig. 5. Scanning area of the drone.
17REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
To move the drone in the (X, Y) plane, the
current position of the drone is obtained, and the
relative position concerning its initial position
is calculated. This allows for left and right
movement, enabling horizontal control of the
drone's position. Fig.6 shows a diagram of the
drone’s movement. The initial position is defined
as P0, from which the first displacement moves
the drone above the Y-axis and to the right
along the X-axis by 10 meters. In the second
displacement, P1, the Y- coordinate remains the
same, and the drone moves only along the X-axis
to the right. In the third displacement, P2, the
drone moves below the Y-axis and to the left
along the X- axis. Finally, in P3, it maintains the
same Y-coordinate and moves to the left along
the X-axis, forming a square as shown in Fig. 6.
Due to the simulation being wireless, different
propagation models based on the log-distance
model can be configured. Some models that
Mininet-WiFi allow configuration are specified
in Table III with their different exponents that
can be configured.
To simulate video acquisition by the drones,
video streaming was performed using the VLC
media player with the Real-Time Streaming
Protocol (RTSP), which operates in conjunction
with the Real-Time Transport Protocol (RTP)
for multimedia data delivery. To verify that the
network was correctly transmitting and receiving
video traffic, packets at the base station were
analyzed using Wireshark. The captured RTP
packets had a payload size of 475 bytes.
The use of real-time video streams as traffic
load is consistent with previous studies on UAV
communications, where video streaming is used
as a representative application due to its high
bandwidth consumption and sensitivity to channel
variations. It has been shown that, in IEEE 802.11
links used by drones, throughput fluctuations,
latency, and jitter have a direct impact on service
continuity and perceived video quality [27], [28].
For this reason, the evaluation of metrics such as
throughput, RTT, jitter, and packet loss is essential
when analyzing communication architectures for
real-time aerial surveillance.
General topology can be observed using the
Mininet-WiFi Graph tool, as shown in Fig. 7, to
verify that the topology was configured correctly,
The Table I shows the main configured parameters
for each drone and AP.
E. Mininet WiFi and Coppelia Sim
To enable communication between Mininet WiFi
and Coppelia Sim, a socket server was configured
to create a communication channel between them
for sending data. Table II shows the configured
parameters for establishing communication
between Mininet WiFi and Coppelia Sim and the
propagation model configured.
Fig. 6. Drone movement around the area.
Table. I. main network parameters
Table. II. CommuniCation parameters
Table. III. table of log distanCe model exponents
Configuration Parameter
Height Drone 5m
Conical Aperture (Camera) 60°
APs Coverage Area 50m2
Drone Coverage Area 20m2
Dron motor step 100
Displacement Value 0.005
Configuration Parameter
IP Address Server 127.0.0.1
Port 12345
Propagation Model
Exponential
long-distance
3
Exp Environment Description
2 Free-Space Ideal model with no obstacles or
reflections
3 Standard Urban City Urban areas with scattered
buildings
4 Densely Built Urban Area High density of buildings and
obstructions
To move the drone in the (X, Y) plane, the
current position of the drone is obtained, and the
relative position concerning its initial position
is calculated. This allows for left and right
movement, enabling horizontal control of the
drone's position. Fig.6 shows a diagram of the
drone’s movement. The initial position is defined
as P0, from which the first displacement moves
the drone above the Y-axis and to the right
along the X-axis by 10 meters. In the second
displacement, P1, the Y- coordinate remains the
same, and the drone moves only along the X-axis
to the right. In the third displacement, P2, the
drone moves below the Y-axis and to the left
along the X- axis. Finally, in P3, it maintains the
same Y-coordinate and moves to the left along
the X-axis, forming a square as shown in Fig. 6.
Due to the simulation being wireless, different
propagation models based on the log-distance
model can be configured. Some models that
Mininet-WiFi allow configuration are specified
in Table III with their different exponents that
can be configured.
To simulate video acquisition by the drones,
video streaming was performed using the VLC
media player with the Real-Time Streaming
Protocol (RTSP), which operates in conjunction
with the Real-Time Transport Protocol (RTP)
for multimedia data delivery. To verify that the
network was correctly transmitting and receiving
video traffic, packets at the base station were
analyzed using Wireshark. The captured RTP
packets had a payload size of 475 bytes.
The use of real-time video streams as traffic
load is consistent with previous studies on UAV
communications, where video streaming is used
as a representative application due to its high
bandwidth consumption and sensitivity to channel
variations. It has been shown that, in IEEE 802.11
links used by drones, throughput fluctuations,
latency, and jitter have a direct impact on service
continuity and perceived video quality [27], [28].
For this reason, the evaluation of metrics such as
throughput, RTT, jitter, and packet loss is essential
when analyzing communication architectures for
real-time aerial surveillance.
General topology can be observed using the
Mininet-WiFi Graph tool, as shown in Fig. 7, to
verify that the topology was configured correctly,
The Table I shows the main configured parameters
for each drone and AP.
E. Mininet WiFi and Coppelia Sim
To enable communication between Mininet WiFi
and Coppelia Sim, a socket server was configured
to create a communication channel between them
for sending data. Table II shows the configured
parameters for establishing communication
between Mininet WiFi and Coppelia Sim and the
propagation model configured.
Fig. 6. Drone movement around the area.
Table. I. main network parameters
Table. II. CommuniCation parameters
Table. III. table of log distanCe model exponents
Configuration Parameter
Height Drone 5m
Conical Aperture (Camera) 60°
APs Coverage Area 50m2
Drone Coverage Area 20m2
Dron motor step 100
Displacement Value 0.005
Configuration Parameter
IP Address Server 127.0.0.1
Port 12345
Propagation Model
Exponential
long-distance
3
Exp Environment Description
2 Free-Space Ideal model with no obstacles or
reflections
3 Standard Urban City Urban areas with scattered
buildings
4 Densely Built Urban Area High density of buildings and
obstructions
18REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
where the Base Station is represented by ap5,
where the metrics were collected to evaluate
network performance.
Fig. 8 presents a real-time console capture from
the Ryu controller during STP convergence. The
bridge with DPID 1000000000000002 (AP2) is
not the root once superior BPDUs arrive. Its ports
2 and 3 are first listed as DESIGNATED_PORT/
BLOCK, temporarily stopping traffic while the
algorithm recalculates the tree.
Subsequent lines show AP2 designating port 2 as
ROOT_PORT and moving it through the LISTEN
to LEARN to FORWARD sequence, whereas
port 3 remains NON_DESIGNATED_PORT /
LISTEN, preventing loops on that segment.
F. Spanning Tree Protocol Configuration
The STP protocol avoids looping between APs,
allowing redundancy-free communication on the
links. This protocol is implemented within the
RYU controller. The AP with the lowest priority
is defined as the root bridge, which is the central
point to build the loop-free tree. Each bridge has
a unique identifier called the bridge ID, which is
made up of the MAC address and a previously
configured priority value. STP has three types of
ports, which are root port, designated port, and
non-designated port. Through these ports, the
APs will be able to send, receive, or block packets,
avoiding loops.
Table IV shows the STP configuration, setting the
priority of each AP. The assigned value is entered
in hexadecimal format, where the lowest value
configured is 0x5000. Once the priorities have
been established, a table is defined that contains the
MAC addresses of each port, avoiding unnecessary
flooding. When the topology changes, the STP
protocol updates the port status on each AP and
changes the port type, thus deleting the flows
previously installed on the AP and cleaning up the
entries in its MAC table. If a packet arrives and the
destination MAC address is known to the AP, the
packet is forwarded directly. If the MAC address is
not known, the AP floods by broadcast until it learns
the new MAC address and can send the packet.
Other bridges display similar transition ports with
the lowest path cost becoming ROOT_PORT or
DESIGNATED_PORT, advance to the LEARN
state for MAC table population, and finally reach
FORWARD once the topology stabilizes, while
higher-cost ports stay blocked. Overall, the log
illustrates how STP systematically suppresses
redundant paths, elects forwarding interfaces, and
restores full connectivity without broadcast storms.
For this section Iperf tool was used to generate
UDP traffic with a fixed bandwidth of 10 Mbps.
Fig. 7. Mininet-WiFi of the topology.
Fig. 8. Ryu controller STP convergence.
Table. IV. table of log distanCe model exponents
AP MAC Address Status Priority
1 0x1 Bridge 0x7000
2 0x2 Bridge 0x5000
3 0x3 Bridge 0x6000
4 0x4 Bridge 0x8000
5 0x5 Bridge 0xa000
IV. Results
where the Base Station is represented by ap5,
where the metrics were collected to evaluate
network performance.
Fig. 8 presents a real-time console capture from
the Ryu controller during STP convergence. The
bridge with DPID 1000000000000002 (AP2) is
not the root once superior BPDUs arrive. Its ports
2 and 3 are first listed as DESIGNATED_PORT/
BLOCK, temporarily stopping traffic while the
algorithm recalculates the tree.
Subsequent lines show AP2 designating port 2 as
ROOT_PORT and moving it through the LISTEN
to LEARN to FORWARD sequence, whereas
port 3 remains NON_DESIGNATED_PORT /
LISTEN, preventing loops on that segment.
F. Spanning Tree Protocol Configuration
The STP protocol avoids looping between APs,
allowing redundancy-free communication on the
links. This protocol is implemented within the
RYU controller. The AP with the lowest priority
is defined as the root bridge, which is the central
point to build the loop-free tree. Each bridge has
a unique identifier called the bridge ID, which is
made up of the MAC address and a previously
configured priority value. STP has three types of
ports, which are root port, designated port, and
non-designated port. Through these ports, the
APs will be able to send, receive, or block packets,
avoiding loops.
Table IV shows the STP configuration, setting the
priority of each AP. The assigned value is entered
in hexadecimal format, where the lowest value
configured is 0x5000. Once the priorities have
been established, a table is defined that contains the
MAC addresses of each port, avoiding unnecessary
flooding. When the topology changes, the STP
protocol updates the port status on each AP and
changes the port type, thus deleting the flows
previously installed on the AP and cleaning up the
entries in its MAC table. If a packet arrives and the
destination MAC address is known to the AP, the
packet is forwarded directly. If the MAC address is
not known, the AP floods by broadcast until it learns
the new MAC address and can send the packet.
Other bridges display similar transition ports with
the lowest path cost becoming ROOT_PORT or
DESIGNATED_PORT, advance to the LEARN
state for MAC table population, and finally reach
FORWARD once the topology stabilizes, while
higher-cost ports stay blocked. Overall, the log
illustrates how STP systematically suppresses
redundant paths, elects forwarding interfaces, and
restores full connectivity without broadcast storms.
For this section Iperf tool was used to generate
UDP traffic with a fixed bandwidth of 10 Mbps.
Fig. 7. Mininet-WiFi of the topology.
Fig. 8. Ryu controller STP convergence.
Table. IV. table of log distanCe model exponents
AP MAC Address Status Priority
1 0x1 Bridge 0x7000
2 0x2 Bridge 0x5000
3 0x3 Bridge 0x6000
4 0x4 Bridge 0x8000
5 0x5 Bridge 0xa000
IV. Results
19REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
Fig. 9. Average bandwidth and throughput versus number of drones.
Table. V. metriCs for different numbers of drones
This value represents the maximum transmission
rate, which means the sender will attempt to
transmit packets at that rate. The metrics that were
considered for our project are throughput, packet
loss, and jitter. Table II presents a summary of
all metrics obtained by increasing the number of
drones. Round-trip time (RTT) delay, however,
will be measured separately using TCP flows to
provide an accurate characterization of end-to-end
latency. To obtain the averages, a total of twenty
measurements were taken for each case presented
in this project.
A. Metrics
1) Round-Trip Time (RTT), commonly referred
to as network delay or latency, is the total time
it takes for a data packet to travel from a source
host to a destination host and back again.
RTT is typically measured in milliseconds
(ms) and includes all delays encountered along
the network path, such as processing delays at
intermediate nodes, queuing delays, propagation
delays, and transmission delays.
2) Bandwidth: Refers to the maximum data-
carrying capacity of the wireless link,
expressed in megabits per second (Mbits/s). It
represents the theoretical upper limit on how
much information can be transmitted over the
channel in one second.
3) Jitter: refers to the variation in latency (packet
delay) experienced by data packets traveling
across a network. Unlike RTT, which measures
the average round-trip delay, jitter specifically
captures fluctuations and inconsistencies in
packet delivery times.
4) Packet Loss: refers to the comparison between
the total packets received in comparison to the
packets sent using a UDP protocol.
5) Throughput is a key performance metric in
networking, defined and represents how data
can be successfully transmitted from one node
to another in a given amount of time.
B. Comparison of metrics between different
numbers of drones
Table V summarizes the metrics obtained for
different numbers of drones. The data in Table V
shows two distinct behaviors. First, the average
bandwidth remains virtually constant: it increases
slightly when going from three to five drones and
returns to a very similar value with seven drones, so
that no significant variation is observed. The same
is true for jitter and RTT, which remain within
a narrow range without significant degradation.
In contrast, effective throughput grows almost
linearly as the number of drones increases,
reflecting the expected aggregation of throughputs
while the links do not reach saturation.
As shown in Fig. 9, the effective throughput grows
almost linearly as the number of drones increases,
reflecting the expected aggregation of throughputs
while the links don’t reach saturation.
As shown in Fig. 10, both Jitter and RTT remain
highly stable as the number of drones increases.
Jitter stays below 1 ms in all cases, indicating
consistent packet timing even as additional nodes
are introduced. RTT also exhibits only a slight
increase, rising from 8.53 ms with three drones to
8.99 ms with seven drones. This steady behavior
suggests that the network maintains low latency
Metric 3 Drone
Avg 5 Drone
Avg 7 Drone
Avg
Bandwidth (Mbits/s) 6.93 7.99 7.75
Jitter (ms) 0.55 0.61 0.58
Packet Loss (%) 21.34 22.20 24.43
Throughput (Mbits/s) 2.75 5.20 7.21
RTT delay (ms) 8.53 8.62 8.99
Fig. 9. Average bandwidth and throughput versus number of drones.
Table. V. metriCs for different numbers of drones
This value represents the maximum transmission
rate, which means the sender will attempt to
transmit packets at that rate. The metrics that were
considered for our project are throughput, packet
loss, and jitter. Table II presents a summary of
all metrics obtained by increasing the number of
drones. Round-trip time (RTT) delay, however,
will be measured separately using TCP flows to
provide an accurate characterization of end-to-end
latency. To obtain the averages, a total of twenty
measurements were taken for each case presented
in this project.
A. Metrics
1) Round-Trip Time (RTT), commonly referred
to as network delay or latency, is the total time
it takes for a data packet to travel from a source
host to a destination host and back again.
RTT is typically measured in milliseconds
(ms) and includes all delays encountered along
the network path, such as processing delays at
intermediate nodes, queuing delays, propagation
delays, and transmission delays.
2) Bandwidth: Refers to the maximum data-
carrying capacity of the wireless link,
expressed in megabits per second (Mbits/s). It
represents the theoretical upper limit on how
much information can be transmitted over the
channel in one second.
3) Jitter: refers to the variation in latency (packet
delay) experienced by data packets traveling
across a network. Unlike RTT, which measures
the average round-trip delay, jitter specifically
captures fluctuations and inconsistencies in
packet delivery times.
4) Packet Loss: refers to the comparison between
the total packets received in comparison to the
packets sent using a UDP protocol.
5) Throughput is a key performance metric in
networking, defined and represents how data
can be successfully transmitted from one node
to another in a given amount of time.
B. Comparison of metrics between different
numbers of drones
Table V summarizes the metrics obtained for
different numbers of drones. The data in Table V
shows two distinct behaviors. First, the average
bandwidth remains virtually constant: it increases
slightly when going from three to five drones and
returns to a very similar value with seven drones, so
that no significant variation is observed. The same
is true for jitter and RTT, which remain within
a narrow range without significant degradation.
In contrast, effective throughput grows almost
linearly as the number of drones increases,
reflecting the expected aggregation of throughputs
while the links do not reach saturation.
As shown in Fig. 9, the effective throughput grows
almost linearly as the number of drones increases,
reflecting the expected aggregation of throughputs
while the links don’t reach saturation.
As shown in Fig. 10, both Jitter and RTT remain
highly stable as the number of drones increases.
Jitter stays below 1 ms in all cases, indicating
consistent packet timing even as additional nodes
are introduced. RTT also exhibits only a slight
increase, rising from 8.53 ms with three drones to
8.99 ms with seven drones. This steady behavior
suggests that the network maintains low latency
Metric 3 Drone
Avg 5 Drone
Avg 7 Drone
Avg
Bandwidth (Mbits/s) 6.93 7.99 7.75
Jitter (ms) 0.55 0.61 0.58
Packet Loss (%) 21.34 22.20 24.43
Throughput (Mbits/s) 2.75 5.20 7.21
RTT delay (ms) 8.53 8.62 8.99
20REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
and does not encounter congestion effects that
would significantly impact temporal performance.
Fig. 11 shows that packet loss increases steadily
as the number of drones grows. The loss rate rises
from 21.34% with three drones to 22.20% with
five drones and reaches 24.43% with seven drones.
This upward trend suggests a gradual increase
in channel contention and interference, which
results in a higher probability of packet drops as
additional nodes share the wireless medium.
Table VI also summarizes the metrics obtained
for the three-drone scenario under different
propagation exponent values (i.e., exp = 2, 3,
4). The average bandwidth remains around 7
Mbit/s (7.11, 6.93, and 7.45 Mbit/s). The channel’s
conditions mainly affect latency and temporal
stability, while aggregate throughput and loss
rate remain virtually constant in this range of
exponents, as we can observe in Fig. 12.
The channel conditions mainly affect latency and
temporal stability, while aggregate throughput and
loss rate remain virtually constant in this range of
exponents.
As shown in Fig. 13, RTT increases significantly
as the propagation exponent grows, rising from
12.57 ms (exp = 2) to 14.30 ms (exp = 3) and
reaching 18.53 ms at exp = 4. Jitter also increases,
although to a lesser extent, going from 0.507 ms to
0.55 ms and 0.748 ms, respectively. These trends
indicate a noisier channel with greater attenuation,
whose clearest impact is reflected in the delays
and temporal variability of the link.C. Comparison of metrics between different
propagation exponents
The Log-Distance model has different exponents
that refer to different environments, which were
specified in Table III. In this way, a comparative
analysis of the metrics in different environments
was performed, as shown in Table VI.
As shown in Fig. 14, packet loss presents only
slight variations as the propagation exponent
Fig. 10. Average Jitter and RTT versus number of drones
Fig. 12. Average bandwidth and throughput versus propagation
exponent
Fig. 13. Average Jitter and RTT versus number of drones
Fig. 11. Average Packet Loss versus number of drones
Table. VI. metriCs average for 3 drones for different propagation
exponents
Propagation
Exponent
Bandwidth
(Mbits/s) Jitter(ms) Packet
Loss (%)
Throughput
(Mbits/s)
RTT
delay
(ms)
exp = 2 7.11 0.507 31.00 2.76 12.57
exp = 3 6.93 0.55 32.00 2.75 14.30
exp = 4 7.45 0.748 31.33 2.76 18.53
and does not encounter congestion effects that
would significantly impact temporal performance.
Fig. 11 shows that packet loss increases steadily
as the number of drones grows. The loss rate rises
from 21.34% with three drones to 22.20% with
five drones and reaches 24.43% with seven drones.
This upward trend suggests a gradual increase
in channel contention and interference, which
results in a higher probability of packet drops as
additional nodes share the wireless medium.
Table VI also summarizes the metrics obtained
for the three-drone scenario under different
propagation exponent values (i.e., exp = 2, 3,
4). The average bandwidth remains around 7
Mbit/s (7.11, 6.93, and 7.45 Mbit/s). The channel’s
conditions mainly affect latency and temporal
stability, while aggregate throughput and loss
rate remain virtually constant in this range of
exponents, as we can observe in Fig. 12.
The channel conditions mainly affect latency and
temporal stability, while aggregate throughput and
loss rate remain virtually constant in this range of
exponents.
As shown in Fig. 13, RTT increases significantly
as the propagation exponent grows, rising from
12.57 ms (exp = 2) to 14.30 ms (exp = 3) and
reaching 18.53 ms at exp = 4. Jitter also increases,
although to a lesser extent, going from 0.507 ms to
0.55 ms and 0.748 ms, respectively. These trends
indicate a noisier channel with greater attenuation,
whose clearest impact is reflected in the delays
and temporal variability of the link.C. Comparison of metrics between different
propagation exponents
The Log-Distance model has different exponents
that refer to different environments, which were
specified in Table III. In this way, a comparative
analysis of the metrics in different environments
was performed, as shown in Table VI.
As shown in Fig. 14, packet loss presents only
slight variations as the propagation exponent
Fig. 10. Average Jitter and RTT versus number of drones
Fig. 12. Average bandwidth and throughput versus propagation
exponent
Fig. 13. Average Jitter and RTT versus number of drones
Fig. 11. Average Packet Loss versus number of drones
Table. VI. metriCs average for 3 drones for different propagation
exponents
Propagation
Exponent
Bandwidth
(Mbits/s) Jitter(ms) Packet
Loss (%)
Throughput
(Mbits/s)
RTT
delay
(ms)
exp = 2 7.11 0.507 31.00 2.76 12.57
exp = 3 6.93 0.55 32.00 2.75 14.30
exp = 4 7.45 0.748 31.33 2.76 18.53
21REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
Fig. 14. Average Packet Loss versus number of drones
changes. The loss rate increases from 31.00% at
exp = 2 to 32.00% at exp = 3 and then decreases to
31.33% at exp = 4. Overall, the variations remain
small across the evaluated exponents.
Similar behavior is described in studies of wireless
communications with UAVs, which show that the
trajectory loss exponent and line-of-sight (LoS)
or non-line-of-sight (NLoS) conditions directly
influence signal attenuation, signal-to-noise ratio,
and link coverage probability [11], [14].
These studies show that as the environment
becomes more obstructed or the propagation
exponent increases, the channel degrades and
it becomes more difficult to maintain reliable
links with comparable quality levels, resulting
in an overall degradation of communication
performance. In our scenario, this effect is
reflected in the increase in average RTT values
and in the greater temporal variability observed
when higher propagation exponents are used.
This work analyzes the performance of an
Unmanned Aerial Vehicle (UAV) network for
real-time video transmission based on SDN.
A network topology was implemented using
Mininet-WiFi, Coppelia Sim, a Ryu controller
responsible for managing packet forwarding in the
network, and the STP protocol to prevent loops
within the backbone network, which consists of
four strategically placed Access Points (APs).
The main metrics obtained are throughput,
packet loss, jitter, and RTT (Round-Trip Time).
The results show that as the number of drones
increases, the network consumes more bandwidth,
V. Conclusions
starting with an initial bandwidth of 6.93 Mbps
with 3 drones and increasing to 7.75 Mbps with
7 drones. As the number of drones increases,
this value will continue to rise, which could
saturate the links and cause information loss,
and the controller would start to fail at managing
the network. When increasing the number of
drones, both the throughput and the packet loss
increase, resulting in a percentage of 21.43% when
considering only 3 drones, rising to 24.43% with
7 drones. Meanwhile, the RTT parameter has a
value of 8.53 ms with three drones compared to
having seven drones where the delay value is 8.99
ms; in this case, there is a small change that does
not affect the transmission; this value depends on
the distance at which each drone is positioned, so
its value could increase if the distance is very far.
If the metrics are analyzed considering different
propagation model exponent values, the RTT
and packet loss values vary depending on the
environment analyzed, showing a considerable
increase from working with an exponent of 2 with
an RTT of 12.57ms to a value of 18.53ms when
considering an exponent of 4. However, the value
of the throughput is the same for all exponent
values; this is because this metric is not affected
by the propagation environment.
The behavior of the analyzed parameters depends
on the efficiency of the management of the RYU
controller and the STP protocol, which are
responsible for avoiding loops and managing
packet transmission between the drones and the
APs. However, this study is limited to having a
single controller; as future work, it is suggested to
incorporate multiple controllers for better traffic
management and network optimization.
[1] O. S. Oubbati, A. Lakas, F. Zhou, M. Güneş,
and M. B. Yagoubi, “A survey on position-based
routing protocols for Flying Ad hoc Networks
(FANETs),” Veh. Commun., vol. 10, pp. 29–56,
2017.
[2] O. S. Oubbati, A. Lakas, F. Zhou, M. Güneş,
and M. B. Yagoubi, “A survey on position-based
routing protocols for Flying Ad hoc Networks
V References
Fig. 14. Average Packet Loss versus number of drones
changes. The loss rate increases from 31.00% at
exp = 2 to 32.00% at exp = 3 and then decreases to
31.33% at exp = 4. Overall, the variations remain
small across the evaluated exponents.
Similar behavior is described in studies of wireless
communications with UAVs, which show that the
trajectory loss exponent and line-of-sight (LoS)
or non-line-of-sight (NLoS) conditions directly
influence signal attenuation, signal-to-noise ratio,
and link coverage probability [11], [14].
These studies show that as the environment
becomes more obstructed or the propagation
exponent increases, the channel degrades and
it becomes more difficult to maintain reliable
links with comparable quality levels, resulting
in an overall degradation of communication
performance. In our scenario, this effect is
reflected in the increase in average RTT values
and in the greater temporal variability observed
when higher propagation exponents are used.
This work analyzes the performance of an
Unmanned Aerial Vehicle (UAV) network for
real-time video transmission based on SDN.
A network topology was implemented using
Mininet-WiFi, Coppelia Sim, a Ryu controller
responsible for managing packet forwarding in the
network, and the STP protocol to prevent loops
within the backbone network, which consists of
four strategically placed Access Points (APs).
The main metrics obtained are throughput,
packet loss, jitter, and RTT (Round-Trip Time).
The results show that as the number of drones
increases, the network consumes more bandwidth,
V. Conclusions
starting with an initial bandwidth of 6.93 Mbps
with 3 drones and increasing to 7.75 Mbps with
7 drones. As the number of drones increases,
this value will continue to rise, which could
saturate the links and cause information loss,
and the controller would start to fail at managing
the network. When increasing the number of
drones, both the throughput and the packet loss
increase, resulting in a percentage of 21.43% when
considering only 3 drones, rising to 24.43% with
7 drones. Meanwhile, the RTT parameter has a
value of 8.53 ms with three drones compared to
having seven drones where the delay value is 8.99
ms; in this case, there is a small change that does
not affect the transmission; this value depends on
the distance at which each drone is positioned, so
its value could increase if the distance is very far.
If the metrics are analyzed considering different
propagation model exponent values, the RTT
and packet loss values vary depending on the
environment analyzed, showing a considerable
increase from working with an exponent of 2 with
an RTT of 12.57ms to a value of 18.53ms when
considering an exponent of 4. However, the value
of the throughput is the same for all exponent
values; this is because this metric is not affected
by the propagation environment.
The behavior of the analyzed parameters depends
on the efficiency of the management of the RYU
controller and the STP protocol, which are
responsible for avoiding loops and managing
packet transmission between the drones and the
APs. However, this study is limited to having a
single controller; as future work, it is suggested to
incorporate multiple controllers for better traffic
management and network optimization.
[1] O. S. Oubbati, A. Lakas, F. Zhou, M. Güneş,
and M. B. Yagoubi, “A survey on position-based
routing protocols for Flying Ad hoc Networks
(FANETs),” Veh. Commun., vol. 10, pp. 29–56,
2017.
[2] O. S. Oubbati, A. Lakas, F. Zhou, M. Güneş,
and M. B. Yagoubi, “A survey on position-based
routing protocols for Flying Ad hoc Networks
V References
22REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
(FANETs),” Veh. Commun., vol. 10, pp. 29–56,
2017.
[3] A. Agbeyangi, J. Odiete, and A. Olorunlomerue,
“Review on UAVs used for Aerial Surveillance,”
Journal of Multidisciplinary Engineering
Science and Technology (JMEST), vol. 3, 2016,
pp. 2458–9403.
[4] N. Dilshad, J. Hwang, J. Song, and N. Sung,
“Applications and challenges in video
surveillance via drone: A brief survey,” in
2020 International Conference on Information
and Communication Technology Convergence
(ICTC), 2020, pp. 728–732.
[5] S. Hayat, E. Yanmaz, and R. Muzaffar, “Survey on
Unmanned Aerial Vehicle Networks for Civil
Applications: A Communications Viewpoint,”
IEEE Communications Surveys & Tutorials, vol.
18, no. 4, pp. 2624–2661, 2016, doi: 10.1109/
COMST.2016.2560343.
[6] İ. Bekmezci, O. K. Şahingöz, and Ş. Temel,
“Flying Ad-Hoc Networks (FANETs): A
Survey,” Ad Hoc Networks, vol. 11, pp. 1254–
1270, 2013, doi: 10.1016/j.adhoc.2012.12.004.
[7] M. Alharthi, A.-E. M. Taha, and H. S.
Hassanein, “An architecture for software defined
drone networks”, in ICC 2019 - 2019 IEEE
International Conference on Communications
(ICC), 2019, pp. 1–5.
[8] Z. Zhao et al., “Software-defined unmanned
aerial vehicles networking for video
dissemination services”, Ad Hoc Netw., vol. 83,
pp. 68–77, 2019.
[9] S. Tomovic, M. Pejanovic-Djurisic, and I.
Radusinovic, “SDN-based mobile networks:
Concepts and benefits,” Wirel. Pers. Commun.,
vol. 78, no. 3, pp. 1629–1644, 2014.
[10] A. Kumar, R. Krishnamurthi, A. Nayyar, A. K.
Luhach, M. S. Khan, and A. Singh, “A novel
Software-Defined Drone Network (SDDN)-
based collision avoidance strategies for on-
road traffic monitoring and management,” Veh.
Commun., vol. 28, no. 100313, p. 100313, 2021.
[11] Y. Zeng, R. Zhang, and T. J. Lim, “Wireless
Communications with Unmanned Aerial
Vehicles: Opportunities and Challenges,”
IEEE Communications Magazine, vol.
54, no. 5, pp. 36–42, 2016, doi: 10.1109/
MCOM.2016.7470933.
[12] S.-Y. Wang, C.-C. Wu, and C.-L. Chou,
“Constructing an optimal spanning tree over a
hybrid network with SDN and legacy switches,”
in 2015 IEEE Symposium on Computers and
Communication (ISCC), 2015, pp. 502–507.
[13] A. Phadke, F. A. Medrano, C. N. Sekharan, and
T. Chu, “Designing UAV Swarm Experiments:
A Simulator Selection and Experiment Design
Process,” Sensors, vol. 23, no. 17, Art. no. 7359,
2023, doi: 10.3390/s23177359.
[14] M. Mozaffari, W. Saad, M. Bennis, Y.-H.
Nam, and M. Debbah, “A tutorial on UAVs for
wireless networks: Applications, challenges,
and open problems,” arXiv [cs.IT], 2018.
[15] Bor-Yaliniz and H. Yanikomeroglu, “The New
Frontier in RAN Heterogeneity: Multi-Tier
Drone-Cells,” IEEE Commun. Mag., vol. 54,
no. 11, 2016, pp. 48–55.
[16] W. Xia, Y. Wen, C. H. Foh, D. Niyato, and
H. Xie, “A survey on software-defined
networking,” IEEE Commun. Surv. Tutor, vol.
17, no. 1, pp. 27–51, 2015.
[17] K. Benzekki, A. El Fergougui, and A. Elbelrhiti
Elalaoui, “Software‐defined networking (SDN):
a survey: Software-defined networking: a
survey,” Secur. Commun. Netw., vol. 9, no. 18,
pp. 5803–5833, 2016.
[18] A. T. Albu-Salih, “Performance evaluation of
Ryu controller in software defined networks,” J.
Al-Qadisiyah Comput. Sci. Math., vol. 14, no.
1, p. Page 1-7, 2022.
[19] K. Kaur, J. Singh, and N. S. Ghumman,
"Mininet as software defined networking
testing platform," in Proc. Int. Conf. Commun.,
Comput. & Syst. (ICCCS), Aug. 2014, pp. 139–
142.
[20] R. R. Fontes, S. Afzal, S. H. B. Brito, M. A. S.
Santos, and C. E. Rothenberg, “Mininet-WiFi:
Emulating software-defined wireless networks,”
in 2015 11th International Conference on
Network and Service Management (CNSM),
2015, pp. 384–389.
[21] “Robot simulator CoppeliaSim: create,
compose, simulate, any robot - Coppelia
Robotics,” Coppeliarobotics.com. [Online].
Available: https://www.coppeliarobotics.com/.
[Accessed: 25-Jul-2025].
[22] V. S. Anusha, G. K. Nithya, and S. N. Rao,
“A comprehensive survey of electromagnetic
propagation models”, in 2017 International
(FANETs),” Veh. Commun., vol. 10, pp. 29–56,
2017.
[3] A. Agbeyangi, J. Odiete, and A. Olorunlomerue,
“Review on UAVs used for Aerial Surveillance,”
Journal of Multidisciplinary Engineering
Science and Technology (JMEST), vol. 3, 2016,
pp. 2458–9403.
[4] N. Dilshad, J. Hwang, J. Song, and N. Sung,
“Applications and challenges in video
surveillance via drone: A brief survey,” in
2020 International Conference on Information
and Communication Technology Convergence
(ICTC), 2020, pp. 728–732.
[5] S. Hayat, E. Yanmaz, and R. Muzaffar, “Survey on
Unmanned Aerial Vehicle Networks for Civil
Applications: A Communications Viewpoint,”
IEEE Communications Surveys & Tutorials, vol.
18, no. 4, pp. 2624–2661, 2016, doi: 10.1109/
COMST.2016.2560343.
[6] İ. Bekmezci, O. K. Şahingöz, and Ş. Temel,
“Flying Ad-Hoc Networks (FANETs): A
Survey,” Ad Hoc Networks, vol. 11, pp. 1254–
1270, 2013, doi: 10.1016/j.adhoc.2012.12.004.
[7] M. Alharthi, A.-E. M. Taha, and H. S.
Hassanein, “An architecture for software defined
drone networks”, in ICC 2019 - 2019 IEEE
International Conference on Communications
(ICC), 2019, pp. 1–5.
[8] Z. Zhao et al., “Software-defined unmanned
aerial vehicles networking for video
dissemination services”, Ad Hoc Netw., vol. 83,
pp. 68–77, 2019.
[9] S. Tomovic, M. Pejanovic-Djurisic, and I.
Radusinovic, “SDN-based mobile networks:
Concepts and benefits,” Wirel. Pers. Commun.,
vol. 78, no. 3, pp. 1629–1644, 2014.
[10] A. Kumar, R. Krishnamurthi, A. Nayyar, A. K.
Luhach, M. S. Khan, and A. Singh, “A novel
Software-Defined Drone Network (SDDN)-
based collision avoidance strategies for on-
road traffic monitoring and management,” Veh.
Commun., vol. 28, no. 100313, p. 100313, 2021.
[11] Y. Zeng, R. Zhang, and T. J. Lim, “Wireless
Communications with Unmanned Aerial
Vehicles: Opportunities and Challenges,”
IEEE Communications Magazine, vol.
54, no. 5, pp. 36–42, 2016, doi: 10.1109/
MCOM.2016.7470933.
[12] S.-Y. Wang, C.-C. Wu, and C.-L. Chou,
“Constructing an optimal spanning tree over a
hybrid network with SDN and legacy switches,”
in 2015 IEEE Symposium on Computers and
Communication (ISCC), 2015, pp. 502–507.
[13] A. Phadke, F. A. Medrano, C. N. Sekharan, and
T. Chu, “Designing UAV Swarm Experiments:
A Simulator Selection and Experiment Design
Process,” Sensors, vol. 23, no. 17, Art. no. 7359,
2023, doi: 10.3390/s23177359.
[14] M. Mozaffari, W. Saad, M. Bennis, Y.-H.
Nam, and M. Debbah, “A tutorial on UAVs for
wireless networks: Applications, challenges,
and open problems,” arXiv [cs.IT], 2018.
[15] Bor-Yaliniz and H. Yanikomeroglu, “The New
Frontier in RAN Heterogeneity: Multi-Tier
Drone-Cells,” IEEE Commun. Mag., vol. 54,
no. 11, 2016, pp. 48–55.
[16] W. Xia, Y. Wen, C. H. Foh, D. Niyato, and
H. Xie, “A survey on software-defined
networking,” IEEE Commun. Surv. Tutor, vol.
17, no. 1, pp. 27–51, 2015.
[17] K. Benzekki, A. El Fergougui, and A. Elbelrhiti
Elalaoui, “Software‐defined networking (SDN):
a survey: Software-defined networking: a
survey,” Secur. Commun. Netw., vol. 9, no. 18,
pp. 5803–5833, 2016.
[18] A. T. Albu-Salih, “Performance evaluation of
Ryu controller in software defined networks,” J.
Al-Qadisiyah Comput. Sci. Math., vol. 14, no.
1, p. Page 1-7, 2022.
[19] K. Kaur, J. Singh, and N. S. Ghumman,
"Mininet as software defined networking
testing platform," in Proc. Int. Conf. Commun.,
Comput. & Syst. (ICCCS), Aug. 2014, pp. 139–
142.
[20] R. R. Fontes, S. Afzal, S. H. B. Brito, M. A. S.
Santos, and C. E. Rothenberg, “Mininet-WiFi:
Emulating software-defined wireless networks,”
in 2015 11th International Conference on
Network and Service Management (CNSM),
2015, pp. 384–389.
[21] “Robot simulator CoppeliaSim: create,
compose, simulate, any robot - Coppelia
Robotics,” Coppeliarobotics.com. [Online].
Available: https://www.coppeliarobotics.com/.
[Accessed: 25-Jul-2025].
[22] V. S. Anusha, G. K. Nithya, and S. N. Rao,
“A comprehensive survey of electromagnetic
propagation models”, in 2017 International
23REVISTA PERSPECTIVASVOL. 8, N˚1 / ENERO - JUNIO 2026 / e-ISSN: 266-6688
Conference on Communication and Signal
Processing (ICCSP), 2017, pp. 1457–1462.
[23] “Effective management of handover process
in mobile communication network”, J. Adv.
Technol. Eng. Res., vol. 2, núm. 6, 2016.
[24] E. Amiri y R. Javidan, “A new method for
layer 2 loop prevention in software defined
networks”, Telecommunication Syst., vol. 73,
n.º 1, pp. 47–57, julio de 2019. Accedido el 31
de julio de 2025. [En línea]. Disponible: https://
doi.org/10.1007/s11235-019-00594-4
[25] M. Erdelj, M. Król, and E. Natalizio, “Wireless
Sensor Networks and Multi-UAV Systems for
Natural Disaster Management,” Computer
Networks, vol. 124, pp. 72–86, 2017, doi:
10.1016/j.comnet.2017.05.021
[26] B. Van den Bergh, A. Chiumento, and S. Pollin,
“Ultra-Reliable IEEE 802.11 for UAV Video
Streaming: From Network to Application,” in
Advances in Ubiquitous Networking 2, UNet
2016, Lecture Notes in Electrical Engineering,
vol. 397, Springer, 2016, pp. 637–647, doi:
10.1007/978-981-10-1627-1_50.
[27] S. Kacianka and H. Hellwagner, “Adaptive Video
Streaming for UAV Networks,” 2015, doi:
10.1145/2727040.2727043.
Conference on Communication and Signal
Processing (ICCSP), 2017, pp. 1457–1462.
[23] “Effective management of handover process
in mobile communication network”, J. Adv.
Technol. Eng. Res., vol. 2, núm. 6, 2016.
[24] E. Amiri y R. Javidan, “A new method for
layer 2 loop prevention in software defined
networks”, Telecommunication Syst., vol. 73,
n.º 1, pp. 47–57, julio de 2019. Accedido el 31
de julio de 2025. [En línea]. Disponible: https://
doi.org/10.1007/s11235-019-00594-4
[25] M. Erdelj, M. Król, and E. Natalizio, “Wireless
Sensor Networks and Multi-UAV Systems for
Natural Disaster Management,” Computer
Networks, vol. 124, pp. 72–86, 2017, doi:
10.1016/j.comnet.2017.05.021
[26] B. Van den Bergh, A. Chiumento, and S. Pollin,
“Ultra-Reliable IEEE 802.11 for UAV Video
Streaming: From Network to Application,” in
Advances in Ubiquitous Networking 2, UNet
2016, Lecture Notes in Electrical Engineering,
vol. 397, Springer, 2016, pp. 637–647, doi:
10.1007/978-981-10-1627-1_50.
[27] S. Kacianka and H. Hellwagner, “Adaptive Video
Streaming for UAV Networks,” 2015, doi:
10.1145/2727040.2727043.