Static and Dynamic Routing Protocols in Network

Security Attacks
1st Lovedeep Abrol 2nd Parul Gahelot
Computer Science & Engineering Computer Science & Engineering
Chitkara University Institute of Chitkara University Institute of
Engineering and Technology Engineering and Technology
Punjab, India Punjab, India
[email protected] [email protected]

Abstract— Network devices use communication protocols use routing that is configured manually to travel by the
to transfer information by physical or wireless technologies, network. On the other hand, in dynamic routing R1, R2, R3,
including routers and access points. Networking components and R4 are routers that use algorithms to determine efficient
are used to software monitor, manage and secure the network. paths to travel by the network.
The router is quite trained by the use of reinforcement
learning which can make all the hardest decisions for making
algorithms. Routing selects IP paths, also known as Internet
Protocol, and those packages are sent from their starting
point. Routers can make the decision to route packets along
with the network paths. An innovative routing system for
FANET is provided by an enhanced form of AntHocNet.
Dynamic routing protocols are more efficient in medium and
large networks because they require most of the computing
power, which is why smaller networks rely on the static
routing protocol. In this paper, we will discuss routing
algorithms, dynamic routing networks, UAV protocol, RPL in
the security of networks, and Flying Ad hoc networks.

Keywords— Dynamic routing networks, Flying ad hoc

networks, Mobile ad hoc networks, Routing protocol for low-
power and lossy networks, Unmanned aerial vehicles, Wireless Fig. 1. Static and dynamic routing
sensor networks.
Dynamic routers detect the shortest and fastest paths. It
I. INTRODUCTION detects on the basis of how long the packets will take to
Networking is exchanging and transporting data reach their destination. By Wang, W. et al. [2], Dynamic
between the nodes by the medium of informative data. Routing Networks (DRNets) can reduce the size of
Networking provides multiple uses in different fields or parameters and can reach an efficiency of inference, and
organizations as well as individual users. By the study by supports applications that require runtime control. DRNets
Zhang, C. et al. [25], Monitoring and controlling the reduce redundant computing which helps in improving
plethora of network components has become impossible due efficiency and it is due to the customizable architecture of
to the rising diversity and complexity of mobile network DRNets.
topologies.
Every organization needs network security to protect its
data or information from the cyber-threats. By Dalal, S. et
al. [11], Security in networks is basically a set of
configurations and rules to protect the accessibility and
integrity of networks. It is applied in different fields such as
firewall protection, network access control, cloud security,
virtual private networks, and secure web gateway.
An empirical examination of IoT-based smart city
automation that is then integrated for assessment has been
used to verify routers. By the study of Garraghan, P. et al.
[26], The router's design incorporates into account all of
these factors and optimizes numerous factors at once, such
as reaction time, Routing maintains and initializes routing
tables for the process of path determination. A router works Fig. 2. The number of increasing network devices and hackers
at the network layer in the OSI model. Routing has two By looking at Figure 2, It is clearly visible that the
types mainly, static routing and dynamic routing.
number of hackers is increasing every year due to an
Below in Figure 1, there is shown D1 and D2 are increase in the statistics of network devices. By network
devices, and in static routing R1 and R2 are routers and they devices, Hackers can always make use of network
equipment for their own gain and launch phishing attacks. secure. A routing protocol for the IoT is called Routing
There are numerous technological devices connected Protocol for Low-Power and Lossy Networks (RPL). The
online globally every second. So many cyber scams and role of RPL is to maintain communication of one-to-one
cyber threats are increasing every day, and that is why to and one-to-many. RPL has experienced many security
secure the network more securely. threats in data traffic. RPL protocol is also used to detect
RPL inside attacks using algorithms such as Machine
Several routing protocols were developed that increase Learning, they help to detect as well counter those threats.
the efficiency of the networks. According to the study by RPL security is used for the prevention of external
Muzammal, S. M. et al. [19], To secure communication in adversaries from entering the networks.
network systems, one needs to make sure that it is entirely

II. LITERATURE REVIEW

TABLE 1. A comparative study in different research papers

Author Name and Proposed Method Definition Advantage

Year Solution
Ignacio Cases et al, Recursive Natural A route that is covered through Quicker and produces less overhe
2019 [1] routing language another learned path from the routing ad
network inference information base for both the next
hop and the destination network

Shaofeng Cai et al, Dynamic Neural It allows routers to choose routes out Better in terms of adaptability and
2021 [2] routing architecture of logical network topology changes scaling
network search that occur in real-time
Ignacio Cases et al, Regularization Stochastic Regularization is the process of It prevents overfitting and less
2019 [3] and gradient training the model sufficiently adherence
composition in descent allowing it to generalize to data that
modules it has never seen before
Shoujin Wang et al, Mixture- System-based It is designed to determine each It determines each item's purpose
2019 [4] channel recommender item's goals and direct it into and assigns them to the
purpose system particular channels appropriate channels
routing
network
Muhammad Z. Wireless Low power In order to track the system, a huge Because it is scalable, any new
Hussain & Zurina M. sensor network networks number of wireless devices are nodes or devices can be added at
Hanapi, 2021 [5] deployed on ad hoc trends over an any time
infrastructure-free wireless network

Hua Yang & Zhiyong Dynamic Mobile ad hoc This protocol is used for routing in Only paths between nodes that
Liu, 2019 [6] source routing network and wireless networks called meshes require communication are kept
Flying ad hoc
network
Ali Seyfollahi & Ali RPL based on Low powered It serves as a routing system It is capable of supporting a
Ghaffari, 2021 [7] machine networks designed for low-power, frequently variety of connection levels
learning packet-loss-prone wireless networks

Matthew Riemer et al, Neural Multi-task It is a set of algorithms designed to A neural network makes decision
2017 [8] network learning find hidden connections in a piece of s and doesn't need to be program
data using a method that imitates med again

Zoubir Mammeri, Routing Reinforcement routing technology for WiFi Reinforcement learning is inventi
2019 [9] protocols learning networks that uses little power and is ve, flexible, and doesn't need a lot
based on frequently prone to loss of packets of labeled data
reinforcement
learning
Dylan Smyth et al, Security-aware Software- A unique variance-based error It is intended to stop assaults that
2023 [10] programmable defined detection technology supports this could have avoided control
switch networking function to offer a layered protection aircraft defenses

Table 2 given above displays the methods and advantages of different existing research papers.
 The modularity faced by machine learning architectures devices. By Hussain, M. Z. et al. [5], This method is used
in solving problems by forming them again from past for enhancing networks' functioning and it requires fewer
solutions, Tamkin, A. et al. [1] introduced Re-cursive resources. It helps network devices interact with sensors and
Routing Networks (RRNS) to address and optimize the connect to the nodes. The utilization of these sensors in
functions. applications is done by Routing Protocol for Low-Power
and Lossy Networks (RPL). RPL performs better in packet
Due to high inference costs, deployment of neural delivery rate and end-to-end latency. Cyber-Physical
networks is restricted. Wang, W. et al. [2] introduced the Systems (CPS) and the Internet of Things (IoT) work to
backbone network and purpose DRNets which have develop factories that increase efficiency and deepened
efficient inference. downtime to produce better products.
Wang, S. et al. [4] proposes the mixture-channel model Neural Architecture Search (NAS) iteratively searches
which more precisely represents a session by for the network architecture and in each iteration, the
accommodation of multi-purpose subsets of items. To architecture searches the algorithms to train and evaluate.
detect every item and assign items to the corresponding Wang, W. et al. [2] showed that during the inference, NAS-
channels, Wang, S. et al. [4] designed mixture-channel searched models have more accuracy and efficiency. For
purpose routing networks (MCPRNs). measuring the number of operations in a deep learning
Routing system modules face such challenges which are network, the FLOPs method is used in a test of computing
not faced by non-routed networks. Riemer, M. et al. [3] units while considering the time. MobileNets reduces the
studied all those challenges and identified the relationship parameter size and cost in mobile devices measured in
between collapse and overfitting and which holds FLOPs with depth-wise convolution.
optimization of the algorithm. A wide class of recursive neural networks is provided by
Wireless Sensor Networks (WSN) face threats the Neural Module Network (NMN). It provides the
simultaneously. Observation of Hussain, M. Z. et al. [5], platform for building deep networks with a dynamic
there’s no research conducted on security methods showing structure Hu, R. et al. [15]. Combining NMN modules may
insights and challenges and a wide range of factors. be trained on various architectures to provide compositional
behavior. The NMN has also concentrated on the
The mobility of nodes for higher speed became a great foundations of applications that assess questions in plain
challenge to FANET routing technologies. Yang, H. et al. language, use that analysis to create layouts, and offer
[6] found for solving this problem neural network based on results as a network applied to some world representations
dynamic source routing (DSR). to produce a response.
MTL is used to improve efficiency but it often had System-based recommender system (SBRS) method is
problems in the reduction of task interference. To address used to tackle the gaps in static recommender systems.
this problem, Rosenbaum, C. et al. [8] introduced a neural Wang, S. et al. [4] stated that SBRS is effective for single-
network and training algorithm. purpose sessions that keep items dependent on each other
Software-defined network (SDN) seems to poison the due to its single-based channel. Most of the existing SBRS
topology with multiple attacks. Smyth, D. et al. [10] comes form a session with a single purpose. Because of memory
up with a Security-Aware Programmable which is designed decay in Recurrent Neural Network (RNN), it fed to items
for the prevention of attacks that also supports anomaly without missing information from previous items. RNN’s
detection to provide a defense of layers to defeat those memory remembers all the information and uses the same
poisoning attacks. parameters as it performs the same task.
In New RPL models, there is the complexity of new Figure 3 given below shows there are devices D1, D2,
attacks. Ghaffari, A. et al. [7] studied RPL and comes up and, D3 connecting to the internet through the DDos traffic,
with an intrusion detection system in RPL that is based on and then the internet faces harmless traffic connecting to the
Deep Learning. router and cloud as well which helps in the identification of
malicious activities.
Reinforcement learning (RL) is for solving optimization
problems in routing networks such as distributed systems.
Mammeri, Z. [9] studied RL and shows characteristics that
present existing routing protocols based on RL.
III. METHODOLOGY
In our methodology, we will mainly focus on methods
in network architecture that identify malicious activity by
facing harmless traffic through the internet which is
connected by a router, and implement it by neural network
for immediate identification algorithm. In current most of
the existing research papers, there is only about improving
performance and efficiency not preventing the routing
networks from malicious attacks. The strength of this
method is to prevent malicious activities by identifying Fig. 3. The connection of devices to the cloud through the router
them. However, this method is that it is mainly suitable for
DDoS attacks. IV. ALGORITHMS IN ROUTING
LPN is intended for effective utilization as well as In the past, mobile ad hoc networks (MANETs) were
effective routing services in finite resources of network used more and more, but now flying ad hoc networks
(FANETs) are being used in the air. According to research With the development of the web in the last t years,
by Yang, H. et al. [6], FANET routing has an extremely mobility services like GPS-routing applications have
hard architecture. The fundamental component of both expanded. According to research by Cabannes, T. et al.
wired and wireless networks is a routing mechanism. There [16], these mobility services have not helped to relieve
are three types of route-finding strategies: hybrid, reactive, traffic congestion. Therefore, the usage of distributed
and proactive. security solutions becomes crucial for MANET security.
Reactive Routing: The reactive routing protocol does V. UAV ROUTING PROTOCOLS
not need to keep track of the routing information for every
Every UAV in the UAV ad Hoc network is involved in
node; instead, it simply generates routes whenever one of
sending data to the FANETs system. According to research
the sources needs to talk.
by Khan, M. A. et al. [13], the backbone UAV serves as a
Proactive Routing: According to a wired system, gateway through a station or a gateway that transports both
proactive routing requires each node to have a routing table low- and high-power wireless communication devices.
with information on how to go from one node to another.
UAV routing networks have prevalent tools. By Meng,
Such routers are not appropriate for FANETs because the
D. et al. [17] uncertain signal fading, rapid response, and
rapid mobility of nodes in FANETs frequently alters the
finite budget are the aspects to bridge the gap of required
topology of the network.
multi-source data of UAVs in routing networks.
Hybrid Routing: Both active and on-demand routing
Multi-Layer UAV: These networks consist of
is used by the hybrid routing method. It employs rules to
heterogeneous UAVs form. The communication is held in
distinguish between domains' reactive and inactive routing
it by the lower layer between the backbone of the UAV. By
and separates networks into domains.
the study of Larrieu, N. et al. [23], these UAVs periodically
DSR protocol: It is a reactive routing protocol that share information on the zone structure to survey with their
extracts information on topology from the source router. In respective neighbors.
the setup node's network interface mode, it is possible to
Multi-Group UAV: Both a centralized network and an
obtain more topology which is kept in a cache to increase
ad hoc network work together to integrate a multi-Group
the lookup's hit rate of the route.
UAV network. Ad hoc backbone connections link several
Route Maintenance: It is used to make the routing UAVs to the station in a centralized way.
table exist in source nodes. Throughout the transmission
Surveillance: The category that includes applications
procedure data, the router which is developed must be
that needed UAVs as flying cameras is surveillance and
continually upheld.
monitoring. According to the research by Bujari, A. et al.
Route Discovery: It determines if the node has a target [14], the jobs often include capturing real-time photos,
and data needed to initialize the info. In order to set up the videos, or audio from flying objects so that one may go
link, the original node is sent a response. When there isn’t forward and gather sensitive information.
any target then that checks the routing information to the
Environmental sensing: In the category of
destination node is cached.
environmental sensing, UAVs function as sensors that
Decision Making: In routing, it is limited in algorithms gather environmental data in a designated region.
of hard decision-making. It allows recursion in the sense. Applications may require sensor data in this situation to
Each block is in charge of managing the information flow evaluate certain environmental conditions.
inside it using a common gating mechanism. By the study
Randomized mobility: The fundamental models for
of Buntine, W. et al. [12], the recurrent units don't care
network research are those that involve randomized
about the blocks' internal architecture.
movement. They represent several mobile nodes, each of
Security in MANETs and FANETs:- whose activities is completely independent of the others
and prior deeds.
Ad hoc networks allow for the use of attacks like "Man
in the Middle," which allow attackers to intercept data Time-dependent mobility: Models for time- and
transmission and derive useful information. By the study of space-dependent mobility. This mobility model's category
Aliouat, M. et al. [21], maintaining the anonymity of aims to prevent abrupt changes in speed and direction.
communicating nodes is crucial to ensuring the stability of Different mathematical formulae can be used to carry out
a safe routing protocol. By the study by Khan, I. U. et al. the motion change smoothly.
[24], FANETs are now a reality because of important
Path-planned mobility: Types of mobility with a plan.
advancements made in the field of wireless communication
These mobility models offer a predetermined route pattern.
technology.
UAVs randomly switch to another pattern or repeat the
For routing apps, new traffic patterns are created that same one as if in a loop after they reach the end of a
are referred to as cut-through traffic. By Hurley-Smith, D. predetermined pattern.
et al. [19], To keep hostile nodes from entering the network,
access control in security is necessary. It is not practical to VI. COMPARATIVE ANALYSIS
establish a single point of command in the MANET due to Table 2 given below displays a comparison study,
considering the network’s dynamic nature according to the demonstrating that our suggested work discusses a number
study by Kumar, S. et al. [22]. of elements while others just examine a small number.
 TABLE 2. A comparative analysis of existing research papers

Factors Ignacio Muhammad Zunnurain Hua Yang and Muhammad Syeda M. This Paper
Cases et Hussain and Zurina Zhiyong Liu [6] Asghar Khan Muzammal
al. [1] Mohd Hanapi [5] et al. [13] et al. [18]
Algorithms 0 0 1 1 0 1
discussion
Network attacks 0 1 0 0 1 1

UAV protocols 0 1 0 1 0 1

LPN application 0 1 0 0 0 1

Network 1 1 0 0 0 1
architecture
Security in 0 0 0 0 0 1
MANETs and
FANETs
Total Score 1 4 1 2 1 6

VII. THE MAJOR ATTACKS ON NETWORK Sinkhole attack: When a changed DIO message is
broadcast, a sinkhole attack could happen. Advertisement
The network layer handles safe data transfer between
of a fake route causes nearby nodes to go in the direction
the perception and application layers as well as the
of the bad node.
processing of detected data. The study of Verma, A. et al.
[20], utilizes transformed data from the network layer and VIII. CONCLUSION
provides it according to the requirements of the user.
This paper discussed the characteristics and algorithms
There are so many types of routing attacks. By the study of various forms of routing networks. The investigation
of Muzammal, S. M. et al. [18], RPL attacks cause by the discovered that the dynamic routing protocol makes better
properties and functionalities of the RPL protocol. WSNs decisions regarding the transfer of packets. FANET
have an advanced property called IoT which inherits network security and significant network attacks are
routing attacks from the operation of WSN. discussed. FANET supports establishing connections to
these routing devices while maintaining network
DDoS attack: Several sites simultaneously launch a performance and connectivity.
distributed denial of service (DDoS) attack known as
DDoS. In contrast to DoS attacks, need several According to some researchers, mechanisms for
compromised nodes to bombard. The device sends a server network security have been proposed, and solutions are in
or network component to just one target using messages or the works. The research is for educational purposes and
requests for connection with the goal to slow it down or lacks adequate deployment or testing.
possibly crash it. In the future, further research on the routing networks
Packet mistreating attack: The router starts can be done on the neural network routing algorithms that
processing the packet improperly after being injected with can apply artificial intelligence to optimize routing better
malicious information. in FANETs and also need to enhance the security for
routers to protect from malicious attacks as well.
Routing table poisoning speaks of an unauthorized or
malicious alteration to the router's routing table. By REFERENCES
inputting incorrect table entries, the network might suffer [1] Cases, I., Rosenbaum, C., Riemer, M., Geiger, A., Klinger, T.,
significant damage as a result of this assault. Tamkin, A., ... & Kart-tunen, L. (2019, June). Recursive routing
networks: Learning to compose modules for language
Hit-and-run attack: To determine if the Internet is up understanding. In Proceedings of the 2019 Conference of the North
American Chapter of the Association for Computational Linguistics:
and running. during this attack, the hacker injects malicious Human Language Technologies, Volume 1 (Long and Short Papers)
packets into the router. If so, the hacker transmits further (pp. 3631-3648).
malicious packets in an effort to damage the router. [2] Cai, S., Shu, Y., & Wang, W. (2021). Dynamic routing networks. In
Proceedings of the IEEE/CVF Winter Conference on Applications
Persistent attacks: The attack is quite damaging and of Computer Vision (pp. 3588-3597).
has a very severe character. The router may stop working [3] Rosenbaum, C., Cases, I., Riemer, M., & Klinger, T. (2019).
if malicious packets are continuously injected. Routing networks and the challenges of modular and compositional
computation. arXiv preprint arXiv:1904.12774.
Sybil attack: It is carried out when a false node [4] Wang, S., Hu, L., Wang, Y., Sheng, Q. Z., Orgun, M., & Cao, L.
assumes many identities belonging to real nodes. The Sybil (2019, January). Modeling multi-purpose sessions for next-item
attack version known as the "clone ID attack" involves the recommendations via mixture-channel purpose routing networks. In
International Joint Conference on Artificial Intelligence.
attacking node copying the identity of a legitimate node. International Joint Conferences on Artificial Intelligence.
[5] Hussain, M. Z., & Hanapi, Z. M. (2023). Efficient Secure Routing [25] Zhang, C., Patras, P., & Haddadi, H. (2019). Deep learning in mobile
Mechanisms for the Low-Powered IoT Network: A Literature and wireless networking: A survey. IEEE Communications surveys
Review. Electronics, 12(3), 482. & tutorials, 21(3), 2224-2287.
[6] Yang, H., & Liu, Z. (2019). An optimization routing protocol for [26] Gill, S. S., Garraghan, P., & Buyya, R. (2019). ROUTER: Fog
FANETs. EURASIP Journal on Wireless Communications and enabled cloud based intelligent resource management approach for
Networking, 2019(1), 1-8. smart home IoT devices. Journal of Systems and Software, 154, 125-
[7] Seyfollahi, A., & Ghaffari, A. (2021). A review of intrusion 138.
detection systems in RPL routing protocol based on machine [27] Kiran, V., Sardana, A., & Kaur, P. (2022, April). Performance
learning for internet of things applications. Wireless Analysis of Various Routing Algorithms on Different Wimax
Communications and Mobile Computing, 2021, 1-32. Networks. In 2022 2nd International Conference on Advance
[8] Rosenbaum, C., Klinger, T., & Riemer, M. (2017). Routing Computing and Innovative Technologies in Engineering (ICACITE)
networks: Adaptive selection of non-linear functions for multi-task (pp. 1661-1664). IEEE.
learning. arXiv preprint arXiv:1711.01239. [28] Datta, P., Tanwar, S., Panda, S. N., & Rana, A. (2020, June).
[9] Mammeri, Z. (2019). Reinforcement learning based routing in Security and issues of M-Banking: A technical report. In 2020 8th
networks: Review and classification of approaches. Ieee Access, 7, International Conference on Reliability, Info-com Technologies and
55916-55950. Optimization (Trends and Future Directions)(ICRITO) (pp. 1115-
1118). IEEE.
[10] Smyth, D., Scott-Hayward, S., Cionca, V., McSweeney, S., &
O’Shea, D. (2023). SECAP Switch—Defeating Topology Poisoning
Attacks Using P4 Data Planes. Journal of Network and Systems
Management, 31(1), 28.
[11] Dalal, S., Manoharan, P., Lilhore, U. K., Seth, B., Simaiya, S.,
Hamdi, M., & Raahemi-far, K. (2023). Extremely boosted neural
network for more accurate multi-stage Cyber attack prediction in
cloud computing environment. Journal of Cloud Computing, 12(1),
1-22.
[12] Zaremoodi, P., Buntine, W., & Haffari, G. (2018, July). Adaptive
knowledge sharing in multi-task learning: Improving low-resource
neural machine translation. In Proceedings of the 56th Annual
Meeting of the Association for Computational Linguistics (Volume
2: Short Papers) (pp. 656-661).
[13] Khan, M. A., Safi, A., Qureshi, I. M., & Khan, I. U. (2017,
November). Flying ad-hoc networks (FANETs): A review of
communication architectures, and routing protocols. In 2017 First
international conference on latest trends in electrical engineering
and computing technologies (INTELLECT) (pp. 1-9). IEEE.
[14] Bujari, A., Palazzi, C. E., & Ronzani, D. (2017, June). FANET
application scenarios and mobility models. In Proceedings of the 3rd
Workshop on Micro Aerial Vehicle Networks, Systems, and
Applications (pp. 43-46).
[15] Hu, R., Andreas, J., Rohrbach, M., Darrell, T., & Saenko, K. (2017).
Learning to reason: End-to-end module networks for visual question
answering. In Proceedings of the IEEE international conference on
computer vision (pp. 804-813).
[16] Cabannes, T., Sangiovanni, M., Keimer, A., & Bayen, A. M. (2019).
Regrets in routing networks: Measuring the impact of routing apps
in traffic. ACM Transactions on Spatial Algorithms and Systems
(TSAS), 5(2), 1-19.
[17] Meng, D., Xiao, Y., Guo, Z., Jolfaei, A., Qin, L., Lu, X., & Xiang,
Q. (2021). A data-driven intelligent planning model for UAVs
routing networks in mobile Internet of Things. Computer
Communications, 179, 231-241.
[18] Muzammal, S. M., Murugesan, R. K., & Jhanjhi, N. Z. (2020). A
comprehensive review on secure routing in internet of things:
Mitigation methods and trust-based approaches. IEEE Internet of
Things Journal, 8(6), 4186-4210.
[19] Hurley-Smith, D., Wetherall, J., & Adekunle, A. (2017).
SUPERMAN: security using pre-existing routing for mobile ad hoc
networks. IEEE Transactions on Mobile Computing, 16(10), 2927-
2940.
[20] Verma, A., & Ranga, V. (2020). Security of RPL based 6LoWPAN
Networks in the Internet of Things: A Review. IEEE Sensors
Journal, 20(11), 5666-5690.
[21] Alouache, L., Nguyen, N., Aliouat, M., & Chelouah, R. (2019).
Survey on IoV routing protocols: Security and network architecture.
International Journal of Communication Systems, 32(2), e3849.
[22] Kumar, S., Goyal, M., Goyal, D., & Poonia, R. C. (2017, December).
Routing protocols and security issues in MANET. In 2017
international conference on infocom technologies and unmanned
systems (trends and future directions)(ICTUS) (pp. 818-824). IEEE.
[23] Maxa, J. A., Mahmoud, M. S. B., & Larrieu, N. (2017). Survey on
UAANET routing protocols and network security challenges. Ad
Hoc & Sensor Wireless Networks.
[24] Khan, I. U., Qureshi, I. M., Aziz, M. A., Cheema, T. A., & Shah, S.
B. H. (2020). Smart IoT control-based nature inspired energy
efficient routing protocol for flying ad hoc network (FANET). IEEE
Access, 8, 56371-56378.