1

Design Guidelines for Blockchain-Assisted

5G-UAV Networks
Moayad Aloqaily, Member, IEEE, Ouns Bouachir, Member, IEEE, Azzedine Boukerche, Fellow, IEEE,
Ismaeel Al Ridhawi, Senior Member, IEEE

F
arXiv:2007.15286v1 [cs.NI] 30 Jul 2020

Abstract—Fifth Generation (5G) wireless networks are designed to of achieving the intended tasks with the aid of fog and
meet various end-user Quality of Service (QoS) requirements through cloud computing [1]. Most of the time-sensitive critical
high data rates (typically of Gbps order) and low latencies. Coupled decisions can now be made at the fog or the resource-rich
with Fog and Mobile Edge Computing (MEC), 5G can achieve high edge without full reliance on the cloud. Such an approach
data rates, enabling complex autonomous smart city services such as
introduced the concept of fog-to-cloud (F2C) and fog-to-fog
the large deployment of self-driving vehicles and large-scale Artificial
Intelligence (AI)-enabled industrial manufacturing. However, to meet the
(F2F) computing [2].
exponentially growing number of connected IoT devices and irregular However, in today’s smart city environments, and with
data and service requests in both low and highly dense locations, the the ever-growing numbers in IoT devices, especially in UDN
process of enacting traditional cells supported through fixed and costly environments, 5G networks will face limitations due to
base stations requires rethought to enable on-demand mobile access unpredictable and non-uniform loading [3]. As such, if the
points in the form of Unmanned Aerial Vehicles (UAV) for diversified supplied network capacity does not match the throughput
smart city scenarios. This article envisions a 5G network environment demand, end-user QoS requirements will not be met. To
that is supported by blockchain-enabled UAVs to meet dynamic user solve such an issue in regards to the disproportion between
demands with network access supply. The solution enables decentral-
the irregular demand and the limitations in access availabil-
ized service delivery (Drones as a Service) and routing to and from
end-users in a reliable and secure manner. Both public and private
ity, UDN could be assisted through mobile serving stations,
blockchains are deployed within the UAVs, supported by fog and cloud namely, unmanned aerial vehicles (UAVs), or simply drones.
computing devices and datacenters to provide wide range of complex UAV mobile base stations (BSs) will improve the QoS of
authenticated service and data availability. Particular attention is paid to the wireless network by increasing network capacity and
comparing data delivery success rates and message exchange in the improving the coverage area [4][5]. Such mobile BSs can
proposed solution against traditional UAV-supported cellular networks. be moved to highly dense locations close to the end-users.
Challenges and future research are also discussed with highlights on Drones also have the potential of replacing or complement-
emerging technologies such as Federated Learning.
ing, to say the least, cellular networks in high interference
conditions or locations where it is not economically feasible
Index Terms—Blockchain, UAV, Drone, cyberphysical systems, dis-
to install a permanent infrastructure. For instance, in cases
tributed ledger, Federated Learning, 5G.
where there are few active end-users, deploying UAVs may
sufficiently serve the users’ needs without reliance on high-
1 I NTRODUCTION end cellular BSs. Moreover, running a few drones in those
locations might reduce the overall power consumption (i.e.
The rapid growth in communication traffic between Internet by switching on and off the drones according to the number
of Things (IoT) devices has revolutionized the current com- of active requests compared to an all-time active fixed
munication framework, leading to smaller cell architectures cellular BS).
and ultra-dense networks (UDNs). The fifth generation (5G) Although mitigating the disproportion between irregu-
network was designed to meet stringent user quality of lar end-user requests and the available 5G access through
service (QoS) requirements in terms of higher data rate an increase in the number of fixed and mobile stations, such
and low latency. Many smart city applications have been over-provisioning leads to more complex service provider
deployed as a result of this ever-evolving communication management issues. To this end, we investigate the feasi-
technology. For instance, smart intelligent autonomous self- bility and significance of a novel blockchain-assisted 5G-
driving and service provisioning vehicles are now capable UAV network over other state-of-the-art UAV-supported
5G networks. The solution relies on a decentralized ser-
• M. Aloqaily is with Al Ain University, UAE. E-mail: maloqaily@ieee.org vice delivery and routing technique to and from end-users
• O. Bouachir is with College of Technological Innovation, Zayed Univer- through a 5G-supportive UAV network. Both public and
sity, UAE. E-mail: ouns.bouachir@zu.ac.ae
• A. Boukerche is with University of Ottawa, Ottawa, ON, Canada. E-mail: private blockchains are deployed in a smart city to ensure
boukerch@uottawa.ca secure communication between participants. Mobile UAVs
• I. Al Ridhawi is with Kuwait College of Science and Technology, Kuwait. are deployed in ultra-dense areas suffering from network
E-mail: i.alridhawi@kcst.edu.kw
congestion and decreased QoS to provide efficient and reli-
 2

Cloud
Data processing, analysis and storage

5G

5G

5G

5G

Fig. 1. An Overview of the Blockchain-Assisted 5g-Drone Network Scenario

able data routing. UAVs communicate with both fog/edge Coupled with the most advanced paradigms of com-
devices and cloud datacenters to ensure extensive coverage munication (i.e. 5G), data storage (using edge, fog, cloud,
of a multitude of smart city services and access to re- blockchain technologies) and processing (e.g. ML and AI),
sources. In addition, we identify challenges associated with DaaS is able to provide a big range of smart services
the blockchain-assisted 5G-UAV network such as energy (figure 1). Indeed, drones are able to perform any task, on-
consumption and reliability. We demonstrate through simu- demand, using the different communication infrastructures
lation results that the proposed solution can be beneficial, in deployed around the city. The various types of collected
terms of data delivery success and minimal data exchange. data, the users’ requests, or the transactions among the
service providers can be processed and stored by edge
2 D RONES AS A S ERVICE (DAA S) or cloud computing, allowing the extraction of valuable
information that can be very useful in enhancing quality
2.1 Operations
of experience, safety and life in the smart city.
With the remarkable fast advances in AI and machine
learning (ML), extraordinary capabilities are designed em- 2.1.1 Services
powering the concept of smart autonomous systems such as
Thanks to their 3D movement flexibility, UAVs do not have
robots and autonomous vehicles. UAVs, or simply, drones,
complex traversing restrictions as seen in roads. They can
are another example of systems that may benefit from these
move in various directions and fly over inaccessible and
advances. Used for the first time in the second world war by
dangerous locations. Various applications can benefit from
the American army, in an attempt to use a manned aircraft in
this feature. Those applications can be grouped into three
an unmanned manner, the name “drones” have always been
different classes based on the task allocated to the aircraft:
coupled with military. Nowadays, UAVs are available with
diverse sizes and capabilities and used in civil domains for • Flying IoT device. Equipped with diverse types of
various’ applications. They perform at different automation sensors (e.g. gas and temperature sensors), a drone
levels, ranging from fully controlled (no automation), to can reach inaccessible and dull locations and get
fully autonomous that can be reflected by a swarm of drones better visibility than any other IoT device to collect
co-operating together to fulfill a common mission. accurate data about a situation. For instance, drones
As shown in Figure 1, UAVs can be active actors in can fly around big buildings, mountains, bridges,
all types of smart city services and applications. They can hazardous areas (e.g. after natural disasters), etc. An
be used with police and fire stations for emergency situa- Example of sensed information the weather forecast
tions, as well as in smart transport management, healthcare, information, which can be forwarded, in real-time,
agriculture, industry, smart grid, environment control, and to the closest edge device, to the cloud, or stored
much more. in the drone’s memory and collected by the ground
 3

station at the end of the exploratory trip. UAVs are uously exchanging data between the drones, they can create
considered a cost-effective flying IoT devices used an idea about the situation and update their tasks based
to collect truthful information for various applica- on that. For instance, each drone may change its path plan
tions. Surveillance, monitoring, environment mea- based on other aircraft trajectories or task results.
surements are popular examples of IoT applications-
based drones. 2.2 UAV Networks Challenges
• Mobile flying base station Due to the expansion
in the number of connected devices, especially in As explained above, all UAV operations are based on a
crowded areas, communication networks are facing continuous exchange of data between the drone and various
many challenges to meet all these simultaneous re- surrounding devices such as other drones, base stations,
quests. Many researchers have focused on this issue vehicles, infrastructure, etc. A drone exchanges diverse
and have proposed to add mobile small cells (MSC) types of messages: information related to its task (e.g. ge-
that can support the connection provided by the ographic location or pilot instructions), gathered IoT data,
mega-cell. [6] Flying few meters above the users, and much more [10]. These messages can be generated by
UAVs offer an important coverage area that can be the drone, sent to the drones or simply relayed to other
provided by a flying mobile small cell. Also, UAVs devices through the drone. The communication system used
can be used as base stations or relays to connect by the drones faces several challenges:
devices in areas where no communication infrastruc- • Fast delivery: To guarantee the best task perfor-
ture is used (p2p networks, rural areas and after mance, drones should exchange information about
hazardous events). surrounding events (or pilot instructions) on time,
• Flying robots Advances in technology are cre- with the shortest delay, to make instant reactions.
ating intelligent systems based on unprecedented Also, some of the forwarded IoT data may be critical
paradigms. AI and ML are creating a revolution and require low transmission delay.
in all technological sectors by providing cutting- • Trust and privacy: drones may perform as a swarm
edge devices such as robots that are able to collect or individually and they may relay information ex-
information about their surrounding environment, changed between other devices. The privacy of the
analyze the current situation, take decisions and per- exchanged messages, the participating aircraft iden-
form various tasks in real-time. UAVs are considered tity verification and authentication are critical issues
flying robots that, based on innovative algorithms, that should be taken into consideration.
can replace humans in many dull missions [7]. Exam- • Security: Drones, operating in the sky, communicate
ples are the amazon delivery and flying ambulance thorough wireless communication technologies, and
projects. may be vulnerable to several privacy and security
risks. Drones can be easily controlled by hackers to
2.1.2 UAV Automation Levels manipulate their tasks, to send wrong data and to
Drones can perform various operations and tasks under target data accountability, data integrity, data autho-
different levels of automation that can go from fully con- rization, and reliability.
trolled drones to fully autonomous. The level of automation • QoS: drones exchange various types of data such as
is defined by how the drone can monitor the progress of its video surveillance (real-time traffic) and emergency
mission. events that require immediate reactions [8]. These
• Zero automation level: A pilot can control the messages have different requirements in term of QoS
drone’s operations from the ground by sending in- such as throughput, reliability and delay, that may
structions for each movement and action. In this change with time.
situation, the aircraft should obey all the received • Energy consumption: Most of the drones are
directions passively. equipped with limited batteries that have short life-
• Hybrid automation level: the UAV’s automation time. The communication exchanges, the computing
level can increase progressively, reducing the hu- processes running on the UAV and the carried pay-
man interposition during the mission. Some deci- load increase the energy consumption and reduce the
sions may be taken locally by the drones, based on lifetime of the UAV.
surrounding events, while some others are received
from the ground controller.
• Full automation level: based on advances in AI and 3 B LOCK C HAIN FOR D RONES
ML algorithms, drones have full control of their op- 3.1 BlockChain Overview
erations including trajectory planning, sudden move-
Generally speaking, blockchain (BC) is a distributed and
ment and data exchanging with the ground station.
tamper resistant ledger that does not rely on a centralized
To make fast and wise decisions, drones should rely
authority to establish trust, with a core layer mechanism
on their sensors and the data received from the sur-
for decentralized trust management. Since the introduction
rounding (e.g. other drones, vehicles, infrastructure,
of Bitcoin, blockchain is used in applications such as cryp-
etc.).
tocurrency, secure storage and asset transfer.
Drones having a certain automation level, and may The simplest form of blockchain is a linked list of
perform as a swarm of cooperative agents [8][9]. By contin- “blocks” that contains various types of information. For
 4

TABLE 1
Comparison between different drone communication platforms. ↑ means comparatively high, ↓ means comparatively low

Characteristics Basic UAV Networks 5G UAV Networks BC Assisted 5G UAV Networks

Communication Limited range that depends on the Extended range based on 5G Extended range based on 5G
range used technology (i.e. Zigbee or infrastructure infrastructure
Wifi)
QoS Fair data delivery speed with Very fast data delivery with high Very fast data delivery with high
limited throughput specially with throughput and reliability level throughput and reliability level
Zigbee that cannot support video
messages
Identity No UAV identity verification No UAV identity verification UAV identity is checked so not any
verification drone can participate in the mission
Data security ↓ ↓ Data is stored in BC that provides
security
Privacy ↓ ↓ Data is stored in BC that provides
privacy
Trust ↓ ↓ Trust is provided since all
participant’s identities are verified,
their privacy is guaranteed and
data storage is secure
Decentralization Centralized system Centralized System Decentralized System
Resource Basic, depends on the used 5G resource management Advanced, managing 5G resources
management technology using BC
Power ↓ ↓ Thanks to the resource
efficiency management features, this system
provide UAVs power efficiency
Scalability ↓↓ ↓ Better scalability
Intelligence
UAV Restricted, within the ↑ ↑
movement communication range
flexibility
Autonomy Limited Limited Very advanced
Swarm Limited Limited Advanced
operations

example, transactions that record Bitcoin money transfers. blockchains are usually an order of magnitude faster and
Each participant stores a local blockchain, and uses some transaction costs are lower due to the reduced number
form of a “consensus” mechanism to establish its exact order of processing nodes. The drawback of blockchain is that
and content. Popular mechanisms include proof-of-work identity/permission management is not as scalable and
(PoW), proof-of-stake (PoS) and practical Byzantine-fault- decentralized.
tolerance (PBFT), each of which has different trade-off and There are also hybrid solutions that combine private and
suitable usage scenarios [11][12]. Blockchain technology has public blockchains, referred to as “consortium blockchains”.
been envisioned because of its robustness in providing trust
and anonymity in any commodity trading such as energy
3.2 BlockChain for Drones and 5G
[13].
3.2.1 Drones and 5G
Each block is “chained” to the previous one using a
cryptographical data structure called a hash pointer, and With the deployment of 5G, many services can benefit from
participants jointly verify all transactions by examining the the advantages of this technology in terms of data rate and
content and the accompanying hash pointers. Thus, false energy consumption. 5G and UAVs have a mutual relation-
transactions are rejected if enough participants are cor- ship. Each technology can provide a range of advantages to
rect. Such mechanisms ensure the system is tamper-proof, the other that may have a significant impact on the quality
since an adversary cannot persuade correct participants to of various smart city services:
switch to an incorrect branch of the blockchain. This is also • 5G in service of UAVs: As mentioned previously,
how blockchain intuitively establishes decentralized trust the biggest challenges for UAV networks are fast
[14][15]. data delivery, QoS and energy consumption. 5G is
The original intention of blockchain was to create a designed to provide these features: highest data rate,
public network similar to the models of Bitcoin, Ethereum better QoS and it was proved that it is an energy-
and Litecoin. But due to performance limitations private aware technology. This makes 5G a promising solu-
blockchains have also been developed. The difference be- tion to make UAVs communicate together and with
tween public and private blockchains depends on the per- the different smart city’s components.
missions of the participants. Anyone is free to join a public • UAVs in service of 5G: cellular networks suffer
blockchain, whereas private versions require an author- from their signal quality degradation due to several
ity to approve membership. Therefore, public types are reasons such as interference, vast number of con-
known as permission-less, and private types as permis- nected devices, big buildings and their thick metallic
sioned. The benefit of identity management is that private infrastructure [6]. Also, some areas may be outside
 5

The use of blockchain-assisted drones in this architecture

Blockchain Cloud
Data Storage improves data security across several layers. Blockchain
Public BC provides a level of security that limits external hacking and
Data Machine Behavior and Decision
Analysis Learning Data Extraction Making
data retrieval when a drone physically falls into the wrong
Reward hands. Moreover, it validates and ensures the identities of
participating UAVs performing a task. Thus, this architec-
Smart ​Fog/edge Consensus
Contract Data Mining ture provides trust among drones and service beneficiaries.
Process
Two types of blockchains can be seen from the architec-
Data Filtering Data ture and are used according to different scenarios (private or
Private BC
Encryption
public) where service providers use public or private keys.
Any Service Data Public keys are used when the service is visible to all parties
while private keys belong to service providers for specific
5G Drones as a Service
services. Blockchain authenticates communications among
UAVs by tracking all their transactions, makes it available
Smart City Components to all network nodes, and achieves data integrity (i.e. cryp-
tography) to provide tamper-resistance. The blockchain is
decentralized with no centralized authority to access or
Fig. 2. System Architecture. modify transactions. All network participants must reach
a consensus to validate transactions in a secure way.
The proposed architecture is secure by the principles of
of the cellular coverage like rural areas or after haz- blockchain, where no centralized authentication mechanism
ardous situations. The fast growing IoT traffic is very such as the public key is used by the UAVs for identifying
challenging for static base stations in 5G networks. each other. For example, when providers mutually agree
Thanks to their flexible mobility and low-cost, UAVs to providing services for a specific area, they also share
can serve as MSC that enhances the cellular network the pool of secret keys which are the private keys for the
coverage and provide a support for 5G infrastructure drones used to exchange load information while executing
empowering new 5G applications. a drone’s service task. As such, drones and service providers
need to execute drone smart contracts. Blockchain can be used
3.3 Concept to enhance the performance of the UAV-5G communication
network by proposing a spectrum trading platform between
Nowadays, communication is no longer traditional. It’s not the aerial and terrestrial systems allowing a secure spectrum
only that we have mobile phones communicating, but they sharing between the network operator’s and UAV service
also connected to the cloud, edge, and fog, as shown in providers. This feature provides privacy based on a dis-
Figure 1. Such a heterogeneous environment is sensitive to tributed sharing information platform and limits the risk
many issues; trust is the most important among all. On- of unauthorized spectrum utilization by malicious nodes.
demand trust in ultra-dense wireless networks is not an easy
task. It requires high speed communication, identification,
and authentication. Drones are capable of communicating 3.4 Advantages
with each other through trust, to relay information, regard- The proposed blockchain-assisted 5G-UAV network pro-
less of the owner (e.g. service provider). ML and AI are vides several advantages and improves the QoE of drone
two techniques that can be used to identify intrusive traffic. services. These benefits are summarized as follows:
Moreover, blockchain can be used to certify the true identity
of drones. • Better network QoS in terms of delay, throughput
This section presents an architecture that secures the and reliability that are essential requirements in UAV
utilization of drones as on-demand flying nodes for inter- service’s to provide the best QoE.
service operability between multiple service providers by • Data privacy and security based on BC distributed
employing the features of machine learning and blockchain. ledger.
• Identity of participating drones is verified for each
Figure 2 presents various components and strategies new transaction, providing trust and anonymity that
for utilizing the use of 5G and blockchain-based drones. protects from malicious attacks trying to join the
Further, Table 2.2 providers an overview of the character- UAV network.
istics and implications of the blockchain-based drones and • Distributed platform-based immutability, trans-
compares the deployment issues with the centralized and parency and traceability that increases trustworthi-
traditional drone-assisted wireless networks. ness between agents and enables the deployment of
Drones are facilitating smart city sub-component ser- high automation systems.
vices through the 5G infrastructure. Data processing and • Better resource and spectrum management that al-
storage is usually handled at the edge or the cloud. lows to control the interference in UAV-5g-based
Blockchain can be used to integrate drones with fog and networks which provide better scalability, accessibil-
cloud site by insuring data integrity and secure communi- ity and power efficiency. This advantage allows to
cation. This in return makes the system more reliable and create a trusted scheme for interference monitoring
trustworthy. in flying small cells.
 6

4 P ERFORMANCE E VALUATION : A S CENARIO the environment, the gap in success rate between the pro-
posed and traditional methods increase. For instance, when
To test the effectiveness of the proposed model, analysis has
comparing N2N-UAV with BC to N2N-BS with 100 mobile
been performed through Network Simulator 3 (NS-3). To
nodes in the network, we see that the success rate for N2N-
maintain uniformity, all experiments have been conducted
UAV with BC is almost 77% compared to 55% for N2N-BS.
on i7-6500U CPU @ 2.5 GHz with 8 GB of RAM. Simulator
Moreover, the success rate for node-to-node communication
settings are summarized in Table II. In the simulation sce-
without reliance on blockchain is 60%. Such results indicate
nario, a test is conducted where up to 100 mobile devices
that with high mobility cases, like in UDNs, UAVs are
are scattered along a network area of 1500m x 1500m, 20
capable of adjusting their location to meet service demands
mobile drones fly over them to form a relay and data de-
in accordance with mobile node density. For instance, when-
livery platform, and one next generation node B (gNB) base
ever node density increases in a certain location within the
station is placed to cover the entire simulation environment.
network, one or more extra UAVs move to that location to
Communication between UAVs is conducted using a flying
ensure continuous and reliable data delivery. The proposed
ad-hoc network topology (FANET). The UAVs are randomly
technique not only delivers data in a timely manner, but also
redistributed to mobile node dense locations throughout the
assures that data communication is authenticated to ensure
simulation environment at predefined time intervals. Such
high delivery success rates.
a technique would represent the movement of UAVs to
locations of high-dense mobile nodes to aid in the process
of all-time connectivity and faster data delivery. Mobile 100

Packet delivery success rate (%)

nodes and UAVs are equipped with two communication 95
interfaces, allowing them to have node-to-drone communi- 90
cation (N2D), node-to-BS communication (N2B), and drone- 85
to-drone communication (D2D). Three solutions were com-
80
pared, namely, i) traditional node-to-node communication
using the BS (N2N-BS), ii) node-to-node communication 75
using both UAVs and BS without blockchain (N2N-UAV 70
without BC), and iii) node-to-node communication using 65
both UAVs and BS with blockchain (N2N-UAV with BC). 60 N2N-BS
The tests conducted focused mainly on data delivery success N2N-UAV without BC
rate and message exchange overhead.
55 N2N-UAV with BC
50
10 20 30 40 50 60 70 80 90 100
TABLE 2
Simulator Settings Number of Mobile Nodes

Simulation Parame- Numerical Values Fig. 3. Comparing the overall packet delivery success rate among the
ters three different techniques.
IEEE 802.11n (for N2D and D2D)
Communication pro- LTE gNB with 1 Gbps bandwidth
tocol -(for N2B and D2B)
4.2 Message Overhead
Area 1500 m X 1500 m
Number of fixed BSs 1 It is well-known that with any short-range communication
Number of UAVs 20 technique, the number of packets sent from the source to
Number of end-users 10 – 20 mobile nodes the destination increase in accordance to the number of
UAVs placement Random – uniform distribution (initially)
Mobility model Random – waypoint (for mobile nodes) relay nodes introduced. Moreover, by using blockchain the
Drone flying heights 50 m number of packets increase more to ensure that commu-
Simulation Duration 60 s nication between UAVs is authenticated. Thus, there is al-
CBR packet length 512 bytes ways a trade-off between an increase in the communication
CBR packet interval 10 ms
0.1 W (for UAVs)
overhead and system efficiency in terms of reduced delay,
Transmission power power usage reduction and much more. For example, an
0.01 W (for mobile nodes)
increase in the network traffic as a result of taking a differ-
ent data delivery route might speed up the data delivery
process if the BS or another primary UAV are busy with
4.1 Authenticated Data Delivery Success Rate handling other data delivery requests. We see from Figure
Authenticated data delivery success rate is the percentage 4 that when comparing a UAV communication method
of positive responses received for all packets sent from the (N2N-UAV) to a traditional cellular communication method
source mobile node to the destination mobile node. Results (N2N-BS), the N2N-BS technique outperforms N2N-UAV
depicted in Figure 3 show that the proposed solution, in terms of reduced network traffic overhead, where the
namely, N2N-UAV with BC outperforms traditional com- overall number of sent messages is 1200 with 100 mobile
munication techniques and non-authenticated N2N-UAV nodes network density. When relying on a UAV solution,
state-of-the-art solutions. It is important to note that without the overall number of exchanged messages is 2150 for N2N-
proper data authentication (with the aid of blockchain), UAV without BC and 2650 for N2N-UAV with BC. Although
sensitive data may be intercepted by hostile UAVs. From results show that a traditional N2N-UAV technique outper-
the figure, we see that as the node density increases within forms the our proposed solution that relies on blockchain,
 7

the difference in message exchange is modest compared to UAVs or instructions received from the pilot), IoT
the gains achieved in terms of authentication and privacy. messages generated by the UAV and other messages
generated from other devices and relayed by the
3,000 flying drone.
N2N-BS • System Infrastructure: Deployment of the 5G infras-
N2N-UAV without BC tructure is under process in many countries and may
2,500
Number of Messages

N2N-UAV with BC take years before having a full 5G coverage.

2,000 • Regulations and standards: Regulation is a big chal-
lenge for both blockchain and UAVs. New rules
1,500 should be defined to regulate the involvement of
drones in cities for safety and privacy reasons and to
1,000 decide entities and service providers that may have
access to the data in the blockchain. Part of these
500 rules should regulate the relationship between the
various service providers.
• Novel UAV services:the proposed system allows to
10 20 30 40 50 60 70 80 90 100 achieve high automation and intelligence level that
Number of Mobile Nodes is the key to designing novel and advanced UAV
services.
Fig. 4. Total number of packets exchanged in the network for data • Innovative consensus algorithms: advanced consen-
delivery between source and destination nodes using the three different sus algorithms should be designed, considering the
data delivery techniques various types of services provided by the aircraft.
• Off-chain blockchain storage:As mentioned above,
UAVs exchange various types of data. Some of the
5 C HALLENGES AND F UTURE R ESEARCH data may be too large to be stored in the blockchain
efficiently or requires frequent modification or dele-
BC- assisted 5G UAV network empowers drone services
tion. An off-chain blockchain storage should be pro-
by providing security, privacy, trust and better resource
vided to solve this issue and enhance the system
management. However, the deployment of this system may
performance.
face a few challenges:
• Dense number of nodes: various challenges
• Scalability and QoS: Blockchain faces performance such as location identification, channel selection,
challenges in terms of scalability, throughput and interference-management and the availability of the
delay due to the rapidly growing volume of data line of sight.
exchanged by such systems. Indeed, data mining • Location and Divisibility: One of the primary chal-
requires several verification’s performed by each lenges for the proposed model is the mutual agree-
blockchain participant. When the data generation ment on the location and divisibility of services be-
rate is faster than the mining process, the perfor- tween the service providers.
mance of the overall system is affected in terms of
throughput and delay. Scalability is one of the critical
challenges causing QoS degradation. 6 C ONCLUSION
• Federated Learning UAVs: 5G provides a unique This article presented research guidelines for a 5G-UAV
communication infrastructure to run sophisticated network that provides fast, reliable and secure service de-
smart applications of autonomous systems. Flying livery to end-users of smart cities, namely, Drones as a
close to the end-users, drones can be utilized as relay Service (DaaS). With the aid of blockchain, UAVs can act as
devices to forward messages, and to support edge mobile access points, routing entities or resource providers
servers. Moreover, through federated learning mecha- in a decentralized manner. With the exponential growth
nisms, UAVs can also help in processing the collected in the number of connected IoT devices, data and service
data and sharing the learned model to the fog/cloud acquisition may be irregular depending on the location and
servers where all the received models are aggregated time. Thus, it may be more beneficial both economically
and compared for decision making. Adopting such and in terms of service quality to run such a blockchain-
an approach adds extra constraints on computation, assisted UAV network instead of a permanent fixed and
which requires optimization of the UAV resources over-provisioned cellular network. The solution supports
(CPU, battery, node election and formation) based service delivery either with the aid of a 5G network or
on efficient task allocation and scheduling (sensing, in a totally decentralized manner using only UAVs. Both
communication and computing), in addition to many public and private blockchains are deployed with the UAVs.
other mechanisms that help in reducing energy con- Moreover, with the support of fog and cloud computing
sumption and increase the operation lifetime. resources, data and services are efficiently delivered. A case
• Data type diversity: drones are exchanging various study was performed using a network simulator to compare
types of messages that should be treated differently. the data delivery success rate and number of messages
Some of exchanged data is used to guide the drones exchanged using the proposed solution and other state-of-
to fulfill their tasks (i.e. message exchange between the-art UAV-supported data delivery techniques.
 8

R EFERENCES Moayad Aloqaily (S’12, M’17) received the

Ph.D. degree in electrical and computer en-
[1] I. A. Ridhawi, M. Aloqaily, B. Kantarci, Y. Jararweh, and H. T. gineering from the University of Ottawa, Ot-
Mouftah, “A continuous diversified vehicular cloud service avail- tawa, ON, in 2016. He was an instructor in
ability framework for smart cities,” Computer Networks, vol. 145, the Systems and Computer Engineering Depart-
pp. 207 – 218, 2018. ment at Carleton University, Ottawa, Canada.
[2] O. Bouachir, M. Aloqaily, L. Tesng, and A. Boukerche, “Blockchain He is working with Gnowit Inc.as a Senior
and fog computing for cyber-physical systems: Case of smart Researcher and Data Scientist since 2016.
industry,” in Computer, 10.1109/MC.2020.2996212. IEEE, 2020. He is also the managing director of xAnalyt-
[3] S. Andreev, V. Petrov, M. Dohler, and H. Yanikomeroglu, “Future ics Inc., Ottawa, ON, Canada. Currently, he is
of ultra-dense networks beyond 5g: Harnessing heterogeneous with the Faculty of Engineering, Al Ain Uni-
moving cells,” IEEE Communications Magazine, vol. 57, no. 6, pp. versity, United Arab Emirates. His current research interests include
86–92, June 2019. AI and ML, Connected and Autonomous Vehicles,Blockchain Solu-
[4] F. Tang, Z. M. Fadlullah, N. Kato, F. Ono, and R. Miura, “Ac- tions, and Sustainable Energy and Data Management. He is an
poca: Anticoordination game based partially overlapping chan- IEEE member and actively working on different IEEE events. He has
nels assignment in combined uav and d2d-based networks,” IEEE chaired and co-chaired many IEEE conferences and workshops such
Transactions on Vehicular Technology, vol. 67, no. 2, pp. 1672–1683, as BCCA2020,PEDISWESA-ISCC2020, ITCVT-NOMS2020, E2NIoT-
Feb 2018. IWCMC2020, ICCN-INFOCOM19,AICSSA’19, BAT-FMEC’19’20. He is
[5] S. Garg, G. S. Aujla, N. Kumar, and S. Batra, “Tree-based a guest editor in many journals including International Journal of Ma-
attack-defense model for risk assessment in multi-uav networks,” chine Learning and Cybernetics, Elsevier IPM Journal, Springer JONS,
IEEE Consumer Electronics Magazine, vol. 8, no. 6, pp. 35–41, 2019. Springer Cluster Computer, Internet Technology Letters,Transaction on
[Online]. Available: https://doi.org/10.1109/MCE.2019.2941345 Telecommunications Technologies, Security and Privacy, and IEEE Ac-
[6] S. Zafar, S. Jangsher, M. Aloqaily, O. Bouachir, and J. B. Othman, cess. He is an Associate Editor with IEEE Access, Cluster Computing,
“Resource allocation in moving small cell network using deep Security and Privacy. He is a Professional Engineer Ontario (P.Eng.).
learning based interference determination,” in 2019 IEEE 30th
Annual International Symposium on Personal, Indoor and Mobile Radio
Communications (PIMRC), Sep. 2019, pp. 1–6.
[7] N. Bousias, S. Papatheodorou, M. Tzes, and A. Tzes, “Collab-
orative visual area coverage using aerial agents equipped with
ptz-cameras under localization uncertainty,” in 2019 18th European
Control Conference (ECC). IEEE, 2019, pp. 1079–1084.
[8] O. Bouachir, M. Aloqaily, F. Garcia, N. Larrieu, and T. Gayraud,
“Testbed of qos ad-hoc network designed for cooperative multi- Ouns Bouachir (M’18) is an assistant professor
drone tasks,” in Proceedings of the 17th ACM International Sympo- of computer engineering in the college of Tech-
sium on Mobility Management and Wireless Access, 2019, pp. 89–95. nological Innovation at Zayed University, UAE.
[9] D. He, H. Liu, S. Chan, and M. Guizani, “How to govern the non- She has a PhD degree in computer engineer-
cooperative amateur drones?” IEEE Network, vol. 33, no. 3, pp. ing from the University Paul Sabatier (Toulouse
184–189, May 2019. 3), Toulouse, France. She also holds an engi-
[10] X. Liang, J. Zhao, S. Shetty, and D. Li, “Towards data assurance and neering degree in Telecommunications from the
resilience in iot using blockchain,” in MILCOM 2017-2017 IEEE Higher School of Communications (Sup’Com),
Military Communications Conference (MILCOM). IEEE, 2017, pp. Tunis, Tunisia. Prior to joining Zayed University,
261–266. she worked as an assistant professor and post-
[11] L. Tseng, L. Wong, S. Otoum, M. Aloqaily, and J. B. Othman, doc researcher at Canadian University Dubai.
“Blockchain for managing heterogeneous internet of things: A Her current research activities are directed toward Connected and
perspective architecture,” IEEE Network, vol. 34, no. 1, pp. 16–23, Autonomous Vehicles, Blockchain Solutions, Internet of Things (IoT),
2020. Energy Optimization, AI and ML. She is an IEEE member and part of
[12] M. Aloqaily, A. Boukerche, O. Bouachir, F. Khalid, and S. Jangsher, the technical program committee for numerous journals, conferences
“An energy trade framework using smart contracts: Overview and and workshops like BCCA2020, ITCVT-NOMS2020, PEDISWESA-
challenges,” IEEE Network, 2020. ISCC2020, and AICSSA2019.
[13] F. Ali, M. Aloqaily, O. Alfandi, and O. Ozkasap, “Cyberphysi-
cal blockchain-enabled peer-to-peer energy trading,” in Computer,
10.1109/MC.2020.2991453. IEEE, 2020.
[14] B. Bera, D. Chattaraj, and A. K. Das, “Designing secure blockchain-
based access control scheme in iot-enabled internet of drones
deployment,” Computer Communications, vol. 153, pp. 229–249,
2020.
[15] O. Alfandi, S. Otoum, and Y. Jararweh, “Blockchain solution for Azzedine Boukerche (F’15) is a Distinguished
iot-based critical infrastructures: Byzantine fault tolerance,” in University Professor and holds a Canada Re-
Network Operations and Management Symposium. IEEE/IFIP, 2020. search Chair Tier-1 position at the University
of Ottawa, Canada. He has received the C.
Gotlieb Computer Medal Award, the Ontario Dis-
tinguished Researcher Award, the Premier of
Ontario Research Excellence Award, the G. S.
Glinski Award for Excellence in Research, the
IEEE Computer Society Golden Core Award,
the IEEE CS-Meritorious Award, the IEEE TCPP
Technical Achievement and Leadership Award,
the IEEE ComSoc ASHN Leadership and Contribution Award, and the
University of Ottawa Award for Excellence in Research. His current
research interests include wireless ad hoc and sensor networks, mobile
computing, and performance evaluation of large-scale distributed and
mobile systems. He has published extensively in these areas. He is a
Fellow of the Engineering Institute of Canada, a Fellow of the Canadian
Academy of Engineering, and a Fellow of the American Association for
the Advancement of Science.
 9

Ismaeel Al Ridhawi (M’09, SM’19) received his Ontario (P.Eng). He’s a Senior IEEE member with many peer-reviewed
BASc, MASc, and Ph.D degrees in Electrical publications in highly ranked magazines, journals and conference pro-
and Computer Engineering from the University ceedings. He is an associate and guest editor in many journals and
of Ottawa, Canada, in 2007, 2009, and 2014 re- has organized a number of IEEE conferences over the years. He has
spectively. He is an Assistant Professor of Com- also served as session chair for a number of symposiums and was
puter Engineering at Kuwait College of Science part of the technical program committee for numerous journals and
and technology and a researcher in the field of conferences. His current research interests include service delivery and
wireless communications. He has also worked provisioning in fog and cloud computing, quality of service monitoring
at the American University of the Middle East in for wireless networks, MEC network management, and service-specific
Kuwait as an Assistant Professor from 2014 to overlay networks.
2019. He is a registered professional engineer in