JOURNAL OF LATEX CLASS FILES, VOL. XX, NO.

X, AUGUST 20XX 1

Convergence of Blockchain and Edge Computing

for Secure and Scalable IIoT Critical Infrastructures
in Industry 4.0
Yulei Wu, Senior Member, IEEE, Hong-Ning Dai, Senior Member, IEEE, Hao Wang, Member, IEEE

Abstract—Critical infrastructure systems are vital to underpin The industrial control system (ICS) is the heart of a critical
the functioning of a society and economy. Due to ever-increasing infrastructure [9], [10]. It is mainly responsible for supervi-
number of Internet-connected Internet-of-Things (IoTs) / Indus- sory control and data collection (SCADA), monitoring the
trial IoT (IIoT), and high volume of data generated and collected,
security and scalability are becoming burning concerns for processes and control flows of system information in industry.
critical infrastructures in industry 4.0. The blockchain technology The wide adoption of Internet-connected IoT devices has pre-
is essentially a distributed and secure ledger that records all sented a variety of challenging issues to critical infrastructures.
the transactions into a hierarchically expanding chain of blocks. First, ICS was originally designed mainly for a proprietary
Edge computing brings the cloud capabilities closer to the and closed infrastructure without considering too much about
computation tasks. The convergence of blockchain and edge
computing paradigms can overcome the existing security and security issues, as traditional critical infrastructures are sort
scalability issues. In this paper, we first introduce the IoT/IIoT of isolated and are not vulnerable to cyberattacks. With these
critical infrastructure in industry 4.0, and then we briefly present infrastructures being connected to the Internet through IoTs,
the blockchain and edge computing paradigms. After that, we a wide range of cyberattacks, including distributed denial-of-
show how the convergence of these two paradigms can enable service (DDoS), malware, breach attack, Brute force attack,
secure and scalable critical infrastructures. Then, we provide a
survey on state-of-the-art for security and privacy, and scalability Man-in-the-middle attack, SQL injection, and phishing, are
of IoT/IIoT critical infrastructures. A list of potential research threatening the operation of ICS to provision normal support
challenges and open issues in this area is also provided, which for services [11], [12], [13], [14]. In addition, ICS is in a
can be used as useful resources to guide future research. position for data acquisition in critical infrastructures. The
Index Terms—Blockchain, Edge Computing, Critical Infras- compromised ICS by cyber attackers may create potential risks
tructure, Industry 4.0, Internet of Things. for the leakage of data privacy [15], [16]. Second, scalability
is another challenge which ICS was not originally designed
to solve. Given the remarkable increase in the number of
I. I NTRODUCTION
IoT devices and the volume of data they are collecting and
Critical infrastructure systems have been used to underpin analysing, the traditional centralised manner for data collection
the functioning of a society and economy. They range from and analysis is becoming the bottleneck of ICSs [17]. A
traditionally-defined physical assets to a more broad definition decentralised way is inevitably needed to fulfill the emerging
of modern assets in the sectors of electricity, gas, water supply, requirements of ICSs in support of advanced critical infras-
agriculture, public health, transportation, security services, tructures in industry 4.0.
telecommunication, etc [1], [2]. This transition is largely due The emerging blockchain and edge computing paradigms
to the ever-increasing usage of Internet-of-Things (IoTs) and are promising technologies that can tackle the above challeng-
their significant support for critical infrastructure systems in ing issues, in terms of security and scalability considerations
the era of industry 4.0 [3], [4], [5], [6]. The international data of critical infrastructures. The blockchain technology has
corporation (IDC) has forecast that there will be an estimate of emerged as a novel secure computing paradigm without the
41.6 billion connected IoT devices, generating 79.4 zettabytes need of any centralised authority in a networked system [18],
(ZB) in 20251 . IoTs have become indispensable parts of criti- [19], [20], [21]. It is a distributed consensus scheme that
cal infrastructures in industry 4.0, creating intelligent services allows transactions to be securely stored and verified. In
such as smart grid and offering a range of advantages for cost terms of security and privacy, the blockchain is created and
savings and efficiencies [7], [8]. maintained securely through the use of asymmetric cryptog-
raphy with crowd computing in a peer-to-peer manner. The
This work was partially supported by the Engineering and Physical Sci-
ences Research Council of United Kingdom under Grant No. EP/R030863/1. zero-knowledge proof has been leveraged to increase privacy
(Corresponding authors: Y. Wu and H.-N. Dai) protection in the blockchain system [22]. Edge computing is a
Y. Wu is with the College of Engineering, Mathematics and Phys- decentralised computing infrastructure that brings computing
ical Sciences, University of Exeter, Exeter, EX4 4QF, U.K. e-mail:
[email protected] and storage capabilities closer to the location where it is
H.-N. Dai is with Faculty of Information Technology, Macau University of needed [23], [24], [25]. In terms of privacy protection, data
Science and Technology, Macau. email: [email protected] does not have to be transferred to the remote cloud for
H. Wang is with Department of Computer Science, Norwegian University
of Science and Technology, Gjøvik, Norway email: [email protected] computation and storage. Blockchain can therefore inevitably
1 https://www.idc.com/getdoc.jsp?containerId=prUS45213219 compensate the security concerns and enhance the privacy
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 2

Government Financial services

Waste management

Transportation
Energy Public health

Manufacturing

Water management

Food and agriculture

Base station (Access point, IoT Gateway)

Sensor (IoT device, smart meter)

Computing facility (edge computing)

Fig. 1. Critical infrastructures with IoT.

TABLE I
A SUMMARY OF NATIONAL CRITICAL INFRASTRUCTURES IN DIFFERENT COUNTRIES .

Critical Infrastructures
Country Number of Sectors
The Same Infrastructures The Different Infrastructures
Chemicals, Civil Nuclear, Defence,
United Kingdom 13 Emergency Services, Space, Food
Financial Services, Chemicals, Dams, Information Technology,
Government Facilities, Commercial Facilities2 , Critical Manufacturing,
Communications, Energy, Health, Defence Industrial Base, Nuclear Reactors,
United States of America 16 Transportation Systems, and Water Materials and Waste, Food
Canada 10 Information and Communication Technology, Food
Australia 8 Food
Singapore 9 Infocomm, Media

protection of edge computing. The convergence of these two Section VII discusses potential challenges and open issues.
technologies is vital to provide necessary computation and Finally, Section VIII concludes this paper.
storage for IoTs, while guaranteeing the security and scala-
bility of critical infrastructures in industry 4.0 [26]. II. C RITICAL I NFRASTRUCTURES IN I NDUSTRY 4.0
Many research has been conducted to tackle security and Critical infrastructures refer to those vital assets, facilities,
privacy, and scalability issues of IoT based on blockchain and systems, sites, networks, information, people, processes, ei-
edge computing technologies. A systematic study of this area ther physical or virtual, that are necessary to underpin the
can take a further step to contribute to the research of IoT functioning of an economy and society3 , as shown in Fig. 1.
critical infrastructures in industry 4.0. With such a motivation, It also includes those functions, sites and organisations that
this paper introduces the critical infrastructure in industry are not critical to the maintenance of essential services upon
4.0 in Section II. Section III presents the technologies of which daily life depends, but that needs to be protected due
blockchain and edge computing, followed by how they can be to potential risks to the public (e.g., civil nuclear sites).
converged to provide necessary support for secure and scalable Different countries have their own definitions of national
critical infrastructures as in Section IV. Sections V and VI 2 This includes a wide range of sites that draw large crowds of people for
review and discuss state-of-the-art for security and privacy, shopping, business, entertainment, or lodging.
and scalability of IoT critical infrastructures, respectively. 3 https://www.cpni.gov.uk/critical-national-infrastructure-0
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 3

critical infrastructures. Table I shows a summary of critical To cope with the above issues, the proposed solutions for
infrastructures in United Kingdom, United States of America, critical infrastructures in industry 4.0 need to consider the
Canada, Australia, and Singapore. following factors:
Traditional critical infrastructures were quite isolated and • Security. Appropriate security mechanisms need to be in
were mainly vulnerable to physical attacks e.g., via an infected place to safeguard the IoT devices, the computing infras-
USB drive [27]. For example, the damage to the nuclear pro- tructure, the communication infrastructure, and various
gram of Iran by Stuxnet, a malicious computer worm, probably data running over these infrastructures.
via an infected USB drive, has caused significant damage to • Privacy. Many control and maintenance decisions are
industrial centrifuges used for enrichment of uranium4 . Due to learnt from data, e.g., fault prediction, detection and
digital transformation of industry 4.0 driven by smart factories, localisation are carried out based on many advanced data
big data and machine learning, critical infrastructures are analytics methods [34], [41]. The data usage needs to be
equipped with a dramatic increasing number of IoT devices, transparent, and sensitive data should not be transmitted
or industrial IoT (IIoT) in the context of industry 4.0 [28], outside its local network region.
[29], creating the so-called critical infrastructure with IoT or • Scalability. Data analytics methods rely on data collec-
IoT critical infrastructure. tion, storage and processing. The delay of these processes
Such a digital transformation of critical infrastructures is need to meet the stringent requirements of critical infras-
a double-edged sword. On the one hand, IoT critical infras- tructures and need to be scalable with the increase in the
tructures are in greater risks of being exposed of its internal size of the infrastructure.
structure, due to the connection to the Internet through massive
IoT devices via open standard protocols [30], [31], [32], [33].
III. B LOCKCHAIN AND E DGE C OMPUTING
On the other hand, it provides more useful information that
can be used for better maintenance of the system. For example, The emerging blockchain and edge computing technologies
Airbus launched a digital manufacturing initiative called the have exhibited excellent features that can cope with the above
“Factory of the Future”5 . Due to the complex process of issues mentioned in Section II. In this section, how the two
building a commercial airliner, many things that may go wrong technologies work and how they can handle these issues will
during the manufacturing process and may further endanger be presented.
passenger safety. To mitigate these potential risks, Airbus
equipped sensors in its machines. Through the collected data, A. Blockchain
a set of useful actions (e.g., anomaly prediction, detection and A blockchain is essentially a distributed and secure ledger
localisation) are performed for proactive maintenance [34], that records all the transactions into a hierarchically expanding
[35]. Faults can be repaired by engineers before escalating chain of blocks [42], [43], [44], [45], [46]. Each block in the
to a more serious error that may stop service provision. blockchain is linked to its previous block through the hash
According to the nature of how IoT critical infrastructures value of the parent block, except for the first block, usually
work, a number of components of the infrastructure are called the genesis block which does not have a parent block.
vulnerable to cyberattacks, including the following aspects: New blocks can be committed to a blockchain only upon
• Industry devices. There are a large number of already- their successful completion of the competition enforced by
deployed devices that are difficult to upgrade or patch, a consensus algorithm [47], [48]. Each block consists of the
making critical infrastructures inflexible for efficient han- following components (see Fig. 2):
dling of potential faults and attacks. In contrast, the new • Previous hash, which is the hash of the parent block.
IoT devices are connected to the Internet, and therefore, • Timestamp, recording the current time in seconds.
they are vulnerable to cyberattacks and can be easily • Nonce, starting from 0 and increasing for every hash
compromised [6], [36]. calculation.
• Communication infrastructure. IoT devices can now con- • Merkle Root, which is the hash of all the hashes of all
nect with other devices, including other IoT devices, com- the transactions in the block.
puting and storage devices, through open medium such • Transactions (Tx), which is the transactions executed
as cellular and Wi-Fi connections using open standard during a given period of time.
protocols [37]. The communication infrastructure itself
The blockchain technology possesses many features [42],
is also vulnerable to cyberattacks and the communication
[49], [50], [51] that are useful to tackle security, privacy and
may be eavesdropped [38].
scalability issues of critical infrastructures in industry 4.0,
• Computing infrastructure. Critical infrastructures were
including
using centralised cloud computing, where all the data
• Decentralisation. A blockchain validates a new block in
need to be transferred to the cloud data centre for process-
ing. This creates the potential risk of privacy leakage [39], a decentralised way without any centralised third-party
[40]. authority. In principle, every network user (node) can
participate in this validation. This trustfulness validation
4 https://www.europarl.europa.eu/RegData/etudes/BRIE/2019/642274/ process is essentially to complete a consensus procedure
EPRS BRI(2019)642274 EN.pdf via competition amongst all the involved users, and this
5 http://e-lass.eu/media/2018/02/TTG-ZAL.pdf can be achieved by consensus algorithms, such as proof
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 4

Block Header Data Block Header Data Block Header Data

Timestamp Nonce

Timestamp Nonce

Timestamp Nonce
Tx_1

Tx_1

Tx_1
……

……

……
Previous Hash

Previous Hash

Previous Hash
Merkle Root

Merkle Root

Merkle Root
Genesis block …

Tx_2

Tx_2

Tx_2
Tx_n

Tx_n

Tx_n
Tx_3

Tx_3

Tx_3
Block_1 Block_2 Block_m

Fig. 2. An example blockchain structure.

Network Operator A
Edge Computing
Network Operator B
Node (ECN)
Macro BS IoT Gateway
5G

Smart Home

Roadside Access
Point
Network Operator D

Offloading
Network Operator C

End devices with ECN

Fig. 3. Typical scenarios with edge computing.

of work (PoW) [52], proof of stake (PoS) [53], proof of This preserves a certain level of user privacy [56].
burn (PoB) [53], Byzantine fault tolerance (BFT) [54], • Accountability/non-repudiation. Blockchain systems have
and practical BFT (PBFT) [55]. a digital signature scheme. Transaction initiator signs a
• Immutability. Each block in a blockchain has the hash message with her private key before issuing it out, and the
of its previous (parent) block. Any changes to the parent recipient of this signed message uses the sender’s public
block invalidates all the subsequent blocks. In addition, key to prove the validity of the message. The transaction
the Merkle root is the hash of all the hashes of all the initiator therefore cannot be denied its signed transaction.
transactions in the block. Any modification to any trans- • Automation. Blockchain systems allow smart con-
actions in a block, after the block has been successfully tracts [57], [58], where approved contractual clauses
committed into a blockchain, will result in a new Merkle are transformed into executable computer programs and
root. The falsification to any transactions can therefore are executed automatically when a certain condition is
be easily detected. satisfied. The execution of each contract statement will be
• Transparency. Every user of a blockchain system can recorded as an immutable transaction in the blockchain.
access and interact with the blockchain network.
• Pseudonymity. As blockchain addresses are allowed to B. Edge Computing
be anonymous, users cannot access identification infor- With the dramatic increase in the number of IoT/IIoT de-
mation of the users who have made those transactions. vices, centralised cloud computing is becoming more difficult
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 5

to satisfy various quality-of-service (QoS) requirements of Cloud Computing

diversified industrial applications [59]. Edge computing has

been introduced to bring the computation capability closer

Blockchain Systems
to a computation task, in order to reduce network latency Edge Computing

AI Services
and save bandwidth resources towards the remote cloud data
centre [60], [61].
Edge computing nodes (ECNs) possess different functions IoT/IIoT Devices

and have different computation capacities according to their

location distance with end users [62], as shown in Fig. 3.
ECNs can be deployed at the macro base stations, providing
main data computation and storage capacities. ECNs can also Fig. 4. A layered architecture for IoT/IIoT with the convergence of blockchain
be deployed in a house in the smart home scenario, providing and edge computing.
extra computing power for smart home IoT devices. They
can also mounted at the roadside for the case of smart trans-
portation, significantly reducing the response time for delay-
sensitive applications such as autonomous driving. ECNs can
also be deployed at IoT/IIoT gateways for industrial scenar- pre-processing with light computation. Edge computing can
MEN
ios, providing data collection and aggregation functionalities. be embedded within an IIoT device, deployed in a house, an
ECNs can even be deployed at the end devices, performing office building, a micro base station, and even a macro base
data pre-processing [63], [64]. station; it provides necessary local computation and/or storage
capabilities to satisfy stringent QoS requirements of many
IV. C ONVERGENCE OF B LOCKCHAIN AND E DGE IIoT applications, such as ultra-low latency for autonomous
C OMPUTING FOR S ECURE AND S CALABLE C RITICAL vehicles and ultra-high reliability for remote surgery. In ad-
I NFRASTRUCTURES dition, edge computing allows local data to be processed
locally, without being transferred to the remote cloud. In
ECNs are operating by different third-party operators and
contrast, cloud computing has more computing power but it
are being deployed in a decentralised way, making it difficult
is located far away from computation tasks, and therefore the
to ensure the same level of security, transparency, and privacy
communication with cloud data centres can incur additional
preservation [65], [66]. The blockchain technology can essen-
network latency and consume more network bandwidth. The
tially overcome the shortcomings of edge computing. In addi-
blockchain system can ensure the security and transparency,
tion, edge computing can provide necessary local computing
and enhance privacy and scalability, of the above three layers
capabilities for computation tasks of blockchain systems, e.g.,
in the critical infrastructure. AI services provide the ability
smart contract execution and consensus procedure. Therefore,
for data processing at IoT/IIoT devices, edge computing and
the convergence of blockchain and edge computing paradigms
cloud computing layers.
can enable the following features that are crucial towards
secure and scalable critical infrastructures in industry 4.0:
Let us take an example to facilitate the understanding of
• Security. All transaction data in IoT/IIoT with edge com- the layered architecture shown in Fig. 4. A drone in the
puting are enforced automatically by smart contracts and IoT/IIoT Devices layer is monitoring weather conditions and
added to a blockchain upon successfully committed to a needs to transmit the pre-processed weather data to the weather
block. Security mechanisms can be easily implemented station. Due to limited computing resources and energy-
by smart contracts. efficiency considerations at the drone, part of the collected
• Privacy. Data can be collected and handled locally by data need to be offloaded to edge servers at the base station
edge computing. Data that is required to be transmitted for processing, in the Edge Computing layer. All transactions
outside where it originates, has a certain level of privacy are recorded by a blockchain, where the involved devices such
protection by virtue of blockchain’s pseudonymity mech- as the drone and the base station are miners. If the local
anism. computing resources of miners are limited, miners can offload
• Scalability. Both blockchain and edge computing their computing works, such as achieving an agreement by
paradigms are decentralised schemes. In other words, consensus algorithms, to more powerful computing facilitates.
they can be smoothly and readily converged without For example, drones can offload the computing works to edge
introducing additional scalability issues. servers, and the computing works at an edge server can be
Fig. 4 shows a layered architecture for IoT/IIoT in industry offloaded to other edge servers or the Cloud Computing layer.
4.0 with the convergence of blockchain and edge computing The AI Services layer provides necessary AI models to make
paradigms. The architecture consists of four layers: IoT/IIoT intelligent decisions, e.g., when is the best time for offloading.
devices, edge computing, cloud computing and blockchain
systems. IIoT devices are the smart devices in IoT/IIoT en- In what follows, the state-of-the-art that consider the con-
vironment, such as robotic arms in smart factories, smart vergence of blockchain and edge computing, to ensure the
farm sensors in smart agriculture, and smart thermometer in security and scalability of critical infrastructures, will be
smart home. They are responsible for data acquisition and investigated and discussed.
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 6

Cloud server
Data analysis Monitoring Resource
scheduling

Cloud computing layer

Computing layer
Blockchain node
Edge and cloud Data caching Pricing and incentive
orchestration and storage mechanism

Blockchain network layer Blockchain

Network layer
Edge node
Network Network Network
access control softwarization Services

Pico BS IoT Gateway WiFi AP

Macro BS
Blockchain edge layer
Communication layer

Spectrum Mobile services

ID Management
Management Management
IoT Device layer Reader

Fig. 5. Security and privacy solutions brought by the convergence of blockchain and edge computing.

V. S ECURITY AND P RIVACY OF II OT C RITICAL the decentralisation of blockchain systems can simplify the
I NFRASTRUCTURES IN I NDUSTRY 4.0 ID management process and lower the administration costs.
The integration of blockchain with edge computing has the Specifically, an IoT device can register, revoke and expire
potential to secure IIoT critical infrastructures and also protect its ID in the decentralised blockchain-based ID management
privacy-sensitive data in the infrastructure. Fig. 5 presents systems which can guarantee the trust without the necessity
an overview for the state-of-the-art of security and privacy of a third party. Meanwhile, the privacy/security risks of the
solutions brought by the convergence of blockchain and edge centralised systems can be eliminated or mitigated thanks
computing. In particular, the IoT device layer, blockchain edge to the temper-proof and non-repudiation characteristics of
layer and cloud computing layer correspond to the IoT/IIoT blockchains [68]. Moreover, the decentralisation brought by
devices, edge computing and cloud computing, respectively, blockchain can also help to reduce the SPF risk in the
as presented in Section IV. It is worth mentioning that the centralised systems. In addition, blockchains can also ensure
blockchain network layer plays a crucial role in connecting the anonymity of IoT devices since the generated addresses of
ECNs and cloud servers in the same plane via an overlay IoT devices can only be used to interact with each other in
network (i.e., the P2P network). We summarise the counter- the system.
measures to guarantee security and privacy in IIoT critical On the other hand, edge computing can also disburden
infrastructures in the communication, network and computing the centralised ID management systems by offloading tasks
layers, which are illustrated as follows. (e.g., registration, revoking and updating) to distributed ECNs.
Integration with blockchain can further improve the trans-
parency and security of ID management. All ID data stored in
A. Communication layer blockchain become traceable and immutable so as to improve
The integration of blockchain and edge computing can the overall security of IIoT. There are several proposals on IoT
protect the security and privacy for IIoT in identification ID management. For example, recent work [69] investigated
management and radio spectrum management. the usage of blockchain for Lithium-ion (Li-ion) battery man-
1) Identification management: The proliferation of diverse agement. Moreover, Guo et al. [70] adopted an integration of
IoT devices poses the challenges in identification (ID) man- blockchain and edge computing to achieve the trust of access
agement of IoT devices [67]. However, the incumbent ID control across diverse IoT systems.
management systems are not scalable with the explosion of 2) Spectrum management: We have experienced the radio
heterogeneous IoT devices. Meanwhile, the centralised IoT ID spectrum shortage due to the ever-growing demands on wire-
management can inevitably lead to low efficient bureaucratic less bandwidth driven by massive IoT devices and diverse IoT
processes, huge administrative costs, vulnerability to malicious applications [71], [72]. Blockchain also brings opportunities
attacks and susceptibility to single-point-failure (SPF) and in radio spectrum management for IoT devices. Blockchain is
privacy breaches. essentially a distributed database (a.k.a. a distributed ledger),
The advent of blockchain as well as edge computing which can record the usage of radio spectrum and enforce
technologies can potentially solve the above drawbacks of the effective spectrum access mechanisms. For example, the
the centralised ID management systems. On the one hand, work [73] investigated the adoption of blockchain as dis-
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 7

tributed spectrum database to achieve a fine-grained access feature of this system (mainly owing to distributed blockchain
control on radio spectrum, through which users may share nodes) can improve the scalability of the entire system. In
radio spectrum through appropriate settings of licenses in contrast to conventional network access control, blockchain-
blockchain. based network access control has the following advantages: 1)
Moreover, the built-in incentive and pricing mechanisms decentralization of network access control so as to improve
of blockchain can help to simplify the spectrum trading pro- the interoperability across the entire IoT system, 2) fine-
cess. The work [74] discussed the possibilities of leveraging grained access control can be achieved by the traceability of
blockchain for radio spectrum auction. Smart contracts running blockchain and ABE schemes.
on top of blockchain can also automate the spectrum auction The advances of blockchain-based smart contracts also
and trading. Meanwhile, a blockchain-based spectrum sharing foster the flexibility of network access control. In particular,
framework for 5G communications was proposed in [75]. In the work [80] presented a smart contract-based access control
particular, a smart contract regularising the terms for spectrum scheme to achieve flexible network access control. In this
sharing as well as the payments is given. The underlying scheme, there are three types of smart contracts for regis-
blockchain can also protect the privacy of trading parties. tration, judgement and multiple access control properties. An
3) Mobile services management: In 5G and beyond 5G implemented prototype also demonstrated the effectiveness of
communication systems, the management of massive services the proposed framework. Moreover, Islam and Madria [81]
subscribers is becoming a challenge due to the difficulty proposed an attribute-based network access control scheme
in coordinating the fragmented heterogeneous networks and based on smart contracts running on top of Hyperledger Fabric,
the growing administrative cost for handling various mobile which is a permissioned blockchain. Experiments on an IoT
services. Blockchain technology can potentially address these testbed were conducted to further verify the effectiveness of
emerging issues in mobile services management in 5G and the proposed scheme. In addition, in [82], the distributed
beyond 5G networks. trust in Internet-of-Vehicles (IoV) can be achieved through
As shown in Fig. 5, blockchain is essentially a middleware the consensus mechanism of blockchain.
to connect distributed mobile networks together. Consequently, 2) Network softwarisation: In order to cater for the growing
diverse communication networks can be integrated together demands of diverse IoT applications, the network softwari-
to offer a seamless mobile service to users. Moreover, the sation has drawn extensive attention recently [83]. Typical
built-in smart contracts on top of blockchain can also auto- network softwarisation technologies mainly include software-
mate the service subscription, suspension, modification and defined networks (SDN) [84], [85], network functions virtual-
termination, thereby reducing the administrative expenditure. isation (NFV) [86], [87] and network slicing [88], [89].
The study [76] presented a blockchain-based roaming man- The advent of SDN can fulfill the flexible and scalable
agement system for cellular networks, providing users with connections of massive IoTs while most of the existing SDN
an ubiquitous roaming service. Furthermore, the positioning solutions that are centralised are susceptible to SPF or mali-
accuracy is also a critical issue in mobile services, especially cious attacks. The introduction of blockchain can decentralise
for vehicular networks. The study [77] proposed a blockchain- SDN schemes thereby improving the reliability of SDN-based
based framework on collaborative positioning. Particularly, IoT systems. The recent study [90] presented a blockchain-
blockchain can ensure the data provenance of positioning data. based decentralised SDN solution, which can effectively solve
the handover authentication problem. Moreover, the work
of [91] also presented a blockchain-based SDN scheme for
B. Network layer the IoV scenario, in which blockchain was adopted to achieve
In the network layer, the incorporation of blockchain into decentralisation and trustworthiness of multiple network enti-
edge computing can protect the security in aspects of network ties and SDN was leveraged to guarantee the effectiveness of
access control and network softwarisation. We next illustrate network management.
the research advances in these two aspects in detail. Meanwhile, the provision of NFV technologies can facilitate
1) Network access control: One of the major obstacles in the diverse services for IoT applications while both security
IoT data sharing is the reciprocity absence across diverse IoT and trust among multiple virtualised network entities pose
systems. The introduction of blockchain to the IoT ecosystem a challenge in popularising NFV to IoT communities. The
can not only improve the interoperability [42] but also provide convergence of blockchain with NFV can potentially overcome
fine-grained access control of various components in IoT these challenges. The study presented in [92] investigated the
systems. The work [78] presented a blockchain-based access integration of blockchain with NFV to secure NFV orchestra-
control for IoT systems. In particular, a fine-grained access tion functions so as to achieve traceable and non-repudiated
control based on attribute-based encryption (ABE) can ensure services. Moreover, the work [93] harnessed the auditability
the permission attribute updating in time. As an example, and incentive mechanism of blockchain to design a reverse
the revoked users cannot access the on-chain data, which auction scheme to solve the competition of virtual network
nevertheless can be visible to authorised users. Moreover, functions (VNF) services providers.
the work [79] presented a decentralised network management Network slicing accompanying by SDN and NFV tech-
system for IoT on top of blockchain. Even though the per- nologies can fulfill the diverse demands of various IoT ap-
formance improvement over the conventional IoT network plications via partitioning the entire physical network into
management systems is not significant, the decentralisation multiple segregated network planes. The work [94] presented
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 8

a blockchain-based broker mechanism for IoT devices in 5G tasks from IoT devices to ECNs. In addition, Jiao et al. [105]
networks, in which network resources can be securely leased to presented an auction model to analyse the trading procedure
end users in a privacy-protected manner. Moreover, the study between cloud/edge services vendors and blockchain miners.
presented in [95] showed that the introduction of blockchain Moreover, the work [106] presented an overview on using
to network slicing can further improve the reliability of the blockchain for cloud services exchange in a cloud market.
content sharing in information-centric networks (ICN) [96]. Furthermore, authors in [107] presented a cloud-edge orches-
3) Network services: The integration of network ac- tration framework to coordinate crowdsensing tasks in mobile
cess control and network softwarisation mechanisms with IoT scenarios. In this framework, a cloud server playing a role
blockchain can offer unified network services. On the one as a controller can gather sensing data from ECNs, which out-
hand, blockchain and smart contracts can enable the flexible source sensing tasks to mobile IoT devices to collect sensing
network access control. On the other hand, the combination of data. An auction mechanism was proposed to incentivise par-
blockchain with network softwarisation technologies can also ticipatory workers (i.e., IoT devices). Similarly, the work [108]
facilitate the network management. Consequently, the provi- presented a blockchain-based mobile crowdsensing system for
sion of secure and ubiquitous network services is envisioned IIoT. In contrast to conventional mobile crowdsensing systems,
for critical infrastructures. the decentralisation of blockchain can further enhance the
There are several representative network services based reliability and security of the system.
on blockchain. In [97], a blockchain-based network storage 2) Data caching and storage: The explosion of IoT data
service was presented. In particular, a blockchain-based data poses challenges in data management, especially in data stor-
auditing scheme integrated with the bi-linear pairing crypto- age and data analytics [118]. Cloud computing can offload
graphic mechanism was devised to ensure the data integrity. storage and processing burdens at IoT devices while it also
Meanwhile, Aujla et al. [98] presented a framework of inte- brings the challenges in data privacy and security protection.
grating SDN and blockchain to offer flexible network services. Edge computing can undertake storage and processing at
Particularly, blockchain-enabled SDN can mitigate the attacks ECNs in approximation to users, thereby improving context-
such as malware and denial-of-service (DoS) attacks. More- awareness and protecting data privacy.
over, the study [99] investigated the integration of blockchain The in-depth integration of blockchain and edge/cloud can
with ECNs to provide trusted edge services. Incentive schemes further preserve IoT data privacy and security. In particular,
that are embedded with smart contracts can incentivise ECNs the work [109] presented a blockchain-based data management
to contribute to edge services. system for IoT, in which both edge and cloud computing facil-
ities are integrated with blockchain to guarantee effective data
sharing. Harnessing the non-repudiation and anti-tampering
C. Computing layer characteristics of blockchain, Xu et al. [110] proposed a
The integration of blockchain and edge computing can blockchain-based data sharing system to support a diversity
solve the following security issues in the computing layer. We of edge applications. Experimental results further verified the
discuss the research advances as follows. effectiveness of the proposed scheme.
1) Edge and cloud orchestration: IoT data has typically There are other studies on investigating the adoption of
been uploaded to remote clouds for storage, analysis and blockchain in other edge computing scenarios. In particular,
interpretation [100]. However, cloud services have typically the work [111] investigated to leverage blockchain to achieve
been owned by untrustworthy third parties, which may misuse the trust of multiple ECNs, which can temporarily store pop-
IoT data or unintentionally disclose the privacy-sensitive data ular contents (a.k.a. caches) so as to improve user experience.
to others. Moreover, it may cause considerable end-to-end Blockchain can also be used in the video streaming scenario.
delay to upload IoT data to remote clouds. The advent of Liu et al. [112] presented a blockchain-based video streaming
edge computing [101] can overcome the drawbacks of cloud framework with edge computing. Meanwhile, a three-stage
computing through offloading computation and storage tasks Stackelberg game was used to investigate the interaction
to ECNs, which are close to users. Thus, edge computing can among the users, base stations and video providers. Moreover,
essentially complement with cloud computing to better serve the work [113] exploited the merits of blockchain such as anti-
IoT. tempering and decentralisation to achieve the fast repairing of
The effective edge and cloud orchestration is a necessity data storage nodes in IIoT environment.
for IoT ecosystem while it also poses a number of security 3) Pricing and incentive mechanisms in computing: The
and privacy challenges especially in the trustless and het- IIoT critical infrastructure consists of diverse computing facil-
erogeneous computing environment [102]. There are a few ities, such as IoT nodes, ECNs and cloud servers with different
studies to guarantee trust and security of edge and cloud computing capabilities and storage capacities. It is crucial to
orchestration enabled by blockchain. In particular, Xiong et motivate diverse computing nodes to participle in computing
al. [103] modeled the interactions among cloud servers, ECNs and storage tasks. In addition, many consensus algorithms of
and blockchain miners as a multi-leader multi-follower game, blockchain also require substantial computing contributions
which is essentially a computationally-complex problem while from some computing nodes (i.e., miners). Therefore, the
authors successfully solved the problem by an Alternating Di- pricing and incentive mechanisms become a challenge in the
rection Method of Multipliers (ADMM) approach. Meanwhile, IIoT critical infrastructure.
the study of [104] investigated to disburden blockchain mining Many recent studies aim at addressing this issue. Kang
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 9

TABLE II
S UMMARY OF SECURITY AND PRIVACY SOLUTIONS ENABLED BY BLOCKCHAIN AND EDGE COMPUTING

Perspectives Issues References

• Identification management [67], [68], [69], [70]
Communication layer • Spectrum management [73] [74] [75]
• Mobile services management [76] [77]
• Network access control [78] [79] [80] [81] [82]
Network layer • Network softwarization [90] [91] [92] [93] [94] [95]
• Network services [97] [98] [99]
• Edge and cloud orchestration [102] [103] [104] [105] [106] [107] [108]
Computing layer • Data caching and storage [109] [110] [111] [112] [113]
• Pricing and incentive mechanisms [114] [115] [116] [117]

et al. [114] proposed a two-stage strategy to mitigate the servers and data storage servers which have strong comput-
collusion of blockchain miners. In particular, a contract theory ing/storage capabilities may store the entire blockchain and
was introduced to incentivise miners to contribute to the also be responsible for the computationally-intensive tasks
block verification. Meanwhile, the study [115] investigated the such as mining and executing machine learning/deep learning
incentive mechanisms in ECNs providing blockchain miners algorithms. However, ECNs and IoT devices may only store
with computing services. Particularly, a two-stage Stackelberg partial blockchains (e.g., hash values of the blockchain or
game model was used to analyse the interactions between a subset of blocks) due to the limited computing/storage
ECNs and miners. Moreover, the study [116] investigated a capabilities. It is a critical issue to ensure the consistency of
mechanism to promote the consensus propagation across the blockchain data across the entire network.
blockchain network. Furthermore, a credit-based approach was
devised in [117] to achieve the computing resource trading VI. S CALABILITY OF II OT C RITICAL I NFRASTRUCTURES
between ECNs and blockchain-enabled IoT nodes. IN I NDUSTRY 4.0

In addition to security and privacy concerns of IIoT critical

D. Summary infrastructures, the scalability has limited the wide adoption of
The integration of blockchain and edge computing can IIoT. The integration of edge computing and blockchain can
address the security and privacy concerns in critical infras- improve the scalability of IIoT. First, edge computing can offer
tructures of IIoT in communication, network and computing ubiquitous computing facilities to blockchain and IIoT nodes.
layers. Table II summarises the state-of-the-art solutions in Second, blockchain being a middleware across different IIoT
different aspects. systems can enhance the security and privacy of both ECNs
and IIoT. In this section, we discuss the scalability of IIoT
Manufacturing Power grid Transportation
critical infrastructures mainly in two aspects: 1) the intrinsic
scalability of IIoT and 2) the scalability of blockchain.
Applications

Camera
Industrial

Smart meter EV

Hash A. Scalability of IIoT

IoT Gateway value Macro BS Pico BS
With the vision of ubiquitous connections everywhere and
Partial blockchain data Partial blockchain data elastic access for everything, IIoT has the stringent require-
Cloud computing layer

ment on the scalability in critical infrastructures [119]. How-

Blockchain data
Blockchain data ever, as shown in Fig. 7, the scalability of IIoT is affected
Cloud Server Cloud Server
by heterogeneous IIoT devices, diverse IIoT networks, and
Sensor (IoT device)
Blockchain data
massive IIoT data [120]. In particular, IIoT consists of various
ECN
Wireless link
Data storage IoT devices such as RFID tags, sensors, controllers, and
robot arms, which are connected through wired networks or
Fig. 6. An example to illustrate the working of the integration of blockchain
wireless networks. The heterogeneity of IIoT devices exhibits
and edge/cloud computing. in both hardware (e.g., ICs and sensors) and software (e.g.,
operating systems and firmware). In addition to heterogeneous
With respect to the practical realisation of the solutions, the IIoT devices, the networks connecting various IIoT devices
integration of blockchain and edge/cloud computing facilities also have different protocols across the entire protocol stack.
should be decomposed into diverse computing facilities, which For example, near-field communication (NFC), back-scatter
are distributed across the entire IIoT critical infrastructure. communications, and Bluetooth have often been adopted for
Fig. 6 depicts an example of the integration of blockchain short-range communications, while WiFi, Low Power WAN
and edge/cloud computing facilities. In this scenario, cloud and cellular communications (e.g., 4G and 5G) have been
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 10

nologies to edge computing and network softwarisation has

Heterogeneous become an inevitable trend to further improve the scalability
Diverse IIoT
IIoT devices
Edge computing
networks of IIoT. In particular, Pan et al. [129] proposed a framework
with integration of blockchain and ECNs to enhance IoT. The
Massive IIoT
Network smart contracts running on top of blockchain can automate
softwarisation
Data the regulation of IoT devices, consequently enhancing the
reliability. Moreover, the work [130] presented a blockchain-
Blockchain based framework to improve the scalability with consideration
of heterogeneous IoT devices. To address the transaction
Scalability of IIoT Solutions
throughput bottleneck of blockchain systems, the authors also
designed a new consensus scheme as well as a space-structured
Fig. 7. Scalability of IIoT and solutions. chain structure. The resource limitation of IoT devices also
leads to the difficulty of the adoption of blockchain to IIoT
scenarios. The work [131] proposed a solution to address this
used to connect IoT devices over a longer distance [121], issue by localising blockchain peers, thereby improving the
[122]. Moreover, a proliferation of massive structured and non- scalability.
structured IIoT data also leads to the difficulties in data storage
and analytics.
B. Scalability of blockchain
On the other hand, the heterogeneous IIoT critical in-
frastructures across different business sectors or government The scalability of current blockchain technologies6 is still
departments have led to difficulties in information sharing and far from meeting the demand of IIoT. One of the earlier
reciprocal operations among different IIoT systems, conse- analyses on the scalability of blockchain was conducted by
quently leading to the difficulty in reaching the scalability. Croman et. al. [132], which pointed out the large throughput
Meanwhile, the insufficient resources of IIoT devices, as well gap between Bitcoin (7 transactions/second maximum) and
as heterogeneity of IIoT devices and networks, are also the root the mainstream payment processor such as Visa credit card
causes of many security and reliability vulnerabilities [123]. (2000 transactions/second on average). The authors proposed
Moreover, the massive volume of IIoT data has often been a decomposition of the Bitcoin system into 5 abstraction
uploaded to remote clouds, which have nevertheless been layers (planes): Network, Consensus, Storage, View, and Side.
possessed by trustless third parties, consequently leading to In each layer, the authors reviewed different approaches to
privacy leakage risks. Security, privacy and reliability vulner- improve the scalability. One limitation of this work was that
abilities of existing IIoT critical infrastructures also increase it heavily focused on cryptocurrencies.
the difficulty in achieving the scalability. More recently in 2018, Dinh et. al. [133] presented a
The fusion of edge computing, network softwarisation and more comprehensive overview of different blockchain systems
blockchain technologies can offer solutions to the scalability w.r.t. data processing and performance. The authors described
of IIoT, as shown in Fig. 7. First, the recent advances in edge another decomposition into 4 layers: Application, Consen-
computing and network softwarisation can potentially address sus, Execution engine, and Data model. They performed
the scalability challenges of IIoT critical infrastructures. In extensive experiments, with their benchmarking framework
particular, the work of [124] presented an edge computing- BLOCKBENCH, to evaluate the throughput, latency, scalabil-
based attestation systems for IoT devices. An implemented ity, fault tolerance, and security metrics on three representative
prototype demonstrated the scalability of the entire system. blockchain systems, namely Ethereum, Parity, and Hyper-
Meanwhile, Togou et al. [125] presented a decentralised SDN ledger. Their evaluation showed that current blockchains’
to achieve the scalability of large-scale networks in contrast to performance is “far below what a state-of-the-art database
the conventional centralised SDN solutions. The existing SDN system can offer”. One limitation of this work, despite its
solutions have bottlenecks at network controllers, which often extensive coverage, paid insufficient attention to IoT applica-
have the limited computing capabilities. To address this issue, tions. One reason can be in 2018, the IoT-oriented blockchains
the work [126] proposed an integration of edge and cloud were less developed compared to general or crypto-currency
computing facilities to overcome the computing bottlenecks blockchains.
of SDN controllers. One of the most important scalability Here we summarise in Table III the approaches that
metrics is the latency. The recent work [127] presented a have been proposed in the community for the scalability of
study of integrating SDN, NFV and network slicing to achieve blockchains.
the ultra-low latency in 5G networks. Moreover, both edge Very recently in 2020, Lao et. al. [146] presented a sur-
and cloud computing facilities have been deployed to the vey of IoT applications in blockchain systems. The survey
core network to improve the computing capabilities, thereby did cover the direct acyclic graph (DAG) based distributed
reducing the delay. Furthermore, the work [128] presented ledger technology (DLT), along with the fast growing IOTA
an optimisation scheme for network slicing recovery and
6 Strictly speaking, some technologies reviewed here are currently preferred
reconfiguration, thereby improving the system reliability and
to be called distributed ledger technology, instead of blockchain. For example,
scalability. the network topology of IOTA Tangle is not a chain structure, but a graph
On the other hand, the introduction of blockchain tech- structure.
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 11

TABLE III cross-chain synchronisation protocols to ensure satisfactory

S UMMARY OF APPROACHES TO IMPROVING THE SCALABILITY OF global consensus. Atomic cross-chain swaps by Herlihy [139],
BLOCKCHAINS
modelled in a directed graph structure, enables the exchange of
Aspects Works assets across multiple (unrelated) blockchains. Delegating the
DAG Tangle [134], [135]
execution of some transactions from the main blockchain to a
Structure
set of sidechains [140], or called parachains in Polkadot [141]
Federated
[133], [136], [137], [138] is another well known approach. A recent work [142] on
blockchains
Multiple chains [139], [140], [141], [142], [143]
the multi-chain structure, in the context of industrial Internet,
proposed a node-clustering strategy to reduce cross-chain
Consensus Stellar [136], Ripple [137], CAPER [138]
interactions and improve throughput. The delayed-replication
Sharding [144] algorithm by Hellings and Sadoghi [143] aimed to improve
Database tech.
Transaction the efficiency of processing read-only workloads.
[145]
reordering
The database community recently proposed several solid
works in transitioning database technologies to blockchain sys-
tems with notable improvements on scalability. In SIGMOD
Tangle [134] implementation. However, this survey mainly 2019, Dang et. al. [144] presented a scalable blockchain based
focused on architecture, consensus, and traffic modeling, and on sharding and achieved a throughput of over 3000 trans-
thus, the coverage on the scalability is insufficient, and the actions/second. Sharma et. al. [145] successfully employed a
authors did not review critically the weaknesses of IOTA well-known database technique, transaction reordering, and
Tangle. increased the throughput of successful transactions with a
DAG-based DLT has higher throughput and scalability com- factor of 12x and decreased the average latency to almost half.
pared to the original chain-structure-based blockchains [147]. Research works integrating blockchain with wireless net-
It is due to the fact that DAG is unidirectional with no works, IIoT, and cloud/edge computing started emerging very
ring structure, which ensures high efficiency in searching and recently. Sun et. al. [149] presented a blockchain-enabled
communication. A typical DAG implementation is the IOTA wireless IoT model and a search algorithm aiming to find
Tangle. IOTA Tangle recently grows fast especially in indus- the optimal deployment of full function nodes under a given
tries. IOTA, being a public permissionless distributed ledger, node density and transaction throughput. Liu et. al. [150]
is designed for IoT applications to support high-frequency also looked at the wireless IoT systems and proposed a new
micro transactions. Considering those micro transactions are blockchain system that considers the heterogeneity, resource
normally low value transactions, there are no explicit trans- constraints, and dynamics (frequent join/leave due to on-off
action fees in IOTA. When a new transaction arrives, it will switching or mobility reasons) of IoT devices. Their prototype
select two previous transactions (called tips) to validate. This achieves a peak throughput of 3400+ transactions/second.
means IOTA, though having higher throughput, follows a Xiong et. al. [151] and Yao et. al. [152] studied resource
PoW protocol. The tip selection algorithm (TSA), being the management that allows IIoT devices to offload computational
key component in the IOTA consensus, is an active research tasks to cloud/fog providers. Xiong et. al. [153] and Wu et.
topic. IOTA Tangle proposed two TSA algorithms in their al. [154] proposed blockchain systems that offloads the miners’
original white paper [134], the random selection algorithm and PoW computation tasks to the mobile edge computing (MEC)
the random-walk-based Monte Carlo Markov Chain (MCMC) network.
selection algorithm. But as recognised by themselves and From the above review on the scalability, we observe that
follow-up improvements [135], IOTA Tangle is still vulnerable the blockchain is developing towards multiple chains, mod-
to the parasite chain attack, which could cause damage to the elled in graph structures, with hierarchical consensus protocols
immutability and irreversibility of the ledger. In this aspect, and more database techniques. We see the synergy between
IOTA Tangle still needs critical improvements to meet the this trend and the development of edge/cloud computing,
high demands in security and fault tolerance from IIoT critical which has been envisioned in our proposed layered architec-
infrastructure applications. ture for IoT/IIoT with convergence of blockchain and edge
As commented by [133], the main-stream consensus pro- computing in Section IV.
tocols based on PBFT [148] are still communication bound,
thus they have hard limits in scalability. Stellar [136] and VII. P OTENTIAL C HALLENGES AND O PEN I SSUES
Ripple [137] can be called federated blockchain systems. Such Although blockchain and edge computing technologies have
blockchains divide the network into smaller groups called been used to ensure secure and scalable IoT/IIoT critical
federates [133] or quorums [136], and each federate maintains infrastructures, there are still many on-going challenges and
local consensus. Local consensus can be propagated to the open issues that need to be considered in future research. In
whole network, and the global consensus can be reached under this section, we discuss a set of issues, in terms of architecture,
certain conditions. The parallel executing federates improve secure infrastructure, and scalable infrastructure.
throughput. CAPER [138] further elaborated and proposed 3
global consensus protocols. Interestingly, CAPER adopts the A. Architecture of IIoT critical infrastructures
DAG structure for the distributed ledger. 1) Standard application programming interface for appli-
One step further is to employ multiple chains with relevant cation developers: Edge computing is in a position to serve
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 12

diversified applications, and each application has its own range of use cases in critical infrastructures, e.g., industrial
ecosystem that uses the platform that may be ecosystem- manufacturing and a variety of other sectors. The network
specific7 . For example, the platforms and application program- economy models are crucial for the success of the ecosystem
ming interfaces (APIs) for transportation systems should be of these sectors. Existing solutions in this area seldom consider
different from those required for government facilities. Each network economy models, and thus they are not sustainable
country has many critical infrastructures covering a range and practical solutions. How to design a practical solution
of sectors, e.g., financial services, energy, health, etc (see for the convergence of edge computing and blockchain, by
more details in Table I). A robust edge computing framework considering network economy factors e.g. pricing mechanisms
solution should be able to provide standard northbound APIs of real-world applications, is still an open issue. Game theory
for application developers from different platforms and flexibly is a versatile tool that has been useful to make decisions related
deploy necessary functionalities, along with the advanced to network economy [25]. However, network environment is
networking technologies, e.g., SDN and NFV. It should also be becoming much more dynamic than ever, we must bring in
able to accommodate various southbound transmission proto- the tools (e.g., AI) that can capture the features of this ever-
cols between IoT/IIoT devices and the cloud. In addition, edge changing environment to help game theory do a better job.
cloud may belong to different operators. An efficient east-west
data transmission protocol may be needed for communication
B. Secure IIoT critical infrastructures
between different network operators. The design of these
protocols should be coupled with the stringent requirements 1) Security vulnerabilities of IIoT devices: The resource
of diversified IoT/IIoT applications and security guarantee limitations of IIoT devices have often been the root causes
mechanisms (e.g., blockchain) in critical infrastructures. of the security vulnerabilities. On the one hand, the limited
2) Integrated networking, computing, storage and power computational capability and battery capacity lead to the
resource allocation: Resource allocation is an important re- difficulty of deploying computationally-complicated encryp-
search topic in edge computing, which depends on many fac- tion algorithms at IIoT devices. Meanwhile, the failure of
tors, including energy consumption, power allocation in energy upgrading or patching IoT firmware also results in the IIoT
renewable networks [155], [156], computing capabilities at the devices being vulnerable to malicious attacks [157]. It is
IoT/IIoT devices, the edge and the cloud, the key components reported in [158] that blockchain-enabled smart contracts can
of emerging network architectures, wireless communications, automate the IoT firmware upgrading procedure through the
to name a few. Many studies only consider one or two of contract clauses (i.e., instructions) built-in IoT devices since
these factors. For example, most studies consider energy- the date of production. Moreover, the recent work [159] also
computing trade-off for computation offloading solutions. It presented a blockchain-based solution via monitoring software
is challenging to have an integrated networking, computing, status of IoT devices. Blockchain can store snapshots of IoT
storage and power resource allocation scheme that is useful software status to monitor and detect any malicious activities
for a practical use case. For example, how to incorporate the (e.g., a backdoor firmware upgrading).
in-network caching of ICN, which was designed to reduce 2) Security vulnerabilities of blockchain: Although
the delay for content retrieval, into the resource allocation blockchain has the advantages in security enhancement of
solutions of edge computing, is still an open issue. the IoT ecosystem, the intrinsic security vulnerabilities of
3) Decentralised network management: The convergence blockchain systems also prevent blockchain from being
of edge computing and blockchain is a decentralised network widely adopted in IIoT critical infrastructures. For example,
in nature. The traditional centralised or hierarchical network it is reported in [160] that the failure of properly-configuring
management would not work efficiently. In addition, the surge gas costs of Ethereum Virtual Machine (EVM) may lead to
in data volume that will come from the massive IoT/IIoT Ethereum suffering from DoS attacks. Moreover, blockchain-
devices enabled by 5G has made edge computing more based domain name system (BDNS) can be abused by
difficult to manage. Furthermore, edge computing has been cyberattackers to conduct intrusion attacks [161] due to the
coupled with advanced networking technologies, e.g., SDN, anonymity of BDNS. The recent progress in big data analytics
NFV and network slicing, for efficient service deployment on blockchain data brings the opportunities to remedy the
and network control. The complexity of edge computing due security vulnerabilities of blockchain. For example, the recent
to these factors, coupled with the introduction of blockchain, work [162] presented a framework to collect blockchain data
has made the network management a hard task. Decentralised and detect various attacks occurring on blockchain.
network management is definitely a trend, but how to design 3) Integration of AI to secure IIoT critical infrastructures:
an efficient solution that considers the above factors and can Massive data has been generated from the entire IIoT critical
be interworked/integrated with the management framework infrastructures from the communication layer to the computing
of emerging networking technologies, e.g., management and layer. Big data analytics (BDA) on IIoT critical infrastruc-
orchestration (MANO) in NFV, is still a challenging issue. tures can classify abnormal behaviours, detect and recognise
4) Network economy: The convergence of edge computing intrusions as well as malicious attacks [163]. Meanwhile,
and blockchain plays an important role in enabling a wide BDA on the operational data of IIoT critical infrastructures
can also help to identify the performance bottlenecks and
7 https://www.ericsson.com/491e83/assets/local/reports-papers/ericsson- make proactive actions (like tuning performance metrics).
technology-review/docs/2020/next-generation-cloud-edge-ecosystems.pdf Moreover, BDA on blockchain can be beneficial to pinpoint
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 13

vulnerabilities of blockchain as analysed above. However, the edge/cloud computing facilities to enhance the scalability of
heterogeneity and diversity of IIoT critical infrastructure data IIoT ecosystem. In addition, the introduction of blockchain to
also pose the challenges in data analytics [164]. The recent IIoT critical infrastructures can improve the interoperability
advances in AI have brought opportunities to address the above (i.e., reciprocal operations) among different IIoT systems.
issues. Moreover, the massive data generated in IIoT can be used to
First, the integration of AI with cloud computing can identify the performance bottlenecks or abnormal activities so
process massive IoT data and extract valuable information. as to improve the scalability of IIoT [181]. In the future, the
Second, AI can empower ECNs and IoT devices with in- fusion of AI with the above technologies can further improve
telligence [165], [166]. Due to the resource limitation, IoT the scalability and elasticity of IIoT ecosystems.
devices may possess the limited intelligence. The intelligence 2) Scalability of blockchain: The scalability of blockchain
bestowed to ECNs by AI that is named as edge intelligence in itself is a big open problem, which prevents the adoption of
can serve an important complement to IoT devices [167]. this technology in many real-world application domains. There
For example, the work [168] proposed an amalgamation of is still no tangible scalable solution for IoT applications. IOTA
blockchain with edge computing, in which a deep reinforce- Tangle, being the largest and most successful one, is still far
ment learning (DRL) [169] was proposed to achieve the from satisfactory, with several critical vulnerabilities, includ-
dynamic resource scheduling. Meanwhile, authors in [170] ing being not deterministic, relying on a central coordinator
adopted a DRL method to allocate both computing resources to avoid security attacks, and being susceptible to the parasite
and blockchain operations in an adaptive manner. Moreover, chain attack. The direction in coupling multiple chains, in
the work [171] presented a DRL method to optimize network graph structures, with hierarchical consensus protocols seems
slicing in 5G networks. Furthermore, deep learning approaches promising, but obviously calls for large amount of research
can help to identify these malicious attacks through analysing efforts.
the activity reports and suggest relevant countermeasures. 3) Coordination across disciplines: The momentum and
For example, the work [172] showed that deep learning can interest shown in scalability from different communities (e.g.,
analyse the network traffic to identify the attacks. database, network, and high performance computing), while
4) Data privacy preservation: In addition to cloud servers, contributing good knowledge and insights in this important
both IIoT devices and ECNs are vulnerable to privacy leakage issue, expose the fragmented and un-coordinated nature of
risks. On the one hand, it is shown in a recent work [173] that these efforts from different angles. The layered architecture
user privacy during the spectrum auction can be breached. On with the convergence of blockchain, IIoT, and edge/cloud
the other hand, the privacy leakage risks exist when raw data computing is a strong push towards coordinated research
collected from IIoT devices is sent to untrustworthy ECNs, efforts, by linking the strengths from different communities,
which can be hijacked or misused by attackers. Consequently, for scalable and secure solutions. This convergence opens up
data stored at ECNS can be stolen or misused. many research opportunities. A good coordination between
Recent advances in differential privacy [174], homomorphic the blockchain layer and cloud/edge computing and/or IIoT
encryption [175] and federated learning [176], [177] bring devices in the architecture (as shown in Fig. 4) has the
the opportunities in offering privacy protection in IIoT critical potential of lifting the scalability of blockchain, and in general
infrastructures [178]. In particular, the work [174] presented a the critical infrastructure applications to a new level.
joint framework of blockchain, differential privacy and feder-
ated learning to protect data privacy in IIoT. Feng et al. [175] VIII. C ONCLUSION
presented a privacy preservation method based on tucker Critical infrastructures, also known as national critical in-
decomposition on top of blockchain for IIoT. The authors frastructures in the United Kingdom, are becoming vulnerable
in [176] proposed using federated learning to train machine to cyberattacks due to wide adoption of Internet-connected
learning models locally, which can be finally aggregated into IoT/IIoT devices in industry 4.0. Security and scalability are
a global model while the data privacy can be preserved. therefore becoming burning concerns for this “modern” critical
Besides federated learning and cryptographic algorithms, infrastructures. In this paper, we introduced a layered architec-
the advent of recent machine learning and deep learning ture for IoT/IIoT critical infrastructures with the convergence
technologies can also potentially address the privacy concern. of blockchain and edge computing. The state-of-the-art of
For example, Alkadi et al. [179] presented a blockchain- security and privacy solutions, and scalability solutions, for
based framework with deep learning approaches to identify the IoT/IIoT infrastructures were reviewed and discussed. Despite
intrusion attacks while preserving data privacy. Moreover, the numerous efforts have done, there are still many on-going
work [180] introduced a privacy-aware deep learning method, challenges and open issues that need to be considered to ensure
which allows the collaboration of multiple nodes to train deep the success of critical infrastructures in era of industry 4.0.
neural networks while preserving data privacy. We then provided a range of potential research challenges and
open issues at the end of this paper to guide the future research
C. Scalable critical infrastructures in this area.
1) Scalability of IIoT: The scalability of IIoT is influenced
by the heterogeneity of IIoT devices and the diversity of R EFERENCES
IIoT networks. Recent studies have demonstrated the effec- [1] T. Wang, M. Z. A. Bhuiyan, G. Wang, M. A. Rahman, J. Wu, and
tiveness of the integration of SDN, NFV, network slicing and J. Cao, “Big data reduction for a smart city’s critical infrastructural
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 14

health monitoring,” IEEE Communications Magazine, vol. 56, no. 3, [23] S. Deng, H. Zhao, W. Fang, J. Yin, S. Dustdar, and A. Y. Zomaya,
pp. 128–133, 2018. “Edge intelligence: The confluence of edge computing and artificial
[2] M. Chen, Y. Jiang, N. Guizani, J. Zhou, G. Tao, J. Yin, and K. Hwang, intelligence,” IEEE Internet of Things Journal, pp. 1–1, 2020.
“Living with i-fabric: Smart living powered by intelligent fabric and [24] L. U. Khan, I. Yaqoob, N. H. Tran, S. M. A. Kazmi, T. N. Dang, and
deep analytics,” IEEE Network, pp. 1–8, 2020. C. S. Hong, “Edge computing enabled smart cities: A comprehensive
[3] A. Petropulu, K. I. Diamantaras, Z. Han, D. Niyato, and S. Zonouz, survey,” IEEE Internet of Things Journal, pp. 1–1, 2020.
“Contactless monitoring of critical infrastructure [from the guest ed- [25] J. Zhang, Y. Wu, G. Min, F. Hao, and L. Cui, “Balancing energy
itors],” IEEE Signal Processing Magazine, vol. 36, no. 2, pp. 19–21, consumption and reputation gain of uav scheduling in edge computing,”
2019. IEEE Transactions on Cognitive Communications and Networking, pp.
[4] L. Russell, R. Goubran, F. Kwamena, and F. Knoefel, “Agile iot 1–1, doi: 10.1109/TCCN.2020.3004592, 2020.
for critical infrastructure resilience: Cross-modal sensing as part of [26] H. Liu, P. Zhang, G. Pu, T. Yang, S. Maharjan, and Y. Zhang,
a situational awareness approach,” IEEE Internet of Things Journal, “Blockchain empowered cooperative authentication with data trace-
vol. 5, no. 6, pp. 4454–4465, 2018. ability in vehicular edge computing,” IEEE Transactions on Vehicular
[5] J. M. Mcginthy and A. J. Michaels, “Secure industrial internet of things Technology, pp. 1–1, 2020.
critical infrastructure node design,” IEEE Internet of Things Journal, [27] J. Jang-Jaccard and S. Nepal, “A survey of emerging threats in
vol. 6, no. 5, pp. 8021–8037, 2019. cybersecurity,” Journal of Computer and System Sciences, vol. 80,
[6] I. Stellios, P. Kotzanikolaou, M. Psarakis, C. Alcaraz, and J. Lopez, “A no. 5, pp. 973 – 993, 2014, special Issue on Dependable and
survey of iot-enabled cyberattacks: Assessing attack paths to critical Secure Computing. [Online]. Available: http://www.sciencedirect.com/
infrastructures and services,” IEEE Communications Surveys Tutorials, science/article/pii/S0022000014000178
vol. 20, no. 4, pp. 3453–3495, 2018. [28] D. Georgakopoulos, P. P. Jayaraman, M. Fazia, M. Villari, and
[7] Y. Wu, H. Huang, C.-X. Wang, and Y. Pan, 5G-enabled internet of R. Ranjan, “Internet of things and edge cloud computing roadmap for
things. CRC Press, 2019. manufacturing,” IEEE Cloud Computing, vol. 3, no. 4, pp. 66–73, 2016.
[8] Y. Wu, F. Hu, G. Min, and A. Zomaya, Big Data and Computational [29] H. Yang, A. Alphones, W. Zhong, C. Chen, and X. Xie, “Learning-
Intelligence in Networking. CRC Press, 2017. based energy-efficient resource management by heterogeneous rf/vlc
for ultra-reliable low-latency industrial iot networks,” IEEE Transac-
[9] R. Beerens, S. C. N. Thissen, W. C. M. Pancras, T. M. P. Gommans,
tions on Industrial Informatics, pp. 1–1, 2019.
N. van de Wouw, and W. P. M. H. Heemels, “Control allocation for an
industrial high-precision transportation and positioning system,” IEEE [30] D. Ma, G. Lan, M. Hassan, W. Hu, and S. K. Das, “Sensing, computing,
Transactions on Control Systems Technology, pp. 1–8, 2019. and communications for energy harvesting iots: A survey,” IEEE
Communications Surveys Tutorials, pp. 1–1, 2019.
[10] X. Jiang, Z. Pang, M. Luvisotto, R. Candell, D. Dzung, and C. Fis-
[31] C. K. Wu, K. F. Tsang, Y. Liu, H. Zhu, H. Wang, and Y. Wei,
chione, “Delay optimization for industrial wireless control systems
“Critical internet of things: An interworking solution to improve service
based on channel characterization,” IEEE Transactions on Industrial
reliability,” IEEE Communications Magazine, vol. 58, no. 1, pp. 74–79,
Informatics, pp. 1–1, 2020.
2020.
[11] F. Pan, Z. Pang, M. Luvisotto, M. Xiao, and H. Wen, “Physical-
[32] J. R. Foerster, X. Costa-Perez, and R. V. Prasad, “Communications for
layer security for industrial wireless control systems: Basics and future
iot: Connectivity and networking,” IEEE Internet of Things Magazine,
directions,” IEEE Industrial Electronics Magazine, vol. 12, no. 4, pp.
vol. 3, no. 1, pp. 6–7, 2020.
18–27, 2018.
[33] C. Luo, J. Ji, Q. Wang, X. Chen, and P. Li, “Channel state information
[12] R. Paes, D. C. Mazur, B. K. Venne, and J. Ostrzenski, “A guide to prediction for 5G wireless communications: A deep learning approach,”
securing industrial control networks: Integrating it and ot systems,” IEEE Trans. Netw. Sci. and Eng., vol. PP, no. 99, pp. 1–1, Jun. 2018.
IEEE Industry Applications Magazine, vol. 26, no. 2, pp. 47–53, 2020.
[34] Y. Zuo, Y. Wu, G. Min, C. Huang, and K. Pei, “An intelligent anomaly
[13] C. Shen, C. Liu, H. Tan, Z. Wang, D. Xu, and X. Su, “Hybrid- detection scheme for micro-services architectures with temporal and
augmented device fingerprinting for intrusion detection in industrial spatial data analysis,” IEEE Transactions on Cognitive Communications
control system networks,” IEEE Wireless Communications, vol. 25, and Networking, pp. 1–1, 2020.
no. 6, pp. 26–31, 2018. [35] C. Huang, Y. Wu, Y. Zuo, K. Pei, and G. Min, “Towards
[14] Y. Qin, Q. Zhang, C. Zhou, and N. Xiong, “A risk-based dynamic experienced anomaly detector through reinforcement learning,” in
decision-making approach for cybersecurity protection in industrial AAAI Conference on Artificial Intelligence, 2018. [Online]. Available:
control systems,” IEEE Transactions on Systems, Man, and Cybernet- https://www.aaai.org/ocs/index.php/AAAI/AAAI18/paper/view/16048
ics: Systems, pp. 1–8, 2018. [36] G. Falco, C. Caldera, and H. Shrobe, “Iiot cybersecurity risk modeling
[15] M. Hassan, A. Gumaei, S. Huda, and A. Almogren, “Increasing the for scada systems,” IEEE Internet of Things Journal, vol. 5, no. 6, pp.
trustworthiness in the industrial iot networks through a reliable cyber- 4486–4495, 2018.
attack detection model,” IEEE Transactions on Industrial Informatics, [37] Z. Zhou, J. Gong, Y. He, and Y. Zhang, “Software defined machine-to-
pp. 1–1, 2020. machine communication for smart energy management,” IEEE Com-
[16] M. Li, D. Hu, C. Lal, M. Conti, and Z. Zhang, “Blockchain-enabled munications Magazine, vol. 55, no. 10, pp. 52–60, 2017.
secure energy trading with verifiable fairness in industrial internet of [38] Y. Yang, Y. Chen, W. Wang, and G. Yang, “Securing channel state
things,” IEEE Transactions on Industrial Informatics, pp. 1–1, 2020. information in multiuser mimo with limited feedback,” IEEE Transac-
[17] Y. Zhang, Z. Zheng, H. . Dai, and D. Svetinovic, “Guest editorial: tions on Wireless Communications, pp. 1–1, 2020.
Special section on “blockchain for industrial internet of things” in ieee [39] S. R. Pokhrel, Y. Qu, and L. Gao, “Qos-aware personalized privacy with
transactions on industrial informatics,” IEEE Transactions on Industrial multipath tcp for industrial iot: Analysis and design,” IEEE Internet of
Informatics, vol. 15, no. 6, pp. 3514–3515, 2019. Things Journal, pp. 1–1, 2020.
[18] D. Puthal, N. Malik, S. P. Mohanty, E. Kougianos, and C. Yang, “The [40] T. Hoang, A. A. Yavuz, and J. Guajardo Merchan, “A secure searchable
blockchain as a decentralized security framework [future directions],” encryption framework for privacy-critical cloud storage services,” IEEE
IEEE Consumer Electronics Magazine, vol. 7, no. 2, pp. 18–21, 2018. Transactions on Services Computing, pp. 1–1, 2019.
[19] M. Zhaofeng, W. Lingyun, W. Xiaochang, W. Zhen, and Z. Weizhe, [41] C. Huang, G. Min, Y. Wu, Y. Ying, K. Pei, and Z. Xiang, “Time
“Blockchain-enabled decentralized trust management and secure usage series anomaly detection for trustworthy services in cloud computing
control of iot big data,” IEEE Internet of Things Journal, pp. 1–1, 2019. systems,” IEEE Transactions on Big Data, pp. 1–1, 2017.
[20] B. Yin, Y. Wu, T. Hu, J. Dong, and Z. Jiang, “An efficient collaboration [42] H.-N. Dai, Z. Zheng, and Y. Zhang, “Blockchain for internet of things:
and incentive mechanism for internet of vehicles (iov) with secured A survey,” IEEE Internet of Things Journal, vol. 6, no. 5, pp. 8076–
information exchange based on blockchains,” IEEE Internet of Things 8094, 2019.
Journal, vol. 7, no. 3, pp. 1582–1593, 2020. [43] J. Xie, H. Tang, T. Huang, F. R. Yu, R. Xie, J. Liu, and Y. Liu, “A survey
[21] Y. Wu, H.-N. Dai, H. Wang, and K.-K. R. Choo, “Blockchain-based of blockchain technology applied to smart cities: Research issues and
privacy preservation for 5g-enabled drone communications,” IEEE challenges,” IEEE Communications Surveys Tutorials, vol. 21, no. 3,
Network, 2020. pp. 2794–2830, 2019.
[22] D. Gabay, K. Akkaya, and M. Cebe, “Privacy-preserving authentication [44] M. Wu, K. Wang, X. Cai, S. Guo, M. Guo, and C. Rong, “A
scheme for connected electric vehicles using blockchain and zero comprehensive survey of blockchain: From theory to iot applications
knowledge proofs,” IEEE Transactions on Vehicular Technology, pp. and beyond,” IEEE Internet of Things Journal, vol. 6, no. 5, pp. 8114–
1–1, 2020. 8154, 2019.
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 15

[45] K.-K. R. Choo, Z. Yan, and W. Meng, “Editorial: Blockchain in indus- Situation Awareness of Smart Cities,” IEEE Network, vol. 34, no. 2,
trial iot applications: Security and privacy advances, challenges, and pp. 202–209, 2020.
opportunities,” IEEE Transactions on Industrial Informatics, vol. 16, [68] W. Viriyasitavat, L. D. Xu, Z. Bi, and D. Hoonsopon, “Blockchain
no. 6, pp. 4119–4121, 2020. technology for applications in internet of things—mapping from system
[46] L. Cheng, J. Liu, G. Xu, Z. Zhang, H. Wang, H. Dai, Y. Wu, and design perspective,” IEEE Internet of Things Journal, vol. 6, no. 5, pp.
W. Wang, “Sctsc: A semicentralized traffic signal control mode with 8155–8168, 2019.
attribute-based blockchain in iovs,” IEEE Transactions on Computa- [69] T. Kim, J. Ochoa, T. Faika, A. Mantooth, J. Di, Q. Li, and Y. Lee, “An
tional Social Systems, vol. 6, no. 6, pp. 1373–1385, 2019. overview of cyber-physical security of battery management systems
[47] H. Pourbabak, Q. H. Alsafasfeh, and W. Su, “A distributed consensus- and adoption of blockchain technology,” IEEE Journal of Emerging
based algorithm for optimal power flow in dc distribution grids,” IEEE and Selected Topics in Power Electronics, pp. 1–1, 2020.
Transactions on Power Systems, pp. 1–1, 2020. [70] S. Guo, X. Hu, S. Guo, X. Qiu, and F. Qi, “Blockchain meets edge
[48] Y. Xiao, N. Zhang, W. Lou, and Y. T. Hou, “A survey of distributed computing: A distributed and trusted authentication system,” IEEE
consensus protocols for blockchain networks,” IEEE Communications Transactions on Industrial Informatics, vol. 16, no. 3, pp. 1972–1983,
Surveys Tutorials, pp. 1–1, 2020. 2020.
[49] R. Zhang, R. Xue, and L. Liu, “Security and privacy on blockchain,” [71] G. Min, Y. Wu, and A. Y. Al-Dubai, “Performance modelling and
ACM Comput. Surv., vol. 52, no. 3, Jul. 2019. [Online]. Available: analysis of cognitive mesh networks,” IEEE Transactions on Commu-
https://doi.org/10.1145/3316481 nications, vol. 60, no. 6, pp. 1474–1478, 2012.
[50] J. Xie, F. R. Yu, T. Huang, R. Xie, J. Liu, and Y. Liu, “A survey on [72] Y. Wu, G. Min, and A. Y. Al-Dubai, “A new analytical model for
the scalability of blockchain systems,” IEEE Network, vol. 33, no. 5, multi-hop cognitive radio networks,” IEEE Transactions on Wireless
pp. 166–173, 2019. Communications, vol. 11, no. 5, pp. 1643–1648, 2012.
[51] T. Salman, M. Zolanvari, A. Erbad, R. Jain, and M. Samaka, “Security [73] M. B. H. Weiss, K. Werbach, D. C. Sicker, and C. E. C. Bastidas,
services using blockchains: A state of the art survey,” IEEE Commu- “On the application of blockchains to spectrum management,” IEEE
nications Surveys Tutorials, vol. 21, no. 1, pp. 858–880, 2019. Transactions on Cognitive Communications and Networking, vol. 5,
[52] G. Ramezan and C. S. Leung, “An analysis of proof-of-work based no. 2, pp. 193–205, 2019.
blockchains under an adaptive double-spend attack,” IEEE Transactions [74] Y.-C. Liang, Blockchain for Dynamic Spectrum Management. Singa-
on Industrial Informatics, pp. 1–1, 2020. pore: Springer Singapore, 2020, pp. 121–146.
[53] S. N. G. Gourisetti, M. Mylrea, and H. Patangia, “Evaluation and [75] Z. Zhou, X. Chen, Y. Zhang, and S. Mumtaz, “Blockchain-empowered
demonstration of blockchain applicability framework,” IEEE Transac- secure spectrum sharing for 5g heterogeneous networks,” IEEE Net-
tions on Engineering Management, pp. 1–15, 2019. work, vol. 34, no. 1, pp. 24–31, 2020.
[54] T. Distler, C. Cachin, and R. Kapitza, “Resource-efficient byzantine [76] A. Refaey, K. Hammad, S. Magierowski, and E. Hossain, “A
fault tolerance,” IEEE Transactions on Computers, vol. 65, no. 9, pp. Blockchain Policy and Charging Control Framework for Roaming in
2807–2819, 2016. Cellular Networks,” IEEE Network, vol. 34, no. 3, pp. 170–177, 2020.
[77] C. Li, Y. Fu, F. R. Yu, T. H. Luan, and Y. Zhang, “Vehicle Position Cor-
[55] S. Gao, T. Yu, J. Zhu, and W. Cai, “T-pbft: An eigentrust-based
rection: A Vehicular Blockchain Networks-Based GPS Error Sharing
practical byzantine fault tolerance consensus algorithm,” China Com-
Framework,” IEEE Transactions on Intelligent Transportation Systems,
munications, vol. 16, no. 12, pp. 111–123, 2019.
pp. 1–15, 2020.
[56] T. Liu, J. Ge, Y. Wu, B. Dai, L. Li, Z. Yao, J. Wen, and H. Shi, “A new
[78] G. Yu, X. Zha, X. Wang, W. Ni, K. Yu, P. Yu, J. A. Zhang, R. P. Liu,
bitcoin address association method using a two-level learner model,”
and Y. J. Guo, “Enabling attribute revocation for fine-grained access
in Algorithms and Architectures for Parallel Processing, S. Wen,
control in blockchain-iot systems,” IEEE Transactions on Engineering
A. Zomaya, and L. T. Yang, Eds. Cham: Springer International
Management, pp. 1–18, 2020.
Publishing, 2020, pp. 349–364.
[79] O. Novo, “Scalable Access Management in IoT Using Blockchain:
[57] W. Zou, D. Lo, P. S. Kochhar, X. D. Le, X. Xia, Y. Feng, Z. Chen, and
A Performance Evaluation,” IEEE Internet of Things Journal, vol. 6,
B. Xu, “Smart contract development: Challenges and opportunities,”
no. 3, pp. 4694–4701, 2019.
IEEE Transactions on Software Engineering, pp. 1–1, 2019.
[80] Y. Zhang, S. Kasahara, Y. Shen, X. Jiang, and J. Wan, “Smart contract-
[58] Y. Liu, F. R. Yu, X. Li, H. Ji, and V. C. M. Leung, “Blockchain and based access control for the internet of things,” IEEE Internet of Things
machine learning for communications and networking systems,” IEEE Journal, vol. 6, no. 2, pp. 1594–1605, 2019.
Communications Surveys Tutorials, pp. 1–1, 2020. [81] M. A. Islam and S. Madria, “A Permissioned Blockchain Based Access
[59] O. Osanaiye, S. Chen, Z. Yan, R. Lu, K. R. Choo, and M. Dlodlo, Control System for IOT,” in 2019 IEEE International Conference on
“From cloud to fog computing: A review and a conceptual live vm Blockchain (Blockchain), 2019, pp. 469–476.
migration framework,” IEEE Access, vol. 5, pp. 8284–8300, 2017. [82] Z. Yang, K. Yang, L. Lei, K. Zheng, and V. C. M. Leung, “Blockchain-
[60] Z. Zhao, G. Min, W. Gao, Y. Wu, H. Duan, and Q. Ni, “Deploying based decentralized trust management in vehicular networks,” IEEE
edge computing nodes for large-scale iot: A diversity aware approach,” Internet of Things Journal, vol. 6, no. 2, pp. 1495–1505, 2019.
IEEE Internet of Things Journal, vol. 5, no. 5, pp. 3606–3614, 2018. [83] A. A. Barakabitze, A. Ahmad, R. Mijumbi, and A. Hines, “5G
[61] L. Cui, D. Su, Y. Zhou, L. Zhang, Y. Wu, and S. Chen, “Edge network slicing using SDN and NFV: A survey of taxonomy,
learning for surveillance video uploading sharing in public transport architectures and future challenges,” Computer Networks, vol. 167,
systems,” IEEE Transactions on Intelligent Transportation Systems, p. 106984, 2020. [Online]. Available: http://www.sciencedirect.com/
doi: 10.1109/TITS.2020.3008420, 2020. science/article/pii/S1389128619304773
[62] M. Chen, Y. Miao, H. Gharavi, L. Hu, and I. Humar, “Intelligent traffic [84] W. Miao, G. Min, Y. Wu, H. Wang, and J. Hu, “Performance modelling
adaptive resource allocation for edge computing-based 5g networks,” and analysis of software-defined networking under bursty multimedia
IEEE Transactions on Cognitive Communications and Networking, traffic,” ACM Trans. Multimedia Comput. Commun. Appl., vol. 12,
vol. 6, no. 2, pp. 499–508, 2020. no. 5s, Sep. 2016. [Online]. Available: https://doi.org/10.1145/2983637
[63] S. Guo, Y. Dai, S. Guo, X. Qiu, and F. Qi, “Blockchain meets edge [85] G. Wang, Y. Zhao, J. Huang, and Y. Wu, “An effective approach to
computing stackelberg game and double auction based task offlfloading controller placement in software defined wide area networks,” IEEE
for mobile blockchain,” IEEE Transactions on Vehicular Technology, Transactions on Network and Service Management, vol. 15, no. 1, pp.
pp. 1–1, 2020. 344–355, 2018.
[64] B. Yin, H. Yin, Y. Wu, and Z. Jiang, “Fdc: A secure federated deep [86] X. Cheng, Y. Wu, G. Min, and A. Y. Zomaya, “Network function
learning mechanism for data collaborations in the internet of things,” virtualization in dynamic networks: A stochastic perspective,” IEEE
IEEE Internet of Things Journal, vol. 7, no. 7, pp. 6348–6359, 2020. Journal on Selected Areas in Communications, vol. 36, no. 10, pp.
[65] Y. Cheng, J. Zhang, L. Yang, C. Zhu, and H. Zhu, “Distributed 2218–2232, 2018.
green offloading and power optimization in virtualized small cell [87] W. Miao, G. Min, Y. Wu, H. Huang, Z. Zhao, H. Wang, and C. Luo,
networks with mobile edge computing,” IEEE Transactions on Green “Stochastic performance analysis of network function virtualization in
Communications and Networking, vol. 4, no. 1, pp. 69–82, 2020. future internet,” IEEE Journal on Selected Areas in Communications,
[66] D. Liu, Z. Yan, W. Ding, and M. Atiquzzaman, “A survey on secure vol. 37, no. 3, pp. 613–626, 2019.
data analytics in edge computing,” IEEE Internet of Things Journal, [88] X. Cheng, Y. Wu, G. Min, A. Y. Zomaya, and X. Fang, “Safeguard
vol. 6, no. 3, pp. 4946–4967, 2019. network slicing in 5g: A learning augmented optimization approach,”
[67] B. Hamdaoui, M. Alkalbani, T. Znati, and A. Rayes, “Unleashing the IEEE Journal on Selected Areas in Communications, vol. 38, no. 7,
Power of Participatory IoT with Blockchains for Increased Safety and pp. 1600–1613, 2020.
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 16

[89] H. Wang, Y. Wu, G. Min, J. Xu, and P. Tang, “Data-driven dynamic [109] Z. Xiong, Y. Zhang, N. C. Luong, D. Niyato, P. Wang, and N. Guizani,
resource scheduling for network slicing: A deep reinforcement learning “The Best of Both Worlds: A General Architecture for Data Man-
approach,” Information Sciences, vol. 498, pp. 106 – 116, 2019. agement in Blockchain-enabled Internet-of-Things,” IEEE Network,
[Online]. Available: http://www.sciencedirect.com/science/article/pii/ vol. 34, no. 1, pp. 166–173, 2020.
S0020025519303986 [110] C. Xu, K. Wang, P. Li, S. Guo, J. Luo, B. Ye, and M. Guo, “Making big
[90] C. Wang, Y. Zhang, X. Chen, K. Liang, and Z. Wang, “Sdn-based data open in edges: A resource-efficient blockchain-based approach,”
handover authentication scheme for mobile edge computing in cyber- IEEE Transactions on Parallel and Distributed Systems, vol. 30, no. 4,
physical systems,” IEEE Internet of Things Journal, vol. 6, no. 5, pp. pp. 870–882, 2019.
8692–8701, 2019. [111] Q. Xu, Z. Su, and Q. Yang, “Blockchain-based trustworthy edge
[91] J. Gao, K. O. O. Agyekum, E. B. Sifah, K. N. Acheampong, Q. Xia, caching scheme for mobile cyber-physical system,” IEEE Internet of
X. Du, M. Guizani, and H. Xia, “A Blockchain-SDN enabled Internet Things Journal, vol. 7, no. 2, pp. 1098–1110, 2020.
of Vehicles Environment for Fog Computing and 5G Networks,” IEEE [112] Y. Liu, F. R. Yu, X. Li, H. Ji, and V. C. M. Leung, “Decentralized
Internet of Things Journal, pp. 1–1, 2019. resource allocation for video transcoding and delivery in blockchain-
[92] G. A. F. Rebello, I. D. Alvarenga, I. J. Sanz, and O. C. M. B. Duarte, based system with mobile edge computing,” IEEE Transactions on
“BSec-NFVO: A Blockchain-Based Security for Network Function Vehicular Technology, vol. 68, no. 11, pp. 11 169–11 185, 2019.
Virtualization Orchestration,” in ICC 2019 - 2019 IEEE International [113] W. Liang, Y. Fan, K. Li, D. Zhang, and J. Gaudiot, “Secure data storage
Conference on Communications (ICC), May 2019, pp. 1–6. and recovery in industrial blockchain network environments,” IEEE
[93] M. F. Franco, E. J. Scheid, L. Z. Granville, and B. Stiller, “BRAIN: Transactions on Industrial Informatics, pp. 1–1, 2020.
blockchain-based reverse auction for infrastructure supply in virtual [114] J. Kang, Z. Xiong, D. Niyato, D. Ye, D. I. Kim, and J. Zhao,
network functions-as-a-service,” in 2019 IFIP Networking Conference “Toward Secure Blockchain-Enabled Internet of Vehicles: Optimizing
(IFIP Networking), 2019, pp. 1–9. Consensus Management Using Reputation and Contract Theory,” IEEE
[94] B. Nour, A. Ksentini, N. Herbaut, P. A. Frangoudis, and H. Moungla, Transactions on Vehicular Technology, vol. 68, no. 3, pp. 2906–2920,
“A Blockchain-Based Network Slice Broker for 5G Services,” IEEE 2019.
Networking Letters, vol. 1, no. 3, pp. 99–102, 2019. [115] Z. Chang, W. Guo, X. Guo, Z. Zhou, and T. Ristaniemi, “Incentive
[95] V. Ortega, F. Bouchmal, and J. F. Monserrat, “Trusted 5G Vehicular Mechanism for Edge Computing-based Blockchain,” IEEE Transac-
Networks: Blockchains and Content-Centric Networking,” IEEE Vehic- tions on Industrial Informatics, pp. 1–1, 2020.
ular Technology Magazine, vol. 13, no. 2, pp. 121–127, June 2018. [116] J. Kang, Z. Xiong, D. Niyato, P. Wang, D. Ye, and D. I. Kim, “Incen-
[96] J. Ge, S. Wang, Y. Wu, H. Tang, and Y. E, “Performance tivizing Consensus Propagation in Proof-of-Stake Based Consortium
improvement for source mobility in named data networking based on Blockchain Networks,” IEEE Wireless Communications Letters, vol. 8,
global–local fib updates,” Peer-to-Peer Networking and Applications, no. 1, pp. 157–160, 2019.
vol. 9, no. 4, pp. 670–680, Jul 2016. [Online]. Available: [117] Z. Li, Z. Yang, S. Xie, W. Chen, and K. Liu, “Credit-Based Payments
https://doi.org/10.1007/s12083-015-0353-z for Fast Computing Resource Trading in Edge-Assisted Internet of
[97] Y. Xu, C. Zhang, G. Wang, Z. Qin, and Q. Zeng, “A Blockchain- Things,” IEEE Internet of Things Journal, vol. 6, no. 4, pp. 6606–
enabled Deduplicatable Data Auditing Mechanism for Network Storage 6617, 2019.
Services,” IEEE Transactions on Emerging Topics in Computing, pp. [118] H.-N. Dai, R. C.-W. Wong, H. Wang, Z. Zheng, and A. V. Vasilakos,
1–1, 2020. “Big data analytics for large-scale wireless networks: Challenges and
opportunities,” ACM Computing Surveys (CSUR), vol. 52, no. 5, pp.
[98] G. S. Aujla, M. Singh, A. Bose, N. Kumar, G. Han, and R. Buyya,
1–36, 2019.
“BlockSDN: Blockchain-as-a-Service for Software Defined Network-
[119] Y. Liu, H.-N. Dai, Q. Wang, M. K. Shukla, and M. Imran,
ing in Smart City Applications,” IEEE Network, vol. 34, no. 2, pp.
“Unmanned aerial vehicle for internet of everything: Opportunities
83–91, 2020.
and challenges,” Computer Communications, vol. 155, pp. 66 –
[99] B. Wu, K. Xu, Q. Li, S. Ren, Z. Liu, and Z. Zhang, “Toward
83, 2020. [Online]. Available: http://www.sciencedirect.com/science/
Blockchain-Powered Trusted Collaborative Services for Edge-Centric
article/pii/S0140366419318754
Networks,” IEEE Network, vol. 34, no. 2, pp. 30–36, 2020.
[120] M. Raza, N. Aslam, H. Le-Minh, S. Hussain, Y. Cao, and N. M. Khan,
[100] H. Ko, J. Lee, and S. Pack, “Spatial and Temporal Computation “A critical analysis of research potential, challenges, and future direc-
Offloading Decision Algorithm in Edge Cloud-Enabled Heterogeneous tives in industrial wireless sensor networks,” IEEE Communications
Networks,” IEEE Access, vol. 6, pp. 18 920–18 932, 2018. Surveys Tutorials, vol. 20, no. 1, pp. 39–95, 2018.
[101] T. Taleb, K. Samdanis, B. Mada, H. Flinck, S. Dutta, and D. Sabella, [121] R. Rondón, A. Mahmood, S. Grimaldi, and M. Gidlund, “Under-
“On multi-access edge computing: A survey of the emerging 5g standing the performance of bluetooth mesh: Reliability, delay, and
network edge cloud architecture and orchestration,” IEEE Communi- scalability analysis,” IEEE Internet of Things Journal, vol. 7, no. 3,
cations Surveys & Tutorials, vol. 19, no. 3, pp. 1657–1681, 2017. pp. 2089–2101, 2020.
[102] R. Yang, F. R. Yu, P. Si, Z. Yang, and Y. Zhang, “Integrated blockchain [122] J. Haxhibeqiri, I. Moerman, and J. Hoebeke, “Low Overhead Schedul-
and edge computing systems: A survey, some research issues and ing of LoRa Transmissions for Improved Scalability,” IEEE Internet of
challenges,” IEEE Communications Surveys & Tutorials, vol. 21, no. 2, Things Journal, vol. 6, no. 2, pp. 3097–3109, 2019.
pp. 1508–1532, 2019. [123] H. Xie, Z. Yan, Z. Yao, and M. Atiquzzaman, “Data collection for
[103] Z. Xiong, J. Kang, D. Niyato, P. Wang, and H. V. Poor, “Cloud/edge security measurement in wireless sensor networks: A survey,” IEEE
computing service management in blockchain networks: Multi-leader Internet of Things Journal, vol. 6, no. 2, pp. 2205–2224, 2019.
multi-follower game-based admm for pricing,” IEEE Transactions on [124] M. M. Rabbani, J. Vliegen, J. Winderickx, M. Conti, and N. Mentens,
Services Computing, vol. 13, no. 2, pp. 356–367, 2020. “SHeLA: Scalable Heterogeneous Layered Attestation,” IEEE Internet
[104] W. Chen, Z. Zhang, Z. Hong, C. Chen, J. Wu, S. Maharjan, Z. Zheng, of Things Journal, vol. 6, no. 6, pp. 10 240–10 250, 2019.
and Y. Zhang, “Cooperative and distributed computation offloading for [125] M. A. Togou, D. A. Chekired, L. Khoukhi, and G. Muntean, “A
blockchain-empowered industrial internet of things,” IEEE Internet of hierarchical distributed control plane for path computation scalability in
Things Journal, vol. 6, no. 5, pp. 8433–8446, 2019. large scale software-defined networks,” IEEE Transactions on Network
[105] Y. Jiao, P. Wang, D. Niyato, and K. Suankaewmanee, “Auction mecha- and Service Management, vol. 16, no. 3, pp. 1019–1031, 2019.
nisms in cloud/fog computing resource allocation for public blockchain [126] F. P. Lin and Z. Tsai, “Hierarchical Edge-Cloud SDN Controller System
networks,” IEEE Transactions on Parallel and Distributed Systems, With Optimal Adaptive Resource Allocation for Load-Balancing,”
vol. 30, no. 9, pp. 1975–1989, 2019. IEEE Systems Journal, vol. 14, no. 1, pp. 265–276, 2020.
[106] S. Xie, Z. Zheng, W. Chen, J. Wu, H.-N. Dai, and M. Imran, [127] D. A. Chekired, M. A. Togou, L. Khoukhi, and A. Ksentini, “5G-
“Blockchain for cloud exchange: A survey,” Computers & Electrical Slicing-Enabled Scalable SDN Core Network: Toward an Ultra-Low
Engineering, vol. 81, p. 106526, 2020. Latency of Autonomous Driving Service,” IEEE Journal on Selected
[107] Q. Xu, Z. Su, M. Dai, and S. Yu, “APIS: Privacy-Preserving Incentive Areas in Communications, vol. 37, no. 8, pp. 1769–1782, 2019.
for Sensing Task Allocation in Cloud and Edge-Cooperation Mobile [128] R. Wen, G. Feng, J. Tang, T. Q. S. Quek, G. Wang, W. Tan, and
Internet of Things with SDN,” IEEE Internet of Things Journal, pp. S. Qin, “On robustness of network slicing for next-generation mobile
1–1, 2019. networks,” IEEE Transactions on Communications, vol. 67, no. 1, pp.
[108] J. Huang, L. Kong, H. Dai, W. Ding, L. Cheng, G. Chen, X. Jin, 430–444, 2019.
and P. Zeng, “Blockchain based mobile crowd sensing in industrial [129] J. Pan, J. Wang, A. Hester, I. Alqerm, Y. Liu, and Y. Zhao, “EdgeChain:
systems,” IEEE Transactions on Industrial Informatics, pp. 1–1, 2020. An Edge-IoT Framework and Prototype Based on Blockchain and
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 17

Smart Contracts,” IEEE Internet of Things Journal, vol. 6, no. 3, pp. blockchain networks,” IEEE Internet of Things Journal, vol. 6, no. 3,
4719–4732, June 2019. pp. 4585–4600, 2018.
[130] Y. Liu, K. Wang, K. Qian, M. Du, and S. Guo, “Tornado: En- [152] H. Yao, T. Mai, J. Wang, Z. Ji, C. Jiang, and Y. Qian, “Resource trading
abling blockchain in heterogeneous internet of things through a space- in blockchain-based industrial internet of things,” IEEE Transactions
structured approach,” IEEE Internet of Things Journal, vol. 7, no. 2, on Industrial Informatics, vol. 15, no. 6, pp. 3602–3609, 2019.
pp. 1273–1286, 2020. [153] Z. Xiong, Y. Zhang, D. Niyato, P. Wang, and Z. Han, “When mobile
[131] S. Biswas, K. Sharif, F. Li, B. Nour, and Y. Wang, “A Scalable blockchain meets edge computing,” IEEE Communications Magazine,
Blockchain Framework for Secure Transactions in IoT,” IEEE Internet vol. 56, no. 8, pp. 33–39, 2018.
of Things Journal, vol. 6, no. 3, pp. 4650–4659, 2019. [154] L. Wu, L. Li, X. Li, Y. Yu, L. Zhang, M. Pan, and Z. Han, “Resource
[132] K. Croman, C. Decker, I. Eyal, A. E. Gencer, A. Juels, A. Kosba, allocation in blockchain system based on mobile edge computing
A. Miller, P. Saxena, E. Shi, E. G. Sirer et al., “On scaling decentralized networks,” in 2019 11th International Conference on Wireless Com-
blockchains,” in International conference on financial cryptography munications and Signal Processing (WCSP). IEEE, 2019, pp. 1–6.
and data security. Springer, 2016, pp. 106–125. [155] C. Luo, S. Guo, S. Guo, L. T. Yang, G. Min, and X. Xie, “Green
[133] T. T. A. Dinh, R. Liu, M. Zhang, G. Chen, B. C. Ooi, and J. Wang, “Un- communication in energy renewable wireless mesh networks: Routing,
tangling blockchain: A data processing view of blockchain systems,” rate control, and power allocation,” IEEE Trans. Paral. and Distr. Syst.,
IEEE Transactions on Knowledge and Data Engineering, vol. 30, no. 7, vol. 25, no. 12, pp. 3211–3220, Dec. 2014.
pp. 1366–1385, 2018. [156] C. Luo, G. Min, F. R. Yu, Y. Zhang, L. T. Yang, and V. C. M. Leung,
[134] S. Popov, “The tangle, IOTA whitepaper,” IOTA, Tech. Rep.[Online]. “Joint relay scheduling, channel access, and power allocation for green
Available: https://iota. org/IOTA Whitepaper.pdf, Tech. Rep., 2018. cognitive radio communications,” IEEE J. Sel. Areas in Comm., vol. 33,
[135] A. Cullen, P. Ferraro, C. King, and R. Shorten, “On the resilience no. 5, pp. 922–932, May 2015.
of dag-based distributed ledgers in iot applications,” IEEE Internet of [157] G. Xu, W. Wang, L. Jiao, X. Li, K. Liang, X. Zheng, W. Lian, H. Xian,
Things Journal, 2020. and H. Gao, “SoProtector: Safeguard Privacy for Native SO Files in
[136] Á. Garcı́a-Pérez and M. A. Schett, “Deconstructing stellar consensus,” Evolving Mobile IoT Applications,” IEEE Internet of Things Journal,
in 23rd International Conference on Principles of Distributed Systems vol. 7, no. 4, pp. 2539–2552, 2020.
(OPODIS 2019), 2020. [158] Z. Zheng, S. Xie, H.-N. Dai, W. Chen, X. Chen, J. Weng, and M. Imran,
[137] F. Armknecht, G. O. Karame, A. Mandal, F. Youssef, and E. Zenner, “An overview on smart contracts: Challenges, advances and platforms,”
“Ripple: Overview and outlook,” in International Conference on Trust Future Generation Computer Systems, vol. 105, pp. 475–491, 2020.
and Trustworthy Computing. Springer, 2015, pp. 163–180. [159] S. He, W. Ren, T. Zhu, and K. R. Choo, “BoSMoS: A Blockchain-
[138] M. J. Amiri, D. Agrawal, and A. E. Abbadi, “Caper: a cross-application Based Status Monitoring System for Defending Against Unauthorized
permissioned blockchain,” Proceedings of the VLDB Endowment, Software Updating in Industrial Internet of Things,” IEEE Internet of
vol. 12, no. 11, pp. 1385–1398, 2019. Things Journal, vol. 7, no. 2, pp. 948–959, Feb. 2020.
[139] M. Herlihy, “Atomic cross-chain swaps,” in Proceedings of the 2018 [160] T. Chen, X. Li, Y. Wang, J. Chen, Z. Li, X. Luo, M. H. Au,
ACM symposium on principles of distributed computing (PODC 2018), and X. Zhang, “An adaptive gas cost mechanism for ethereum to
2018, pp. 245–254. defend against under-priced dos attacks,” in International Conference
[140] A. Back, M. Corallo, L. Dashjr, M. Friedenbach, G. Maxwell, on Information Security Practice and Experience. Springer, 2017, pp.
A. Miller, A. Poelstra, J. Timón, and P. Wuille, “Enabling 3–24.
blockchain innovations with pegged sidechains,” URL: http:// www. [161] Z. Huang, J. Huang, and T. Zang, “Leopard: Understanding the Threat
opensciencereview. com/ papers/ 123/ enablingblockchain-innovations- of Blockchain Domain Name Based Malware,” in Passive and Active
with-pegged-sidechains, vol. 72, 2014. Measurement, A. Sperotto, A. Dainotti, and B. Stiller, Eds. Cham:
[141] G. Wood, “Polkadot: Vision for a heterogeneous multi-chain Springer International Publishing, 2020, pp. 55–70.
framework,” [Online]. Available: https://www.win.tue.nl/∼mholende/ [162] T. Chen et al., “SODA: A Generic Online Detection Framework for
seminar/references/ethereum polkadot.pdf, Tech. Rep., 2016. Smart Contracts,” in The Network and Distributed System Security
[142] S. Li, H. Xiao, H. Wang, T. Wang, J. Qiao, and S. Liu, “Blockchain Symposium (NDSS), 2020, pp. 1–17.
dividing based on node community clustering in intelligent manufac- [163] M. Chen, Y. Cao, R. Wang, Y. Li, D. Wu, and Z. Liu, “Deepfocus:
turing cps,” in 2019 IEEE International Conference on Blockchain Deep encoding brainwaves and emotions with multi-scenario behavior
(Blockchain 2019). IEEE, 2019, pp. 124–131. analytics for human attention enhancement,” IEEE Network, vol. 33,
[143] J. Hellings and M. Sadoghi, “Coordination-free byzantine replication no. 6, pp. 70–77, 2019.
with minimal communication costs,” in 23rd International Conference [164] X. Wang, L. T. Yang, Y. Wang, L. Ren, and M. J. Deen, “Adtt: A
on Database Theory (ICDT 2020), 2020. highly-efficient distributed tensor-train decomposition method for iiot
[144] H. Dang, T. T. A. Dinh, D. Loghin, E.-C. Chang, Q. Lin, and B. C. Ooi, big data,” IEEE Transactions on Industrial Informatics, pp. 1–1, DOI:
“Towards scaling blockchain systems via sharding,” in Proceedings of 10.1109/TII.2020.2967768, 2020.
the 2019 International Conference on Management of Data, 2019, pp. [165] R. Wang, M. Chen, N. Guizani, Y. Li, H. Gharavi, and K. Hwang,
123–140. “Deepnetqoe: Self-adaptive qoe optimization framework of deep net-
[145] A. Sharma, F. M. Schuhknecht, D. Agrawal, and J. Dittrich, “Blurring works,” IEEE Network, 2020.
the lines between blockchains and database systems: the case of hy- [166] J. Kang, Z. Xiong, D. Niyato, S. Xie, and J. Zhang, “Incentive Mecha-
perledger fabric,” in Proceedings of the 2019 International Conference nism for Reliable Federated Learning: A Joint Optimization Approach
on Management of Data, 2019, pp. 105–122. to Combining Reputation and Contract Theory,” IEEE Internet of
[146] L. Lao, Z. Li, S. Hou, B. Xiao, S. Guo, and Y. Yang, “A survey of iot Things Journal, vol. 6, no. 6, pp. 10 700–10 714, 2019.
applications in blockchain systems: Architecture, consensus, and traffic [167] Z. Zhou, X. Chen, E. Li, L. Zeng, K. Luo, and J. Zhang, “Edge
modeling,” ACM Computing Surveys (CSUR), vol. 53, no. 1, pp. 1–32, Intelligence: Paving the Last Mile of Artificial Intelligence With Edge
2020. Computing,” Proceedings of the IEEE, vol. 107, no. 8, pp. 1738–1762,
[147] F. M. Benčić and I. P. Žarko, “Distributed ledger technology: 2019.
Blockchain compared to directed acyclic graph,” in 2018 IEEE 38th [168] K. Zhang, Y. Zhu, S. Maharjan, and Y. Zhang, “Edge Intelligence
International Conference on Distributed Computing Systems (ICDCS). and Blockchain Empowered 5G Beyond for the Industrial Internet of
IEEE, 2018, pp. 1569–1570. Things,” IEEE Network, vol. 33, no. 5, pp. 12–19, 2019.
[148] M. Castro, B. Liskov et al., “Practical byzantine fault tolerance,” in [169] Z. Yan, J. Ge, Y. Wu, L. Li, and T. Li, “Automatic virtual network em-
OSDI, vol. 99, no. 1999, 1999, pp. 173–186. bedding: A deep reinforcement learning approach with graph convolu-
[149] Y. Sun, L. Zhang, G. Feng, B. Yang, B. Cao, and M. A. Imran, tional networks,” IEEE Journal on Selected Areas in Communications,
“Blockchain-enabled wireless internet of things: Performance analysis vol. 38, no. 6, pp. 1040–1057, 2020.
and optimal communication node deployment,” IEEE Internet of Things [170] J. Luo, Q. Chen, F. R. Yu, and L. Tang, “Blockchain-enabled software-
Journal, vol. 6, no. 3, pp. 5791–5802, 2019. defined industrial internet of things with deep reinforcement learning,”
[150] Y. Liu, K. Wang, K. Qian, M. Du, and S. Guo, “Tornado: En- IEEE Internet of Things Journal, pp. 1–1, 2020.
abling blockchain in heterogeneous internet of things through a space- [171] Z. Xiong, Y. Zhang, D. Niyato, R. Deng, P. Wang, and L. Wang, “Deep
structured approach,” IEEE Internet of Things Journal, 2019. Reinforcement Learning for Mobile 5G and Beyond: Fundamentals,
[151] Z. Xiong, S. Feng, W. Wang, D. Niyato, P. Wang, and Applications, and Challenges,” IEEE Vehicular Technology Magazine,
Z. Han, “Cloud/fog computing resource management and pricing for vol. 14, no. 2, pp. 44–52, 2019.
JOURNAL OF LATEX CLASS FILES, VOL. XX, NO. X, AUGUST 20XX 18

[172] Y. Chen, Y. Zhang, S. Maharjan, M. Alam, and T. Wu, “Deep Learning Hao Wang [Member, IEEE] is an Associate Pro-
for Secure Mobile Edge Computing in Cyber-Physical Transportation fessor in the Department of Computer Science in
Systems,” IEEE Network, vol. 33, no. 4, pp. 36–41, 2019. Norwegian University of Science and Technology,
[173] Y. Chen, Z. Ma, Q. Wang, J. Huang, X. Tian, and Q. Zhang, “Privacy- Norway. He has a Ph.D. degree (2006) and a
preserving spectrum auction design: Challenges, solutions, and research B.Eng. degree (2000), both in computer science
directions,” IEEE Wireless Communications, vol. 26, no. 5, pp. 142– and engineering, from South China University of
150, 2019. Technology, China. His research interests include
[174] P. C. Mahawaga Arachchige, P. Bertok, I. Khalil, D. Liu, S. Camtepe, big data analytics, industrial internet of things, high
and M. Atiquzzaman, “A trustworthy privacy preserving framework performance computing, and safety-critical systems.
for machine learning in industrial iot systems,” IEEE Transactions on He served as a TPC co-chair for IEEE DataCom
Industrial Informatics, pp. 1–1, 2020. 2015, IEEE CIT 2017, ES 2017, and IEEE CPSCom
[175] J. Feng, L. T. Yang, R. Zhang, and B. S. Gavuna, “Privacy Preserving 2020, and a senior TPC member for CIKM 2019. He is the Chair for Sub-TC
Tucker Train Decomposition over Blockchain-Based Encrypted Indus- on Healthcare of IEEE Industrial Electronics Society Technical Committee on
trial IoT Data,” IEEE Transactions on Industrial Informatics, pp. 1–1, Industrial Informatics.
2020.
[176] Y. Zhan, P. Li, Z. Qu, D. Zeng, and S. Guo, “A Learning-based
Incentive Mechanism for Federated Learning,” IEEE Internet of Things
Journal, pp. 1–1, 2020.
[177] J. Kang, Z. Xiong, D. Niyato, Y. Zou, Y. Zhang, and M. Guizani,
“Reliable federated learning for mobile networks,” IEEE Wireless
Communications, vol. 27, no. 2, pp. 72–80, 2020.
[178] M. Chen and Y. Hao, “Label-less learning for emotion cognition,” IEEE
Transactions on Neural Networks and Learning Systems, pp. 1–11,
2019.
[179] O. Alkadi, N. Moustafa, B. Turnbull, and K. R. Choo, “A Deep
Blockchain Framework-enabled Collaborative Intrusion Detection for
Protecting IoT and Cloud Networks,” IEEE Internet of Things Journal,
pp. 1–1, 2020.
[180] X. Liu, H. Li, G. Xu, S. Liu, Z. Liu, and R. Lu, “PADL: Privacy-aware
and Asynchronous Deep Learning for IoT Applications,” IEEE Internet
of Things Journal, pp. 1–1, 2020.
[181] F. Estrada-Solano, O. M. Caicedo, and N. L. S. Da Fonseca, “NELLY:
Flow Detection Using Incremental Learning at the Server Side of SDN-
Based Data Centers,” IEEE Transactions on Industrial Informatics,
vol. 16, no. 2, pp. 1362–1372, 2020.

Yulei Wu [Senior Member, IEEE] is a Senior Lec-

turer with the Department of Computer Science,
College of Engineering, Mathematics and Physical
Sciences, University of Exeter, United Kingdom.
He received the B.Sc. degree (First Class Hon-
ours) in Computer Science and the Ph.D. degree
in Computing and Mathematics from the Univer-
sity of Bradford, United Kingdom, in 2006 and
2010, respectively. His expertise is on intelligent
networking, and his main research interests include
computer networks, networked systems, software
defined networks and systems, network management, and network security
and privacy. His research has been supported by Engineering and Physical
Sciences Research Council of United Kingdom, National Natural Science
Foundation of China, University’s Innovation Platform and industry. He
is an Editor of IEEE Transactions on Network and Service Management,
IEEE Transactions on Network Science and Engineering, Computer Networks
(Elsevier) and IEEE ACCESS. He is a Fellow of the HEA (Higher Education
Academy).

Hong-Ning Dai [Senior Member, IEEE] is cur-

rently with Faculty of Information Technology at
Macau University of Science and Technology as
an associate professor. He obtained the Ph.D. de-
gree in Computer Science and Engineering from
Department of Computer Science and Engineering
at the Chinese University of Hong Kong. His cur-
rent research interests include Internet of Things,
blockchain, and big data analytics. He has served as
editors for Ad Hoc Networks (Elsevier), Connection
Science (Taylor & Francis), and IEEE Access, guest
editors for IEEE Transactions on Industrial Informatics, IEEE Transactions
on Emerging Topics in Computing, and IEEE Open Journal of the Computer
Society.