Lightweight fog based solution for

privacy-preserving in IoT using blockchain

1st Ali Haleem Alkhazaali 2nd Oğuz ATA

Electrical and Computer Engineering Department Information Technology Depatment
Altinbas university Altinbas university
Istanbul,Turkey Istanbul, Turkey
https://orcid.org/0000-0001-7992-0288 https://orcid.org/0000-0003-4511-7694

Abstract—Internet of things (IoT) mainly depends on clouds

to process and store their data. Clouds cannot handle the volume
and velocity of data generated by IoT. IoT is delay-sensitive and
resources limited. Fog computing proposed endorsing the internet
of things (IoT) demands. Fog computing extends the cloud com-
puting service to the edge of the network. Fog utilization reduces
response time and network overhead while maintaining security
aspects. isolation and operating system (OS) dependency achieved
by using virtualization. Blockchain proposed to solve the security
and privacy of fog computing. Blockchain is a decentralized,
immutable ledger. fog computing with blockchain proposed as
an IoT infrastructure. Fog computing adopted with lightweight
blockchain in this proposed work. This adaptation endorses the
IoT demands for low response time with limited resources. This
paper explores system applicability. Varies from other papers that Fig. 1: investment in IoT worldwide in billion dollars [18].
focus on one factor such as privacy or security—applicability of
the proposed model achieved by concentration different IoT needs
and limits. Response time and ram usage with 1000 transactions
did not encroach 100s and 300MiB in the proposed model. resources. Adaptation of IoT with Cloud did not work well as
Index Terms—Lightweight Blockchain, IoT, Fog computing, it is with these applications. Looking closer to the IoT specific
RAFT consensus needs, we can realize that two-layer IoT -Cloud paradigm
can hardly handle communication, data analysis, and decision-
I. I NTRODUCTION making demands. Cooperation of IoT and Cloud in such a
The internet of things (IoT) inspired our life by their poten- model will confront limitations in scalability, latency, and
tial application changing things into smart devices. IoT now overhead network issues. These issues led by the massive
days used in wide range of application such as smart homes volume, velocity, and variety of data IoT produce that requires
[1], smart cities [2] [3], smart transport [4] [5] , defense and transferring, storage, process, and decision-making. Moving
public security [6] [7], wearable devices [8], e-health [9] [10], this volume of data from IoT to the Cloud will be ineffective
augmented reality [11] [12], industry [13] [14], public service or impossible. Furthermore, time-sensitive and critical privacy
[15] and smart grid [16]. The IoT attributes lead to rapidly applications will suffer because of the distance between Cloud
relay and growth in IoT devices. IDC estimate that around and IoT. A timely response is critical in many applications,
42 billion devices will be connected to the internet generating such as in fire alarm sensors and patient monitoring systems
approximately 80 zettabytes of data [17]. IoT market size in [19] disclosure of privacy is a significant concern too.
billion dollar [18]shown in Figure (1) present that there is The aforementioned issues have stepped up the demand for
more relay and investment on IoT sector in the next years. new paradigms that move the computing power and decision-
Traditionally Clouds used for their high computing power, making closer to the IoT [20]. Edge computing introduced to
a wide range of provided services, and its high availability. overcome the above issues [21]. One of the edge computing
Many sectors like business, industry, education, and gov- is Fog computing. Fog computing, when compared with other
ernment, rely on Cloud, such as Azure, Google, and IBM. edge computing paradigms, is considered the best solution
These sectors benefit from clouds, on-demand services, and since it fits many applicable cases and is considered a more
general method. One of the earliest and important definitions
978-1-7281-9352-6/20/$31.00 ©2020 IEEE. of Fog computing can be found in open Fog, which adopted

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
 consensus algorithms. Lightweight consensus algorithms lead
to reduce response time and efficient resource use.
In this work, the IoT needs are considered for low latency,
faster response time, and low energy consumption while
maintaining security and privacy aspects. The response time
as a direct indicator of the quality of Experience (QoE)
is measured. Blockchain and transport layer secure (TLS)
protocol adopted to ensure mutual authentication, integrity, and
privacy. The proposed system efficiently utilizes fog and IoT
resources.
Fig. 2: IoT, Fog and Cloud layers. The rest of this paper is arranged as follows in section
II related work done in the field of blockchain and Fog
computing for IoT. Section III motivation then section IV
by IEEE standards [22]. Fog computing gain recognition and background. The Propose system is presented in section V
interest form academic and industry field [22] [23]. Archi- and, finally, the performance analysis in section VI.
tectures, algorithms, network application, configuration, and
feature of Fog computing discussed in [22] [23] [24] [25] from
many sides and show that Fog computing is suitable for many
applications. The fog computing paradigm consists mainly of
three layers: IoT, Fog, and Cloud, as shown in Figure (2), Fog
work as a bridge between IoT and Cloud.
Cooperate Cloud with Fog to the best benefits of their
unique characteristics. Cloud has tremendous computing re-
sources and power consume, Fog, on the other hand, has
moderated computing resources, and low power consumes
[26]. Fog and Cloud cannot replace each other. Fog nodes
can be used to preprocess data generated from IoT devices
while sending an exhausting query to the Cloud. [27]. Fog
computing now being used in many applications [23] shown
in Figure (3).
Cloud computing provides security in a centralized manner
while in Fog computing, the security issues must be handled
locally, such as lack of encryption and man-in-middle attack
[28]. Decentralized paradigms that use peer-to-peer network-
ing increase the benefits of Fog computing. IoT heterogeneity
and dynamicity nature must be considered. For the previous
reason, the adaptation of blockchain with Fog computing can
bring security, accountability, traceability, and encryption to Fig. 3: Fog computing applications.
the network and will let cross-vendor and platform connection
faceable.
II. R ELATED WORK
Blockchain is a distributed database technology that elim-
inates the central system and third party while dealing with There are many works done to address and overcome chal-
privacy and security challenges. Blockchain consists of a time- lenges facing IoT, such as the adaptation of Blockchain, Fog
stamped approved block string. Block string connected using computing, edge computing, and Cloud computing Individu-
hash function techniques. Every node in the blockchain net- ally or collectively with IoT models. [31], proposed human
work maintains the same copy of the ledger. Figure (4) show activity recognition base on Fog and Cloud using blockchain
the blockchain block structure. The first stable blockchain framework, improving video classification for monitoring and
application is introduced by bitcoin [29] and then smart recognition of human activity in e-health care. [32], present
contract feature is added by Ethereum [30] , which lead to a trust management model to improve privacy and security
speared blockchain out of the finical sector in an endless for the cyber-physical system by utilizing a smart contract.
use case. Nowadays, blockchain is not only a cryptocurrency. [33], designed Fog-IoT energy-aware scheme showing better
Blockchain application inspired many sectors, even banks performance in terms of energy; They consider the quality
seeking for adaptation of blockchain in their work. The third of service (QoS) and employed Blockchain to establish the
generation of blockchain is consortium blockchain, which al- system. [34], proposed a model that uses IoT, AI, ML, Fog
lows the connection between entities that know each other but computing, Cloud computing, and Blockchain as a healthcare
do not trust each other. This type of blockchain use lightweight solution with high QoS, scalability, low response time, and

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
robustness patient-user need. [35], proposed layered architec- enhances the adaptation of the proposed system in real-world
ture of Fog and Cloud to process the internet of underwater applications.
things (IOUT) data securely by using Blockchain. they reached
an efficient and secure model but with some computational
overhead. [36] , they suggested ensuring the authenticity of
education certification using Blockchain. In [37] , they use
Blockchain for copyright information management, learning
outcomes verification benefiting from transparency and non-
tampering feature of it. In [38], their work focused on devel-
oping IoT architecture for big data analysis supported by AI
and Blockchain.
In [39] , efficient energy trading based on Blockchain, as it
is an essential scope inindustrial IoT (IIoT) discussed.Authors
in [40] proposed a time synchronization base on Blockchain
to reduce attacks on IoT devices. [41] they propose a three- Fig. 4: Blockchain block structure.
layer model consisting of IoT, Fog node, and Cloud layers.
To ensure availability and reliability, the employed software-
III. M OTIVATION
defined network (SDN) to manage and control Fog nodes.
In [42] , they propose a lightweight Blockchain named LSB IoT resources limited; therefore, they rely on an external
to achieve scalability and provide security and privacy. In source to process and store their data. IoT expected to gen-
LSB digital signature, public-key cryptography and one-way erate enormous data that need transfer and process. Clouds
hashing used. However, IoT managed in a centralized manner used widely for this purpose with a two-layer paradigm.
by connecting to one node named block manager, serving as a This paradigm suffers from high response time and network
hub providing key sharing between home devices; this leads to congestion. The motivation behind this work summarized as
a single point of failure. They did not show how the local block below:
manager will be synchronized with overlay block managers 1) 1) Utilizing Fog computing allows bringing the re-
who need more experiments to address issues in scalability sources nearest the network edge; this reduces response
and storage. Overall, their architecture fits specific domains time and network congestion.
rather than being a generalized model. 2) Lightweight Blockchain used to add security, account-
In [43] they use Blockchain and changeable public keys to ability, and traceability.
design secure infrastructure to vehicle. Their work has a single 3) Using Blockchain trims the needs of a third party.
point of failure, high overhead, and high latency. 4) Small-size and low-cost devices help for broader sepa-
The study in [44] focused on employing Blockchain and ration of Fog computing nodes like a single-board com-
smart contracts to supervision access to data generated by puter (SBC). Efficient resource usage by the proposed
IoT devices. Ethereum blockchain platform utilization leads model helps to fit the SBC requirement.
to high overhead and latency. 5) Overcoming the heterogeneity in IoT by employing vir-
tualization to benefit from its specification like isolation,
Blockchain and homomorphic used in [45] to process data
cross-vendor, and operating system dependency.
without disclosing them to preserve privacy. However, 10s as
block intervals considered high, and this is not suitable for IV. BACKGROUND
real-time IoT applications, which Limiting its efficiency. IoT inspired our daily life transfer things to interact with
Authors proposed [46] using Blockchain and certificate less the surrounding environment by sensing, actuating, and com-
scheme to build distributed storage for IoT data. They did not municating with humans and other things. IoT applications
show how to manage the privacy of saved data and Suffers spread in a wide range of use cases. Data generated by
from high computation needs. IoT devices will grow in tremendous velocity and volume.
In [47] they propose dynamic mining strategy were several IoT topology can be classified into two types things centric
nodes in mine mode and other loading data. This work is and cloud-centric [48]. Things centric focus on enhancing
specific problems focused rather than generalized problems user experience [49]. Cloud-centric focus on service and data
since Fog-IoT faces many issues that need to be categorized processing in IoT [50] [51].most of IoT application relay
and solve. on cloud-centric benefiting from cloud computational and
The above works are domain-specific or need high compu- virtualization capability [52] . IoT benefits from clouds [20]
tation. They mostly present limited parameters of the system with efficient and orchestrated use of resources planned service
used, for example, considering only response time or security deployment and anywhere network access. With its capability,
but ignoring network and computation overhead. Applicability this system introduces some limitations, like High latency,
of the system requires the attention of many aspects. Efficient network traffic, and bandwidth consumption, which affect
resource utilization while maintaining security and privacy accurateness decision-making. In IoT, it is mandatory to fulfill

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
low latency since many applications are time-sensitive; on the Consensus algorithms attempt to solve the Byzantine Generals
other hand, cloud need rapidly upgrade to handle the huge Problem [70]. Different blockchain uses different consensus
traffic generated by IoT, which is cost expensive [53]. The algorithm. Each algorithm has its specification table (III)
need arises to bring computing power nearest to IoT devices. present some consensus protocols properties.
Fog computing [54] proposed by cisco as an extension of cloud For this work Hyperledger fabric [76] [77] consortium
[55] moving the cloud feature to the network edge. Fog adds blockchain used with RAFT [78] [73] consensus protocol.
a middle layer between IoT and cloud. This layer benefit IoT These choose made for the following characteristic:
[54] [55] [56] in location awareness, geographical distributed, 1) IoT Participants know each other but do not trust each
resource heterogeneity and more. The comparison of cloud other, so lightweight consortium blockchain used to
and fog computing can be found in table (I). Overall, fog establish a secure peer-to-peer(P2P) network without the
node response time is lower than cloud, and short-duration data need for a third party.
storage in fog increases the decision-making accuracy. security 2) RAFT’s high scalability and throughput with efficient
and privacy in cloud-managed in a centralized manner. Privacy resource utilization are advisable for IoT. RAFT can
and security must be managed locally in fog computing, which tolerate up to 50% of nodes crash fault [73]. For
is an issue that needs to be solved. consortium blockchains, nodes are verified, members.
TABLE I: Comparison of fog and cloud specification. Hence, it is more important to solve crash faults than
Byzantine faults [73].
Feature Cloud computing Fog computing 3) Hyperledger fabric use channels that add a high level of
Architecture Centralize Decentralized
Latency and response time High Low privacy.
Cost of deployment High Low
Geographic coverage Global Local V. P ROPOSED SYSTEM
Fault Tolerant Sensitive The system model component and roles are explained in
Scalability Average High
Data storage Long term Short duration section (A) system model. Workflow and the relation between
component explained in section (B) System description. Fi-
Blockchain proposed to solve privacy and security issues nally, the performance is shown and discuss how every metric
[57] [58].Blockchain, defined as a peer-to-peer(P2P) system, impact IoT requirement rhythm.
has an immutable distributed ledger eliminating the need for Docker containers employed to virtualize OP, FCN, and
a third party that can be used to solve security and privacy in RN. Docker containers employed for their efficient power
fog. Blockchain use cryptography [59] methods to secure their consumption [79], and the beneficial virtualization specifi-
connection and link blocks using their hash figure (4) show cation such as isolation, cross-vendor, and operating system
how hash used to establish such a link. More about blockchain dependency.
and cryptography can be found in [60] [61] [62] [63] [64] A. System model
[65]. blockchain can be classified upon their participation
policy, data transparent and architecture into three class public The proposed system is shown in figure (5) consisting of
blockchain, private blockchain, and consortium blockchain. sections. There are six system components.
Comparative between blockchain types shown in table (II). In 1) Registration Node (RN): authority that is trusted by all
public blockchain [66] anyone can join the network and partic- nodes. RN register Fog computing nodes (FCN), IoT
ipate in the mining process such as bitcoin [29] and Ethereum devices, Organizer Peer (OP), and track illegal behav-
[30]. In private blockchain [67] only invited entities can join iors. Registered entity provided with digital certification
the permissioned network. All the entities participate in the and private, public keys pair. The registration is done
mining process. Consortium blockchain [68] is permissioned through a secure channel.
and only selected nodes engage in the mining process. A 2) Fog Computing Nodes (FCN): full nodes where ledger
consortium blockchain is semi-decentralized or decentralized stored and smart contract installed and executed this
rather than distributed. In the consortium, blockchain entities Reduces the system’s complexity and reduces the cost
know each other but do not trust each other. of deployment by lowering the number of nodes storing
Heart of the blockchain is a consensus protocol. The consen- the public ledger. Every FCN is responsible for the IoT
sus is a mechanism to determine some conditions approved, devices in its section. FCN generates the read and writes
so agreement authorized to add a block to the ledger [69]. set for the requests based on the endorsing policy.
3) Organizer Peer (OP): run the consensus algorithm and
write the result to the ledger. They do not maintain a
TABLE II: public, private and consortium blockchain. copy of the ledger. These nodes receive transactions
Feature Public Private Consortium from all regions and then generate blocks for transac-
Network architecture Distributed Semi Decentralized Semi Decentralized tions which fulfill the endorsement policy. New block
Consensus participation All nodes All nodes Selected nodes
Join the network Anyone Invited only Invited only generated when block generating threshold reached.
Scalability High Medium Medium 4) IoT devices: sensors and actuators generating data to be
Efficiency Low High High processed.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
 TABLE III: Comparison of the consensus algorithm
Consensus protocol POW POS [71] dPOS [71] PoET [72] PBFT RAFT [73]
Decentralization Fully Fully Fully/semi Semi Semi Semi
Scalability High High High High Low [74] High
Latency High Medium Medium Low Low Low
Computing overhead High Medium Medium Low Low Low
Network overhead Low Low N/A Low High N/A
Dedicated hardware No No No Yes [75] No No
Adversary tolerance 51% max evil nodes 51% max evil nodes 51% validator N/A 33% faulty replicas 50% crash fault

TABLE IV: policy and threshold

Policy Threshold
Endorsement policy AND(’Section1.member x’, . . . ,’Section x.member x’)
Batch timeout = 1s
OR
Block generating
Max transaction = 10
threshold
OR
Max size = 98 MB

TABLE V: Notation and Abbreviation

RN Registration node
FCN Fog Computing Nodes
OP Organizer Peer
BN Blockchain network
E Elliptic Curve
P large prime number
b point in E
PK public key
SK private key
r random number
H Hash function

1) Initialization step: in this step, RN generates the needed

Fig. 5: System framework including BN, FCN, OP and RN.
digital material for system work. RN uses elliptic curve
cryptography by following these steps:
5) Channels: to add more privacy, Hyperledger fabric offers a) RN chooses Elliptic Curve E with b as a point in
using channels. Different channels have different ledgers E and P as a large prime number.
and endorsement policy. Nodes can join many channels, b) RN chooses r ∈[1, 1-P] as a random number
but they need permission to join these channels. OP representing the private key (SK).
serves all channels but with each channel ledger and c) RN calculates the public key (PK) by PK=SK. r.
endorsing the policy. In this work, one channel used. d) RN test the collision-resistant hash functions H1
The endorsement policy and block generating threshold and H2.
used in this proposed model shown in table (IV). e) RN publishes (PK, H1, H2, E, b, P).
6) Blockchain network (BN): it consists of FCN and OP in 2) Registration step: In this step, all entity required to
all regions sharing the same channel. register before they can participate in the proposed
An endorsement policy defined as a Boolean relationship system. We assume that all communication is done
between selected peers that approve the transaction result, through the secure channel:
which collected to build agreement on adding transactions a) The entity (FCN, IoT device, OP) send their iden-
to a new block. For example (node 1 from section 1 and tification ID to RN.
node 1from sections 2) or (one node from each section). The b) RN checks the ledger for ID if match found RN
endorsement policy illustrates in figure (9) show the one node redirect the entity to change its ID else wise RN
from each section relationship. Notation and Abbreviation are generate the private key (SK) and public key (PK).
shown in table (V). c) RN transmits the material to the entity throw the
same secure channel.
B. System description d) Entity store the materials for further process and
In this section, the system consisting of the Initialization the public key published.
step, Registration step, Authentication and service step, and 3) Authentication and service step:
Consensus step is presented. Figure (6) show the proposed a) Request signed with the entity’s certification and
system workflow process. sent to the required FCN according to the endors-

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
 Fig. 8: generating new block threshold.

Fig. 6: proposed system workflow.

Fig. 9: endorsement policy.

ing policy.
b) The following checks are done by FCN:
e) If the request is updating, the entity will collect the
i) request was not previously submitted (replay- request-response from the FCN until the endorse-
attack protection). ment policy fulfilled. The user sends the request,
ii) signature is valid. and request-response containing the read-write set
c) If the request passes step (b), FCN runs the request and the FCN signature to the OP. Figure (7) show
against the smart contract but without updating the the request flow digram.
ledger. If there is no conflict with other current 4) Consensus step: in the proposed model, the RAFT
requests, the result, called a read-write set, sent consensus algorithm considered to writ the ledger. OP
back to the entity after signing it by the FCN as a can be in one of the following states: leader, follower, or
request-response. candidate. The leader takes the responsibility to arrange
d) The entity collects request response from the FCN the request in blocks and broadcast them. In RAFT [78],
until the endorsement policy fulfilled. The entity at any time, there is one leader. If the entity send request
checks the FCN signature on the request-response. to the follower, it redirects the request to the leader.
If the request type is querying, the process ends The candidate is the state to elect a new leader. RAFT
here, and the entity does not send the request- consensus algorithm can tolerate up to 50% nodes of
response to the OP. crash fault [73]. For consortium blockchains, nodes are
verified, members. Hence, it is more important to solve
crash faults than Byzantine faults [73].
a) users from the previous step send the request and
request-response to the OP leader.
b) OP leader checks the endorsement policy, signa-
tures, and read-write set.
c) Generate new block when one of the new block
generating threshold fulfilled. Figure (8) shows the
block generating process:
i) Max number of requests reached.
ii) Max Time for collecting requests reached.
iii) Max block size in bytes reached.
d) All the FCN receive the new block and validate
Fig. 7: request flow diagram. the right set, request, signatures, and endorsement
policy.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
 VI. PERFORMANCE EVALUATION
This section analyzes the performance of the proposed sys-
tem. Computational overhead, communication overhead, and
response time. CPU, RAM, and cash usage as an indicator of
computational cost. Communication overhead measurement is
done utilizing network input and output. Finally, the response
time is measured in two-point. First point the time organizer
peer takes to create a block. The second point broadcast time
blocked taken to reach Fog Computing node (FCN).
The simulation of the proposed system was done ten times (a) OP. (b) FCN.
for normalization. Experiment samples range from 100 to 1000
requests, with 100 request steps. Fig. 12: Cash usage OP vs FCN.

A. computational overhead
In this part, we present the CPU, RAM, and Cash used transactions, RAM 300MiB Figure (11-a), and 200MiB cash
by Organizer Peer (OP) and Fog Computing Node (FCN). usage Figure (12-a). These numbers at 1000 transaction which
Organizer unites run the consensus process; the consensus show that the OP and FCN can work with low resources
is the core of the blockchain system. Most of the overhead devices.
happens at this point. RAFT used as a lightweight consensus
algorithm. B. Communication overhead
These measurements will focus on the network input and
output as they will show the traffic generated between system
entities. Figure (13-a) presents network traffic as input to the
OP, which is mandatory in designing the network and choosing
devices.

(a) OP. (b) FCN.

Fig. 10: CPU usage OP vs FCN.

(a) input. (b) output.

Fig. 13: OP network input and output.

(a) OP. (b) FCN.

Fig. 11: RAM usage OP vs FCN.

In Figures (10), (11),(12) minimum, maximum, and the

average consumption of resources in OP and FCN presented.
With the maximum CPU usage is 1.25% Figure (10-b) sat-
urated after 700 requests, RAM usage 400MiB Figure (11- (a) input.
(b) output.
b), and cash around 300MiB Figure (12-b) for the FCN. OP
approximately stable around 3.5% Figure (10-a) after 600 Fig. 14: FCN network input and output.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
 For the OP and FCN network, input/output presented in VII. C ONCLUSION
Figures (13- 14). Numbers show that the highest peak for Fog and blockchain integration serve the IoT demands for
the network input/output did not reach 100MiB. The traffic low response time while preserving security and privacy. This
between entities allows the system to work in a typical network integration needs more research to address improvement and
with no need for particular network devices to handle data development need in the infrastructure, hardware, and tech-
traffic. nologies used in both IoT and blockchain such as consensus
algorithm, virtualization, scalability, and storage. Fog comput-
C. Response time
ing can play a mandatory role in the IoT infrastructure as a
Using a lightweight consensus algorithm reduces resource middle layer between Cloud and IoT, bringing the computation
usage, response time, and commit time, which is the potential resources nearest to the network edge. In this work, Fog
point for non-financial with no token application of blockchain computing used to lower the response time. Utilization of
such as in the IoT. Applications when the security of the Consortium Blockchain with transport layer security protocol
blockchain is needed but faster than public ones. The response preserve security and privacy and trim the need for a third
time stabile around 100ms. party. RAFT is a scalable and lightweight consensus protocol,
meets the IoT needs when used in consortium blockchain
where nodes are verified, members. OP FCN virtualized
using Docker Containers to overcome the heterogeneity. Low
response time with efficient power and resource is the result of
this combination. In the future, using AI and machine learning
can be combined to enhance data processing and decision
making. We planned to do more experiments in real-world
applications. Overall, the experiment shows that the model
is suitable for the IoT application and has low overhead and
latency.
(a) commit time. (b) response time.
R EFERENCES
Fig. 15: FCN commit vs response time.
[1] R. Sarmah, M. Bhuyan, and M. H. Bhuyan, “SURE-H: A Secure IoT
Enabled Smart Home System,” in 2019 IEEE 5th World Forum Internet
On average, it took 60ms to generate a new block in OP, and Things, pp. 59–63, IEEE, apr 2019.
the average response time is 100ms with 1000 transactions. [2] R. Lee, R.-y. Jang, M. Park, G.-y. Jeon, J.-k. Kim, and S.-h. Lee,
“Making IoT Data Ready for Smart City Applications,” in 2020 IEEE
The RAFT consensus algorithm for its unique specification Int. Conf. Big Data Smart Comput., pp. 605–608, IEEE, feb 2020.
produces a stable response time. This stability with these low [3] Y. Zhang, Z. Xiong, D. Niyato, P. Wang, and Z. Han, “Information
rates mandatory for IoT applications. Commit and response Trading in Internet of Things for Smart Cities: A Market-Oriented
Analysis,” IEEE Netw., vol. 34, no. 1, pp. 122–129, 2020.
time presented in Figure (15).
[4] J. Ma, S. Feng, X. Li, X. Zhang, and D. Zhang, “Research on the Internet
of Things Architecture for Intelligent Passenger Transportation Services
D. Compare the proposed work and its Application,” in 2019 4th Int. Conf. Electromechanical Control
Technol. Transp., pp. 194–197, IEEE, apr 2019.
The proposed system is compare to [80] in this part. They [5] S. N. Shukla and T. A. Champaneria, “Survey of various data collection
presented a lightweight anonymous distributed vehicular au- ways for smart transportation domain of smart city,” in Proc. Int. Conf.
IoT Soc. Mobile, Anal. Cloud, I-SMAC 2017, pp. 681–685, IEEE, feb
thentication mechanism using blockchain and fog computing. 2017.
They used blockchain and cryptography to authenticate cross- [6] P. Fraga-Lamas, T. Fernández-Caramés, M. Suárez-Albela, L. Castedo,
data centers. Table (VI) presents a comparison of some critical and M. González-López, “A Review on Internet of Things for Defense
and Public Safety,” Sensors, vol. 16, p. 1644, oct 2016.
specifications for the proposed system and [80]. Overall, the [7] A. Buzachis, M. Fazio, A. Galletta, A. Celesti, and M. Villari, “Infras-
proposed system shows better security and scalability with less tructureless IoT-as-a-service for public safety and disaster response,”
response time and overhead. 2019.
[8] F. John Dian, R. Vahidnia, and A. Rahmati, “Wearables and the Internet
of Things (IoT), Applications, Opportunities, and Challenges: A Survey,”
TABLE VI: compare the proposed model with [80] 2020.
Feature proposed model [80] [9] K. Wang, Y. Shao, L. Xie, J. Wu, and S. Guo, “Adaptive and Fault-
Resists Replay Yes No Tolerant Data Processing in Healthcare IoT Based on Fog Computing,”
smartcontract Yes N/A IEEE Trans. Netw. Sci. Eng., vol. 7, pp. 263–273, jan 2020.
Supports Confidentiality Yes Yes [10] G. Xu, “IoT-Assisted ECG Monitoring Framework With Secure Data
Supports Integrity Yes Yes Transmission for Health Care Applications,” IEEE Access, vol. 8,
Supports Forward Secrecy Yes No pp. 74586–74594, 2020.
two-way authentication Yes No [11] M. A. Ankireddy, A. V. Rajath, M. Ruthwik Ganesh, and M. Anuradha,
consesus RAFT PBFT “Augmented reality rendered for IoT applications,” in 2019 IEEE 16th
Anonymity No Yes India Counc. Int. Conf. INDICON 2019 - Symp. Proc., pp. 1–4, IEEE,
scalability High Low dec 2019.
max response time for 1000 request ≈105ms ≈450ms [12] R. Khanna and V. M, “Augmented Reality Based IOT Controller,” in
2019 Int. Conf. Vis. Towar. Emerg. Trends Commun. Netw., pp. 1–5,
IEEE, mar 2019.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
[13] Q. Duan, D. Sun, G. Li, G. Yang, and W.-W. Yan, “IoT-enabled service IEEE Int. Conf. Hot Information-Centric Networking, HotICN 2018,
for crude-oil production systems against unpredictable disturbance,” pp. 207–211, IEEE, aug 2019.
IEEE Trans. Serv. Comput., pp. 1–1, 2020. [38] S. K. Singh, S. Rathore, and J. H. Park, “BlockIoTIntelligence: A
[14] X. Zhang, X. Chen, J. K. Liu, and Y. Xiang, “DeepPAR and DeepDPA: Blockchain-enabled Intelligent IoT Architecture with Artificial Intelli-
Privacy Preserving and Asynchronous Deep Learning for Industrial IoT,” gence,” Futur. Gener. Comput. Syst., sep 2019.
IEEE Trans. Ind. Informatics, vol. 16, pp. 2081–2090, mar 2020. [39] Z. Guan, X. Lu, N. Wang, J. Wu, X. Du, and M. Guizani, “Towards
[15] Y. Ma, K. Ping, C. Wu, L. Chen, H. Shi, and D. Chong, “Artificial secure and efficient energy trading in IIoT-enabled energy internet: A
Intelligence powered Internet of Things and smart public service,” Libr. blockchain approach,” Futur. Gener. Comput. Syst., no. xxxx, 2019.
Hi Tech, vol. 38, pp. 165–179, jun 2019. [40] K. Fan, S. Sun, Z. Yan, Q. Pan, H. Li, and Y. Yang, “A blockchain-
[16] F. Khan, M. A. B. Siddiqui, A. U. Rehman, J. Khan, M. T. S. A. Asad, based clock synchronization Scheme in IoT,” Futur. Gener. Comput.
and A. Asad, “IoT Based Power Monitoring System for Smart Grid Syst., vol. 101, pp. 524–533, 2019.
Applications,” 2020. [41] A. Muthanna, A. A. Ateya, A. Khakimov, I. Gudkova, A. Abuarqoub,
[17] International Data Corporation, “The Growth in Connected IoT Devices K. Samouylov, and A. Koucheryavy, “Secure and reliable IoT networks
Is Expected to Generate 79.4ZB of Data in 2025, According to a New using fog computing with software-defined networking and blockchain,”
IDC Forecast,” 2019. J. Sens. Actuator Networks, vol. 8, no. 1, 2019.
[18] Statista, “IOT market size worldwide 2017-2025 — Statista,” 2019. [42] A. Dorri, S. S. Kanhere, R. Jurdak, and P. Gauravaram, “LSB: A
[19] M. Nardelli, S. Nastic, S. Dustdar, M. Villari, and R. Ranjan, “Osmotic Lightweight Scalable Blockchain for IoT security and anonymity,” J.
Flow: Osmotic Computing + IoT Workflow,” IEEE Cloud Comput., Parallel Distrib. Comput., vol. 134, pp. 180–197, 2019.
vol. 4, no. 2, pp. 68–75, 2017.
[43] A. Dorri, M. Steger, S. S. Kanhere, and R. Jurdak, “BlockChain:
[20] “Cisco pushes IoT analytics to the extreme edge with mist computing -
A Distributed Solution to Automotive Security and Privacy,” IEEE
Rethink.”
Commun. Mag., vol. 55, no. 12, pp. 119–125, 2017.
[21] W. Z. Khan, E. Ahmed, S. Hakak, I. Yaqoob, and A. Ahmed, “Edge
computing: A survey,” Futur. Gener. Comput. Syst., vol. 97, pp. 219– [44] S. C. Cha, J. F. Chen, C. Su, and K. H. Yeh, “A Blockchain Connected
235, aug 2019. Gateway for BLE-Based Devices in the Internet of Things,” IEEE
[22] OpenFog Consortium Architecture Working Group, “OpenFog Refer- Access, vol. 6, pp. 24639–24649, 2018.
ence Architecture for Fog Computing,” OpenFog, no. February, pp. 1– [45] L. Zhou, L. Wang, Y. Sun, and P. Lv, “BeeKeeper: A Blockchain-Based
162, 2017. IoT System with Secure Storage and Homomorphic Computation,” IEEE
[23] S. Misra and S. Sarkar, “Theoretical modelling of fog computing: a Access, vol. 6, pp. 43472–43488, 2018.
green computing paradigm to support IoT applications,” IET Networks, [46] R. Li, T. Song, B. Mei, H. Li, X. Cheng, and L. Sun, “Blockchain
vol. 5, pp. 23–29, mar 2016. for Large-Scale Internet of Things Data Storage and Protection,” IEEE
[24] M. Mukherjee, L. Shu, and D. Wang, “Survey of fog computing: Trans. Serv. Comput., vol. 12, pp. 762–771, sep 2019.
Fundamental, network applications, and research challenges,” IEEE [47] S. Tuli, R. Mahmud, S. Tuli, and R. Buyya, “FogBus: A Blockchain-
Commun. Surv. Tutorials, vol. 20, no. 3, pp. 1826–1857, 2018. based Lightweight Framework for Edge and Fog Computing,” J. Syst.
[25] C. Mouradian, D. Naboulsi, S. Yangui, R. H. Glitho, M. J. Morrow, and Softw., vol. 154, pp. 22–36, 2019.
P. A. Polakos, “A Comprehensive Survey on Fog Computing: State- [48] K. Shafique, B. A. Khawaja, F. Sabir, S. Qazi, and M. Mustaqim,
of-the-Art and Research Challenges,” IEEE Commun. Surv. Tutorials, “Internet of things (IoT) for next-generation smart systems: A review
vol. 20, no. 1, pp. 416–464, 2018. of current challenges, future trends and prospects for emerging 5G-IoT
[26] F. Jalali, K. Hinton, R. Ayre, T. Alpcan, and R. S. Tucker, “Fog Scenarios,” IEEE Access, vol. 8, pp. 23022–23040, 2020.
computing may help to save energy in cloud computing,” IEEE J. Sel. [49] N. Usman, Q. Javaid, A. Akhunzada, K. K. R. Choo, S. Usman, A. Sher,
Areas Commun., vol. 34, no. 5, pp. 1728–1739, 2016. M. Ilahi, and M. Alam, “A Novel Internet of Things-Centric Framework
[27] Z. Wen, R. Yang, P. Garraghan, T. Lin, J. Xu, and M. Rovatsos, “Fog to Mine Malicious Frequent Patterns,” 2019.
orchestration for internet of things services,” IEEE Internet Comput., [50] A. Kaur and S. K. Sood, “Cloud-Centric IoT-Based Green Framework
vol. 21, no. 2, pp. 16–24, 2017. for Smart Drought Prediction,” IEEE Internet Things J., vol. 7, pp. 1111–
[28] I. Stojmenovic and S. Wen, “The Fog computing paradigm: Scenarios 1121, feb 2020.
and security issues,” 2014 Fed. Conf. Comput. Sci. Inf. Syst. FedCSIS [51] P. Arulanthu and E. Perumal, “An intelligent IoT with cloud centric
2014, vol. 2, pp. 1–8, 2014. medical decision support system for chronic kidney disease prediction,”
[29] S. Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System,” SSRN 2020.
Electron. J., vol. 39, no. 1, pp. 53–67, 2008. [52] L. S. Jayashree and G. Selvakumar, “Cloud Solutions for IoT,” in Get.
[30] V. Buterin, “A next-generation smart contract and decentralized appli- Started with Enterp. Internet Things Des. Approaches Softw. Archit.
cation platform,” Etherum, no. January, pp. 1–36, 2014. Model., pp. 31–48, Cham: Springer International Publishing, 2020.
[31] N. Islam, Y. Faheem, I. U. Din, M. Talha, M. Guizani, and M. Khalil, [53] A. Yousefpour, C. Fung, T. Nguyen, K. Kadiyala, F. Jalali, A. Ni-
“A blockchain-based fog computing framework for activity recognition akanlahiji, J. Kong, and J. P. Jue, “All one needs to know about fog
as an application to e-Healthcare services,” Futur. Gener. Comput. Syst., computing and related edge computing paradigms: A complete survey,”
vol. 100, pp. 569–578, nov 2019. J. Syst. Archit., vol. 98, pp. 289–330, sep 2019.
[32] P. Kochovski, S. Gec, V. Stankovski, M. Bajec, and P. D. Drobintsev,
[54] Y. Yang, X. Luo, X. Chu, and M.-T. Zhou, “Fog Computing Architecture
“Trust management in a blockchain based fog computing platform with
and Technologies,” in Fog-Enabled Intell. IoT Syst., pp. 39–60, Cham:
trustless smart oracles,” Futur. Gener. Comput. Syst., vol. 101, pp. 747–
Springer International Publishing, 2020.
759, dec 2019.
[55] Cisco Systems, “Fog Computing and the Internet of Things: Extend the
[33] A. Toor, S. ul Islam, N. Sohail, A. Akhunzada, J. Boudjadar, H. A.
Cloud to Where the Things Are,” Www.Cisco.Com, p. 6, 2016.
Khattak, I. U. Din, and J. J. Rodrigues, “Energy and performance aware
fog computing: A case of DVFS and green renewable energy,” Futur. [56] K. Kaur and M. Sachdeva, “Fog Computing in IOT: An Overview of
Gener. Comput. Syst., vol. 101, pp. 1112–1121, 2019. New Opportunities,” in Lect. Notes Electr. Eng., vol. 605, pp. 59–68,
[34] S. Tuli, S. Tuli, G. Wander, P. Wander, S. S. Gill, S. Dustdar, R. Sakel- 2020.
lariou, and O. Rana, “Next generation technologies for smart healthcare: [57] J. Sengupta, S. Ruj, and S. Das Bit, “A Comprehensive Survey on
challenges, vision, model, trends and future directions,” Internet Technol. Attacks, Security Issues and Blockchain Solutions for IoT and IIoT,”
Lett., vol. 3, no. 2, p. e145, 2020. J. Netw. Comput. Appl., vol. 149, p. 102481, jan 2020.
[35] M. A. Uddin, A. Stranieri, I. Gondal, and V. Balasurbramanian, “A [58] Y. Yu, Y. Li, J. Tian, and J. Liu, “Blockchain-Based Solutions to Security
lightweight blockchain based framework for underwater iot,” Electron., and Privacy Issues in the Internet of Things,” IEEE Wirel. Commun.,
vol. 8, no. 12, 2019. vol. 25, pp. 12–18, dec 2018.
[36] M. Han, D. Wu, Z. Li, Y. Xie, J. S. He, and A. Baba, “A novel [59] T. M. Fernandez-Carames and P. Fraga-Lamas, “A Review on the Use of
blockchain-based education records verification solution,” SIGITE 2018 Blockchain for the Internet of Things,” IEEE Access, vol. 6, pp. 32979–
- Proc. 19th Annu. SIG Conf. Inf. Technol. Educ., pp. 178–183, 2018. 33001, 2018.
[37] Q. Liu, Q. Guan, X. Yang, H. Zhu, G. Green, and S. Yin, “Education- [60] J. Liu and Z. Liu, “A Survey on Security Verification of Blockchain
Industry Cooperative System Based on Blockchain,” in Proc. 2018 1st Smart Contracts,” IEEE Access, vol. 7, pp. 77894–77904, 2019.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.
[61] I. Homoliak, S. Venugopalan, Q. Hum, and P. Szalachowski, “A Security
Reference Architecture for Blockchains,” in 2019 IEEE Int. Conf.
Blockchain, pp. 390–397, IEEE, jul 2019.
[62] M. Saad, J. Spaulding, L. Njilla, C. Kamhoua, S. Shetty, D. H. Nyang,
and D. Mohaisen, “Exploring the Attack Surface of Blockchain: A
Comprehensive Survey,” IEEE Commun. Surv. Tutorials, pp. 1–1, 2020.
[63] S. Cho and C. Jeong, “A blockchain for media: Survey,” in ICEIC 2019
- Int. Conf. Electron. Information, Commun., pp. 1–2, IEEE, jan 2019.
[64] T. Salman, M. Zolanvari, A. Erbad, R. Jain, and M. Samaka, “Security
services using blockchains: A state of the art survey,” IEEE Commun.
Surv. Tutorials, vol. 21, no. 1, pp. 858–880, 2019.
[65] W. Stallings and L. Brown, Computer Security: Principles and Practice.
USA: Prentice Hall Press, 4th ed., 2018.
[66] M. Belotti, N. Bozic, G. Pujolle, and S. Secci, “A Vademecum on
Blockchain Technologies: When, Which, and How,” IEEE Commun.
Surv. Tutorials, vol. 21, no. 4, pp. 3796–3838, 2019.
[67] S. Pahlajani, A. Kshirsagar, and V. Pachghare, “Survey on Private
Blockchain Consensus Algorithms,” in Proc. 1st Int. Conf. Innov. Inf.
Commun. Technol. ICIICT 2019, pp. 1–6, IEEE, apr 2019.
[68] Y. Jiang and S. Ding, “A high performance consensus algorithm for
consortium blockchain,” in 2018 IEEE 4th Int. Conf. Comput. Commun.
ICCC 2018, pp. 2379–2386, IEEE, dec 2018.
[69] Z. Zheng, S. Xie, H. Dai, X. Chen, and H. Wang, “An Overview of
Blockchain Technology: Architecture, Consensus, and Future Trends,”
in Proc. - 2017 IEEE 6th Int. Congr. Big Data, BigData Congr. 2017,
pp. 557–564, IEEE, jun 2017.
[70] L. Lamport, R. Shostak, and M. Pease, “The Byzantine Generals
Problem,” ACM Trans. Program. Lang. Syst., vol. 4, pp. 382–401, jul
1982.
[71] M. Salimitari and M. Chatterjee, “A Survey on Consensus Protocols in
Blockchain for IoT Networks,” pp. 1–15, sep 2018.
[72] H. sawtooth, “Hyperledger Sawtooth,”
Https://Sawtooth.Hyperledger.Org, 2019.
[73] D. Huang, X. Ma, and S. Zhang, “Performance Analysis of the Raft
Consensus Algorithm for Private Blockchains,” IEEE Trans. Syst. Man,
Cybern. Syst., vol. 50, no. 1, pp. 172–181, 2020.
[74] M. Vukolić, “The Quest for Scalable Blockchain Fabric: Proof-of-Work
vs. BFT Replication,” in BigchainDB (J. Camenisch and D. Kesdoğan,
eds.), vol. 9591 of Lecture Notes in Computer Science, pp. 112–125,
Cham: Springer International Publishing, 2016.
[75] M. Schunter, “Intel Software Guard Extensions,” in Proc. 2016 ACM
Work. Softw. Prot. - SPRO ’16, (New York, New York, USA), pp. 1–1,
ACM Press, 2016.
[76] “Hyperledger Fabric – Hyperledger.”
[77] E. Androulaki, A. Barger, V. Bortnikov, S. Muralidharan, C. Cachin,
K. Christidis, A. De Caro, D. Enyeart, C. Murthy, C. Ferris, G. Lavent-
man, Y. Manevich, B. Nguyen, M. Sethi, G. Singh, K. Smith,
A. Sorniotti, C. Stathakopoulou, M. Vukolić, S. W. Cocco, and J. Yellick,
“Hyperledger Fabric: A Distributed Operating System for Permissioned
Blockchains,” Proc. 13th EuroSys Conf. EuroSys 2018, vol. 2018-Janua,
pp. 1–15, 2018.
[78] D. Ongaro and J. Ousterhout, “In search of an understandable consensus
algorithm,” Proc. 2014 USENIX Annu. Tech. Conf. USENIX ATC 2014,
pp. 305–319, 2019.
[79] R. Morabito, “Power Consumption of Virtualization Technologies: An
Empirical Investigation,” Proc. - 2015 IEEE/ACM 8th Int. Conf. Util.
Cloud Comput. UCC 2015, no. 607728, pp. 522–527, 2015.
[80] Y. Yao, X. Chang, J. Misic, V. B. Misic, and L. Li, “BLA: Blockchain-
assisted lightweight anonymous authentication for distributed vehicular
fog services,” IEEE Internet Things J., vol. 6, no. 2, pp. 3775–3784,
2019.

Authorized licensed use limited to: Birla Institute of Technology & Science. Downloaded on February 13,2023 at 19:21:08 UTC from IEEE Xplore. Restrictions apply.