A Systematic Review of Secure IoT Data Sharing
Thanh Thao Thi Tran
1
, Phu H. Nguyen
2 a
and Gencer Erdogan
2 b
1
University of Oslo, Oslo, Norway
2
SINTEF, Oslo, Norway
Keywords:
IoT, Data Sharing, Security, Privacy, Systematic Review.
Abstract:
The Internet of Things (IoT) is more and more omnipresent. The greater values of the IoT can be realized
by enabling data sharing between different stakeholders. However, one of the biggest challenges is ensuring
security and enabling trust for IoT data sharing. In this paper, we identify state-of-the-art (SotA) approaches
and techniques for secure IoT data sharing. We present high-level results emphasizing the SotA trend and re-
vealing the most addressed domains, as well as more in-depth details such as procedures and methods used to
preserve security in the data sharing environment. The blockchain technology, smart contracts, and InterPlan-
etary File System (IPFS) are among the most widely used approaches. As today’s solutions explore a more
decentralized approach to data sharing, there are several aspects to consider. Based on the findings, we have
identified potential research directions for future work, including the differences between public and private
blockchains, the combination of sharing and analytic, the value of data quality, and the importance of data
management and governance.
1 INTRODUCTION
In recent years, the Internet of Things (IoT) has be-
come omnipresent. The reasons are due to the possi-
bilities of connecting everyday objects, such as cars,
kitchen appliances, and baby monitors, to the Inter-
net, allowing for seamless communication between
processes, people, and things. With billions of IoT
devices currently connected to the Internet, Cisco
expects that 30 billion connected IoT devices will
be available in the IoT market by 2023 (Grossetete,
2018). The IoT enables intelligent and inter-operable
digital and physical things, individuals, and services,
generating a variety of ecosystems capable of en-
abling secure cross-domain interactions.
Recent global important factors suggest further in-
crease and expansion of the use of IoT. These fac-
tors are, amongst others, related to the environmen-
tal aspects and global pandemic diseases, such as the
global sustainability goals and the COVID-19 pan-
demic. Amazon has taken the lead in initiating an ask
of net-zero carbon emissions by 2040, whereas com-
panies like Mercedes Benz and Microsoft have com-
mitted to the Climate Pledge
1
. To achieve such an am-
a
https://orcid.org/0000-0003-1773-8581
b
https://orcid.org/0000-0001-9407-5748
1
https://www.aboutamazon.com/news/sustainability/
bitious goal, companies will need to measure carbon
emissions in coming years. The IoT can, e.g., provide
data-driven analytic, while revolutionizing heavy in-
dustries, and life sustainable industrial activities (Fan-
tana et al., 2013). The advantages of the IoT were,
for example, also demonstrated during the COVID-19
global pandemic (Peerzade, 2021). Besides empow-
ering the control and statistics of the high case rate,
IoT has played a role in monitoring virus-infected
patients. With numerous countries have experienced
lock-downs, the fear of new lock-downs will only in-
crease and further fuel the importance and expansion
of the IoT and its use cases.
Indeed, the IoT is more than just a collection of in-
terconnected devices that provide significant benefits.
Connecting enterprises, consumers, suppliers, or any
other stakeholders in the IoT ecosystems can bring
much higher value. Such greater values come from
sharing data (Jernigan et al., 2016; Nguyen et al.,
2022a) between stakeholders in the IoT ecosystems.
One of the biggest challenges is ensuring security
(Rajmohan et al., 2022; Rajmohan et al., 2020) and
enabling trust for IoT data sharing (Atluri et al., 2020;
Ullah et al., 2021; Siegel et al., 2018) as well as data
quality (Nguyen et al., 2022b; Sen et al., 2022). While
the topic of secure IoT data sharing is becoming more
mercedes-benz-joins-the-climate-pledge
Tran, T., Nguyen, P. and Erdogan, G.
A Systematic Review of Secure IoT Data Sharing.
DOI: 10.5220/0011674200003405
In Proceedings of the 9th International Conference on Information Systems Security and Privacy (ICISSP 2023), pages 95-105
ISBN: 978-989-758-624-8; ISSN: 2184-4356
Copyright
c
2023 by SCITEPRESS Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
95
important, it may still be only the end of the begin-
ning. Few existing studies like (Lo et al., 2019; Al-
Ruithe et al., 2019; De Prie
¨
elle et al., 2020) have ex-
amined related topics of secure IoT data sharing such
as data governance and blockchain solutions, but none
has done a systematic literature review (SLR) of se-
cure IoT data sharing.
Because the IoT is a continuously evolving topic,
it is important to assess the current state-of-the-art ad-
dressing the advantages of IoT data sharing as well
as the security considerations that must be taken into
account. Following the most commonly used guide-
lines (Kitchenham and Charters, 2007), we have con-
ducted an SLR on the state-of-the-art of secure IoT
data sharing, and identified open issues and gaps that
need to be addressed in the future. The contributions
of our SLR are our answers to the following research
questions: RQ1: What are the solutions for secure
IoT data sharing? RQ2: What are the security and
trust aspects of these IoT data sharing approaches?
RQ3: What are the current limitations of the IoT data
sharing, and what are the open issues to be further
investigated? In addition to revealing the most ad-
dressed domains, our high-level results and statisti-
cal numbers emphasize the publication increase and
trends. We did, however, obtain more in-depth infor-
mation on the procedures and methods used to pre-
serve security in the data sharing environment. Using
blockchain technology and smart contracts, as well as
the InterPlanetary File System (IPFS), a decentralized
peer-to-peer storage system, are among them. These
and other important discoveries do all contribute to an
increase in secure IoT data sharing.
The remainder of this paper is structured as fol-
lows. Section 2 gives the details of our SLR approach.
Then, we present the results of our SLR in Section 3,
and compare our study with related work in Section
4. In Section 5, we discuss some possible threats to
validity and give our conclusions in Section 6.
2 REVIEW PROCESS
Using online inquiry features of popular publication
databases is the most notable approach to scan for
primary studies when directing supplemental studies
(Kitchenham and Charters, 2007). We used five pop-
ular publication databases, i.e., IEEE Xplore
2
, ACM
Digital Library
3
, ScienceDirect
4
, Scopus
5
, and Web
2
https://ieeexplore.ieee.org
3
https://dlnext.acm.org
4
https://sciencedirect.com/
5
https://scopus.com
of Knowledge
6
to search for potential primary stud-
ies. These databases contain peer-reviewed articles
and provide advanced search capacities.
Based on our research questions, we identified an
initial set of keywords to construct our search strings.
This initial set of keywords were then used to iden-
tify the top relevant papers in the search engines. Fi-
nally, by reviewing the top relevant papers, we iden-
tified a more fine-grained set of keywords that are re-
lated to the broad topics of data sharing, application
domain, security, and governance. This was an iter-
ative process where we also tested the keywords on
the search engines to see if the top relevant papers ap-
peared using our fine-grained set of keywords. The
final set of search keywords are listed below, and we
structured the keywords by following the guidelines
from (Kitchenham and Charters, 2007), e.g., having
groups of keywords to construct our search string us-
ing Boolean ANDs and ORs. The search query was
adapted to fit the search engine of each publication
database.
(”data sharing” OR ”sharing” OR ”data ex-
change” OR ”context sharing” OR ”context aware
data sharing” OR ”context-sensitive information
sharing” OR ”sharing of data” OR ”sharing data”
OR ”ecosystem” OR ”marketplace” OR ”data mar-
ketplace”)
AND (”Internet of Things” OR ”IoT” OR ”In-
dustry 4.0” OR ”smart cities” OR ”smart city”
OR ”smart contract” OR ”manufacturing” OR ”en-
ergy” OR ”supply chain”)
AND (”access control” OR ”secure” OR ”se-
curity” OR ”trust” OR ”trustworthy” OR ”encryp-
tion” OR ”data security” OR ”secure communica-
tion” OR ”secure data sharing” OR ”context-aware
security” OR ”management” OR ”governance” OR
”protocols” OR ”standards”)
Our selection process is based on the predefined
selection criteria (Section 2.2). We first filtered the
candidate papers based on their titles and abstracts.
When the titles and abstracts were not enough to de-
cide whether to discard or keep the papers, we contin-
ued to skim and scan through the contents of these pa-
pers. When a candidate paper appeared in more than
one database, we kept it, at first, in multiple search
results. Then, we consolidated the outcomes in group
discussions among the authors to cross-check the se-
lection decisions and acquire the first set of primary
studies with no duplicates. The review process is il-
lustrated in Figure 1.
6
https://www.webofscience.com/
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
96
Figure 1: Overview of the search and selection steps.
2.1 Research Questions
This SLR aims to answer three RQs as explained in
Section 1. RQ1 includes four sub-RQs. RQ1.1 - What
is the current trend of publications on secure IoT data
sharing? Answering this question allows us to see
when the importance of secure data sharing became
a topic of interest. RQ1.2 - What are the reported
application domains of IoT data sharing? Answer-
ing this question allows us to identify the domains of
interest in the IoT data sharing ecosystem. RQ1.3 -
What are the purpose and benefits of data sharing
considered in the primary studies? As we are inter-
ested in identifying the domains, we are also inter-
ested in the purpose of why different domains would
like to share data and the advantages that could come
from it. RQ1.4 - What are the architectures for IoT
data sharing in the primary studies? Answering this
question will give a high-level understanding of the
architectures for IoT data sharing before we introduce
in RQ2 the more in-depth findings on how different
mechanisms and approaches have been utilized.
RQ2 has three sub-questions. RQ2.1 - What are
the most common threats and vulnerabilities to IoT
data sharing today? Answering this question would
give an overview of current threats and vulnerabili-
ties that come with IoT data sharing, which need to
be considered. RQ2.2 - What and how are the tech-
niques and approaches used to preserve secure IoT
data sharing? After identifying the most common
threats and vulnerabilities related to our scope, we
extract how different contributions address these con-
cerns. RQ2.3 - What is the role of data management
and governance, and how do their standards, poli-
cies, and guidelines support trusted and secure data
sharing? As data management and governance could
contribute to secure IoT data sharing, the extraction of
what policies that have been supported and how they
preserve the security and privacy aspects is analyzed.
RQ3 does not consist of any sub-RQs. However,
this RQ helps to express the current issues and suggest
possible directions for future research.
2.2 Inclusion and Exclusion Criteria
Because the search strategy produced a wide range of
primary studies with diverse content and outcomes, it
was necessary to establish a set of inclusion and ex-
clusion criteria that all primary papers had to meet.
Our selection procedure was conducted in the most
transparent and unbiased way possible, with all the
primary studies having to meet all the following In-
clusion Criteria (IC):
(IC1) Must address IoT data sharing.
(IC2) Must address IoT data sharing architecture
models.
(IC3) Must consider the security (& trust) aspects
of IoT data sharing.
We excluded papers matching any of the Exclusion
Criteria (EC):
(EC1) Non-peer review publications are excluded.
(EC2) Papers written in any other language than
English will be filtered out and excluded in the
search process.
(EC3) Papers published before year 2010 will be
excluded. According to (Wikipedia, nown), it
is claimed that the IoT was ”born” around year
2008-2009, but did not gain popularity before
year 2010, which is a good starting point to re-
view.
(EC4) Papers less than or equal to six pages (sin-
gle column) and four pages (double column) are
excluded. In general, short papers have shown to
not include all necessary details. We therefore ex-
cluded these papers.
2.3 Search and Selection Strategy
The database search discovered 55 primary studies re-
lated to our topic (see Figure 1). Seven of these stud-
ies were discovered through the manual search and
used as the test set for the database search. With the
combination of both the database and manual search,
we were left with a final set of 55 primary studies.
3 RESULTS
Table 1 presents the fifty-five (55) selected primary
studies. We conducted in-depth evaluations and anal-
yses of these studies in order to answer the research
questions as follows.
A Systematic Review of Secure IoT Data Sharing
97
Table 1: The primary studies*.
# Year PV Title (click to open the corresponding publication)
#1 2017 C Towards Blockchain-based Auditable Storage and sharing of IoT data
#2 2017 J Secure and Efficient Data Sharing with Atribute-based Proxy Re-encryption Scheme
#3 2017 J IoT data privacy via blockchains and IPFS
#4 2017 C Big Data Model of Security Sharing Based on Blockchain
#5 2018 J A Peer-to-Peer Architecture for Distributed Data Monetization in Fog Computing Scenarios
#6 2018 J Continuous Patient Monitoring with a Patient Centric Agent: A Blockchain Architecture
#7 2018 C Towards a Decentralized Data Marketplace for Smart Cities
#8 2018 C Providing Context Aware Security for IoT Environments Through Context Sharing Feature
#9 2018 C A Blockchain-Based Framework for Data Sharing With Fine-Grained Access Control in Decentralized Storage Systems
#10 2018 J Smart-toy-edge-computing-oriented data exchange based on blockchain
#11 2019 C Security and Privacy of Electronic Health Records: Decentralized and Hierarchical Data Sharing using Smart Contracts
#12 2019 J Accelerating Health Data Sharing: A Solution Based on the Internet of Things and Distributed Ledger Technologies
#13 2019 J MedChain: Efficient Healthcare Data Sharing via Blockchain
#14 2019 C Toward a Decentralized, Trust-Less Marketplace for Brokered IoT Data Trading Using Blockchain
#15 2019 C Blockchain for Secure and Efficient Data Sharing in Vehicular Edge Computing and Networks
#16 2019 C Towards Multi-party Policy-based Access Control in Federations of Cloud and Edge Microservices
#17 2019 C BlendMAS: A Blockchain-Enabled Decentralized Microservices Architecture for Smart Public Safety
#18 2019 C BPIIoT: A Light-Weighted Blockchain-Based Platform for Industrial IoT
#19 2019 C Enabling Industrial Data Space Architecture for Seaport Scenario
#20 2019 C Blockchain based Proxy Re-Encryption Scheme for secure IoT Data Sharing
#21 2019 J IoT Passport: A Blockchain-Based Trust Framework for Collaborative Internet-of-Things
#22 2020 C BEAF: A Blockchain and Edge Assistant Framework with Data Sharing for IoT Networks
#23 2020 C A Blockchain-based Medical Data Marketplace with Trustless Fair Exchange and Access Control
#24 2020 C Blockchain Smart Contract for Scalable Data Sharing in IoT: A Case Study of Smart Agriculture
#25 2020 J Fully Decentralized Multi-Party Consent Management for Secure Sharing of Patient Health Records
#26 2020 J Secure data exchange between IoT endpoints for energy balancing using distributed ledger
#27 2020 J BDSS-FA: A Blockchain-Based Data Security Sharing Platform With Fine-Grained Access Control
#28 2020 J EdgeMediChain: A Hybrid Edge Blockchain-Based Framework for Health Data Exchange
#29 2020 C Decentralized patient-centric data management for sharing IoT data streams
#30 2020 C Blockchain-Based Multi-Role Healthcare Data Sharing System
#31 2020 J Data Sharing System Integrating Access Control Mechanism using Blockchain-Based Smart Contracts for IoT Devices
#32 2020 J Subscription-Based Data-Sharing Model Using Blockchain and Data as a Service
#33 2020 J Towards a remote monitoring of patient vital signs based on iot-based blockchain integrity management platforms in smart
hospitals
#34 2020 C TrustedChain: A Blockchain-based Data Sharing Scheme for Supply Chain
#35 2020 J A multi-layered blockchain framework for smart mobility data-markets
#36 2021 J Blockchain-Driven Trusted Data Sharing With Privacy Protection in IoT Sensor Network
#37 2021 J MedShare: A Privacy-Preserving Medical Data Sharing System by Using Blockchain
#38 2021 J Fortified-Chain: A Blockchain-Based Framework for Security and Privacy-Assured Internet of Medical Things With Ef-
fective Access Control
#39 2021 C A Cooperative Architecture of Data Offloading and Sharing for Smart Healthcare with Blockchain
#40 2021 C ITrade: A Blockchain-based, Self-Sovereign, and Scalable Marketplace for IoT Data Streams
#41 2021 J Proxy re-encryption enabled secure and anonymous IoT data sharing platform based on blockchain
#42 2021 J Medi-Block record: Secure data sharing using block chain technology
#43 2021 J PrivySharing: A blockchain-based framework for privacy-preserving and secure data sharing in smart cities
#44 2021 J AgriOnBlock: Secured data harvesting for agriculture sector using blockchain technology
#45 2021 J BlockHealth: Blockchain-based secure and peer-to-peer health information sharing with data protection and right to be
forgotten
#46 2021 J BCHealth: A Novel Blockchain-based Privacy-Preserving Architecture for IoT Healthcare Applications
#47 2021 J A conceptual IoT-based early-warning architecture for remote monitoring of COVID-19 patients in wards and at home
#48 2021 J A blockchain-based trading system for big data
#49 2021 J MedHypChain: A patient-centered interoperability hyperledger-based medical healthcare system: Regulation in COVID-
19 pandemic
#50 2021 J SmartMedChain: A Blockchain-Based Privacy-Preserving Smart Healthcare Framework
#51 2021 J eHealthChain—a blockchain-based personal health information management system
#52 2021 J A Threshold Proxy Re-Encryption Scheme for Secure IoT Data Sharing Based on Blockchain
#53 2021 J A Blockchain-Based Medical Data Sharing Mechanism with Attribute-Based Access Control and Privacy Protection
#54 2021 J FAST DATA: A Fair, Secure and Trusted Decentralized IIoT Data Marketplace enabled by Blockchain
#55 2021 J Blockchain Assisted Secure Data Sharing Model for Internet of Things Based Smart Industries
PV: Publication venue; J: Journal; C: Conference; * Because of space restriction for this paper, we can not cite the primary studies in the references.
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
98
3.1 High-Level Details (RQ1)
Answering RQ1.1: Figure 2 illustrates the 55 primary
studies that have been published since 2017, which
according to the histogram, have been the most rele-
vant to our specific topic of secure IoT data sharing
(2020: 14; and 2021: 20 papers).
Figure 2: The publication years of the primary studies.
Answering RQ1.2: As an adaptable technol-
ogy, IoT is utilized across multiple areas. We have
divided the application domain according to IDS
Data Space Radar
7
as follows: smart cities, man-
ufacturing, energy, supply chain, automotive, and
cross-domain/other. The application domain labeled
”Cross-domain/other” is the most dominant, repre-
sented by 44 out of 55 primary studies. The papers ad-
dressing cross-domain/other are divided into four sub-
categories: healthcare, surveillance, smart toys, and
generic domains. Healthcare and the generic topic
have a shared first place, being represented by 21 out
of 55 primary studies. Unlike the topics of smart toys
and surveillance, which are each addressed in a single
study. For the other domains, energy is represented
by one paper, while as for supply chain, manufactur-
ing, and automotive, they are presented in two papers
each, and smart city in four.
Answering RQ1.3: All the primary papers share a
common purpose and goal in their studies and work:
to develop a reliable and efficient data sharing solu-
tion that allows data owners and users to securely ex-
change their data while making data sources more ac-
cessible to authorized actors. There are some studies
that only offer a broad overview of the purpose and
benefits, while others delve further into the purpose
and benefits of data sharing applicable specifically to
the domain they address, such as papers #18 and #28
(see Table 1).
Our findings show that the majority of primary pa-
pers with publications addressing the healthcare do-
main deal with data sharing within the same sys-
tem. The work by paper #28, on the other hand, may
be categorized as addressing data exchange in cross-
7
https://internationaldataspaces.org/adopt/data-space-r
adar/
healthcare systems. Doctors are one of the stakehold-
ers, as they give healthcare to patients, which implies
data sharing within the same system. However, there
is another stakeholder referred to as the ”requestor”
of data. As the division of stakeholders is represented
by doctor, patient, and requestor, it indicates that the
”requestor” may be from a different system.
Data sharing also has a significant impact in the
manufacturing industry, which involves many dif-
ferent stakeholders, e.g., customers, employees, and
supply-chain organizations. Paper #18 explains how
the traditional manufacturing environment is com-
plex, with various manufacturing data, e.g., equip-
ment data, which is often stored in separate systems.
Because these systems may belong to multiple ser-
vice providers, and manufacturing organizations may
not have direct control over this type of data and may
be unable to comprehend the true and full value of the
massive amount of data generated. As a result, the
aim and benefits of sharing equipment data are ex-
plored in depth in this paper. The data on the equip-
ment comprises not just their capacity, but also their
status data. Data sharing can empower R&D, making
manufacturing and distribution audits more effective,
which assists production companies in reducing oper-
ating and manufacturing costs.
Answering RQ1.4: The extractions were catego-
rized and filtered based on the data sharing models
that were utilized. To emphasize, the differences be-
tween these three data sharing models are in terms
of data management and governance perspectives. In
domain-specific sharing, only stakeholders and inter-
ests from inside the domain are allowed to participate
in the process. Peer-to-peer refers to the lack of a
middleman; in other words, the absence of a central
system in the solution. A marketplace, in contrast to
peer-to-peer, implies the presence of a central system,
at least to initiate the sharing process. Our findings
reveal that peer-to-peer is the most common model
utilized. To provide a more exact overview, the peer-
to-peer model is addressed in 23 out of 55 primary
studies, domain-specific sharing in 21 out of 55 pri-
mary studies, and the marketplace is represented in
11 out of the 55 primary studies we have analyzed.
In summary, it has been challenging to fully an-
swer RQ1 in terms of today’s solution in the aspects
of analytics. This is due to a lack of information as
many studies do not cover or address the topic of data
sharing in combination with analytic aspects, at least
not in-depth. However, it is important to note that
data analysis comes with great benefits, such as re-
ducing the amount of analysis that other stakehold-
ers must perform. With this in mind, one can debate
whether data analysis before sharing data to a stake-
A Systematic Review of Secure IoT Data Sharing
99
holder is restricting, in the sense that the stakeholder
could have an interest in raw data for their own anal-
ysis tools and goals.
3.2 Low-Level Details (RQ2)
Answering RQ2.1 We evaluated the contributions
against the OWASP top ten for the IoT domain
8
.
The results show that the issues of insecure network
services, insufficient privacy protection, and insecure
data storage and transfer were the most addressed
OWASP issues in regard to our scope of secure IoT
data sharing. Our findings, however, reveal that the
ten OWASP security concerns are not equally ad-
dressed. The majority of studies are focused on the
three security threats outlined, whereas two studies
deal with insecure ecosystem interfaces, one with lack
of secure update mechanism, and six with use of in-
secure or outdated components. Furthermore, weak,
guessable, or hard-coded passwords, lack of device
management, insecure default settings, and lack of
physical hardening have not been considered in any
way. We may conclude that the issues of insecure
network services, insufficient privacy protection, and
insecure data storage and transfer have been the cur-
rent most important factors to consider while using
IoT data sharing. Nevertheless, the evidence gives us
an overview that several OWASP threats and vulnera-
bilities have been neglected and overlooked.
We evaluated each study against security and pri-
vacy terms as well, to underline what security and
privacy aspects are handled in the solution (see Fig-
ure 3). There are 20 studies addressing confidential-
ity; 20 studies addressing integrity; and 13 studies
addressing availability. Authentication, on the other
hand, is addressed in 22 studies, whereas authoriza-
tion is represented in 23 studies. Privacy is the most
frequently mentioned quality, appearing in 42 of the
55 primary studies. According to the extraction, only
six papers explicitly address the CIA triad (Confiden-
tiality, Integrity, and Availability). Because the CIA
triad (Gollmann, 2010; Nguyen et al., 2015) forms
the core foundation for developing secure information
systems, a lack of awareness could lead to threats and
vulnerabilities when applying the various IoT (data
sharing) solutions to everyday life.
Answering RQ2.2: The main contributions of the
studies show that many of the recent solutions for
trusted IoT data sharing leverage the possibilities of
blockchain technology. The findings show that 51 out
of 55 primary studies utilize blockchain technology;
32 studies out of 55 primary studies utilize smart con-
8
https://owasp.org/www-pdf-archive/OWASP-IoT-To
p-10-2018-final.pdf
Figure 3: Security and its subgroups of specification.
tracts; 10 papers address access control; and one pa-
per, namely #19, utilizes the IDSA architecture. Al-
though some articles use blockchain alone, the ma-
jority of the papers use it in combination with smart
contracts, access control, identity management, or a
mix among these.
The utilization of smart contracts varied from the
studies. For example, paper #28 elaborates on how
they use smart contracts to ensure automated regu-
lation of rules and policies that govern access to the
shared health data in a non-deniable way, whereas pa-
per #27 uses the smart contract in two ways, the first
one being for validation, while the other contract is
for decryption purposes. Just like how smart contract
solutions and use cases differ from study to study, so
does access control management. Paper #18 utilizes
a verification node, which is responsible for access
management, with the access control policies written
on the blockchain. In paper #28, they allow sellers
of data to enforce their own access control policy on
their encrypted records.
While 51 primary studies utilize blockchain tech-
nology, 17 of these papers address what kind of
blockchain type they deploy. From the selection of
primary studies, there are 9 studies that are taking
advantage of the Hyperledger Fabric, while 8 stud-
ies are addressing Ethereum. Hyperledger Fabric is
mentioned in papers #22 and #54, and Ethereum in
papers #14, #20, #40, and #48 of the studies with so-
lutions addressing the generic domain. In the health-
care domain, papers #33, #45, #49, #50, #51, and #53
leverage Hyperledger Fabric while papers #25, #28,
and #37 use Ethereum. Paper #43 within the smart
city domain use Hyperledger Fabric, whereas paper
#26 in the domain of the smart grid use Ethereum.
There are many possibilities with each one of the dif-
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
100
ferent blockchain types. However, without any per-
mission and the total transparency of Ethereum, the
cost may affect the performance of scalability and pri-
vacy. Whereas for the Hyperledger, the consensus and
access control have to be well-defined. The health-
care sector is a frequently mentioned domain in our
primary studies, which tends to be associated with
sensitivity and personal information. It is therefore
worth looking into why some studies, specifically pa-
pers #25, #28, and #37 have been utilizing the public
Ethereum platform in this domain, as the Ethereum
platform consists of total transparency.
Paper #28 has implemented a prototype on the
Ethereum blockchain to validate and evaluate the fea-
sibility as well as the performance. They point out
that the Ethereum blockchain consists of a large,
global development community in addition to be-
ing completely open source and supporting a vari-
ety of use cases such as smart contracts and de-
centralized applications. In their solution, all func-
tions executed in the contract are logged in the
ledger. Therefore, they exude an energetic and pos-
itive view of Ethereum’s transparency. As all trans-
actions are recorded, the blockchain is considered
hacker-resistant, which leads to it being more diffi-
cult to commit fraud within the system. Since every-
one can join the Ethereum network, the solution as-
sumes that the health data being shared is encrypted
and anonymized to protect the privacy and real iden-
tity of patients.
As for the contributions by papers #25 and #37,
they do combine the public Ethereum blockchain in
combination with smart contracts as well. For the first
one, a lot of the focus is related to the implementation
of the prototype and, thereafter, the performance eval-
uation. As for the latter one, they do address that even
though their contribution keeps all patient data secure
using public-key infrastructure (PKI) cryptography,
certain sensitive files cannot be shared on the public
Ethereum network due to privacy issues. However,
they do address one possible solution to this problem,
which is to use forks of Ethereum that are permis-
sioned and private, such as Quorum and Hyperledger
Besu.
Not only do the trust aspects and mechanisms such
as blockchain types have an effect on data sharing. It
is also the layer that the data sharing is being done.
We have been following the IoT World Forum Refer-
ence Model
9
, which consists of seven layers. How-
ever, we will divide the seven layers into three layers,
namely perception (physical devices and controllers),
network (connectivity and edge computing), and ap-
9
https://www.cisco.com/c/dam/global/en ph/assets/cis
coconnect/pdf/bigdata/jim green cisco connect.pdf
plication (data accumulation, data abstraction, appli-
cation, collaboration and processes). There is a no-
ticeable layer that stands out from the crowd, namely
the network layer, which is the most leveraged layer
to share data on from our findings across the selection
of our primary studies. The network layer is presented
35 times, followed by the application layer being rep-
resented 17 times, and finally the perception layer be-
ing represented three times. The lack of coverage
in the perception layer, especially the physical layer,
raises the concern over the cross-domain physical-to-
cyber attacks (Lanotte et al., 2020; Yampolskiy et al.,
2012) that could make secure IoT data sharing mean-
ingless.
To leverage the great benefits of data sharing, the
technical aspects of how the data is stored need to be
considered as well. The primary studies show that
the current trend is to remove themselves from a tra-
ditional centralized solution and move to a more de-
centralized solution. With this in mind, the primary
studies we have collected are introducing a distributed
protocol for sharing and storing called the Interplane-
tary File System (IPFS) (IPFS, nown; Steichen et al.,
2018). Compared with cloud storage, IPFS, as the
peer-to-peer storage network it is (IPFS, nown), can
contribute to preventing the problem of a single-point-
of-failure.
More specifically, #3, #5, #7, #9, #24, #25, #27,
#28, #30, #34, #39, #50 and #52 utilize the peer-to-
peer IPFS data storage protocol, while #2, #20, #32,
#38, #41 and #53 still utilize the traditional central-
ized cloud storage. The use of a third-party cloud ser-
vice has been incorporated in some way or another,
either as a primary storage of data, or just a back-up
storage.
When storing the data, studies have pointed out
the huge difference between storing raw data and en-
crypted data directly in their storage system. The find-
ings show that a great number of selected primary
studies have addressed this difference. There are 17
papers that utilize encryption. There are, however,
only 11 papers addressing what type of encryption
they utilize. From this number, study #2, #20, #25,
#48, #52, and #54 utilize proxy re-encryption, while
studies #9, #11, #27, #28, and #37 practice attribute-
based encryption (ABE) in their solutions.
Answering RQ2.3: The data governance establish
the policies and procedures, which the data manage-
ment implements. A few studies look at the neces-
sity for access control from the standpoint of the data
owner, while none of them mention usage control (La-
zouski et al., 2010). A few ideas, in particular, would
allow data owners control over their own data. Papers
#7 and #43 address GDPR, while paper #7 also ad-
A Systematic Review of Secure IoT Data Sharing
101
dresses the California Consumer Privacy Act. How-
ever, most of the primary studies do not go into as
much detail or emphasis on the importance of data
management and governance as we would like to see.
In general, primary studies have revealed relatively
little information about the role and impact of data
management and governance.
Because the value of data sharing is derived from
the data itself, it is important to assess the data’s qual-
ity for the purposes of reuse and valid analysis. DNV
are experts in assurance and risk management. They
are using their expertise to improve safety and perfor-
mance while also setting industry benchmarks, e.g.,
a framework for data quality assessment
10
, which ex-
plains how to assess data quality. The framework em-
phasizes the importance of data quality assessment
and the need for continual monitoring of the qual-
ity. ISO 8000
11
is a standard that provides approaches
for managing, measuring, and improving the quality
of data and information. However, none of the se-
lected primary studies included inputs from DNV or
ISO 8000, or other relevant data quality frameworks,
demonstrating a lack of data quality management in
the context of IoT data sharing.
To wrap up RQ2, there are many mechanisms and
approaches utilized in the scope of secure IoT data
sharing. However, when relating to the security as-
pects, there are only six contributions that cover the
CIA principles. Even if other studies might cover the
aspects of the CIA triad, contributions that lack elab-
oration and explanation have not been included. We
can conclude that we have some shortcomings in the
findings that contribute to our inability to fully answer
RQ2. In particular, RQ2.3, in terms of the impact and
role of data management and governance. This is be-
cause the majority of the primary studies we looked
at did not include any specific standards, policies, or
procedures that might have contributed to their solu-
tion for secure IoT data sharing. Even though GDPR
was mentioned a few times, there was no in-depth ex-
planation of how it was used as a contributing factor
to secure the data sharing process.
3.3 Gaps and Limitations (RQ3)
This section discusses our findings to answer RQ3.
Limitations. A limitation that has been identified
through a number of studies and findings is the lack
of stress testing (limitation1). However, there are only
10
https://rules.dnv.com/docs/pdf/DNV/RP/2017-01/DN
VGL-RP-0497.pdf
11
https://www.iso.org/obp/ui/#iso:std:iso:8000:-61:
ed-1:v1:en
four studies that explicitly address this limitation with
their contribution, namely the contributing paper #3,
#11, #19 and #33. Stress testing means testing against
different use cases and seeing the results of utilizing a
larger scale network. In addition, a variety of the stud-
ies operate on one blockchain platform, but do men-
tion the wish to test their solution on other blockchain
platforms to see if the results could have significant
meaning. As the IoT technology is widely used and
connects a variety of devices within different domains
and areas, it would be useful to see results on the per-
formance when putting the solution under stress.
Furthermore, another limitation of IoT data shar-
ing is the concern about the heterogeneity nature of
IoT (limitation2). As IoT technology makes human
life more connected than ever over the Internet, it also
implies the existence of a wide range of links from
devices to multiple points such as endpoint devices,
applications, and cloud platforms. More specifically,
the studies show that the limitation occurs when com-
bining and adding different technologies could lead to
the growth of complexity.
Another limitation that is addressed is scalability
(limitation3). The definition of scalability, when re-
lated to the domain of computer science, is the mea-
sure of a system’s ability to increase or decrease in
performance and cost in response to changes in ap-
plication and system processing demands (Gartner,
nown). When mentioning scalability as a limitation, it
refers to the growth in the number of participants and
transactions. As a side note, as the complexity of a
system increases, it will have an effect on the scalabil-
ity of the same system. A hand-full of solutions find
it difficult to handle moving parts and components, in
addition to the foundational scalability of a decentral-
ized marketplace. To summarize, because the primary
studies are concerned about the complexity of their
solution, there will also be an indirect concern about
scalability as well.
Last, we have the limitation related to the through-
put and performance of the solutions (limitation4).
One of the papers elaborates on their blockchain
architecture, which is hardly supported by high
throughput performance. More specifically, paper
#10 revealed that the maximum throughput was es-
timated to be a maximum of 50,000 requests per
second. Moreover, for (near) real-time secure IoT
data sharing, especially involving dynamic contexts,
heavy-weight blockchain-based solutions are less ap-
plicable. The advanced context-aware (de Matos
et al., 2018) access control/usage control-based ap-
proaches should be more favorable. Such approaches
can still employ blockchain technology for ensuring
the trust-ability of IoT data sharing by logging only
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
102
the metadata of importance for the data sharing trans-
actions like data quality metrics, the necessary head-
ers of shared data. In general, as the IoT is of-
ten connected to critical infrastructure domains such
as healthcare (Solanas et al., 2014), its performance
would have significant meaning, where the outcome
of the effects could be critical.
Open Issues. Throughout our extraction, paper #2,
#7, #9, #29, and #38 mention more or less the open is-
sues regarding their contributions in their papers. The
other papers did not explicitly mention open issues,
but we still identified and extracted some open issues.
The main findings related to the open issues have been
divided into four topics, namely GDPR, access policy,
cost, and storage.
Paper #29 discusses permission-less data manage-
ment that empowers patients to securely and selec-
tively share their own personal data. This paper dis-
cusses data management but mentions their lack of fo-
cus on how to address the compliance of GDPR; for
instance, the erasing of personal data (issue1). The
open issues were stated by the authors and are specific
to their paper. On the other hand, this is an overall is-
sue that applies to IoT data sharing or IoT security in
general, as elaborated on in our results.
Paper #9 discuss the data storage and sharing
scheme in their contribution. They elaborate that their
solution is safe only if the Ethereum blockchain net-
work and the attribute-based encryption scheme are
safe. They do mention that their contribution does not
include an access policy update (issue2).
Paper #7 on decentralized data marketplace for
smart cities do address the concern of cost (issue3).
The concern about cost is associated with the trans-
actions and the contracts used, which must be mini-
mized to reduce the cost involved in using a decen-
tralized data marketplace. They do, however, propose
a solution that is worth investigating: a special short
data format for storing metadata in the blockchain.
A common topic that turns out to be a recurring
point in many, if not all the studies, is the security is-
sues regarding data storage (issue4). Some of today’s
solutions utilize the cloud server, including paper #2,
#20, #32, #38, #41 and #53. Even though the storage
phase is done in combination with encryption, the so-
lution utilizing a third-party solution still suffers from
a single-point-of-failure.
To summarize, there should be further research on
the security and privacy aspects of today’s data shar-
ing solutions, including the methods and strategies
used to ensure secure IoT data sharing. By looking
into the relationship between various trust and secu-
rity mechanisms and methodologies as well as the se-
curity and privacy evaluation of studies, one may be
able to uncover underlying reasons why, for instance,
the CIA triad is not being addressed as much as de-
sired. Additional future research should also be the
elaboration of the significance and influence of data
management and governance. For example, full anal-
yses of current standards, policies, and guidelines, as
well as explanations of how each standard or policy
helps to data sharing security, should be supplied.
4 RELATED WORK
While the topic of secure IoT data sharing is becom-
ing more important, it may still be only the end of the
beginning. Few existing studies like (Lo et al., 2019;
Al-Ruithe et al., 2019; De Prie
¨
elle et al., 2020) are
relevant to secure IoT data sharing, but none having
the scope of our study, nor answering our questions.
Kuang Lo et al. (Lo et al., 2019) elaborate on the
issues that data management solutions face, as well as
the key issue of single-point-of-failure caused by the
use of centralized management servers. Our study and
(Lo et al., 2019) complement each other by pointing
out technologies as a solution to the single-point-of-
failure, such as blockchain as smart contracts. How-
ever, our work further extracted on the architectural
layers (Rajmohan et al., 2022) used for the execution
of data sharing, who the stakeholders in different do-
mains are, and why data sharing is of interest to adapt
for these stakeholders and domains.
Al-Ruithe et al. (Al-Ruithe et al., 2019) presents
an SLR of data governance and cloud governance in
their use of data. They highlight the need for more
advanced research in data governance, in addition to
suggesting areas for further research within data gov-
ernance, which can be taken into account when con-
ducting our research. However, they do not go into
detail on the implications for IoT data sharing because
their main focus is on data and cloud governance.
The necessity of ecosystem data governance
for data platforms is discussed by Prie
¨
elle et al.
(De Prie
¨
elle et al., 2020). Future research directions
are elaborated, such as the importance of data plat-
form governance in access and usage as a primary
concern. They also emphasize that there is a lack of
research on the many types of benefits that data shar-
ing generates, which is an important future research
direction as well. However, we focus on data sharing
as the primary topic, with data governance and its im-
pact on data sharing as a subtopic. Furthermore, they
do not address various standards, policies, and guide-
lines that have been considered.
A Systematic Review of Secure IoT Data Sharing
103
5 THREATS TO VALIDITY
In this section, we address the various concerns and
risks to the validity of our SLR. The validity analy-
sis is carried out by combining our work experience
with the knowledge gathered from Kitchenham et al.
(Kitchenham and Charters, 2007)’s principles.
One of the pitfalls when performing an SLR can
be related to how the search process has been con-
ducted. More specifically, in regard to the keywords
and queries. The query can be used as proof of the va-
lidity and limitations of the study, as there are an infi-
nite number of other keywords that could be included
in the search query. However, to defend the selec-
tion of our search query, we performed a test case to
have a handful of quality test papers that we believe
are relevant and should be included in our research.
These test papers were used to evaluate the quality of
our search query results. If the search results in the
electronic database X include all the test set papers
published by X, we know that the other papers in the
results are relevant in some way. Otherwise, we may
conclude that there is a lot of noise in our search.
In addition, there was a need for filtering the re-
sults from IEEE Xplore as the result gave a number
greater than 2000. We filtered the results to only pa-
pers related to the ”Internet of Things”. This can be
considered a limitation, as there may be papers in the
original IEEE Xplore result that could be of interest.
Another threat to our study is with regard to the se-
lection of primary studies. The studies that have been
included in our selection of primary studies from our
SLR, by passing our inclusion and exclusion criteria,
in addition to being relevant based on our predefined
taxonomy, can still be doubtful. There might be rel-
evant studies we have overlooked or publications we
have missed out on during our search phase.
Last, the snowballing procedure could be em-
ployed as part of the search process, after the auto-
matic and manual searches have been completed. Of-
ten, this method is used to enrich the set of primary
studies. For possible additional primary studies, the
technique comprises examining the list of references
of the existing primary studies, and new publications
citing them. Even though we did not go through
this process, our database search was extensive and
backed up by the test set from the manual search.
6 CONCLUSIONS
After systematically reviewing 55 primary studies rel-
evant to the scope of secure IoT data sharing, we have
discovered that the topic is getting hotter in recent
years. In the IoT data sharing domain, centralized
cloud-based solutions have traditionally dominated.
This means that the solutions are relying on a sin-
gle trusted third-party. However, there is a growing
interest in utilizing blockchain’s decentralized quali-
ties. From our findings, the objectives behind the use
of blockchain are: the removal of the centralized third
party, immutability, improved data sharing, enhanced
security, and reduced overhead costs in distributed ap-
plications.
Based on the synthesized data from analyzing the
primary studies, we have been able to make the fol-
lowing remarks. (1) To make IoT data sharing more
secure and valuable, based on our findings, we sug-
gest looking into the gap between public versus pri-
vate blockchain, and which is the most preferable
blockchain technology type that should be utilized. It
may also be worth analyzing the utilization of spe-
cific blockchain types in specific domains and the rip-
ple effects that may occur. (2) As our findings show,
there is a lack of research on the combination of shar-
ing and analytics. However, there are numerous pa-
pers addressing the issue, which can occur when the
data owner analyses the raw data and shares this ana-
lyzed data instead of the raw data. More specifically,
there is further need for research to discover whether
receiving already analyzed and processed data or un-
processed data makes a difference to stakeholders in
the IoT ecosystem. (3) A clear standout is the topic
of data management and governance and its lack of
research. Future research should therefore highlight
both the need and importance of data management
and governance, in addition to addressing specific
standards, policies, and guidelines that should be or
are considered as well, and how they increase the se-
curity aspect in data sharing.
The great value of data sharing often derives from
the data itself. Therefore, it is important that the data
being sent and received is of high quality to ensure
that analyses and the use of data are valid. Future re-
search should emphasize data quality in the context of
IoT data sharing, where consideration of ISO 8000 for
data quality could be of interest. We further suggest
collaboration across domains, as the combination of
different data resources with data sharing across do-
mains can enable IoT technology to reach even further
and discover even more possibilities in the future.
ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy
104
ACKNOWLEDGEMENTS
The research leading to these results has received
funding from the European Commission’s H2020
Programme under the grant agreement numbers
958357 (InterQ), and 958363 (DAT4.Zero).
REFERENCES
Al-Ruithe, M., Benkhelifa, E., and Hameed, K. (2019). A
systematic literature review of data governance and
cloud data governance. Personal and Ubiquitous
Computing.
Atluri, V., Hong, Y., and Chun, S. A. (2020). Security, pri-
vacy and trust for responsible innovations and gov-
ernance. In The 21st Annual International Confer-
ence on Digital Government Research, dg.o ’20, page
365–366, New York, NY, USA. Association for Com-
puting Machinery.
de Matos, E., Tiburski, R. T., Amaral, L. A., and Hessel, F.
(2018). Providing context-aware security for iot en-
vironments through context sharing feature. In 2018
17th IEEE International Conference On Trust, Secu-
rity And Privacy In Computing And Communications/
12th IEEE International Conference On Big Data Sci-
ence And Engineering (TrustCom/BigDataSE), pages
1711–1715.
De Prie
¨
elle, F., De Reuver, M., and Rezaei, J. (2020). The
role of ecosystem data governance in adoption of data
platforms by internet-of-things data providers: Case
of dutch horticulture industry. IEEE Transactions on
Engineering Management, pages 1–11.
Fantana, N. L., Riedel, T., Schlick, J., Ferber, S., Hupp, J.,
Miles, S., Michahelles, F., and Svensson, S. (2013).
Iot applications—value creation for industry. Internet
of Things: Converging technologies for smart envi-
ronments and integrated ecosystems, page 153.
Gartner (Unknown). Scalability.
Gollmann, D. (2010). Computer security. Wiley Interdisci-
plinary Reviews: Computational Statistics, 2(5):544–
554.
Grossetete, P. (2018). IoT and the Network: What is the
future?
IPFS (Unknown). What is IoT?
Jernigan, S., Kiron, D., and Ransbotham, S. (2016). Data
sharing and analytics are driving success with iot. MIT
Sloan Management Review, 58(1).
Kitchenham, B. A. and Charters, S. (2007). Guidelines for
performing systematic literature reviews in software
engineering. Technical Report EBSE 2007-001, Soft-
ware Engineering Group.
Lanotte, R., Merro, M., Munteanu, A., and Vigan
`
o, L.
(2020). A formal approach to physics-based attacks
in cyber-physical systems. ACM Trans. Priv. Secur.,
23(1).
Lazouski, A., Martinelli, F., and Mori, P. (2010). Usage
control in computer security: A survey. Computer Sci-
ence Review, 4(2):81–99.
Lo, S. K., Liu, Y., Chia, S. Y., Xu, X., Lu, Q., Zhu, L.,
and Ning, H. (2019). Analysis of blockchain solutions
for iot: A systematic literature review. IEEE Access,
pages 58822–58835.
Nguyen, H.-H., Phung, P. H., Nguyen, P. H., and Truong,
H.-L. (2022a). Context-driven policies enforcement
for edge-based iot data sharing-as-a-service. In 2022
IEEE International Conference on Services Comput-
ing (SCC), pages 221–230.
Nguyen, P. H., Kramer, M., Klein, J., and Traon, Y. L.
(2015). An extensive systematic review on the model-
driven development of secure systems. Information
and Software Technology, 68:62–81.
Nguyen, P. H., Sen, S., Jourdan, N., Cassoli, B., Myrseth, P.,
Armendia, M., and Myklebust, O. (2022b). Software
engineering and ai for data quality in cyber- physi-
cal systems - sea4dq’21 workshop report. SIGSOFT
Softw. Eng. Notes, 47(1):26–29.
Peerzade, E. (2021). The role of IoT sensors in the COVID-
19 fight.
Rajmohan, T., Nguyen, P. H., and Ferry, N. (2020). Re-
search landscape of patterns and architectures for iot
security: A systematic review. In 2020 46th Eu-
romicro Conference on Software Engineering and Ad-
vanced Applications (SEAA), pages 463–470.
Rajmohan, T., Nguyen, P. H., and Ferry, N. (2022). A
decade of research on patterns and architectures for
iot security. Cybersecurity, 5(1):1–29.
Sen, S., Husom, E. J., Goknil, A., Tverdal, S., Nguyen, P.,
and Mancisidor, I. (2022). Taming data quality in ai-
enabled industrial internet of things. IEEE Software,
39(6):35–42.
Siegel, J. E., Kumar, S., and Sarma, S. E. (2018). The future
internet of things: Secure, efficient, and model-based.
IEEE Internet of Things Journal, 5(4):2386–2398.
Solanas, A., Patsakis, C., Conti, M., Vlachos, I. S., Ramos,
V., Falcone, F., Postolache, O., Perez-martinez, P. A.,
Pietro, R. D., Perrea, D. N., and Martinez-Balleste,
A. (2014). Smart health: A context-aware health
paradigm within smart cities. IEEE Communications
Magazine, 52(8):74–81.
Steichen, M., Fiz, B., Norvill, R., Shbair, W., and State, R.
(2018). Blockchain-based, decentralized access con-
trol for ipfs. In 2018 IEEE International Conference
on Internet of Things (iThings) and IEEE Green Com-
puting and Communications (GreenCom) and IEEE
Cyber, Physical and Social Computing (CPSCom) and
IEEE Smart Data (SmartData), pages 1499–1506.
Ullah, I., De Roode, G., Meratnia, N., and Havinga, P.
(2021). Threat modeling—how to visualize attacks
on iota? Sensors, 21(5):1834.
Wikipedia (Unknown). Internet of things.
Yampolskiy, M., Horvath, P., Koutsoukos, X. D., Xue, Y.,
and Sztipanovits, J. (2012). Systematic analysis of
cyber-attacks on cps-evaluating applicability of dfd-
based approach. In 2012 5th International Symposium
on Resilient Control Systems, pages 55–62.
A Systematic Review of Secure IoT Data Sharing
105