A Systematic Literature Review On Cloud Computing Security Threats And Mitigation Strategies

Received March 17, 2021, accepted April 9, 2021, date of publication April 14, 2021, date of current version April 21, 2021.
Digital Object Identifier 10.1109/ACCESS.2021.3073203
A Systematic Literature Review on Cloud
Computing Security: Threats and
Mitigation Strategies
BADER ALOUFFI 1, MUHAMMAD HASNAIN 2, ABDULLAH ALHARBI3, WAEL ALOSAIMI 3,
HASHEM ALYAMI1, AND MUHAMMAD AYAZ 2
1Department of Computer Science, College of Computers and Information Technology, Taif University, Taif 21944, Saudi Arabia
2School of Information Technology, Monash University Malaysia, Subang Jaya 47500, Malaysia
3Department of Information Technology, College of Computers and Information Technology, Taif University, Taif 21944, Saudi Arabia
Corresponding author: Muhammad Hasnain (muhammad.malik1@monash.edu)
ABSTRACT Cloud computing has become a widely exploited research area in academia and industry. Cloud
computing benefits both cloud services providers (CSPs) and consumers. The security challenges associated
with cloud computing have been widely studied in the literature. This systematic literature review (SLR)
is aimed to review the existing research studies on cloud computing security, threats, and challenges. This
SLR examined the research studies published between 2010 and 2020 within the popular digital libraries.
We selected 80 papers after a meticulous screening of published works to answer the proposed research
questions. The outcomes of this SLR reported seven major security threats to cloud computing services.
The results showed that data tampering and leakage were among the highly discussed topics in the chosen
literature. Other identified security risks were associated with the data intrusion and data storage in the
cloud computing environment. This SLR’s results also indicated that consumers’ data outsourcing remains
a challenge for both CSPs and cloud users. Our survey paper identified the blockchain as a partnering
technology to alleviate security concerns. The SLR findings reveal some suggestions to be carried out in
future works to bring data confidentiality, data integrity, and availability.
INDEX TERMS Auditing, cloud computing, cloud models, decryption, encryption, malicious behavior,
intrusion, secured communication.
I. INTRODUCTION
Cloud Computing concept has emerged from the distributed
software architecture. Cloud computed technology is aimed
to provide hosted services over the internet. In recent years,
cloud computing in Information Technology has given rise
to various new user communities and markets [1]. Cloud
computing services are provided from data centers located
in different parts of the world. Microsoft SharePoint and
Google applications are general examples of cloud computing
services.
Security plays an important role in the wider acceptance of
cloud computing services [2]. Existing literature is focused
on different security solutions, including technology and
security policy implementation. The latter study introduced
new attacks on the cloud environment from criminological
perspectives. The proposed solution to these recent attacks
The associate editor coordinating the review of this manuscript and
approving it for publication was Chin-Feng Lai .
is based on criminal theories for the protection of the cloud.
A study [3] identified several security issues affecting cloud
computing attributes. The same research proposes to over-
come the identified problems concerning the security of
cloud. A security guide, developed in this research, enables
the cloud user organizations to be aware of security vulnera-
bilities and approaches to invade them.
Security vulnerabilities and challenges arise from the
usage of cloud computing services. Currently, cloud comput-
ing models are the primary source of these challenges and
vulnerabilities. The intruders exploit the weakness of cloud
models in accessing the users’ private data, by attacking the
processing power of computer systems. The ‘‘Autonomous
Cloud Intrusion Response System’’ (ACIRS) has been
recently proposed to overcome the problem mentioned
earlier [4]. Before this work, the ‘‘Network Intrusion Detec-
tion and Countermeasure Selection system’’ (NICE) [5]
worked on the selection of the best countermeasures to
mitigate the risks to cloud virtual networks. As compared
57792
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 9, 2021

B. Alouffi et al.: Systematic Literature Review on Cloud Computing Security
to NICE, ACRIS is superior in mitigating the risks and
challenges to networks.
There is a widespread use of cloud computing (CC) in
information technology. However, many service owners are
still reluctant to fully adopt the CC as relevant security tech-
nologies are not as yet matured. Thus, literature shows a need
for service providers to invest in CC-associated device secu-
rity [6]. We have found a few studies that show the proposal of
evaluation of cloud computing security. One of these research
studies introduces an ‘‘attack tree map’’ (ATM) to analyze
security vulnerabilities and threats. Research [7] highlights
various facets of CC combined with the trusted computing
platform to provide security services such as confidentiality,
authentication, and integrity.
II. BACKGROUND INFORMATION ON CLOUD
MODELS AND RELATED WORKS
This section presents a short overview of the cloud computing
models. We also offer a summary of related works to our
research topic.
A. CLOUD MODELS
1) SOFTWARE AS A SERVICE (SaaS)
The SaaS model facilitates users to access the software and
other programs in a cloud. Using the SaaS solution eliminates
the need for in-house applications, data storage, and support
for the application administration. Companies pay to use the
SaaS resources on a user basis [8].
2) PLATFORM AS A SERVICE (PaaS)
PaaS is a cloud computing service that supports a full soft-
ware life cycle and allows users to develop cloud applications
and services [9]. Programmers and developers do not need
to purchase their equipment; instead, they use intermediary
equipment and deliver the developed applications to clients
over the internet. In PaaS, an individual or a company is not
required to buy the software and hardware to develop the
applications. Google App Engine, Azure services platform of
Google, Amazon’s relational database services (RDS) are the
key examples of PaaS model.
3) INFRASTRUCTURE AS A SERVICE (Iaas)
IaaS is the cloud computing service delivered in the form
of platform in a virtual environment [10]. Clients are not
required to purchase servers, data centers, network equipment
or space (e.g. Amazon EC2).
4) CONTAINER AS A SERVICE (CaaS)
Based on container virtualization, CaaS has emerged as a
cloud model to resolve application development issues in
the PaaS environment. The CaaS cloud model is aimed to
free the application by making them independent of PaaS
environment specifications [11]. Amazon EC2 Container Ser-
vice (ECS) and Google container engine are examples of
CaaS model.
Studies [8] and [10]–[11] emphasized general cloud secu-
rity models such as SaaS, PaaS, IaaS, and CaaS. Empirical
studies show that these models play an important role in cloud
security from the perspectives of cloud security providers and
clients. However, a successive range of cloud security models
is more important to safeguard the users’ activities on clouds.
If a security model is more secure, a user will be happier to
use the client computing services.
B. RELATED WORK
Information technology has rapid changes in recent years.
Cloud computing has added more promising role of IT
with the addition of storage for users. Cloud computing has
enabled the vendors to rent out their services at hourly rates.
They also rent out the space to users on their physical sys-
tems. However, these services have several security threats
for users. In a report, Cloud Security Alliance revealed that
abuse, insecure interfaces, and nefarious usage were the vul-
nerable threats. These threats have been associated with the
application program interfaces and cloud computing [12].
Information security splits into three main objectives, such
as integrity, confidentiality, and availability. Security threats
to these security goals include a long-term confidentiality
issue because one considers that present and past encryption
schema are not secure. Information leakage vulnerability is
another concern as data is outsourced. Tampering with data
also poses threats to data confidentiality [13].
As new technologies are emerging to meet the users’
demands, there is a significant increase in cloud security
threats. These threats are occurring in the form of several
unseen exploitation through the cloud computing services
and their associated interfaces. It has become essential to
counteract the occurred and potential attacks [14]. Presence
of the insecure interfaces is a big challenge to both cloud users
and cloud service providers.
Cloud services’ security and availability mainly depend
on APIs that involve in data access and data encryption on
clouds. Further research can be undertaken to ensure the
security of these APIs and network interfaces. New security
proposals can meet the protection challenges of services from
intentional and accidental attacks and violation of terms of
services. Furthermore, layered APIs have more complexity
as third body operators use cloud services. Actual propri-
etors cannot access the services. In addition to it, malicious
insiders are common threats to cloud services as they violate
the services’ terms and access the information they are not
permitted. Usually, an employee is the malicious insider who
steels the confidential information that belongs to a com-
pany or its legal users, An inside malicious user can corrupt
the information, particularly in a peer to peer file sharing
systems [15].
Multi-Tenancy is referred as resource sharing and associ-
ated risks of data confidentiality and integrity. Multi-Tenancy
has been called a serious concerning issue for professionals in
cloud computing. Professionals’ understanding about attack
surfaces and attack vectors is most important [16]. In [17],
VOLUME 9, 2021 57793

an increased number of cloud computing service users
resulted in data security and privacy threats.
Cloud users face issues of reliability from big players in
cloud computing. In [18], researchers reported that Amazon’s
EC2 and S3 suffered an outage for 3 hours and 8 hours
in February 2008 and July 2008, respectively. Google’s
Gmail services remained unavailable for 3 hours, which pre-
vented 113 million users from accessing documents through
their Gmail accounts. Microsoft and Amazon offered their
customers access to selecting the geo-spatial location of
companies’ data centers.
1) CLOUD SERVICE USER
Cloud service users, either clients or attackers, must have
users’ authentication privileges to access the cloud services.
Communication between users and cloud resources seeks
the secured channel that keeps the users’ login information
more secure than hypertext transfer protocol (HTTP) users.
Secondly, communication between cloud services and users
must be ensured with synchronization [19].
C. INTRUSION DETECTION SYSTEM
Cloud service manager as a cloud provider must follow the
service level agreements (SLAs) and conform to technical
standards. Cloud data on cloud premises needs to be protected
from physical and crypto-graphical intrusion attacks [20].
Resource starvation or crashing the server may prevent the
legal client from accessing the cloud computing services.
Cloud services would have an intrusion detection system
that identifies users’ vulnerabilities and sends an alert mes-
sage to cloud service admin to take swift action against
the attackers or intruders. Cloud service manager rejects the
access of such intruders if they are found to be real intruders.
An authenticated user is given access to desired data files.
Data encryption approach for both client and server ends
is chosen. To detect intruders’ anomaly behavior through
encrypted traffic, an intrusion detection system discovers
attacks by analyzing a large number of data patterns. In [21],
information extraction from encrypted data traffic is obtained
through features of HTTP traffic and frequency of access.
In the following, we discuss how an intrusion detection com-
ponent is designed and implemented for encrypted traffic.
An intrusion detection system is classified into three tiers as
given in the following.
1) DATA AT REST
The intrusion detection system scans the data from cloud data
storage sources, encrypts the data, and removes it encrypted
when it is found non-trusted. However, fine-tuning and con-
figuration constraints are real concerns for data at rest.
2) DATA IN MOTION
Clients’ real concern lies in data security in moving data
from client to cloud or cloud to client. Network monitoring
of sensitive information via emails, and instant messaging,
a system requires the identification and prevention of log
email information that attempts to leak information from a
client or an organization by the aid of steganography.
3) DATA IN USE
Data transmission from the client-side computer is monitored
via the output peripherals such as USB ports, printers, CDs,
and external storage devices. If the applications access sensi-
tive data, it is filtered before sending it to relevant peripher-
als [22]. A plain-text message transmits information between
a portable device and EC3 unit, and data leakage can occur
when a legal cloud computing services user stores the data
at remote-based storage devices. However, data leakage is
prevented by using the algorithms developed in [23] to iden-
tify the legal cloud storage devices. Cloud database maintains
the list of legal devices and requires the authentication of
portable devices before data exchanges between clouds and
cloud clients. After verification of devices, the cloud client
retrieves and transmits the data. When a cloud client wants to
upload data, the EC3 unit sends a request to connect a cloud
manager. Cloud client passes a plain text denoted by the Pm
to EC3, where a chip of encryption/decryption encrypts the
Pm to a random positive integer k.
Fig. 1 shows us the device authentication process with
several steps. The first step focuses on extracting the device
information, followed by the ID decryption in the second step.
The final step involves the decryption in the ECC integration
unit. Under these seven steps, for the authentication process,
a legal user cannot read, copy, or transmit personal records
to a system because data-writing is disabled on the storage.
Therefore, a legal user even needs the primitive root g from
p and prime number p before accessing the system’s records.
A private key of an integer value that ranges between 2 and
p – 2 is used to publicize the (g, p, B). A seven step pro-
cess involves an authentication process of devices, as shown
in Fig. 1. A new user is asked to fill the form for account
creation. Credentials of a new user are stored in the cloud
system [24]. The choice of elliptic curve cryptography with
(2m) fields is compatible with the modern computers and
logic gates of binary states [25].
FIGURE 1. Authentication process for devices.
4) AUDITING AND UNIT TESTING
Cloud providers should have control over the auditing and
maintenance of cloud services. It can be achieved through the
57794 VOLUME 9, 2021

consecutive monitoring of users’ logs and administrative ses-
sions with those parts of cloud services affected by users. The
trace-back technique is applied to identify users who violate
the policies and laws. Cloud testing is a hope to determine the
balance of advantages and risks of cloud computing services.
Cloud service providers and customers perform cloud testing
at their levels.
On the other hand, the customer is responsible for per-
forming tests to avoid or mitigate any risk to its organization
or customers. In a cloud security model, the cloud services
provider ensures the architecture’s physical security through
auditing procedures and controls to overcome the cold-boot
attacks. The cloud service manager is responsible for setting
the default baseline of security measures for users. Cloud
service managers monitor tenants’ malicious attacks when
they target different services to flood the other tenants’ virtual
machines [26]. In [27], researchers proposed that the pro-
tection of multi-tenancy areas was the major focus of cloud
vendors.
Cloud services are well-known approaches for product
traceability of industrial systems to provide data integration
and sharing services [28]. However, malicious cloud ser-
vices prevent industrial participants from the correct acqui-
sition of traceability of products. Recent research proposes
Acics as consistent and fast auditing schema for the mas-
sive industrial information in the untrusted cloud computing
environment. This schema enables the industrial participants
to play an auditor role for consistency checking of products.
Experimental results show that the proposed Acics compared
to other approaches is efficient in data consistency checking
of small amounts of products at a reasonable cost. The pro-
posed Acics schema shows better read or write latency rate.
This schema has been not tested on large products, and hence
it can be examined in the future works.
Before the latter reviewed study, research [29] highlighted
that cloud computing has an alarming security risk to cus-
tomers’ data. Customers transfer their data to CSP’s pro-
vided storage but remain unaware of evaluating the CSP’s
security controls to protect their sensitive data. The cloud
security alliance (CSA) sets the guideline to gauge the CSP’s
organization’s security controls. It enables a cloud service
user to trust the services of CSPs. The main issue is that
CSA’s questionnaire-based security assessment is not vali-
dated for the responses’ accuracy. A framework proposed in
the same study aims to bridge this gap by using third party
auditors (TPAs) for validating the participants’ responses.
Still, there is no way of using the outcomes of third party
checking for external users. To make users more trusting
in an organization’s cloud services, we need to get their
feedback for further quality evaluation of cloud providers’
services.
5) CLOUD RESOURCES MANAGEMENT
Cloud computing services models concern the management
of cloud resources. Cloud service providers offer clients
resources such as virtual machines, network devices, load
balancers, and firewalls. Resource management is one of the
pressing issues of IaaS [30]. Services availability directly
links to the security of physical equipment and devices at
premises of cloud service providers. Denial of service attack
makes resources unavailable for the intended consumers of
cloud computing services. Attackers get access to protected
resources by exploiting system leakage, bugs, configuration
mistakes, and design flaws [31]. In [26], the denial of service
attack issue is subject to tenant machines and compromised
virtual machines. An inside or outside attacker floods the
virtual machine with the attacked traffic of TCP, UDP, and
‘‘internet control message protocol synchronized’’ (ICMP
SYN) floods. The attacker also chooses to allocate more
resources to virtual machines to generate more attacked traf-
fic. To cope with denial of service from UDP and ICMP
traffic, a client or tenant uses the maximum threshold (λ) for
ICMP and UDP traffic.
FIGURE 2. Cloud security architecture [26].
Fig. 2 shows us the security architecture proposed in [26].
As shown in Fig. 3, tenants may keep their own host-based
security tools (HBST) to run on their virtual machines. They
obtain virtual machines from cloud providers. Thus, monitor-
ing a system through the HBST has good visibility. The other
essential components of the architecture are service provider
attach detection (SPAD) and tenant-specific attack detection
(TSAD) [26].
6) SECURITY ANALYSIS
ECC’s encryption process efficiently encrypts the different
messages by using the varying points of the elliptic curve.
A short key size of 256 bits prevents the attacks of algo-
rithms on the ECC encryption system because the computing
complexity of attacking algorithms is O (2128). Hence, cloud
clients’ IDs and private keys are stored in their smart cards.
Therefore, an illegal user cannot generate a valid digital
signature [32].
The client can also ask the client service provider to drop
data if UDP or ICMP exceeds the threshold and produces
an alarm by the client administrator. The TCP SYN flooding
attacks the tenant and exploits the weaknesses presented in a
three-way handshake process. In this case, an attacker floods
the victim’s machine and overflows the victims through the
half-open connections.
VOLUME 9, 2021 57795

FIGURE 3. Taxonomy of survey literature on cloud computing topics.
7) INTRUSION DETECTION FEATURES
FOR ENCRYPTED TRAFFIC
SlowDos attacks are challenging to researchers and they
need to be detected using the signature-based approaches
for a HTTP encrypted traffic. Recent work [33] introduces
artificial intelligence (AI) based approach to cope with the
challenge mentioned earlier. The proposed work analyses,
processes, and aggregates the data packets to detect the
anomalies dynamically. Both clustering and deep learning
techniques combine to increase the accuracy efficiency of
proposed approach. The proposed approach’s evaluation is
performed on a real testbed to show the performance and
accuracy in detecting the attacks over the encrypted traffic.
8) TAXONOMY OF SURVEY LITERATURE
ON CLOUD COMPUTING
Current survey literature bespeaks most of the security and
privacy concerns and their mitigation strategies (see Table 5).
As observed in this survey paper, a large body of litera-
ture is focused on mitigating the security threats on clouds.
However, the survey paper observes that the current literature
is classified into security issues and other cloud computing
domains. Based on multiple security perspectives, the secu-
rity issues are classified according to the cloud aspects. The
majority of authors of the existing survey papers have paid
attention to security and privacy challenges. Other than secu-
rity and privacy issues, resource scheduling is the widely
examined area of cloud computing. However, security taxon-
omy and attack detection are less explored areas, as shown
in Fig. 3. It indicated that inadequately classified security
challenges could prevail than much spending on security
issues.
Moreover, we present the classification of the cloud com-
puting (CC) combined with other research domains. These
domains include education, IoT, Blockchain technology,
mobile computing, and resilience. Among these domains,
mobile computing is a widely studied research area. However,
two survey papers [SP25] and [SP31] have been published
recently on blockchain technology with cloud computing
topics. The former survey paper is a short conference paper
that does not show a detailed discussion on blockchain tech-
nology’s role in overcoming cloud computing security issues.
The latter survey paper is a combined effort of researchers
on fog computing and blockchain technology. Since several
papers have been published on blockchain technology with
cloud computing, we present a brief posture of blockchain
technology’s role in cloud computing.
III. RESEARCH METHOD
A systematic literature review (SLR) as a research study
method is used to evaluate and interpret the research topic’s
available literature. A systematic literature analysis is an
alternative to an SLR. Kitchenham guidelines are more spe-
cific to conduct this SLR [34]. The SLR protocol for this
study consists of the following subsections.
A. RESEARCH QUESTIONS
The main objective of this research is to examine the security
concerns regarding cloud computing services. This research
also focuses on risk mitigation strategies from significant
existing research studies. To cover the aims and objec-
tives of this SLR, we formulated four research questions as
follows:
1) RQ1: What are the cloud computing security threats
and their mitigation strategies?
2) RQ2: What are the security problems that have not been
addressed by commercial cloud providers?
3) RQ3: What are consumers’ concerns from cloud com-
puting standards and policies implementation?
4) RQ4 What is the role of blockchain technology in the
security of cloud data?
57796 VOLUME 9, 2021

B. SEARCH STRATEGY
This phase is focused on search keywords, electric
sources, reference management tool, and a search process.
We describe each process in the following subsections.
1) SEARCH KEYWORDS
The search keywords and strings have been derived from
designed research questions. We have also included syn-
onyms and alternatives. We took synonym keywords from
the relevant literature on cloud computing security topics.
We show the search keywords in the following.
‘‘Cloud security challenges’’, ‘‘cloud security models’’,
‘‘commercial cloud provides and security challenges’’,
‘‘cloud security mitigation strategies’’, ‘‘cloud security
models’’, ‘‘blockchain technology’’.
2) ELECTRONIC SOURCES
Popular digital libraries have been used to search for papers—
these digital libraries are IEEE Xplore, Springer, ACM,
Google Scholars, and Science Direct. These digital libraries
are the primary source of publications on topics from com-
puter science domain.
3) REFERENCE MANAGEMENT
Using search keywords and their alternative, we found a
large number of studies from the above-mentioned electronic
sources. We used Endnote X9 [35] as a reference man-
agement tool to collate and manage the retrieved material.
It enabled us to easily add and remove the studies whenever
we required them.
4) SEARCH PROCESS
We launched a search process on the digital libraries to
retrieve the relevant literature from journals, conferences,
and books. This search process resulted in more than
1500 studies. We used Endnote to manage the pdf ﬁles and
their respective references, making it easier to read papers.
Next, we applied the selection process to ﬁlter out the irrele-
vant papers.
C. STUDIES INCLUSION AND EXCLUSION CRITERIA
We followed the studies’ inclusion and exclusion criteria as
listed in Table 1.
D. QUALITY ASSESSMENT CRITERIA (QAC)
The QAC criteria is formulated to ensure the quality and
strength of primary studies. A checklist has been used in [36]
to assess the quality of research studies. The QAC checklist is
designed that depends upon the questions related to problems
of the domain area. As given in Table 2, these questions aim to
sort out the relevant studies to include them in the systematic
literature review (SLR).
The quality assessment criteria are applied to determine
the essential studies for evidence in cloud computing secu-
rity concerns and cloud security models. Studies’ analysis
TABLE 1. Studies’ inclusion and exclusion criteria.
TABLE 2. Quality assessment criteria.
is based on the proposed questions, and if a study answers
Yes, then 2 points are awarded to that study. If the answer is
No, then 0 points are awarded and if the answer is partially
then 1 point is allocated to that study. Initially, we gathered
1500 research studies, and next, we applied QAC criteria
to sort out studies that were suited to answer the research
questions. By application of QAC, 71 studies were found to
be answering the research questions. Hence, we retained (5%)
of total studies suitable for this systematic literature review.
E. DATA SYNTHESIS
Data synthesis is aimed to aggregate the evidence from cho-
sen studies to answer the research questions. The proof is
a single piece that provides little evidence compared to the
aggregation of many parts of evidence [36]. Data extracted
in this SLR is qualitative (weaknesses and strengths of cloud
security models or approaches) and quantitative (estimation
and accuracy values of cloud security issues). Extracted data
have been synthesized by using different strategies. Data
about research questions 1-2 have been synthesized through a
narrative method. Data is tabulated so that each study is con-
sistent with cloud security domain, cloud security approach
and focused area of cloud security models. Visualization tools
such as bars and pie charts are also employed to present cloud
security issues and mitigation strategies or approaches in the
enhanced manners. Data of research questions 3-4 highlight
the important cloud security models and its focused issues in
each security model. Bar as a visualization tool is employed
that shows how many studies have focused on certain cloud
security issues.
VOLUME 9, 2021 57797

IV. RESULTS AND DISCUSSION
This section presents our SLR’s results and their discussion
is given in the following subsections.
Table 3 shows us the studies publication year, security
approaches and purpose of the research. In table 3, it has
become clear that cloud computing and its security concerns
have remained the focus of researchers in the last few years.
Second, the chosen studies research focus shows that cloud
computing has advantages for both cloud service providers
and users. However, cloud consumers’ concerns about infor-
mation security have made them rethink before they use cloud
computing services.
A. SECURITY THREATS AND RISK MITIGATION
STRATEGIES (RQ1)
Data loss due to its leakage is a severe threat to cloud security.
Data compromise and modifications occur without keeping
the backup copy by altering or deleting the original informa-
tion. Also, data storage on cloud media has less reliability
because insiders and third parties can access the data. In irre-
sponsible media, the companies’ offering of cloud service are
regarded as fraudulent [37]. Utility based approach can be
used to overcome the latter-mentioned challenge by detecting
the malicious behavior of users. This utility allows users
to recover their data. Unity service is a personal repository
service, which is different from other cloud services.
Cloud services users do not need to become unauthorized
users of services, but internal individuals of CSP organization
present malicious behavior. In studies [20] and [37], we have
found such incidents, which can harm the data security from
cloud service providers. Therefore, trust-building between
client and cloud service providers is emphasized through a
unity model.
The emerging paradigm of cloud computing ensures the
reliability, availability, and scalability while users access the
cloud computing services over the internet. The concept of
software and services is activated on the users’ demands.
However, several additional risks have been perceived by
the productive organization. A hacker can perform tricks to
snatch the confidential information as each and every thing
is kept inside the cloud computing box. Due to evolution in
the security disciplines, hackers can be prevented to have an
illegal access to cloud data. Also, more security measures
can be proposed to provide a comprehensive security to users
regarding their data.The key-splitting and Homo-morphic
encryption can be extended to have new breakthroughs in this
area of research [38]. In [39], the emphasis was on secure data
transfer between customers and cloud computing providers
over the secured communication channels.
A group of researchers proposed an attack model in [40],
where they identified the three most suspicious customers.
Researchers used the datasets from Google and sensed
the questionable behavior of customers. Businesses and
TABLE 3. Publications and their research focus.
57798 VOLUME 9, 2021

individuals get nervous using the cloud systems, as a cloud
is a metaphor in the internet’s perspective. They fear any
unauthorized users can access the information as cloud
computing simply lingers in cyberspace. Respective cloud
computing providers ensure security and privacy in cloud
computing [41]. If the infrastructure of cloud computing is
insecure, the information’s availability and confidentiality
become at risk. Firewall configuration, server patching, and
placement of the intrusion-detection system are enforced. The
use of encryption and authentication processes can ensure
data availability and confidentiality, as shown in Fig. 4.
FIGURE 4. Safety in cloud computing.
A remote office requests for cloud computing services
from vendors to enable the data encryption sent by clients.
The encrypted data is stored at the cloud center. Clients or
remote users of cloud computing services get access to the
encrypted data [42].
Most significantly, cloud computing in financial organiza-
tions is exposed to risk from hackers. Therefore, we have a
security approach of encryption and authentication to reduce
the risk of alteration or information disclosure in transit and
storage. The encryption process provides data security from
prying eyes. Entities trying to retrieve the companies’ pro-
tocol for accessing the information get the scrambled infor-
mation. This approach accomplishes the use of keys in two
ways. Cloud service providers never retain the keys, instead
their clients retain keys locally and ensure that used keys are
destroyed or rotated correctly. Microsoft and Amazon have
made advancements in making the simple storage service
(S3) available for customers through REST and SOAP [18].
Maintenance and distinctiveness are critical issues in the
implementation of the cryptographic scheme. Replication is a
standard in cloud computing for keys’ encryption/decryption
maintenance. Amazon faced a similar matter, and since then,
such a problem has been resolved. However, the lack of
foresight in the process of cryptography results in disastrous
outcomes [43].
FIGURE 5. Identified cloud computing security threats.
Fig. 5 shows us the seven identified types of cloud com-
puting security concerns in research studies. Data tampering
and leakage are the deepest valued concerns for both clients
and cloud computing service providers. These types of cloud
computing security issues encompass cloud computing users
who cannot access cloud computing resources. They are
either attackers, hackers, or users who attempt to access the
cloud computing services for which they are not registered.
Although data intrusion threat has a close resemblance with
the earlier types of data tampering and leakage, intruders use
different means of entering the data clouds. In [41], data con-
fidentiality and availability have been called primary secu-
rity concerns irrespective of other security lapses between
clients and cloud computing services. A secure communica-
tion between clients and cloud computing service providers is
the second-highest score regarding cloud computing security
risks. We have pointed out the IaaS level concerns as a
separate type of cloud computing security risks.
Data tampering and leakage is the real concern of users
in several domains. A cloud-based healthcare system may
have challenges of patients’ data leakage [47]. Forensic data
collection and edge computing environments have severe
security risks, including forensic data removal or log leak-
age. The edge cloud services have their local isolation from
the cloud, and security monitoring becomes inappropriate.
Malicious users target geographically divided manage-
ment [48]. Besides this, log data cannot be stored securely,
and forensic information can be removed from the cloud to
prevent forensic investigation.
The second identified threat is the data intrusion in
cloud computing. While adopting cloud computing services,
an intrusion detection system (IDS) provides a solution to
security threats [49]. Various IDS have been developed to
recognize accurate attacks. A hybrid approach of intrusion
detection is proposed in [50], which suits high traffic data like
clouds. Compared with the prevailing techniques, the hybrid
approach efficiently detects anomalies and provides better
security to increase traffic in a cloud environment. Besides
these developments, a suspicious IDS itself can have adverse
effects for detection in the cloud. SaaS providing companies
include Google, Adobe, Microsoft, Shopify, and HubSpot
that still suffer security, loss of control, and connectivity
VOLUME 9, 2021 57799

requirement issues. Researchers can undertake future
research to address the security issues arising from cloud
computing models, i.e., IaaS.
Data storage, data confidentiality, and data availability in
cloud services are among the other identified threats to cloud
computing security. For example, outsourcing stored data at
the cloud requires an additional security layer to strengthen
the data confidentiality. Research [51] introduced a ‘‘cloud
storage based on ID-based encryption’’ (CS-IBE) with the
one file access policy and a user’s identity proposed to be
used as an encryption key. Although this approach simplifies
the key management issues, it shows limited performance for
data confidentiality.
A recently published work [52] introduces a new cloud
service, namely ‘‘database as a service’’ (DBaaS), where a
service provider has responsibility to grant access of soft-
ware, hardware, and network to users. It would enable them to
access and manage a database. At the same time, a distrustful
service provider keeps the control of database queries. As a
result, security issues emerge in the confidentiality of stored
data during outsourcing of data to a user.
B. COMMERCIAL CLOUD COMPUTING PROVIDERS
AND SECURITY CHALLENGES (RQ2)
A survey study [53] finds that the top five cloud providers,
including Amazon, Azure, Adobe, Google cloud platform,
and VMWare, are efficient in their cloud services’ data secu-
rity feature. Reliability and performance are among other
features. To measure the cloud providers’ trustworthiness is
still an issue for researchers, and a customer cannot judge it
without appropriate tools. Container launched by VMs is an
emerging practice offering security sandboxing. Containers
enable the cloud providers to continue managing their appli-
cations on clouds [54]. Since application management at the
edge is challenging for cloud providers, it is done either ad
hoc or with the platform. When multi-tenant run their appli-
cations on the same host resource, security, and privacy issues
arise from their applications. Existing literature on com-
mercial service providers (CSPs) reveals that cloud service
models are involved in hampering security concerns [55].
Therefore, it is users’ workload based on the sensitive data
that they do not need to outsource to a public cloud directly.
Outsourcing the consumers’ data and addressing the asso-
ciated risks is challenging for both users and cloud service
providers. These risks include shadow-IT, security, control
and transparency, and business continuity [56]. Interoperabil-
ity is another challenge because many consumers are locked
to a single CSP due to interoperability issues.
Amazon AWS ‘‘Identity and Access Management’’
(IAM) [57] operate the access control mechanism and iden-
tity management system. Each customer gets a tenant account
upon the subscription of the Amazon AWS product. To assure
the security of users’ data, they are allocated security cre-
dentials to access AWS resources. However, Amazon IAM is
not so expressive and contains simple featured-based policies
with the limited encryption attributes. AWS and Microsoft
Azure policies clearly state that customers need to address
the security of their data, operating system, application, and
their configuration, identity, and access management in a
shared responsibility. The AWS only considers the hard-
ware, software, and networking facilities security [58]. Cloud
users mistakenly assume that their public IaaS providers
have responsibility for securing their data, operating sys-
tem, and applications [59]. ‘‘Secure by design’’ [58] is an
innovative idea that needs to be integrated with all cloud
services and applications to enable an agile environment for
businesses in the face of security threats and the technology
ecosystem. Thus, researchers can undertake future research to
develop security architecture to ensure built-in security from
developing systems and services, tools, and technologies
across the cloud environment. Before the latter mentioned
studies, a research article [60] focused on software security
requirements and proposed replacement of traditional soft-
ware development with the emerging services. Thus, cloud
providers can share a security service with the distributed
stakeholders.
C. CONSUMERS’ CONCERNS FROM CLOUD
COMPUTING STANDARDS AND THEIR
POLICIES IMPLEMENTATION (RQ3)
Standards and policies are the significant measures for users
and admin of CSPs. These policies and standards also apply
to partners with CSPs. In a cloud model, policies are con-
figured to monitor end-users devices and cloud computing
services [32]. Service level agreements are one of the obsta-
cles that contain the scope of cloud services. Cloud services’
consumers come across while adopting cloud computing
services. Data unavailability, vendor lock-in and insufficient
measures of security create concerns for users. Consumers
show their concern over the lack of interoperability and stan-
dards. Portability features are provided with limited offers.
Therefore, evaluation of SLAs benefits the service providers
in terms of legal actions, while minimal assurance of data pro-
tection for consumers is specified to reflect the consumers’
requirements at the right time [13].
National laws also conflict with the SLAs that enforce
the cloud computing service providers to disclose their
customers’ sensitive information [61]. Cloud computing pro-
cesses and elements like APIs did not follow the cloud com-
puting standards that presented a barrier for clients to switch
cloud suppliers without expense and pain [9]. Although most
vendors have defined their standard mechanisms, they could
not complete the function of writing standards once and run
it everywhere [41]. Data security, privacy, and up-time were
typically called the essential items in [44]. In addition to these
advantages, cloud client did not require the initial capital to
invest in the infrastructure. Cloud clients shift their risks to
cloud computing service providers.
Literature reveals that cloud computing and social media
have become an influencing factor in controlling and mon-
itoring public policies. Social media applications are widely
used and have become a significant source of data generation.
57800 VOLUME 9, 2021

Cloud computing can store the generated data that can map
useful public opinion in different aspects such as issues and
their solutions, advantages and disadvantages, and proposals.
Research in [62] proposed to combine CC services with
social media, and monitor the public policies efficiently.
For the evaluation of the presented approach Tweets data
is used. Thus, data from other platforms such as Facebook
and Instagram have been not used. Although CC has several
advantages for users, it still suffers from unreliable latency,
network communication cost, and mobility support.
Extrinsic concerns such as network, compliance, and infor-
mation security arise while implementing the CC services
in small, medium enterprises (SMEs). Security and cloud
services professionals and SME representatives should con-
sider these issues before implementing cloud services [63].
Decision making in an organization is mainly based on infor-
mation security. On the other hand, cloud users have no
experience of handling their data on cloud storage. It leads to
an ambiguity in the minds of users. Thus, the cloud provider
must ensure users’ information by installing data safety
measures.
Although CC has many deployment and implementation
benefits, organizations adopting them face compliance, trust,
hosting, legal, security, and privacy issues [64]. Policymakers
also state about the inadequate resources and guidelines to
inform decision-makers. A comprehensive framework is pro-
posed to investigate the CC services, their deployment, and
delivery models. The proposed framework has been adopted
by the Saudi Government agencies overseas. Among issues
faced by the adopting organization, data security and privacy
has been considered the leading factor when deciding whether
to adopt it. Legal and policies are other issues that may not
allow the adoption of CC services due to security and privacy
concerns during the implementation.
D. ROLE OF BLOCKCHAIN TECHNOLOGY IN THE
SECURITY OF CLOUD DATA (RQ4)
Security is one of the major challenges of big indus-
tries, including banking, supply chain management (SCM),
electronic health records, and smart applications in recent
years. A research [65] proposes a novel framework to mon-
itor the activities taking place at the particular data evi-
dence. This data evidence comes from data and users’ signa-
ture based on the SHA-256 Cryptographic Hash Algorithm.
A cloud-based ‘‘software defined network’’ (SDN) is created
with the support of a Blockchain controller, cloud server,
and an authentication server (AS). Researchers propose to
register all users with the AS to obtain the secret key.
In this proposed framework, the ‘‘Elliptic Curve Integrated
Encryption Scheme’’ (ECIES) technique encrypts data pack-
ets and transfers them to a cloud server. The experimental
result showed that better performance concerning throughput,
response time, accuracy, and total change security features
has been achieved. Security features in the blockchain-based
cloud include authentication, network security, access mech-
anism, and privacy method. This work is comprehensive
and can be implemented for a real-time application with the
improved security. Traceability of users due to modification
in data has become possible, increasing the reliability and
preserving the users’ data privacy. However, a SDN controller
that may suffer from various DoS and DDoS attacks can be
prevented by extending it in future works. However, we care-
fully require to handle the computational overheads and avoid
the increasing cost.
Resource allocation and task scheduling are challenging
issues to academia and industry. Security during the task
scheduling on a distributed computing environment has been
considered a significant criterion for personalizing cloud ser-
vices. Blockchain technology integration with the cloud clus-
ters helps secure the cloud transactions and access the appli-
cation and data codes [66]. Recent work proposes a novel
blockchain scheduler that is better in performance compared
with other cloud scheduling models. One of the limitations
of the proposed research is its simulation, which could be
near a realistic one but cannot be 100 percent applicable to
real-world case studies. Therefore, researchers can consider
realistic scenarios with different cloud clusters and cloud
technologies to develop a multi-cloud system in future works.
Given the increased number of published works, we cannot
capture all relevant research papers. The possible reason is
that Blockchain technology is booming nowadays among
scholars and industrial professionals. We have selected a
few recently published works to show the integration of
blockchain technology with the cloud computing environ-
ment. However, cloud computing is a separate research area
and much work is in progress on the latter mentioned area.
The majority of research studies [65]–[69] have been focused
on securing the cloud data through the blockchain technology.
The effective use of blockchain technology provides an extra
security layer to data at clouds, and users’ trust remains intact
while outsourcing the information. A legitimate user achieves
the relevant information.
Data sharing systems such as electronic medical
records (EMRs) facilitate users to access the medical record
and store the patients’ information effectively [72]. Moreover,
data exchanges occur between connected mobile devices in
the cloud computing environment. Still, blockchain security
and privacy concerns remain to be settled in future research
works. It is essential to protect the users’ identity. A study [73]
proposed an attribute-based approach to secure the users’
identity. The proposed approach, based on the KUNodes,
has an attribute master key and an attribute signing key.
These attributes’ revocation can be easily achieved by using
the KUNodes algorithm. The proposed approach is security
preserving, unforgeable, and collision resistant.
Cloud-centric computing has several downfalls. One of
them is the issues produced from the unilateral test. This
is because an application is delegated to a user in the
cloud environment. Cloud-centric computing is not scalable
when a single point failure occurs [74]. To overcome these
issues, decentralized computing is utilized in different Inter-
net of Things (IoT) scenarios. Devices connected can be
VOLUME 9, 2021 57801

locally managed instead of using third-body services, thereby
resulting in conveniently available applications and services.
Blockchain technology is optimizing the resource alloca-
tion to customers. Auction algorithms ensure the appropri-
ate allocation to the actual buyers of computing resources.
A study [75] introduced the broker concept in managing and
adjusting the trading market. Besides, the proposed research
introduces edge-cloud computing and blockchain technology
to address the issues of real-time processing and allocation of
resources. The proposed research designed a trading market
where buyers compete with the sellers’ resources. An iterative
auction schema works to collect the buyers’ requirements and
then allocate the resources accordingly.
Trust and security are real issues to the IoT Big Data.
Trust concerns have been addressed by proposing the permis-
sioned blockchain to perform data invoking and data storage
and avoid data tampering from inner-side users and external
attackers [76]. Even if data is tampered with on some partial
nodes, the entire ledger still works and provides security and
trust to users because tampering can’t reach and succeed at
all. However, the proposed approach still has security issues
while releasing the key for encrypting data at the user layer.
Blockchain technology provides the necessary privacy
and security to the communication between fog nodes, IoT
devices, and cloud computing. The research carried out
in [77] is not entirely successful as it is still in progress.
The proposed framework in the latter study has not been
implemented in a real-world environment. This is due to the
big head size issue of blockchain structure. This problem has
not been resolved yet.
Latency demand is another crucial issue for blockchain
technology stakeholders. Different scenarios are used to eval-
uate the proposed blockchain architecture [78], and smart
contracts. Minimized latency is needed when communication
and computational capabilities are transferred close to the
sensor nodes. To ease the network scalability, mobility, and
location awareness, minimized latency is required for the
devices connected in a physical environment. Therefore, low
latency is favorable to cloud computing users.
A security layer is added by blockchain technology to the
data transactions on distributed cloud storage. A trusted third
party is no more required as a smart contract is employed
that ensures the reliability and fairness of data trade. The
proposed research in [79] reduces the cost of transferring
the data on cloud storage. However, blockchain technology
suffers the limitations in communicating the same data con-
tents as claimed by the buyers. An attacker can resell the data
privately using other venues.
We have summarized and presented the proposed
approaches on data security issues regarding cloud com-
puting. Since the last few years, cloud computing with big
data has grown dramatically, and data security remains a
paramount concern. Cloud services availability has conﬂict-
ing constraints for the varied services. A semantic-based
access control (SBAC) approach [80] acquires the ﬁnancial
services in clouds. The proposed approach enables the access
FIGURE 6. Taxonomy of studied literature to answer research questions.
to users to access the various media on different platforms.
It also meets the MDB requirements and also protects sensi-
tive information.
E. TAXONOMY DESCRIPTION
As the volume of literature is large on cloud security, it is
unclear how the current research represents every class of
opinion. Current research fuses some critical viewpoints of
what constitutes the prominent as well as cloud security
57802 VOLUME 9, 2021

TABLE 4. Blockchain based propsoed studies.
challenges. This paper provides a substantial growing liter-
ature on cloud computing security challenges, commercial
cloud services providers, cloud consumers’ concerns, and
blockchain technology. The proposed taxonomy, as shown
in Fig. 6, is based on three levels. Existing literature other
than the survey papers is the top one level followed by four
categories on the second level. Categories at the second level
are based on our proposed research questions (RQ1...RQ4).
On the third level of the proposed taxonomy, we observe
the different research perspectives and contributions towards
answering the research question at the second level. The sec-
ond level of the proposed taxonomy further highlights the
challenges for category 1 (RQ1) and category (RQ3) that
remain open. This taxonomy not only validates our paper’s
contribution; instead, it provides the opportunity for further
discussion. We can extend this taxonomy by adding more
categories at the second level. The fourth level of this tax-
onomy accounts for the focused papers on the respective
categories.
F. CROSS ANALYSIS
Fig. 6 shows a taxonomic analysis and categorizes domains in
a pictorial representation. A taxonomy was developed using
the cloud computing security approaches. This survey paper
further extends to several research directions of the proposed
taxonomy (see Fig. 6). It is crucial to introspect the cur-
rent approaches on cloud security, cloud security addressed
by commercial cloud computing services providers, and
blockchain technology. Security threats have been studied
in the literature than consumers’ concern (42% vs. 17%),
while security techniques with blockchain technology are
studies in 25% of studies. On the other hand, 17% of studies
presented commercial CC service providers.
V. LIMITATIONS
This systematic literature review covers a large number of
studies to answer the designed research questions. We are
sure that a systematic literature review covers the cloud secu-
rity issues, cloud security models, risk mitigation strategies,
security issues to CSPs and users, and the role of blockchain
technology, which are published to date. One of the limita-
tions include the simpliﬁed search keywords used to search
for research articles in this SLR. However, those search key-
words can be further generalized. It is unnecessary that the
authors of research articles were native English speakers,
so there is always a risk that some terms used in research
studies may be misinterpreted. However, this deﬁciency has
been overcome by double-checking the result and discus-
sion sections to verify English language terms’ correct use.
Another limitation of this SLR is excluding very recently
published research studies supposed to be included in this
SLR to answer the RQs but were not included.
VOLUME 9, 2021 57803

TABLE 5. Survey papers on cloud computing topics.
57804 VOLUME 9, 2021

VI. CONCLUSION AND FUTURE WORKS
First, in this SLR, we have reviewed the literature on cloud
computing topics, including cloud security threats and their
mitigation strategies. We identified several security risks
to cloud computing. Data tampering and leakage is one of
the identified risks. Consumers’ trustworthiness, data out-
sourcing, and its associated risks are significant challenges
identified in this SLR. This SLR identified commercial cloud
services providers and highlighted the security issues they
face during cloud services deployment and implementation.
The trustworthiness of cloud users is challenging to con-
sumers of commercial cloud services providers. Data unavail-
ability, insufficient security measures, and vendor lock-in,
lack of interoperability and standards are identified
additionally to above-mentioned issues.
Moreover, we identified that Tweeter data generates and
is used to evaluate the proposed CC approaches. This SLR
identified that researchers had rarely used Facebook and
Instagram data for the evaluation of proposed strategies. Dur-
ing the CC deployment and implementations, data security
and privacy are concerns that a cloud adopting must consider
before using the cloud services. Blockchain technology is
found as an emerging technology to alleviate the security
concerns in the CC environment.
Cloud computing services have significant advantages for
vendors and users but need to bridge security gaps for cloud
users. Overall, this SLR claimed that security was the most
critical issue for users and CSPs. Literature review supported
our claim and thus it is suggested to propose an appro-
priate implementation of cloud computing security policies
and standards. We have presented some recommendations
in Table 4, which can be practiced, and implemented in future
works.
APPENDIX
See Table 5.
REFERENCES
[1] P. T. Jaeger, J. Lin, and J. M. Grimes, ‘‘Cloud computing and information
policy: Computing in a policy cloud?’’ J. Inf. Technol. Politics, vol. 5, no. 3,
pp. 269–283, Oct. 2008.
[2] C. Vidal and K.-K. R. Choo, ‘‘Situational crime prevention and the mit-
igation of cloud computing threats,’’ in Proc. Int. Conf. Secur. Privacy
Commun. Syst. Springer, 2017, pp. 218–233.
[3] N. Khan and A. Al-Yasiri, ‘‘Cloud security threats and techniques to
strengthen cloud computing adoption framework,’’ in Cyber Security and
Threats: Concepts, Methodologies, Tools, and Applications. Hershey, PA,
USA: IGI Global, 2018, pp. 268–285.
[4] H. A. Kholidy, A. Erradi, S. Abdelwahed, and F. Baiardi, ‘‘A risk mitigation
approach for autonomous cloud intrusion response system,’’ Computing,
vol. 98, no. 11, pp. 1111–1135, Nov. 2016.
[5] C.-J. Chung, P. Khatkar, T. Xing, J. Lee, and D. Huang, ‘‘NICE: Net-
work intrusion detection and countermeasure selection in virtual net-
work systems,’’ IEEE Trans. Dependable Secure Comput., vol. 10, no. 4,
pp. 198–211, Jul. 2013.
[6] J.-Y. Park, S.-H. Na, and E.-N. Huh, ‘‘An optimal investment scheme based
on ATM considering cloud security environment,’’ in Proc. 11th Int. Conf.
Ubiquitous Inf. Manage. Commun., Jan. 2017, pp. 1–7.
[7] P. A. Boampong and L. A. Wahsheh, ‘‘Different facets of security in the
cloud,’’ in Proc. 15th Commun. Netw. Simulation Symp., 2012, pp. 1–7.
[8] K. Jamsa, Cloud Computing: SaaS, PaaS, IaaS, Virtualization, Business
Models, Mobile, Security and More. Burlington, MA, USA: Jones &
Bartlett, 2012.
[9] T. Dillon, C. Wu, and E. Chang, ‘‘Cloud computing: Issues and chal-
lenges,’’ in Proc. 24th IEEE Int. Conf. Adv. Inf. Netw. Appl., Apr. 2010,
pp. 27–33.
[10] A. Bouayad, A. Blilat, N. E. H. Mejhed, and M. El Ghazi, ‘‘Cloud com-
puting: Security challenges,’’ in Proc. Colloq. Inf. Sci. Technol., Oct. 2012,
pp. 26–31.
[11] M. K. Hussein, M. H. Mousa, and M. A. Alqarni, ‘‘A placement architec-
ture for a container as a service (CaaS) in a cloud environment,’’ J. Cloud
Comput., vol. 8, no. 1, p. 7, Dec. 2019.
[12] C. S. Alliance, ‘‘Top threats to cloud computing v1. 0,’’ Cloud Secur.
Alliance, Bellingham, WA, USA, White Paper 23, 2010.
[13] S. M. Habib, S. Ries, and M. Muhlhauser, ‘‘Cloud computing landscape
and research challenges regarding trust and reputation,’’ in Proc. 7th Int.
Conf. Ubiquitous Intell. Comput. 7th Int. Conf. Autonomic Trusted Com-
put., Oct. 2010, pp. 410–415.
[14] A. Chonka, Y. Xiang, W. Zhou, and A. Bonti, ‘‘Cloud security defence to
protect cloud computing against HTTP-DoS and XML-DoS attacks,’’ J.
Netw. Comput. Appl., vol. 34, no. 4, pp. 1097–1107, 2011.
[15] B. Halpert, Auditing Cloud Computing. Hoboken, NJ, USA: Wiley, 2011.
[16] H. AlJahdali, A. Albatli, P. Garraghan, P. Townend, L. Lau, and J. Xu,
‘‘Multi-tenancy in cloud computing,’’ in Proc. IEEE 8th Int. Symp. Service
Oriented Syst. Eng., Apr. 2014, pp. 344–351.
[17] M. A. Zardari, L. T. Jung, and M. N. B. Zakaria, ‘‘Hybrid multi-cloud data
security (HMCDS) model and data classification,’’ in Proc. Int. Conf. Adv.
Comput. Sci. Appl. Technol., Dec. 2013, pp. 166–171.
[18] N. Sultan and S. van de Bunt-Kokhuis, ‘‘Organisational culture and cloud
computing: Coping with a disruptive innovation,’’ Technol. Anal. Strategic
Manage., vol. 24, no. 2, pp. 167–179, Feb. 2012.
[19] B. Tomas and B. Vuksic, ‘‘Peer to peer distributed storage and computing
cloud system,’’ in Proc. ITI 34th Int. Conf. Inf. Technol. Interfaces, 2012,
pp. 79–84.
[20] H. Tianfield, ‘‘Security issues in cloud computing,’’ in Proc. IEEE Int.
Conf. Syst., Man, Cybern. (SMC), Apr. 2012, pp. 1082–1089.
[21] A. Yamada, Y. Miyake, K. Takemori, A. Studer, and A. Perrig, ‘‘Intrusion
detection for encrypted Web accesses,’’ in Proc. 21st Int. Conf. Adv. Inf.
Netw. Appl. Workshops (AINAW), 2007, pp. 569–576.
[22] K.-L. Tsai, F.-Y. Leu, and J.-S. Tan, ‘‘An ECC-based secure EMR transmis-
sion system with data leakage prevention scheme,’’ Int. J. Comput. Math.,
vol. 93, no. 2, pp. 367–383, Feb. 2016.
[23] N. Kumar, V. Verma, and V. Saxena, ‘‘A security algorithm for online
analytical processing data cube,’’ Int. J. Comput. Appl., vol. 79, no. 14,
pp. 7–10, Oct. 2013.
[24] N. Tirthani and R. Ganesan, ‘‘Data security in cloud architecture based on
Diffie Hellman and elliptical curve cryptography,’’ IACR Cryptol. ePrint
Arch., vol. 2014, p. 49, 2014.
[25] O. Alowolodu, B. Alese, A. Adetunmbi, O. Adewale, and O. Ogundele,
‘‘Elliptic curve cryptography for securing cloud computing applications,’’
Int. J. Comput. Appl., vol. 66, no. 23, pp. 1–8, 2013.
[26] V. Varadharajan and U. Tupakula, ‘‘Security as a service model for
cloud environment,’’ IEEE Trans. Netw. Service Manage., vol. 11, no. 1,
pp. 60–75, Mar. 2014.
[27] R. Farrell, ‘‘Securing the cloud—Governance, risk, and compliance
issues reign supreme,’’ Inf. Secur. J.: Global Perspective, vol. 19, no. 6,
pp. 310–319, 2010.
[28] J. Cheng, S. Qi, W. Wang, Y. Yang, and Y. Qi, ‘‘Fast consistency auditing
for massive industrial data in untrusted cloud services,’’ in Proc. Great
Lakes Symp. VLSI, Sep. 2020, pp. 381–386.
[29] S. S. Rizvi, T. A. Bolish, and J. R. Pfeffer, ‘‘Security evaluation of cloud
service providers using third party auditors,’’ in Proc. 2nd Int. Conf.
Internet things, Data Cloud Comput., Mar. 2017, pp. 1–6.
[30] S. S. Manvi and G. Krishna Shyam, ‘‘Resource management for infrastruc-
ture as a service (IaaS) in cloud computing: A survey,’’ J. Netw. Comput.
Appl., vol. 41, pp. 424–440, May 2014.
[31] G. Yan, D. Wen, S. Olariu, and M. C. Weigle, ‘‘Security challenges in
vehicular cloud computing,’’ IEEE Trans. Intell. Transp. Syst., vol. 14,
no. 1, pp. 284–294, Mar. 2013.
[32] G. Xu, W. Yu, Z. Chen, H. Zhang, P. Moulema, X. Fu, and C. Lu,
‘‘A cloud computing based system for cyber security management,’’ Int.
J. Parallel, Emergent Distrib. Syst., vol. 30, no. 1, pp. 29–45, 2015.
[33] N. Garcia, T. Alcaniz, A. González-Vidal, J. B. Bernabe, D. Rivera,
and A. Skarmeta, ‘‘Distributed real-time SlowDoS attacks detection over
encrypted traffic using artificial intelligence,’’ J. Netw. Comput. Appl.,
vol. 173, Jan. 2021, Art. no. 102871.
VOLUME 9, 2021 57805

[34] B. Kitchenham, R. Pretorius, D. Budgen, O. P. Brereton, M. Turner,
M. Niazi, and S. Linkman, ‘‘Systematic literature reviews in soft-
ware engineering–a tertiary study,’’ Inf. Softw. Technol., vol. 52, no. 8,
pp. 792–805, 2010.
[35] Endnote X9. Accessed: Jan. 12, 2021. [Online]. Available:
https://endnote.com
[36] J. Wen, S. Li, Z. Lin, Y. Hu, and C. Huang, ‘‘Systematic literature review of
machine learning based software development effort estimation models,’’
Inf. Softw. Technol., vol. 54, no. 1, pp. 41–59, Jan. 2012.
[37] B. H. Kim, W. Huang, and D. Lie, ‘‘Unity: Secure and durable personal
cloud storage,’’ presented at the CCSW ACM Workshop Cloud Comput.
Secur. Workshop, New York, NY, USA, 2012, pp. 31–36.
[38] M. Nojoumian and D. R. Stinson, ‘‘Social secret sharing in cloud comput-
ing using a new trust function,’’ in Proc. 10th Annu. Int. Conf. Privacy,
Secur. Trust, Washington, DC, USA, Jul. 2012, pp. 161–167.
[39] H. Eken, ‘‘Security threats and solutions in cloud computing,’’ in Proc.
World Congr. Internet Secur. (WorldCIS-), London, U.K., Dec. 2013,
pp. 139–143.
[40] N. Deluca. (2012). IT Consultants. Insight on Business Technol-
ogy NSK Inc. [Online]. Available:http://blog.nskinc.com/IT-Services-
Boston/bid/101166/Cloud-Security-Why-Clouds-are-Safe-and-Sound
[41] M. B. Mollah, K. R. Islam, and S. S. Islam, ‘‘Next generation of computing
through cloud computing technology,’’ presented at the Elect. Comput.
Eng. (CCECE), 25th IEEE Canadian Conf., Montreal, QC Canada, 2012.
[42] GED.-i. Ltd. (2013). Cloud Computing. [Online]. Available:
http://www.ged-i.com/—–use-case.htm#Cloudconfig1
[43] A. M. Talib, R. Atan, R. Abdullah, and M. A. A. Murad, ‘‘Security
framework of cloud data storage based on multi agent system architecture:
Semantic literature review,’’ Comput. Inf. Sci., vol. 3, no. 4, pp. 175–186,
Oct. 2010.
[44] S. Sakr, A. Liu, D. M. Batista, and M. Alomari, ‘‘A survey of large scale
data management approaches in cloud environments,’’ IEEE Commun.
Surveys Tuts., vol. 13, no. 3, pp. 311–336, 3rd Quart., 2011.
[45] D. C. Marinescu, Cloud Computing: Theory and Practice. Waltham, MA,
USA: Elsevier, 2013.
[46] J. W. Bos. (2013). Elliptic Curve Cryptography in Practice. [Online].
Available: https://eprint.iacr.org/2013/734.pdf
[47] A. Sajid and H. Abbas, ‘‘Data privacy in cloud-assisted healthcare systems:
State of the art and future challenges,’’ J. Med. Syst., vol. 40, no. 6, p. 155,
Jun. 2016.
[48] K. Janjua, M. A. Shah, A. Almogren, H. A. Khattak, C. Maple, and
I. U. Din, ‘‘Proactive forensics in IoT: Privacy-aware log-preservation
architecture in fog-enabled-cloud using holochain and containerization
technologies,’’ Electronics, vol. 9, no. 7, p. 1172, Jul. 2020.
[49] S. Shamshirband, M. Fathi, A. T. Chronopoulos, A. Montieri, F. Palumbo,
and A. Pescapè, ‘‘Computational intelligence intrusion detection tech-
niques in mobile cloud computing environments: Review, taxonomy,
and open research issues,’’ J. Inf. Secur. Appl., vol. 55, Dec. 2020,
Art. no. 102582.
[50] J. K. Samriya and N. Kumar, ‘‘A novel intrusion detection system using
hybrid clustering-optimization approach in cloud computing,’’ Mater.
Today, Proc., pp. 1–6, 2020, doi: 10.1016/j.matpr.2020.09.614.
[51] A. Bentajer, M. Hedabou, K. Abouelmehdi, and S. Elfezazi, ‘‘CS-IBE:
A data confidentiality system in public cloud storage system,’’ Procedia
Comput. Sci., vol. 141, pp. 559–564, 2018.
[52] M. Rady, T. Abdelkader, and R. Ismail, ‘‘Integrity and confidentiality in
cloud outsourced data,’’ Ain Shams Eng. J., vol. 10, no. 2, pp. 275–285,
Jun. 2019.
[53] M. Alshehri, ‘‘An effective mechanism for selection of a cloud service
provider using cosine maximization method,’’ Arabian J. Sci. Eng., vol. 44,
no. 11, pp. 9291–9300, 2019.
[54] P. Varshney and Y. Simmhan, ‘‘Characterizing application scheduling on
edge, fog, and cloud computing resources,’’ Softw., Pract. Exper., vol. 50,
no. 5, pp. 558–595, May 2020.
[55] H. Zhang, P. Li, Z. Zhou, J. Wu, and X. Yu, ‘‘A privacy-aware virtual
machine migration framework on hybrid clouds,’’ J. Netw., vol. 9, no. 5,
p. 1086, May 2014.
[56] L. Heilig, E. Lalla-Ruiz, and S. Voß, ‘‘Modeling and solving cloud service
purchasing in multi-cloud environments,’’ Expert Syst. Appl., vol. 147,
Jun. 2020, Art. no. 113165.
[57] C. Ngo, Y. Demchenko, and C. de Laat, ‘‘Multi-tenant attribute-based
access control for cloud infrastructure services,’’ J. Inf. Secur. Appl.,
vols. 27–28, pp. 65–84, Apr. 2016.
[58] R. Duncan, ‘‘A multi-cloud world requires a multi-cloud security
approach,’’ Comput. Fraud Secur., vol. 2020, no. 5, pp. 11–12, May 2020.
[59] B. Scott, ‘‘How a zero trust approach can help to secure your AWS
environment,’’ Netw. Secur., vol. 2018, no. 3, pp. 5–8, Mar. 2018.
[60] M. Ramachandran, ‘‘Software security requirements management as an
emerging cloud computing service,’’ Int. J. Inf. Manage., vol. 36, no. 4,
pp. 580–590, Aug. 2016.
[61] G. Adamson, L. Wang, M. Holm, and P. Moore, ‘‘Cloud manufacturing–a
critical review of recent development and future trends,’’ Int. J. Comput.
Integr. Manuf., vol. 30, nos. 4–5, pp. 347–380, 2015.
[62] P. Singh, Y. K. Dwivedi, K. S. Kahlon, R. S. Sawhney,
A. A. Alalwan, and N. P. Rana, ‘‘Smart monitoring and controlling
of government policies using social media and cloud computing,’’ Inf.
Syst. Frontiers, pp. 315–337, Apr. 2019.
[63] S. Gupta, S. C. Misra, N. Kock, and D. Roubaud, ‘‘Organizational, tech-
nological and extrinsic factors in the implementation of cloud ERP in
SMEs,’’ J. Organizational Change Manage., vol. 31, no. 1, pp. 83–102,
Feb. 2018.
[64] A. Albugmi, R. Walters, and G. Wills, ‘‘A framework for cloud computing
adoption by saudi government overseas agencies,’’ in Proc. 5th Int. Conf.
Future Commun. Technol. (FGCT), Aug. 2016, pp. 1–5.
[65] P. Velmurugadass, S. Dhanasekaran, S. Shasi Anand, and V. Vasudevan,
‘‘Enhancing blockchain security in cloud computing with IoT environment
using ECIES and cryptography hash algorithm,’’ Mater. Today: Proc.,
vol. 37, pp. 2653–2659, 2021.
[66] A. Wilczyński and J. Kołodziej, ‘‘Modelling and simulation of security-
aware task scheduling in cloud computing based on blockchain
technology,’’ Simul. Model. Pract. Theory, vol. 99, Feb. 2020,
Art. no. 102038.
[67] J. Cha, S. K. Singh, T. W. Kim, and J. H. Park, ‘‘Blockchain-empowered
cloud architecture based on secret sharing for smart city,’’ J. Inf. Secur.
Appl., vol. 57, Mar. 2021, Art. no. 102686.
[68] H. Huang, X. Sun, F. Xiao, P. Zhu, and W. Wang, ‘‘Blockchain-based
eHealth system for auditable EHRs manipulation in cloud environments,’’
J. Parallel Distrib. Comput., vol. 148, pp. 46–57, Feb. 2021.
[69] Y. Ren, Y. Leng, J. Qi, P. K. Sharma, J. Wang, Z. Almakhadmeh, and
A. Tolba, ‘‘Multiple cloud storage mechanism based on blockchain in
smart homes,’’ Future Gener. Comput. Syst., vol. 115, pp. 304–313,
Feb. 2021.
[70] J. Li, J. Wu, G. Jiang, and T. Srikanthan, ‘‘Blockchain-based public audit-
ing for big data in cloud storage,’’ Inf. Process. Manage., vol. 57, no. 6,
Nov. 2020, Art. no. 102382.
[71] N. Eltayieb, R. Elhabob, A. Hassan, and F. Li, ‘‘A blockchain-based
attribute-based signcryption scheme to secure data sharing in the cloud,’’
J. Syst. Archit., vol. 102, Jan. 2020, Art. no. 101653.
[72] D. C. Nguyen, P. N. Pathirana, M. Ding, and A. Seneviratne, ‘‘Blockchain
for secure EHRs sharing of mobile cloud based E-Health systems,’’ IEEE
Access, vol. 7, pp. 66792–66806, 2019.
[73] Q. Su, R. Zhang, R. Xue, and P. Li, ‘‘Revocable attribute-based sig-
nature for blockchain-based healthcare system,’’ IEEE Access, vol. 8,
pp. 127884–127896, 2020.
[74] Y. He, Y. Wang, C. Qiu, Q. Lin, J. Li, and Z. Ming, ‘‘Blockchain-based
edge computing resource allocation in IoT: A deep reinforcement learn-
ing approach,’’ IEEE Internet Things J., vol. 8, no. 4, pp. 2226–2237,
Feb. 2021.
[75] Z. Li, Z. Yang, and S. Xie, ‘‘Computing resource trading for edge-cloud-
assisted Internet of Things,’’ IEEE Trans. Ind. Informat., vol. 15, no. 6,
pp. 3661–3669, Jun. 2019.
[76] M. Zhaofeng, W. Lingyun, W. Xiaochang, W. Zhen, and Z. Weizhe,
‘‘Blockchain-enabled decentralized trust management and secure usage
control of IoT big data,’’ IEEE Internet Things J., vol. 7, no. 5,
pp. 4000–4015, May 2020.
[77] S. Algarni, F. Eassa, K. Almarhabi, A. Almalaise, E. Albassam, K. Alsubhi,
and M. Yamin, ‘‘Blockchain-based secured access control in an IoT sys-
tem,’’ Appl. Sci., vol. 11, no. 4, p. 1772, Feb. 2021.
[78] T. M. Fernández-Caramés, I. Froiz-Míguez, O. Blanco-Novoa, and
P. Fraga-Lamas, ‘‘Enabling the Internet of mobile crowdsourcing health
things: A mobile fog computing, blockchain and IoT based continuous glu-
cose monitoring system for diabetes mellitus research and care,’’ Sensors,
vol. 19, no. 15, p. 3319, Jul. 2019.
[79] Y. Li, X. Feng, J. Xie, H. Feng, Z. Guan, and Q. Wu, ‘‘A decentralized
and secure blockchain platform for open fair data trading,’’ Concurrency
Comput., Pract. Exper., vol. 32, no. 7, Apr. 2020.
[80] Y. Wang, B. Rawal, and Q. Duan, ‘‘Securing big data in the cloud with
integrated auditing,’’ in Proc. IEEE Int. Conf. Smart Cloud (SmartCloud),
Nov. 2017, pp. 126–131.
57806 VOLUME 9, 2021

BADER ALOUFFI received the Ph.D. degree.
He is currently an Assistant Professor with the
College of Computer and IT, Taif University. His
research interests include software engineering,
temporal logic, runtime veriﬁcation, data science,
and AI.
MUHAMMAD HASNAIN received the M.S.
degree in software engineering from Abasyn Uni-
versity, Pakistan, in 2016. His research interests
include software performance regression testing,
web services scalability improvement, machine
learning, and cloud computing.
ABDULLAH ALHARBI received the Ph.D. degree from the University of
Technology Sydney, Australia. He is currently an Assistant Professor with
the Information Technology Department, Taif University. His research inter-
ests include human–computer interaction, information systems modeling,
big data, streaming data analytics, cybersecurity, and data science.
WAEL ALOSAIMI was born in Saudi Arabia, in 1979. He received the B.Sc.
degree in computer engineering from King Abdulaziz University, in 2002,
and the M.Sc. degree in computer systems security and the Ph.D. degree
in cloud security from the University of South Wales, in 2011 and Novem-
ber 2016, respectively. From 2002 to 2004, he worked with Saline Water
Conversion Corporation (SWCC), as an Instrument and Control Engineer.
He served as a Trainer for Technical and Vocational Training Corporation
(TVTC), until 2008. He joined Taif University, as a Teaching Assistant.
It provides him with a scholarship to pursue his studies in the U.K. Since
2017, he has been an Assistant Professor with the Computer Engineering
Department, Taif University. He has many publications in peer-reviewed
conferences and journals. His research interests include cloud computing,
cloud security, information security, network security, e-health security,
the Internet of Things security, and data science.
HASHEM ALYAMI received the bachelor’s degree in computer science
from Taif University, Saudi Arabia, in 2007, the master’s degree in secure
computer system from the University of Hertfordshire, U.K., and the Ph.D.
degree from the University of Reading, U.K. He is currently an Assistant
Professor with the Computer and Information Technology College, Taif
University. His research interests include cybersecurity, artiﬁcial intelligent,
and data science.
MUHAMMAD AYAZ received the B.S. and M.S.
degrees in computer science. He is currently pursu-
ing the Ph.D. degree with the School of IT, Monash
University, Australia. His research interests
include eHealth, software engineering, databases,
big data computing, and cloud computing.
VOLUME 9, 2021 57807

A Systematic Literature Review On Cloud Computing Security Threats And Mitigation Strategies

More Related Content

Similar to A Systematic Literature Review On Cloud Computing Security Threats And Mitigation Strategies (20)

More from Claire Webber (20)

Recently uploaded (20)

A Systematic Literature Review On Cloud Computing Security Threats And Mitigation Strategies