PRISMA Archetype-Based Systematic Literature Review of Security Algorithms in the Cloud

,


Introduction
As data production and utilization have expanded, so have the difculties and opportunities. To preserve this massive amount of data, a paradigm shift in data storage, security, integrity, and availability is required [1]. Te use of cloud computing can help solve these issues. Cloud computing is the delivery of computer services through the Internet on a pay-as-you-go basis [2].
Te use of cloud computing enables the use of software and infrastructure anywhere on the earth, with all services administered by cloud service providers [3]. Cloud computing has been on the agenda of organizations and governments throughout the world in order to achieve lower operating costs and fexibility of data capabilities, which are believed to be the greatest information technology solutions [4]. Te adoption of the cloud by governments and businesses has resulted in the emergence of several service providers such as Salesforce, Amazon, and Yahoo, with many suppliers such as IBM and Oracle ofering database technical support [5].
Despite the multiple benefts and the drive by other organizations to contribute to the sustainability of cloud computing, data security remains a key concern [6]. Tis is due to the various cloud computing architecture and designs, such as software, hardware, and application programming interfaces [7]. However, because of this disparity in setup, cloud customers and providers face a variety of security challenges [6,7]. Tis is seen in [8] where Imperva warns clients to be on the lookout for a new attack on cloud services known as man-in-the-cloud.

Services and Cloud Deployment Models.
Tis section provides an overview of the various cloud computing service models and a quick assessment of the subject at hand. (SaaS). Software as a service refers to the practice of a third party providing software to several tenants on a pay-per-use basis. Clients from small and big enterprises can access these systems, which only require the deployment of software once. An integrated system that is always in use on the internet may call for both routine modifcations and innovative activities [9].

Platform as a Service (PaaS).
Platform as a Service [10] provides an environment that permits the creation, development, and maintenance of applications. Te produced apps may be immediately planned, improved, and evaluated by the cloud customers, and the development cycle of the applications can be tracked. (IaaS). All other cloud services are based on the foundation of Infrastructure as a Service as indicated in Figure 1. In the typical network design, this takes the role of conventional data centers. Tis concept is used by cloud service providers to ofer platforms on which cloud clients may store their resources [11]. Resources from cloud customers are transferred to IaaS on the assumption that the cloud service provider can maintain the level of service they ofer. Te service-level agreement (SLA), which is connected to the lifespan of the cloud service provider and exhibits the fnancial as well as procedural dynamism related to SLA, serves as a guarantee for the use of IaaS. (CaaS). Developers that utilize this service use a package for all of their programming needs. Te container includes all of the coding requirements, run time, and confgurations needed for the system to operate on a host computer [12]. Te libraries required to run a program are all provided by a container as a service, which eliminates the need for additional virtual systems to supply the necessary libraries as shown in Figure 2. For uploading, setting up, running, scaling, and maintaining the container, they can ofer a complete unit.

Cloud Deployment Models.
Te distinctiveness of gaining access to shared resources in the cloud is determined by the deployment models in cloud computing. Based on this, four models are taken into consideration.

Public
Cloud. According to their shared objective, this kind of cloud enables entities to access data over the internet and makes programs accessible to the group with the aid of cloud servers [13]. Tese clouds are made available to the public at large, are managed by governmental bodies, companies, academics, or a combination of all three, and are hosted by a cloud service provider on their website [14].

Private
Cloud. Private clouds, according to the authors in reference [15], are designed to be used by businesses to carry out work or store employees' details. Such a cloud platform is trademarked since it specifcally saves the entity's extremely sensitive data. Private clouds are utilized as a single entity scheme, and they operate using either new resources or existing technology that is housed on the organization's hardware but is managed by a diferent frm [16].

Hybrid Cloud.
A hybrid cloud is created by combining the strengths and weaknesses of public and private clouds, which has the beneft of improving data security for both infrastructures [17]. Hybrid clouds, according to the authors in [18], assert that their integration calls for a higher level of technical expertise in terms of data gathering, analysis, evaluation, and overall management of hybrid platforms. Hybrid clouds provide several challenges including data governance and security.

Community Cloud.
Tis is a sort of hybrid private cloud and is regarded as a multitenancy platform designed to let businesses use a shared resource [19]. Because they are utilizing the community software to accomplish a single objective, this enables the users to cooperate on a shared project as shown in Figure 3. On this platform, clients are all concerned with shared security as well as the guiding principles of agreement with the delegation of oversight and evaluation to a third party [21].
1.3. Cloud Client. Cloud customers are people or businesses who make use of a cloud service provider's resources. Tey have the absolute right to select the service of their choice and to pay for the services that the service provider really provides to them before their contract expires. Te cloud customer establishes a service level agreement to specify how well the service will be provided [22]. Tese contracts are signed in relation to service quality, privacy, security, and integrity.

Cloud Service Providers. Te term "Cloud Service
Provider" (CSP) refers to organizations that provide cloud clients with computing as a service. Cloud service providers are in charge of overseeing the management of all cloud services and infrastructure [23]. In Software as a Service (SaaS) and Infrastructure as a Service (IaaS) platforms, the cloud service provider is responsible for organizing, arranging, maintaining, and keeping up-to-date applications as well as managing infrastructure in order to give resources to cloud customers. All of the architectural planning and computer infrastructure, including networks, servers, and infrastructure hosting, are provided by the cloud service provider [24]. A detailed functionality of performers in the cloud is depicted in Table 1.

Security Challenges in Cloud Services
1.5.1. Security in Software as a Service. Te cloud computing interaction layer is represented by this. Terefore, all security concerns are data based [26]. Tis is because, at this point, it is up to the cloud client to ensure the necessary security for the data of-loaded by implementing checks on who may access such data as well as the security measures used by the cloud service provider. Te most frequent security concerns relating to software as a service are lack of control, access management, data privacy, and continuous monitoring. Due to the aforementioned problems, it is crucial to take into account how cloud providers and SaaS providers relate to one another in terms of security. Tis necessitates careful examination of the suppliers' security measures.

Security in Platform as a Service (PaaS).
Tere are three layers for the platform as a service, according to [27]. Te layer that connects to software as a service, the middle layer is intended for application storage, and database data run timing management, and the last layer is for back-end operations including network, storage, and CPU storage. Te security concerns are data breaches and security controls as suggested by [28]. As a result of these security concerns, cloud service providers need to make sure that adequate identity and verifcation processes are in place for PaaS.

Security in Infrastructure as a Service (IaaS).
To store their data and optimize their CPU and other functions, cloud clients employ virtual devices at this service level. Based on how frequently the service is accessed, this layer has several security problems. Security issues that have been raised include denial of service attacks, limited control, and compromised identity [29]. Tis necessitates the establishment of appropriate legal provisions and guidelines for cloud clients using IaaS.

Security in Container as a Service (CaaS).
Because of benefts such as being light, quick, easy to deploy, improved resource usage, and version control, the adoption of containers as a service has expanded [30].
Te following are a few security issues with CaaS [30]: well construction of container images and requiring new security methodology. Since clients are permitted to share      [25].

Performer Function
Cloud client An entity that uses the services rendered by a cloud service provider based on pay as you use

Service provider
Any company that renders computing service to cloud clients via the internet and ensures the provision of resources the clients require to attain the satisfaction desired Cloud auditor Tis is a third-party responsible for the evaluation of services rendered to a cloud client. Tis is in the form of assessing performance, system operation, and security Broker of cloud An entity that mediates between cloud client and cloud service provider seeing to quality service and delivery Carrier of cloud Tey ensure connectivity between cloud clients and cloud service providers for the transport of cloud services the same OS, the OS kernels must be more secure to make Container as a Service more secure. Te security issues with the diferent cloud services are indicated in Table 2.
1.6. Cloud Security Issues 1.6.1. Confdentiality. Confdentiality is the prevention of unauthorized parties from accessing data [31]. To accomplish this, several researchers have employed a variety of strategies, ranging from cryptography to the combination of encryption and block division [32]. Examples of security measures implemented to protect data confdentiality include the following: (1) Encryption Using a Biometric Approach. Tis method makes use of a variety of factors, including speech transmissions, eye iris readings, facial recognition, and fngerprint scans. To obtain access to the stored data, these systems are quite diferent and challenging to modify [33].
(2) Using the Classifcation Approach of K-NN. Tis is one of the most sophisticated methods for ensuring data security, and it is regarded as a supervised machine-learning method [34]. Tis is frequently used in pattern recognition, data segmentation, forecasting, and approximation with the goal of choosing insightful data that enables data secrecy.
(3) A Secure Scheme Using HPI. By utilizing HTTP's (hypertext transfer protocol) security protocol, this method enables cloud clients to store data in the cloud [32]. When the data are requested, they are unscrambled after being scrambled and transferred to the cloud. Because of this, the end encryption strategy increases secrecy.

1.7.
Integrity. Data integrity guarantees the accuracy of clients' data by demonstrating that their data are secure and have not been altered using the right cryptographic algorithms [35]. Data integrity is concerned with preventing the loss of consumer's information. Due to the fact that clients often access data from the cloud service provider, this is quite crucial. Data integrity is attained through the following:

Verifcation Based on BLS Signature.
A security method called the Boneh-Lynn-Shacham (BLS) signature is used to verify a signer's identity. Tis approach is built on the fundamentals of an elliptic curve and employs a bilinear pair for authentication. Tis increases its defenses against an index attack. Tis was used in the works of the authors in reference [36], who recognized the conventional faws in privacy-preserving models' ability to guarantee data integrity. Key generation, token generation, challenge, response, and check proof are the fve steps of verifability in their approach. Tis strategy encourages auditing verifcation since it guarantees accurate verifcation.

Blockchain.
Blockchain has been proposed as a cryptosystem alternative for protecting data integrity in the cloud. Tis is seen in the works of the authors in reference [37], where an integrated linear mapping technique was applied. As a result, third party auditors' trust issues are lessened while also saving signifcant computational and connection overheads. To create tags for the sample verifcation, the message is sliced and homomorphically verifed.
1.8. Availability. Data availability guarantees that data stored in the cloud are accessible to its owners. Tis seeks to retrieve data in its entirety. Te cloud client wants to ensure that none of the cloud service provider's internal data failures, device malfunctions, software defects, or other cloud dangers have had any impact on the data [38]. Replication of data is used nowadays to provide data availability, and the following setup by the cloud client automatically replicates data on two or more virtual servers. Amazon S3 and Google Cloud are the two well-known companies that provide multidata duplication at many locations [39].
1.9. Proposed Techniques Used to Secure the Cloud 1.9.1. Firewall. A frewall is implemented to guarantee protection against host and network threats. Tis makes it a useful security technique that may be applied to guarantee cloud security. Te connection of devices can be evaluated and regulated by a frewall [40]. Tis aids in thwarting attacks such as cross-virtual machines (VM) and Economic Denial of Service (EDoS) [41]. Shielding internal nodes from outside threats aids in securing the entry of autonomous architecture and the system's security. Because cloud computing is dynamic, it might reduce the inner and outside security benefts that a frewall provides. Terefore, the external parties use rented instances to run their program [41]. Tis makes using frewall to be noted to maintain the privacy and security of cloud data quite safe.

Encryption.
Data security in the cloud is achieved by using the right cryptographic technique. By using this method, the message is rendered unintelligible [42]. Te encryption key used to carry out the encryption process determines how strong such methods are. Prime factorization, the foundation of an algorithm such as the RSA developed by Rivest, Shamir, and Adleman, is challenging to calculate in discrete logarithmic time. Te cloud is protected by several cryptographic methods, including advanced encryption scheme (AES), DES, and Blowfsh. Blowfsh, AES, SHA1, and DES are just a few of the other integration techniques that are employed. All of them are used to guarantee cloud security. Te most frequent attack on cryptography is thought to be a brute force attack.

Data Masking.
Data masking is the ability to conceal genuine data from their natural structure while maintaining their authenticity to stop data leakage. Tis is viewed as Security and Communication Networks Container as a service (CaaS) (i) Te sharing of OS used by the host possess a security threat (i) Critical concentration on run time is needed (ii) Security of apps in the container (ii) Securing container-to-container activities a bridge connecting the token technique and encryption. By masking some parts of the message that consumers are not meant to view has the feature of hiding the actual data [43]. Based on legal restrictions, this method enables outsourcing, allocation, afliation, and the use of cloud technologies.
Both types of dynamic and static data masking are taken into consideration. Dynamic masking is the use of selective concealment depending on legal considerations for data readers, providing security to sensitive data equivalent to plaintext without any scrambling properties. Data that have been hidden from view thanks to static data masking are irrevocable.

Blockchain (Distributed Ledger Technology).
Blockchain is the ever-evolving technology that writers claim can potentially safeguard data in this era of information growth. Te list of blocks is protected cryptographically since it is organized in hierarchical tiers. Using the connection of widely dispersed computers, their activities are organized in a peer style [44]. It prevents data loss, alteration, or manipulation by storing a duplicate of the mirrored data on each machine in the network. Tis contributes to improving the security of the data being managed.
Several studies, such as those conducted by the authors in references [45,46], have attempted to address these security vulnerabilities. However, the present cryptographic techniques are incapable of withstanding contemporary security threats that target cloud customers and providers due to proportionality between data size and run time, making security a major setback to the full adoption of cloud computing. Again, as linear run times are produced as a result of the relationship between data size and run time, there is excessive CPU engagement creating wear and tear on client and provider equipment. Furthermore, such schemes require additional data transport bandwidth when large data are to be transferred [47].
As a result of the importance of securing data on the cloud, this paper conducts a systematic literature review of various published articles aimed at securing the cloud [48]. Also, this study unravels the type of cryptographic algorithms employed (symmetric, asymmetric, or protocol) [49] to attain cloud data security. Again, the trend of run times of these cryptographic algorithms (linear/nonlinear time), the purpose of these cryptographic algorithms, and cloud security concerns are also investigated by this study.

Identifed Problem.
Cloud computing is a growing and progressive technique to ofer ofshore storage and computing services that has become riskier in terms of security in recent years. Te control and administration of an organization's data and assets are at the mercy of a third party, exposing the data to a variety of vulnerabilities such as confdentiality, privacy, data leakage, data theft, dependability, capacity, and performance assessment. As a result of these security difculties, cryptographic approaches have been proposed by researchers as appropriate tools for ensuring the security of subscribers' data on the cloud. Despite the multiple cryptographic systems suggested by experts, security remains a barrier to cloud computing's widespread adoption. Again, the run time and data sizes of these cryptographic systems are proportional, suggesting that the larger the data size, the longer the execution time, making the algorithm's execution times linear (O(N)). Because data volumes are related to execution time, when large amounts of data are outsourced to the cloud service provider, it puts wear and tear on both the cloud service provider and the cloud client devices. Tis has necessitated a review of published articles from 2016 to 2022 on the most commonly used cryptographic scheme to secure the cloud, the type of cryptographic algorithms used (symmetric, asymmetric, or protocol) to achieve cloud data security, the run times of these cryptographic algorithms (linear or nonlinear time), the purpose of these cryptographic algorithms, cloud security concerns, and cloud security techniques.

Literature Review
Researchers have paid close attention to cloud computing security. Several conferences, including the 2nd International Conference on Electrical, Communication, and Computer Engineering, the 2nd World Congress on Computing and Communication Technologies, and the ACM International Conference Proceeding Series (2020), have focused on cloud computing security. Aside from these, the majority of other publications have committed to publishing cloud computing-related papers aimed at accomplishing cloud security. Tis section discusses a thorough evaluation of works performed by researchers on cloud security.
Te study of the authors in reference [50] looked at the Fusion-based Advanced Encryption Algorithm (FAEA) to ofer a cost-efective, workable security architecture for using big data in the cloud. Te performance of the FAEA approach was compared to that of the Map Reduce Encryption Scheme (MRE) and Hadoop Distributed File System (HDFS) and shows that it performed 98% better in terms of efciency, scalability, and security.
In the work of the authors in reference [51], they challenged the attackers with more advanced security measures using a powerful real-time service-centric feature sensitivity analysis (RSFSA) model. Te RSFSA model examines the sensitivity of various characteristics used by each service at several levels. Te method computes the FLAG value for the user from the provided profle by checking the set of features being accessed at each level and the number of features to which the user has access permissions. Te user has either been given access to the service or not, depending on FLAG's value. On the other hand, the technique maintains several encryption protocols and keys for every feature level. Te technique maintains a set of schemes and keys for each level-specifc feature since the features are grouped at diferent levels.
Muthulakshmi and Venkatesulu [52] proposed a revolutionary customized advanced encryption standard (AES) cryptographic algorithm. Te goal of their technique is to improve the performance of the AES algorithm by Security and Communication Networks shortening the cryptography process. Te notion of a chunk fle system is proposed in an attempt to increase the efciency of AES. Te input fle is chunked into numerous fles, allowing for fast and efcient encryption. A comparative study is performed using the current algorithms' lightweight keyword searchable encryption (LFSE) and cloud key management system (CKMS) to demonstrate the efectiveness of their suggested system. Te time required for key creation, encryption, and decryption is the basis for the comparative efort.
Rupa et al. [53] suggested a homomorphic encryption scheme based on matrix transformations using shifts, rotations, and transpositions of each letter in the plain text's binary transformed ASCII values. Te symmetric cryptography uses the same secret key for both encryption and decoding.
William et al. [25] published a paper in which the researchers proposed combining symmetric and asymmetric methods, which are also processed by the hashing algorithm. Te suggested approach frst turns the provided data into cipher text using an AES algorithm with a key size of 128, 192, or 256 bits. Te AES key is encrypted again using the Elliptical Curve Cryptography (ECC) technique. To construct the message digest, the encrypted text is again put through the Secure Hash Algorithm (SHA) 256 method. Te encrypted message and the encrypted AES key are both transferred over the network, where the encrypted AES key is frst decrypted using the ECC decryption technique, and then the AES decryption is performed using the retrieved key to recover the original plain text. Te SHA digest is used to verify data integrity. Te outcomes are computed for textual and picture datasets.
Data security in terms of attribute-based encryption, access control, and data integrity was evaluated by Rajeswari and Kalaiselvi in their study on cloud data storage [54]. Based on their fndings, they concluded that the compute overhead for data storage and security should be reduced and that cloud security might be improved through verifcation, approval, privacy, and integrity.
El-Attar et al. [55] proposed a hybrid automated approach to preserve secrecy while achieving great efciency during the encryption of huge data. Te suggested technique comprises random key creation utilizing the RSA algorithm to generate private and encrypted keys. Te data to be uploaded are separated into random-size blocks, and for each block, automated sequential cryptography and automated random cryptography are used. Te encrypted blocks are then saved in the cloud. Both sequential and random algorithms rely heavily on AES and DES methods. Both automated systems achieved a high degree of security as well as great efciency during encryption and decryption. Te fndings are also compared with better automated random cryptography based on the S-Box generator. When compared to the prior approaches, this new algorithm produced more efcient outcomes.
Data security was improved by Mani and Devi [56] by using preprocessing before encryption. Te frst level of security is achieved by encoding the provided plain text using the Lucas and Fibonacci series. Te second level of security is then achieved by further compressing the encoded text using the Hufman encoding, and the third level of security is attained by subjecting it to the RSA public key cryptographic algorithm. In the reverse process, the encrypted text is transformed back to the original plain text by being frst decrypted, then decompressed, and lastly decoded.
Khalid Yousif et al. [57] used the (NTRUEncrypt) algorithms in Hadoop to speed up the fle encryption and decryption procedures in their investigation. If HDFS is involved in the Map Task, it will handle both the encryption and decryption procedures. Te suggested protection approach, which employs cryptography, can keep data on the cloud private and safe.
Yang et al. [58] proposed a novel cloud-based parallel block Wiedemann technique for solving large and sparse linear problems over GF (2). Strip partitioning, cyclic partitioning, and modifed strip partitioning are included in the proposed parallel block Wiedemann method to parallelize distinct phases in the block Wiedemann process.
Trough the fusion of two approaches, Kumar [59] devised a cryptographic algorithm to attain security of cloud data. Tese two supplied a signifcantly more secure data security platform, namely, the DNA-based algorithm and the AES algorithm. Data encryption and decryption are performed using the DNA cryptographic technology and the AES method. DNA encryption enables you to encrypt a lot of data with only a small bit of DNA.
In the research study of Suganya and Sasipraba [60], a genetic crossover-based cryptography system was proposed. Tis was a unique encryption technique used to store sensitive and nonsensitive data in a heterogeneous multicloud environment in order to guard against the riskiest activities, such as data breaches, man-in-the-middle attacks, and insider attacks. To increase the security of the data, the fle was encrypted using the suggested prime crossover approach and stored in several cloud settings. Data integrity, confdentiality, and accessibility are not compromised in this way.
Aruna and Mohan [40] developed an efcient probabilistic public key eEncryption (EPPKE) as a cryptographic method to protect data in the cloud. Covariance Matrix Adaptation Evolution Strategies (CMA-ESs) are used to optimize this strategy. By using the Luhn algorithm and BLAKE 2b encapsulation, it guarantees data integrity. Tis allows enhanced protection for data that are sent over the cloud.
According to the previous studies, the use of cryptographic algorithms is the primary emphasis when it comes to guaranteeing data security in the cloud. However, little has been said about the most commonly used cryptographic approach for securing data in the cloud, the type of cryptographic algorithms used (symmetric, asymmetric, or protocol), the trend of the cryptographic algorithm run times (linear or nonlinear time), the purpose of these cryptographic algorithms, and cloud security concerns. Again, none of the materials available are from African academics, showing that there is a big gap in Africa when it comes to security issues in cloud applications.

Methodology
Te number of works regarding cloud data security is taken into account in this study and their interpretation is based on publications on the subject, and a systematic literature evaluation based on PRISMA is produced.

Research Questions.
Te goal of this study is to assess the security concerns with cloud computing. Researchers' security interventions are also taken into account. Tis study takes six objectives into account. Tese goals are as follows: What is the most often used cryptographic technique for cloud data security? Which type of cryptographic algorithms is used to secure data on the cloud? How can cryptographic algorithms encrypt and decrypt cloud data? In terms of linear versus nonlinear time, what is the trend in the execution time of the used cryptographic algorithms? What are the intended aims of these cryptographic schemes? What are some of the security concerns in cloud computing?

Approaches for Accessing Articles.
Tis section focuses on the diferent terms, databases, reference tools, and search methods used in answering the research questions. Tis has been well discussed as follows.

Phrases Used.
To arrive at the various articles utilized in this study, several keyword searching was employed such as "Data security in the cloud," "Cloud security challenges," "Cloud security models," "Cloud providers and security challenges," and "Cloud security mitigation strategies."

Search Processes.
Te popular databases were searched to obtain articles related to the subject at hand. Tis included journal papers, conference proceedings, and books. All 157 papers were downloaded, and they were arranged for reading and referencing convenience using an IEEE reference generator and Mendeley. Te results are then shown in a PRISMA framework as shown in Figure 4 [61]. Te publications were grouped using a selection technique based on their relevance using their center of interest, and 72 of the 157 publications were deemed relevant to the issue under consideration. Te exclusion procedure used for the selection of the papers of interest is as follows: (I) Papers with publication dates earlier than 2015 (II) Papers with no DOI (III) Papers with a concentration on cloud taxonomy (IV) Articles using anonymous citation (V) Papers that concentrate on technology other than cloud computing security.

Results
Te results of the systematic literature review are presented in this part, and their commentary is provided in the supplemental subsections. Table 3 categorizes the list of publications, the purpose of the algorithm, the number of sources and references, and the run time trend of these cryptographic algorithms. Te table provides clear guidelines for academics looking for answers to cloud security issues. Tere is also the concern that, despite the benefts of cloud computing, cloud clients are unwilling to shift to the cloud.

What Is the Most Often Used Cryptographic Technique for
Cloud Data Security? According to Figure 5 and Table 3, the adoption of encryption techniques, which accounts for 16.7% of publications from 2017 to 2021, is the most popular method for ensuring cloud security. Te existing cryptographic techniques and hybrid algorithms were utilized in these encryption schemes. Tis was followed by the usage of encryption models, which represented 9.7% of the total. Te most often used encryption approach was based on the MapReduce layer, which was implemented on a Hadoop platform [62].

Which Type of Cryptographic Algorithms Are Used to
Secure Data on the Cloud? Te many cryptographic algorithms used to safeguard data in the cloud are depicted in Figure 6. Tese are divided into two types, namely, asymmetric and symmetric algorithms. Figure 6 shows that, in 2016, 2% of the published papers utilized in this review were based on both symmetric and asymmetric features. Tis threshold was raised to 5% for asymmetric algorithms and 4% for symmetric algorithms. In 2021, there was considerable growth in the adoption of asymmetric methods to protect data in the cloud, with a proportion of 8% vs. 7% for symmetric algorithms. Asymmetric algorithms fell from 8% in 2021 to 0% in 2022, and symmetric algorithms fell from 7% in 2021 to 1% in 2022.

How Can Cryptographic Algorithms Encrypt and Decrypt
Cloud Data? On the cloud, data encryption and decryption are accomplished in two ways. Te frst method is to encrypt Security and Communication Networks the entire data as a block, which is known as block ciphering, while the second method is to execute the data alphabet by alphabet, which is known as stream ciphering. According to Figure 7, 4% of the algorithms proposed in 2016 utilized a block cipher method, whereas 0% used a stream cipher strategy. However, the use of block cipher algorithms increased to 10% in 2020, with a comparable 2% for stream cipher methods. Te use of block cipher algorithms fell from 10% in 2020 to 9% in 2021, while stream cipher algorithms increased by 7% in 2021.

In Terms of Linear versus Nonlinear Time, What Is the Trend in the Execution Time of the Used Cryptographic
Algorithms? Te execution time trend evaluates an algorithm's performance by measuring how long it takes to encrypt and decrypt data of various sizes [62]. Te execution time of an algorithm is classifed as linear or nonlinear according to its performance evaluation. Linear performance evaluation is noticed when the algorithm's execution time is proportionate to the data size (O(N)). As a result, the larger the data amount, the longer the execution time, as demonstrated by the works of the authors in reference [63]. However, the efciency of the nonlinear method is determined not by data quantity but by the magnitude of the nonce value employed during algorithm execution [64]. According to Figure 8, linear and nonlinear algorithms were used 1% of the time in 2016. Te use of linear algorithms increased by 12% in 2020, representing an 11% rise over the period of 2016, with no interest in nonlinear algorithms remaining at 0%. In 2021, researchers focused heavily on linear algorithms, resulting in 14%, with nonlinear execution time techniques remaining at 0%.

What Are the Intended Aims of Tese Cryptographic
Schemes? Te diferent goals for using the diferent cryptographic techniques are shown in Figure 9. Cryptographic techniques are procedures used to handle cloud security concerns such as data privacy, cloud security, data confdentiality, and data security. According to Figure 9 and Table 3, guaranteeing data security on the cloud accounts for 30.6% of all articles included in this survey from 2016 to 2022. Tis is obvious in the works of the authors in references [50, 52-56, 58, 65] and [66,67] where data security has been the primary priority. Cloud security was next, accounting for 29.2% of all publications included in this systematic literature review backed by the work of the authors in reference [40]. Figure 9 and Table 3 show that just 2% of publications focused on data confdentiality, as suggested by the author in reference [59] in related works. Te procedures used are mostly for cloud penetration testing and anomaly detection.

What Are Some of the Security Concerns in Cloud
Computing? Aside from the benefts of cloud computing, cloud clients and cloud service providers present several concerns. As a result, they are unable to fully transition to the cloud [68]. Tis is obvious in Gartner's categorization of cloud security risk, which is divided into seven segments [69]. Gartner's security concerns are classifed into the following seven categories: (I) Access Control. Tis manages the infow and outfow of access to data by cloud clients (II) Governance. Tis controls clients' data security and integrity (III) Te geographical location of data. Tis controls the siting of data centers to store clients' data (IV) Division of data. Tis defnes the ways to break data into units for storage. (V) Data Recovery. Te ability to recover data in case of a disaster such as a virus attack (VI) Fact-fnding. Tis explains if there is the possibility to investigate any illicit task (VII) Data Availability. Tis is to fnd out if the stored data will be made available anytime they are needed by the cloud client.
Tese seven categorizations of Gartner's category have led to the security challenges depicted in Figure 10.  Science, IEEE Xplore, Science Direct, Hindawi, Google Scholar, and ACM. A thorough analysis of the articles from well-known databases has indicated that many researchers have put in the efort to safeguard the cloud as shown in Table 3. From Table 3, it is evident that 46.58% of the articles used in this study were all directed towards achieving cloud security. Also, 24 of the articles aimed at ensuring data security in the cloud, representing 32.88% of the articles used in this study. Tis indicates that excessive work has been performed concerning the cloud security. However, from Table 3, it is evident that all of the algorithms' run-time trends are linear (O(N)). Tis implies that run times and data sizes are proportionally related. Tis results in excessive engagement of the CPU when large data sizes are executed which cause wear and tear of gadgets. Tis makes algorithms executions time predictable and gives room for hackers to hack systems because of prejudice of the run time. Te linear nature of the proposed algorithms results in higher execution times. Again, except for algorithms that employ hash functions, linear run-time algorithms require signifcant bandwidth to transfer data to the cloud due to the growth in data volumes resulting from data size and run-time relationships. Tis is a major weakness of the majority of proposed algorithms directed towards securing the cloud.

Limitations
Tis comprehensive review of the literature covers various articles that explore the study's objectives. Te authors are confdent that this systematic literature review will cover the type of cryptographic algorithms used (symmetric, asymmetric, or protocol), the trend of run times for these cryptographic algorithms (linear/nonlinear time), the purpose of these cryptographic algorithms, and cloud security concerns published between 2016 and 2022. One of the limitations of this SLR is the use of simple search terms to discover research papers. Such search terms, on the other hand, can be broadened. Again, publications that did not have a digital object identifer (DOI) and were not deemed relevant papers limit our analysis. Another shortcoming of this SLR is that it excludes very recently published research articles that should have been included in this SLR to address the research questions.

Conclusion
A systematic literature review approach was used in this study to review the literature on cloud computing, with a focus on the most commonly used security approach to control security issues in the cloud, the type of encryption algorithms used to secure the cloud, how algorithms encrypt and decrypt data on the cloud, the run-time trends of the algorithms used in the cloud (linear time/nonlinear time), and the intended aims of these security approaches. Many security methods, such as frewalls, data masking, encryption, and blockchain, have been identifed as answers to data security concerns.
Te recognized security challenges included confdentiality and privacy, which may be achieved through encryption. Data integrity as a security concern may be achieved using BLS signature verifcation and blockchain. Data availability has also been noted as a security risk in cloud computing, which may be solved by data replication across several servers. Te cloud migration has the advantages for cloud customers and cloud service providers, but maximizing these profts needs efective and long-term security measures to address the breach of security problems in cloud computing. Tis comprehensive literature study highlighted security as a key barrier to complete cloud computing adoption on the side of both the cloud customer and the cloud service provider. Furthermore, the comprehensive literature analysis revealed that encryption techniques are the best ways to secure the cloud. Te paper states that linear time complexity algorithms account for 90% of the encryption algorithms proposed from 2016 to 2022, making linear (O(N)) time algorithms the most popular and widely used.
However, the present cryptographic techniques are incapable of withstanding contemporary security threats that target cloud customers and providers face due to their linear run times. Because linear run times are dependent on data size, attackers may estimate the time required for each data execution. Furthermore, such encryption techniques need additional data transport bandwidth when large data are to be transferred [47]. When huge volumes of data are sent, the reliance on data proportionality and run times creates wear and tear on the client's and provider's equipment.

Recommendation
(i) Cloud service providers should use nonlinear algorithms as security schemes to ensure the interoperability of devices with lesser specifcations. (ii) Stakeholders in device developing and manufacturing should consider using nonlinear algorithms to ensure security of data on their devices.

Future Works
(i) Tere is limited literature on nonlinear algorithms to secure data on the cloud; as a result, more research should be conducted on nonlinear algorithms (f(x) � b − cn 2 ). (ii) More studies should be conducted on cloud challenges such as confdentiality and privacy, multitenancy, and data reliability.