Cloud Computing Literature Review Computer Science Essay Example

involved. In certain cases, cloud computing may not adequately protect the customer's assets, highlighting a significant drawback of this technology.

[Chang et al, 2005] Cloud computing comprises the storage of customer data in large-capacity data centers. These centers, which utilize servers, centrally store and process the data. The cloud provider takes responsibility for safeguarding customer data and ensuring its availability. To establish trust, a service legal agreement (SLA) is required between the customer and provider. Consistency within the SLA plays a crucial role in building client trust. The overall security of the cloud relies on the implementation of security measures by the cloud itself. Despite numerous computing benefits offered by cloud computing, concerns regarding security persist as a significant barrier.

The mentioned perspectives suggest that data in cloud computing is usually stored and managed by data servers located in data centers, which have the capability to store vast amounts of information. It is essential for cloud providers to ensure the security and effective management of customer's data. To establish trust between the provider and customer, it is necessary to adhere to a legal agreement called SLA (Service Level Agreement).

[McKinsey, 2009] Cloud computing is the specification of IT capabilities, which include hardware, software, and services. It offers computing, networking, and storage capacity through hardware-based services. Certain aspects like data integrity, recovery, and privacy require risk assessment for these clouds. To ensure customer information protection, a security perimeter is established as a trusted boundary for storing and processing valuable data. The network facilitates data transportation and consists of other trusted end hosts. Trusted areas are identified due to the undefined boundaries in cloud computing environments for data storage and

processing. Privacy issues are commonly encountered in both public clouds and due to security concerns.

According to the authors, clouds are made up of hardware-based services that offer computing, networking, and storage capacity. These clouds possess distinct characteristics in areas like data integrity, recovery, and privacy that necessitate risk assessment. To protect valuable customer data, a security border is established to establish a trusted boundary for storage and protection. The network surrounding the cloud enables secure transportation of trusted end hosts, functioning similarly. Although public clouds often raise privacy concerns in cloud computing, they also have their own set of security worries.

[Rob, 2008] The "cloud pyramid" in the field of cloud computing visually represents the concept of cloud computing, which is classified into three sections: cloud application, cloud platform, and cloud infrastructure.

According to [Rob, 2008], the cloud application is the initial segment at the top of the pyramid. It involves the interaction of applications through a web browser in cloud computing. By utilizing the cloud application, there is no longer a requirement to install and run the application on the customer's computer. This eliminates the difficulty for the customer in maintaining both the software and its processes.

The cloud computing pyramid consists of three sections: cloud application, cloud platform, and cloud infrastructure. The first section, cloud application, allows applications to communicate with each other through a web browser. This section is designed to simplify software maintenance and processes for customers.

[Rob, 2008] Typically, the cloud platform provides the theoretical work as a service or the computing platform. It is situated in the middle layer of the cloud pyramid. The cloud platform caters to the requirements by providing animatedly

necessities and configuring and reconfiguring servers.

Based on the authors' perspectives, the cloud platform is utilized for computing the platform and framework as a service. It corresponds to the second level in the cloud computing pyramid. This section is responsible for dynamically meeting server requirements, as well as configuring and reconfiguring servers. It occupies the middle position in the cloud pyramid.

According to Rob Lovell in 2008, cloud infrastructure is an essential component of the cloud pyramid. It utilizes virtualization to provide IT infrastructures. This allows for the separation of hardware into independent and self-contained environments. The cloud infrastructure includes the delivery of web services, farms, cloud centers, and other hardware appliances.

The text below is a quote and should not beor unified as it provides specific information about the source of the text.

In summary, the cloud infrastructure is the final tier of the cloud pyramid and it provides IT infrastructures through virtualization. It enables the creation of self-contained environments and encompasses web services, farms, cloud centers, and other hardware appliances.

[Jensen et al, 2009] explain that when users release their data into the cloud, the cloud is responsible for ensuring the security of this data. However, there are certain operational issues associated with this responsibility.

According to [Jensen et al, 2009], organizations may encounter security vulnerabilities when operating in the cloud, potentially leading to the exposure of important company data and information. The lack of adequate standards for data management and security in the cloud computing environment can contribute to these security concerns. To address this problem, it is important to establish a comprehensive security management system that involves both customers and providers, aiming to protect against

potential hacking attempts by external entities.

Cloud computing is responsible for securing customers' valuable information, according to multiple authors. However, challenges may arise in data transfer to the cloud. The security of operations in cloud computing is a major concern as it can leave an organization's assets vulnerable. The cloud does not guarantee adequate security standards, which could result in potential issues with safeguarding information. To address these concerns, it is essential to establish a security management system between the customer and provider that effectively defends data against hackers.

[Siani et al, 2010] The cloud service provider should be capable of providing a statement about the security of sensitive data and ensuring that confidential user information remains inaccessible to unauthorized individuals.

The authors highlighted the significance of service providers in guaranteeing the security and safeguarding of valuable information, stressing the need to thwart unauthorized individuals from viewing or accessing customers' private data.

[Siani et al, 2010] It is essential for the provider to prevent any modifications or attempts to alter innovative data in order to maintain the integrity and appropriateness of the data stored in the cloud system.

It is crucial to maintain the accessibility of all resources required for data processing, as stated by [Siani et al, 2010]. This necessitates providing the user with access to these resources while also preventing unauthorized interference or malicious actions.

[Siani et al, 2010] The presence of automatic agreements between organizations or authorized individuals and officials is vital, as it allows for confirmation if needed.

Security is essential at multiple levels within organizations, including access to servers, the internet, databases, and programs. Cloud service providers must prioritize data privacy and ensure it is adequately protected.

The

above authors' perspectives suggest that reliability, accessibility, and non-reputation are shared concerns when it comes to cloud computing security. The provider has the ability to restrict modification of innovative data, which must be utilized responsibly in order to preserve integrity within the cloud. It is vital that users have uninterrupted access to all necessary resources for data processing, without any interference from external parties.

[Steve, 2008] In a traditional enterprise environment, there are established tools for guaranteeing security in computer, storage, and network. However, in cloud computing, user data is processed and stored on shared equipment. This means that companies of all sizes can adopt cloud computing, including large corporations as well as medium-sized, small, and startup companies. Nevertheless, there are certain threats unique to cloud security. These primarily revolve around determining who is responsible for various aspects of security and addressing concerns like the potential loss of data, unauthorized access to data, and improper usage of customer data.

Authors have focused on creating various tools in conventional project settings to guarantee security in computer systems, data storage, and networks. However, cloud computing poses distinct security obstacles due to the generation and storage of user data on shared resources. Cloud computing is employed by companies of all scales – big, medium, and small – leading to these challenges arising from specific threats.

[Steve, 2008] The cloud controls the hardware and hypervisors on which the data is stored and applications are run. It is important for the cloud provider to have high security standards.

According to the opinions of various authors, it can be inferred that the management of data and application operations is governed by the cloud using hardware

and hypervisors. Therefore, it is crucial for cloud providers to maintain high security standards.

[Steve, 2008] The data in the cloud environment can be shared by all customers inside the cloud or connected to it. There is a possibility of data access or interference by other users if the barriers between them are compromised.

Based on the aforementioned perspectives, it can be inferred that cloud customers share data in a cloud environment. When barriers between users are removed, there is a possibility for data interference or unauthorized access.

[Mather, 2009] The use of cloud computing techniques raises various legal and regulatory concerns. One potential issue is the restriction on data export within a jurisdiction. When sensitive issues arise in cloud computing, it is important to address and resolve any associated legal and regulatory problems.

Considering the authors' viewpoints, it can be concluded that the implementation of cloud computing techniques in organizations results in the emergence of various legal and regulatory issues. These issues arise as a consequence of the common concerns that are raised.

[Mather, 2009] Many organizations employ the model of perimeter security for strong security at the enterprise network perimeter. However, this model has been increasingly ineffective due to outsourcing, workforce, and high mobility. Now, critical data and applications can be stored in the cloud, even outside the enterprise perimeter.

Authors agree that many organizations use the border security model to protect their enterprise network. However, they also acknowledge that this model has been ineffective due to increased mobility, workforce changes, and outsourcing in recent years.

[Jin Li et al 2008] The unclear (fuzzy) keyword search greatly extends the usability of the system by returning the same files when the

user's search inputs closely match the predefined keywords or when nearby plausible files are determined based on semantic similarity to the keywords, especially when exact matches are ineffective.

According to the authors' perspectives, the fuzzy keyword significantly enhances the system usability. It does this by providing similar files when users enter correctly predefined keywords. However, if the exact match is not successful, the system relies on keyword similarity semantics to find nearby relevant files.

[Li et al, 2008] The importance of fuzzy search has led to a greater emphasis on plain text search in the information retrieval field. This issue has been addressed in traditional information access paradigms by allowing users to search without the need for exact matching and exploring techniques for identifying relevant data based on predicted string matches. As a result, algorithms for estimating string matches can be divided into two main categories: online and offline.

According to the aforementioned perspectives of authors, it can be inferred that in plain text fuzzy keyword search, certain challenges arise in terms of the importance given to fuzzy search in the context of information retrieval. As a result, traditional information access paradigms have addressed this issue by allowing users to search without needing to learn and understand the technique for identifying relevant data based on predictable string matches. The algorithms for estimating these matches can be categorized into two sections: online and offline.

According to [Chakrabart et al, 2006], online techniques find the method of searching without an index to be undesirable due to their limited exploration capabilities. In contrast, offline techniques utilize indexing algorithms, such as suffix trees, metric trees, and q-gram methods, to achieve faster searches.

Initially, it may seem feasible for users to explicitly describe these string matching algorithms in terms of searchable encryption by evaluating trapdoors on the nature support located in an alphabet. However, this simple structure is vulnerable to attacks related to dictionary and statistics, ultimately making it ineffective in ensuring search privacy.

After considering the opinions of authors, it can be concluded that when it comes to online techniques, not using index performance in search is not desirable due to limited functionality. In contrast, offline techniques utilize indexing techniques to make the process faster. These techniques include suffix trees, metric trees, q-gram methods, and others. However, it is crucial to be aware of potential issues associated with index and statistics.

[Curtmola et al, 2006] The previous searchable encryption methods have been extensively evaluated in the field of cryptography. These methods primarily focus on formalizing security definitions and improving efficiency. Initially, each statement in the text is individually encrypted using a two-tier encryption structure. Subsequently, bloom filters are introduced to generate indexes for the data files.

[Bellare et al, 2007] This method involves the creation and storage of trapdoors for all words in a file on the server. To search for a specific word, the user evaluates the trapdoor of that word and sends the search request to the server. Upon receiving the request, the server checks if any bloom filters contain the trapdoor for the queried word and returns the corresponding file recognizers if found.

After considering the perspectives of various authors, it can be concluded that earlier searchable methods are widely used in cryptography scenarios. These methods focus on formalizing security definitions and improving efficiency. In the past, each

statement in the text was encrypted independently using a two-layer encryption structure. Subsequently, bloom filters were introduced to create indexes for data files. This method involved developing and storing trapdoors of all words on the server for each file. To search for a word, the user needs to generate the trapdoor of the word and send it to the server as a search request.

[Waters et al, 2004] Proposed improvements in search techniques involve the implementation of similar "index" strategies. One such strategy involves the creation of a single encrypted confusion table index for the compilation of all files. Each entry in this index contains encrypted file identifiers, and these identifiers correspond to data files that contain the keyword from the index table. Additionally, a trapdoor of a keyword is included. Another approach to achieving searchable encryption, based on public-key cryptography, has also been developed as a complementary technique. However, due to user agreement issues, these techniques are not currently being considered. Nevertheless, they are applicable in cloud computing scenarios where precise keyword search is essential.

According to the authors' perspectives, it can be concluded that similar "index" techniques should be designed in order to achieve a more successful search. This technique creates a single encrypted confusion table index for overall file completion. In the index table, each entry contains an encrypted group of file identifiers that have the same keyword. It also includes the trap door for the keyword. The key-based searchable encryption technique is developed as a balancing approach to the first technique. However, these developments are not widely accepted due to user agreement issues. Nevertheless, they are relevant to cloud computing as these techniques

help in accurate keyword search.

[Shi et al, 2007] The confidential matching data in secure multi-party computation allows dissimilar parties to review a function of their individual data collaboratively without revealing their data to others. These functions involve the connection or approximate private matching of two different sets. However, this technique is often used for secret recovery of correspondent sets. It has been widely employed in database data retrieval and usually involves random calculation complexity.

According to various authors, the use of private correlation data allows different parties to review specific functions of their own data without exposing it to others in secure multi-party calculations. These functions typically involve approximating matches or relationships between two sets. This technique, which frequently employs a method that randomly increases calculation complexity, is commonly used to confidentially recover corresponding sets from a database.

[Chow et al, 2009] Advanced techniques are employed for constructing more sensible and efficient fuzzy keyword searches, focusing on storage and search efficiency. These techniques aim to enhance the simple method of organizing the ambiguous keyword set. The scholars focused on the edit distance case with d=1, while ensuring generalization is not compromised. The calculation remains the same for higher values of 'd'. The design of this technique ensures that it does not compromise search accuracy while managing the ambiguous keyword set.

Wild card based fuzzy set construction is developed to represent correct operations in the same location. Wild card based, gram based, and symbol based fuzzy keywords are employed to address security concerns in cloud computing.

Authors agree that in order to address security issues in cloud computing, it is beneficial to use wild card based, gram based, and

symbol based fuzzy keywords. Advanced techniques offer better and more efficient fuzzy keyword search strategies that improve storage and search efficiency. These techniques largely focus on developing a simple approach for creating fuzzy keyword sets. Specifically, researchers have concentrated on the case of edit distance d=1 without compromising simplification in this method. The computation is also similar for improved principles of 'd'. By using this method, a fuzzy set structure based on wildcards is created to perform accurate operations at the same location.

According to Song et al, 2000, techniques for constructing fuzzy sets based on wildcards are used to represent correct operations at the same position. The fuzzy set of wi is represented as Swi,d={SA?a‚¬A?wi,0, SA?a‚¬A?wi,1,A·A·A· , SA?a‚¬A?wi,d} with an edit distance ‘d’, where SA?a‚¬A?wi,A?a?z represents the set of words wA?a‚¬A?i with A?a?z wildcards. Each wildcard accepts the correct operation for wi, which is fully described by the following equation. For instance, the word CASTLE with a pre-set correct distance of d=1 can be constructed as follows:

The storage in the clouds can be reduced by using a pre-set edit distance of d=1. If the edit distance is set to 2, the representation of size Swi, 2 will change.

When the distance is set to 3, then the size representation will be Swi,3.

By analyzing the given equations, it can be inferred that a fuzzy set structure utilizing wildcards is created to perform operations at the specified location. Each wildcard determines the accurate operation, which is evident from the aforementioned equation. The storage capacity of cloud data can be reduced by implementing a predetermined edit distance of d=1. The size representation is determined by the edit distances.

Another efficient

technique used to build the fuzzy set is gram-based creation [Behm, 2009]. The gram of the string is a substring that serves as a signature for efficient estimated search, and it is used for matching purposes. The representation of sizes is as follows:

To create the fuzzy word set for the keyword CASTLE, you can follow these steps:

{CASTLE, CSTLE, CATLE, CASLE, CASTE, CASTL, ASTLE}

Based on the perspectives of these authors, it can be inferred that constructing a gram-based fuzzy set is one of the most effective techniques for creating a fuzzy search set. This creation of the string is considered a substring and can be used as a signature for efficient estimated search. These grams can be utilized for this purpose.

[Feigenbaum, 2001] presents a symbol-based trie-traverse search scheme aimed at enhancing search efficiency. This scheme involves the creation of a multi-way tree to store a fuzzy keyword set within a limited symbol set. The primary motive behind developing this scheme is to address the presence of trapdoors that share a common prefix and may have common nodes. By improving the symbols involved in a trapdoor search from the root to the leaf where the trapdoor ends, it becomes possible to recover symbols if an empty set is encountered.

Based on the viewpoints of the authors mentioned above, it is possible to conclude that a trie-traverse search scheme is devised to improve search efficiency. To store a fuzzy keyword set using a restricted symbol set, a multiway tree is constructed. The symbols can be retrieved whenever an empty set is encountered at the trapdoor.

[Boneh et al, 2004] The efficient key word search scheme should consider the

following points:

Initially, an unclear keyword set Swi is created by the information holder to generate the index for wi using the wildcard based technique. Subsequently, the information holder must compute the trapdoor set {Twi} for each wi’ AZAµ S wi d, using a secret key Sk shared between the data holder and the certified users. The data holder then encrypts FID wi. Meanwhile, the directory table and locked data files are stored on the cloud server for storage purposes.

The user who has been approved calculates the trapdoor set for searching with the values w and k, which are also used to derive S w, k from the wildcard based fuzzy set structure. Once the user sends this information to the server, the server evaluates it using the locked file recognizer's capabilities when receiving the search request. Finally, the user retrieves the relevant files of interest and unlocks the results that are returned.

According to Bao et al (2008), the process of generating a search request for the keyword 'w' is similar to creating a directory for that keyword. This means that the search request acts as a trapdoor set dependent on S_w,k, rather than a single trapdoor like in the traditional method. By using this approach, the accuracy of the search results can be guaranteed.

Based on the author's views, it can be concluded that there are two key considerations for creating an efficient fuzzy keyword search. The first consideration is that the fuzzy keyword is created by the information holder. The second consideration is that the accepted user computes the trapdoor set for searching with a wildcard-based fuzzy keyword structure.

Cloud computing is a service-oriented IT model

that provides infrastructure and computer resources. It allows businesses to improve their operations easily and efficiently. The accessibility of cloud computing benefits various customers by offering dynamic and scalable resources through the internet. This helps reduce capital and operational costs while fostering economic growth. However, there are also risks associated with cloud computing due to the possible storage of customer data in data centers.

The cloud computing is categorized into three parts: cloud application, cloud platform, and cloud infrastructure. It faces various operational challenges including operational security, privacy, reliability, accessibility, non-reputation, security provision failures, customer attacks, legal and regulation issues, and a broken perimeter security model. The inclusion of the fuzzy keyword feature significantly enhances system usability by providing similar files when users input predefined keywords accurately. Fuzzy search plays a vital role in information retrieval due to its effectiveness in searching plain text.

The conditions of secure multi-party calculation utilize confidential matching data to enable disparate parties to collectively review a function of their own data without disclosing it to others. Advanced techniques are employed to enhance the efficiency and practicality of fuzzy keyword search constructions, encompassing storage and search efficiency considerations. To address security concerns in cloud computing, fuzzy keyword search methods such as wild card, gram-based, and symbol-based approaches are utilized.