8.3.1 Amplified Cloud Security Problems

Amplified cloud security problems are those deriving from the underlying technologies upon which cloud computing is heavily built, such as virtualization, web application servers, and multitenant software architectures. This category of problems includes those originating from failing to adhere to well‐known and commonly established security best practices, which are hard or infeasible to implement in a cloud computing environment.

The most common amplified cloud security problems are the following:

• Misuse of administrator rights and/or activities of malicious insiders: In cloud computing, virtual machines (VMs) are used for managed servers. A CSP is responsible for the underlying host system and always has access to the VMs running on the host through the hypervisor. As of today, it is still hard to detect misuse of administrator rights or the presence of malicious insiders due to a general lack of transparency into the CSP's processes and procedures. This problem may violate core security principles, such as confidentiality, authenticity, authorization, integrity, data protection, accountability, and nonrepudiation.
• Missing transparency of applied security measures: In traditional IT outsourcing, service providers can prove their security compliance to their customers by showing the usage of the baseline security measures through, for example, International Organization for Standardization (ISO) 27001 or Payment Card Industry (PCI) Data Security Standard (DSS) certificates. In cloud computing, not every CSP follows these baseline rules, although today global government agencies and laws are changing in order to require CSPs to follow these regulations. This problem may cause a violation of one or more of the following core security principles: integrity, availability, and data protection. A notable exception is Amazon Web Services (AWS)—one of the first CSPs that has started to follow global security compliance very seriously. More about cloud compliance and regulations will be described in Section 8.4 of this chapter.
• Missing transparency with security incidents: In traditional IT outsourcing, responsibility for security‐incident response is transferred to the service provider, which uses experienced personnel (for example, a computer emergency response team (CERT). When it comes to cloud computing, things are more complicated. In this case, a customer and the CSP have to work together to collect all users' data information generated before and during a security incident. Any concern about the cloud hardware and software infrastructure must be associated with the different cloud resources available to the customer and involved during the incident. Today, a standardized procedure is still missing. In fact, cloud offers available on the market do not offer a transparent process for customers on how security incidents are detected. This problem may affect the following core security principles: data protection, integrity, availability, and nonrepudiation. Section 8.4 of this chapter also explains what the SysAdmin, Audit, Network, Security (SANS) organization is doing to correct this problem, especially when it comes to public cloud services. Section 8.4 also explains how the SAaaS can be used for cloud incident detection and as a possible enabler to perform cloud audits while respecting cloud‐specific characteristics.
• Shared‐technology issues: In the multitenant cloud‐computing model, virtual resources are shared with multiple customers. Furthermore, there may be misconfigured VMs that endanger other resources due to lack of proper isolation. Exploits have already been demonstrated by (Kortchorski and Rutkowska 2009). The increasing code complexity in hypervisor software amplifies this threat. Memory‐cache isolation is another issue since some of the underlying components that make up the cloud infrastructure do not lend themselves to offering proper isolation. For example, graphics processing unit (GPU) and central processing unit (CPU) caches were not designed to offer strong isolation properties in a multitenant architecture. As of today, no CSP discloses information on how shared resources are securely wiped before being reassigned to a different customer. Furthermore, cloud consumers get a default source access point to a VM in the current IaaS. It has been proven that using this default source access point, attackers have an easier way to break through the isolation of shared resources. The Federal Office for Information Security suggested that using certified Common Criteria (CC) compliant hypervisor software (minimum EAL 4) might alleviate this threat (https://www.bsi.bund.de/EN/Topics/Certification/certification_node.html). This problem may affect the following core security principles: integrity, availability, data protection, confidentiality, authentication, and nonrepudiation.
• Data life cycle in case of a provider switch or termination: In cloud computing, due to shared usage of resources, this threat is particularly serious. The CSA states that, in the absence of satisfactory rules defined by the CSP, every cloud consumer needs to impose specific rules for when the contract with the CSP ends. Such rules must clearly regulate how customers' data is exported from the Cloud and how the CSP will securely erase that data at the end of the contract. This problem may affect the data protection and confidentiality core security principles.
• Monitoring service‐level agreements (SLAs): In cloud computing, several multitenant applications running in a virtualized environment need a special technology for monitoring SLAs. New technologies for the hypervisor, virtualized networking, monitoring, etc. must be used. (Patel et al. 2009) proposed a mechanism for managing SLAs in a cloud‐computing environment using the web service‐level agreement (WSLA) framework, developed for SLA monitoring and SLA enforcement in a service‐oriented architecture (SOA). In June 2014, the European Commission (EC) published the Cloud Service Level Agreement Standardization Guidelines (https://ec.europa.eu/digital‐single‐market/en/cloud‐select‐industry‐group‐service‐level‐agreements). More about this standardization is described in Section 8.4 of this chapter. This problem may affect the availability and integrity core security principles.

8.3.2 Cloud‐Specific Security Problems

Cloud‐specific security problems are those that are inherent to the Cloud itself and its characteristics, and not necessarily inherited from the technologies used in the underlying infrastructure. The most common cloud‐specific security problems are the following:

• Unclear data location: Today, many cloud customers do not know in which country their data is saved or processed. Perhaps the only exception is AWS, which exposes to consumers the approximate cloud data center's continental location in which their data is stored (for example, the AWS data center in Northern Ireland). As of today, there is no way to know whether a customer's data has been outsourced by a cloud provider. (Tiwana et al. 2010) proposed a solution to expose the network location of data to applications. There are only a few specific acts and laws related to protecting users' data. Among these is Germany's Data Protection Act, §11 (Mell and Grance 2009; German Parliament 2009). Section 8.4 talks more about different global data‐location compliance rules. The uncertainty related to the geographical location in which a consumer's data is stored may affect the following core security principles: data protection, confidentiality, and availability.
• Abuse and nefarious use of cloud resources: Cloud computing offers many attractive characteristics. Among these is easy, fast access to many virtual supercomputing machines. Unfortunately, these great characteristics also attract malicious users, who find the high‐performance computing infrastructure of a modern cloud system the right platform for attacking other systems. For example, in 2011, malicious users took advantage of AWS and used it to host malware, the Zeus botnet, a phishing Trojan horse that steals banking information within AWS. Cloud hackers can also easily aggregate as many VMs as they need to perform DDoS attacks on a single CSP, which can also affect the cloud consumer. This problem may affect the availability and confidentiality core security principles.
• Missing monitoring: If cloud consumers' data, especially personal data, is at risk of fraud or loss of integrity, it is essential that a CSP be able to detect these risks and eliminate them. As of today, it is not trivial for a CSP to use an information‐policy system that automatically detects security issues and informs customers. When designing a risk analysis that runs a service on a cloud, it is important to include data‐protection measures to secure the cloud environment, such as antivirus software and intrusion detection systems (IDSs), as well as measures for DoS detection and prevention. For large IT environments, the best practice to monitor against these types of attach is to run IDSs with distributed sensors as input feeds. However, this solution is not sufficient and flexible enough for cloud‐computing environments due to their inherent complexity and the dynamic changes driven by users. (Meng et al. 2012) published a work on reliable state monitoring in cloud data centers for better monitoring such complex and dynamic environments. (Doelitzscher et al. 2012) introduced SAaaS as a cloud incident‐detection system built upon intelligent autonomous agents that are aware of underlying business‐driven intercommunicating cloud services. This problem may affect the following core security principles: nonrepudiation, availability, data protection, and confidentiality.
• Unsecure APIs: A CSP offers application programming interfaces (APIs) to cloud consumers in order for them to deploy, control, and manage their cloud resources. Therefore, it is crucial for these APIs to provide access control, encryption, and activity monitoring in order to protect against both malicious and unintentional attempts to bypass a security policy. For instance, a load‐balancing service is a complex architectural layer that can be inserted into a system to improve the performance of a service and increase its availability. When such a complex component is used, it is necessary to carefully examine the overall system in order to make sure that no security holes are made available to attackers. To prevent attacks from being mounted against the system through the use of unsecure APIs, standardized protocols and measurements for secure software development are available. These include Microsoft Secure Development Lifecycle (SDL) and the Open Web Application Security Project (OWASP) Software Assurance Maturity Model (SAMM). The inadvertent use of unsecure APIs is a problem that may violate the following core security principles: confidentiality, integrity, availability, nonrepudiation, data protection, and accountability.
• Missing monitoring of cloud scalability: Scalability is one of the most attractive characteristics of cloud computing. For this reason, it is essential to be able to deal with service usage peaks: for instance, if there is a new update of a popular software program, and a huge number of downloads is expected. Usually, peaks are predictable and confined to specific time frames. Therefore, cloud‐application engineers design their solutions to initiate new instances if a certain threshold is reached to provide service availability. However, this raises two new challenges for cloud security:
  1. Scaling driven up by IaaS business: Since a cloud consumer's infrastructure can vary quickly (for example, by growing and shrinking as in the case of a peak scenario), a monitoring system must take care of all the peak events and the defined scalability thresholds.
  2. Scaling driven up by IaaS attacks: A cloud attacker can control the creation of new cloud instances to the maximum scalability threshold number of allowed requests, which could cause, for example, the distribution of malicious software.

These issues may violate the availability and accountability core security principles.

• Missing interoperability of cloud providers: Today, CSPs are often incompatible with each other, particularly because each CSP uses customized VM formats and proprietary APIs. This may increase the risk of data lock‐in, a phenomenon that prevents data portability across different CSPs. (Loutas et al. 2013) wrote a comprehensive and systematic survey of cloud‐computing interoperability efforts by standardization groups, industrial organizations, and research communities. More details about this work are described in Section 8.4. There are some initial development projects working on this issue. These include the Open Cloud Computing Interface (OCCI), Open Virtualization Format (OVF), OpenStack cloud software Rackspace hosting, and the National Aeronautics and Space Administration (NASA). Moreover, a specific strategy must be agreed upon between provider and customer for regulating data formats, perpetuating logic relations, and total costs if a CSP change occurs. This issue may cause a violation of the availability core security principle.

8.3.3 Correlation of Cloud Security Issues and Research Efforts

In order to better emphasize the complexity and importance of auditing in cloud computing, this section correlates each of the cloud audit issues with the current state of the art in research. Since a cloud consumer is not exposed to the details of how a CSP governs the underlying cloud infrastructure, the only choice for the consumer is to give complete trust to the CSP and hope the CSP applies compliance regulations and data‐protection laws for protecting confidential and sensitive data. Furthermore, unknown or unclear geographic location of data in the Cloud may cause unexpected issues since different jurisdictions enforce different legislation and compliance requirements when it comes to data governance. Following is a list of some of the current research work to alleviate this problem:

• (Ries et al. 2011) proposed a new geographic location approach based on network coordinate systems, and evaluated the accuracy of their solution on the three prevalent cloud deployment models: IaaS, PaaS, and SaaS. Even though a CSP may use additional measures, such as traffic relaying, to hide the location of the resources stored in the Cloud, a high probability of location disclosure is achieved by means of supervised classification algorithms.
• (Vaish et al. 2013) published a mechanism that uses remote‐attestation technology of trusted platform modules. A remote‐attestation technique is used to validate the current location of the data, and the generated result is passed to the user verifier. The fact that the trusted platform module is tamperproof provides the basis for the accuracy of the result.
• (Gondree and Peterson 2013) introduced and analyzed a general framework for authentically binding data to a location while providing strong assurance against CSPs that, either inadvertently or maliciously, attempt to relocate cloud data.
• (Paladi et al. 2014) proposed a mechanism allowing cloud users to control the geographical location of their data, stored or processed in plaintext on the premises of IaaS CSPs. They used trusted computing principles and remote attestation to establish platform state. They also enabled cloud users to constrain plaintext data exclusively to the jurisdictions they specify, by sealing the decryption keys used to obtain plaintext data to the combination of the cloud host's geographic location and platform state.
• Cloud storage permits moving remote data to the centralized data centers, where loss of data integrity can arise. (Sunagar et al. 2014) proposed a study of the problem of ensuring the integrity of data storage in the Cloud.

Cloud auditing can also be used to help detect abuse and nefarious use of cloud resources, although this is still a complex task. (Hamza and Omar 2013) explored and investigated the scope and magnitude of this problem, which is one of the most serious security threats in cloud computing. The authors also presented some of the specific attacks related to this threat, since it constitutes a major restriction for moving business to the Cloud. Following are some of the most common attacks that relate to this problem and the related research work that has been done in order to resolve them:

• Host‐hopping attacks: These attacks can be easily mounted when a CSP has no mechanism to restrict shared access to cloud resources, such as data storage and VMs, and to enforce isolation of different customers or hosts. Failing to enforce customer‐resource separation may facilitate hackers hopping on other hosts, endangering their resources, or interrupting their services, thereby damaging their reputation and impacting their revenue. These types of attacks are particularly common in a public PaaS deployment model, where multiple clients share the same physical machine.
• Malicious insider and abuse of privileges: Since cloud infrastructures are based on multitenancy and shared resources, there is the risk of unauthorized access to customers' confidential data. This may lead to the exposure, leak, or selling of sensitive customer information. This problem may lead to even graver risks if a CSP does not correctly enforce the foundation security rule known as the principle of least privilege (Bishop 2002). In such cases, malicious users may exploit rights that they were not intended to be granted to subvert the integrity of the system or steal confidential data.
• Identity theft attacks: Cloud hackers can easily create temporary accounts with CSPs, use cloud resources, and only pay for the usage of those resources. By doing this, they can try to get access to customer data and sell it, leading to identify theft. This type of attack also occurs when cyber criminals set up a fake cloud and attract users by hosting their sensitive information and providing them with cloud‐based services such as email and web hosting. This is a great catch for stealing customers' identities and financial information.
• Service‐engine attacks: An attractive characteristic of cloud computing is highly customizable platforms. For example, in the IaaS deployment model, attackers can rent a VM to hack the service engine from the inside, and use it to their advantage. In particular, they can try to escape VM isolation to reach other VMs, steal sensitive business information, and compromise data of other cloud customers.

To address the lack of transparent monitoring of a cloud infrastructure, the Cloud Research Lab at Furtwangen University, Germany, in 2012, started to develop SAaaS, a prototype for incident detection in the cloud. SAaaS is an audit solution where techniques of behavioral analysis and anomaly detection are used to distinguish between normal and nefarious use of cloud resources. SAaaS uses intelligent autonomous agents, which support cross‐customer event monitoring within a cloud infrastructure. This work also evaluated which cloud‐specific security problems are addressed by SAaaS. The results of this work shows that autonomous agents and behavioral analysis are sufficient to identify cloud‐specific security problems and can create an efficient cloud‐audit system.

(Meng et al. 2012) proposed state monitoring for detecting critical events and abnormalities of cloud‐distributed systems. When it comes to cloud computing, as the scale of the underlying infrastructure grows and the amount of workload consolidation increases in cloud data centers, node failures and performance interferences (in particular, transient ones) become quite common. Therefore, distributed state‐monitoring tasks are very often exposed to broken communication. This can bring about misleading results and cause problems for cloud consumers who heavily rely on state monitoring to perform automatic management tasks, such as autoscaling. This work introduced a new state‐monitoring approach that addresses this challenge by exposing and handling communication dynamics, such as message delay and loss in cloud‐monitoring environments. This methodology delivers two different characteristics. First, it quantitatively approximates the accuracy of monitoring results to capture uncertainties introduced by messaging dynamics. This feature helps users differentiate trustworthy monitoring results from ones that heavily deviate from the truth, yet significantly improves monitoring utilities compared to simple techniques that invalidate all monitoring results generated in the presence of messaging dynamics. Second, it adapts itself to nontransient messaging problems by reconfiguring distributed monitoring algorithms to minimize monitoring errors. This work demonstrates that, even under severe message loss and delay, this new approach consistently improves monitoring precision and, when applied to cloud application auto‐scaling, outperforms existing state‐monitoring techniques.

In distributed systems, the security of the host platform is critical. Platform administrators use security‐automation methodologies, such as those provided by the Security Content Automation Protocol (SCAP) standards, to check that the outsourced platforms are set up correctly and follow security recommendations, e.g. those provided by governmental or industrial organizations. Nevertheless, users of remote platforms must still have confidence in the platform administrators. (Aslam et al. 2013) proposed a remote platform‐evaluation mechanism that can be used by remote platform users or by auditors, to perform frequent platform‐security audits. The authors analyzed the existing SCAP and Trusted Computing Group (TCG) standards for this solution. They also identified shortcomings and suggested ways to integrate these standards. This platform‐security‐evaluation framework uses the combined effort of SCAP and TCG to address the limitations of each technology when used independently.

Auditing systems are extremely important to make sure that customer data is properly hosted in the cloud. (Yu et al. 2014) investigated the possibility for active adversary attacks in three auditing mechanisms for shared data in the Cloud, including two identity‐privacy‐preserving auditing mechanisms named Oruta and Knox, and a distributed‐storage integrity‐auditing mechanism. The authors showed that these schemes start to become insecure when active adversaries are involved in the cloud storage. In particular, they proved that there are ways for an active adversary to change cloud information without being detected by the auditor. The authors claimed to have found a solution to this downside without sacrificing the benefits of these mechanisms.

CSPs have control over enforcing cloud security in order to ensure the integrity and confidentiality of their customers' data. Cloud‐computing security infrastructure is an extremely important research area and the subject of a consistent body of work by both the academic and industrial research communities. In a cloud environment, resources are under the control of the CSP, and third‐party auditors must ensure data integrity and confidentiality, particularly when data storage is outsourced. (Sathiskumar and Retnaraj 2014) proposed data‐encryption and proxy‐encryption algorithms for CSPs to enable privacy and integrity of outsourced data in cloud‐computing infrastructures.

A service cloud infrastructure that offers web‐service composition improves the accessibility and flexibility of web services hosted on the cloud. Nevertheless, security challenges exist, which include both vulnerabilities due to classic web‐service communication and new, specific issues carried by intercloud communication. Cloud auditing is a complex task, due to the enormous scale of the system, the noncentralized architecture of the Cloud, and the wide range of security issues that need to be taken into account. Of course, there exist security standards, protocols, and auditing methodologies that provide audit logs, which can be subsequently analyzed, but these logs are not always suitable to reveal the type, location, and impact of the security threats that might have occurred. Assuming a cloud infrastructure that explicitly declares the scope of its audit logs, defines the expected auditable events in the cloud, and provides evidence of potential threats, in 2013, researchers introduced the concept of vulnerability diagnostic trees (VDTs) to formally demonstrate vulnerability patterns across many audit trails created within the cloud service. They accounted for attack scenarios based on the allocation of services to a web‐service composition that provides end‐to‐end client‐request round‐trip messaging.

In spite of the large body of research in the area of cloud security and auditing, performed by both the academic and industrial research communities, a collaborative cloud architecture for security and auditing that does not require a third‐party auditor (TPA) has not yet been fully explored. In their cloud‐security research survey, (Waqas et al. 2013) suggested a collaborative cloud architecture that can securely share resources when needed and audit itself without the involvement of a TPA.

8.4.1 Data Integrity

Data integrity is a security principle that establishes that users must not modify data unless they are authorized to do so. For all the cloud‐computing deployment models—IaaS, PaaS, and SaaS—data integrity is the foundation for providing service. Clouds offer a large number of entry and access points. Therefore, enforcing a good authorization mechanism is crucial to maintain data integrity. It is also important to provide third‐party supervision. In fact, analyzing data integrity is the prerequisite to deploying applications securely. (Bowers et al. 2009a) proposed a theoretical framework called Proofs of Retrievability (POR) for verifying remote data integrity by combining error‐correction code and spot‐checking. (Bowers et al. 2009b) also developed a high‐availability and integrity layer (HAIL) by using POR checking the storage of data across many different clouds servers. HAIL can also detect duplication of data copies and understand data availability and integrity. (Schiffman et al. 2010) presented a Trusted Platform Module (TPM) to check data integrity remotely.

When it comes to cloud computing, it is important not only to protect cloud consumers' sensitive data via use of cryptographic routines, but also to protect consumers from malicious behaviors by validating the operations performing computations on the data. (Venkatesa and Poornima 2012) proposed a novel data‐encoding scheme called layered interleaving, designed for time‐sensitive packet recovery in the presence of sudden data loss. This is a high‐speed data‐recovery scheme with minimal loss probability, which uses a forward‐error‐correction scheme to handle data loss. This methodology is highly efficient in recovering data right after sudden losses. In 2013, research started to design a cloud computing security development life‐cycle model to enforce data safety and minimize consumer data exposure risks. A data‐integrity‐verification algorithm eliminates the need for third‐party auditing by protecting static and dynamic data from unauthorized observation, modification, and interference. (Saxena and Dey 2014) proposed work to achieve better data‐integrity verification and help users utilize Data‐as‐a‐Service (DaaS) in cloud computing. This framework is partitioned into three platforms: platinum for storing sensitive data, gold for a medium level of security, and silver for nonsensitive data. The authors have also designed different algorithms for implementing these three platforms. Their results showed that this framework is easy to implement and provides strong security without affecting performance in any significant way.

8.4.2 Data Confidentiality

Cloud computing, among many other offerings, provides high availability and elastic access to resources. Third‐party cloud infrastructures, such as Amazon Elastic Compute Cloud (EC2), are completely transforming the way today's businesses operate. While we all take advantage of the benefits of cloud computing, businesses must realize that there can be serious risks to data security, particularly to data confidentiality, a security principle that establishes that private data cannot be exposed to unauthorized users. Very important factors, such as software bugs, operator errors, and external attacks, can interfere with the confidentiality of sensitive data stored on external clouds, thereby making that data vulnerable to unauthorized access by malicious parties. (Puttaswamy et al. 2011) studied how to improve the confidentiality of application data stored on third‐party computing clouds. They identified all functionally encryptable data—sensitive data that can be encrypted without reducing the functionality of the application on the Cloud. This data will only be stored on the Cloud in encrypted form, accessible exclusively to users with the correct keys. This mechanism protects data confidentiality against unintentional errors and attacks. The authors also described Silverline, a set of tools that automatically (i) recognize all functionally encryptable data in a cloud application, (ii) assign encryption keys to specific data subsets to minimize key‐management complexity while ensuring robustness against key compromise, and (iii) provide transparent data access at the user device while preventing key compromise even from malicious clouds. Via a thorough evaluation, the authors were able to report that numerous web applications heavily use storage and sharing components that do not require raw‐data interpretation. Thus, Silverline can protect the vast majority of data manipulated by such applications, simplify key management, and protect against key compromise. These techniques provided an important first step toward simplifying the complex process of incorporating data confidentiality into cloud applications.

Cloud data storage mainly gives small and medium‐sized enterprises (SMEs) the capability to reduce investments and maintenance of storage servers while still providing high data availability. The majority of SMEs now outsource their data to cloud storage services. User data sent to the Cloud must be stored in the public cloud environment. Data stored in the Cloud might intersperse with other user data, which leads to data‐protection issues in cloud storage. Thus, if the confidentiality of cloud data is broken, serious losses may occur.

Data confidentiality is one of the most important requirements that a CSP must meet. The most used method for ensuring cloud‐storage data protection is encryption. However, encryption by itself does not completely guarantee data protection. (Arockiam and Monikandan 2014) proposed a new way for achieving efficient cloud‐storage confidentiality. The authors used encryption and obfuscation as two different techniques to protect the data stored in the Cloud. Depending on the type of data, encryption and obfuscation can be applied. Encryption can be applied to strings and symbols, while obfuscation can be more appropriate for numeric data types. Combining encryption and obfuscation provides stronger protection against unauthorized users. In addition, (Alomari and Monowar 2014) addressed the problem of portability and secure file sharing in cloud storage. Their work consists of four different components: (i) the encryption/decryption provider (EDP), which performs the cryptographic operations; (ii) a TPA, which traces the EDP operations and audits them; (iii) a key storage provider (KSP), for key management; and (iv) a data storage provider (DSP), which stores user files in an encrypted form. Based on the experimental results, the authors demonstrated that this new encryption ensures data confidentiality and maintains portability and secure sharing of files among users. (Djebaili et al. 2014) listed the latest trends in cloud data outsourcing and the many different threats that may undermine data integrity, availability, and confidentiality unless cloud data centers are properly secured. The authors observed that different schemes addressing data integrity, availability, and confidentiality, and complying with all the security requirements must be in place. These include high scheme efficiency, stateless verification, unbounded use of queries, and retrievability of data. However, very important questions remain, particularly how to use these schemes efficiently and how often data should be verified. Constantly checking data is a clear waste of resources, but checking only periodically increases risks. The authors attempt to resolve this tricky issue by defining the data‐check problem as a noncooperative game, and by performing an in‐depth analysis on the Nash Equilibrium and the underlying engineering implications. Based on the game’s theoretical analysis, the sequence of reactions is to anticipate the CSP's behavior; this leads to the identification of the minimum resource verification requirement and the optimal strategy for the verifier.

8.4.3 Data Availability

It is essential that an organization is ready to respond in case a disaster occurs. A provider hosting a cloud data center must be prepared for such events and make sure that cloud data becomes available within seconds. It is crucial that organizations define strategies for protecting and restoring access to data appropriately according to operational needs. It is not possible to transfer terabytes (TB) or petabytes (PB) of data information through the network in a few seconds right after a disaster has occurred, and it is necessary for data to be saved in different locations in order to facilitate data access in the event of a failure in a given primary location. Different strategies are used by system architects for backing up data. (Cabot and Pizette 2014) concluded that replicated databases in cloud environments are a cost‐effective alternative for ensuring the availability of data in cloud systems, although some other issues arise, such as data synchronization and privacy. Since the cloud‐computing paradigm is based on the concept of shared infrastructure by multitenants, DDoS attacks on a specific target can quickly affect many or even all tenants. By guaranteeing that availability is given the necessary importance, organizations can enable stakeholders to properly assess the risks associated with a specific cloud‐computing model and successfully mitigate those risks in order to obtain the advantages of cloud computing while ensuring continuity of functions. Intrusion protection systems (IPSs) and firewalls are a crucial defense from these attacks, but they miss an important capability: these solutions do not protect the availability of services. Moreover, these technologies can themselves become targets of DDoS attacks.

8.4.4 Data Privacy

Based on an article published at the end of 2014 on searchcloudsecurity.techtarget.com (Wright 2014), the CSA said that cloud data privacy and the issue of defining authority over data would be the top problems for enterprises in both the United States and abroad for the year 2015. In fact, during that time, a legal battle between the United States government and Microsoft over e‐mails located in an offshore data center in Dublin, Ireland was going on. These issues were caused by different data‐privacy regulations among countries. These include the USA Patriot Act (http://www.justice.gov/archive/ll/highlights.htm), which is an act of the U.S. Congress that was signed into law by President George W. Bush on October 26, 2001, and a U.S. search warrant. Ireland's Minister for Data Protection has it made clear that “when governments seek to obtain customer information in other countries, they need to comply with the local laws in those countries.” The US Congress did not clearly express its intent to put a US business in the difficult position of violating the local laws of countries where customer data is saved in order to comply with a US search warrant. Instead of using a search warrant in the Microsoft case, the U.S. government should have followed the procedures of the Mutual Legal Assistance Treaty (MLAT) between the US and Ireland to ask to receive the necessary information from the Irish government in a way that is consistent with Ireland's laws. This episode may have negative impact on cloud investment and, ultimately, on the economy. This example is a perfect illustration of the types of conflicts that may arise in the near future.

Jim Reavis, co‐founder and CEO of the CSA, said that cloud data sovereignty is “probably the number‐one issue” for European enterprises using, or planning to use, cloud services. We must ensure that individuals and organizations can have confidence in the rules and processes that have been put in place to safeguard privacy.” On April 3, 2014 the Cloud Service Alliance announced the launch of the second version of its Privacy Level Agreement (PLA) Working Group (https://cloudsecurityalliance.org/media/news/csa‐announces‐pla‐v‐2). In an effort to help CSPs and future cloud customers objectively evaluate privacy standards, PLA V2, being sponsored by CSA corporate member EMC, aims to provide a clearer and more effective way to communicate to customers regarding the layer of data protection offered by a CSP. The PLA Working Group was originally founded in 2012 with the goal of defining compliance baselines for data‐protection legislation as well as establishing best practices for defining a standard for communicating the level of privacy measures (such as data protection and security) that a CSP agrees to follow while hosting third‐party data. Moreover, the PLA Working Group is composed of independent privacy and data‐protection subject matter experts, privacy officers, and representatives from data protection authorities (DPAs). The PLA Working Group released three core documents over the course of a year (April 2014–April 2015), as follows:

• The first document is the PLA V2, with special emphasis on the European Union (EU) market. The first deliverable of the PLA WG was a transparency tool for the EU market. Based on those initial results, the PLA Working Group is creating a compliance tool that will satisfy the requirements expressed by the Article 29 Working Party and by the Code of Conduct currently development by the European Commission (EC).
• The second document is the Feasibility Study on Certification/Seal based on the PLA. The group will provide a document assessing the feasibility of a Privacy Certification Module (PCM) in the context of the Open Certification Framework (OCF), and establish a roadmap and guidance for its creation and implementation.
• The third document is the PLA Outline for the Global Market. The CSA will expand the scope of the PLA V1 by considering relevant privacy legislation outside the EU.

8.4.5 Dataflows

One of the most attractive, yet most dangerous characteristics of cloud computing is to promote flow and remote storage of data. Within the public cloud‐computing deployment, new applications for data sharing are encouraging internet users to store their data on the Cloud with no mention of any geographical boundaries. The relevant CSPs may have their service centers anywhere in the world. In some jurisdictions, there are laws governing the use of personal data, for example by prohibiting the exportation of personal data to countries that have no enforceable data‐protection law. Data transfers from all countries with national legislation are restricted. These include all the countries in the EU and the European Economic Area (EEA), Argentina, Australia, Canada, Hong Kong, and New Zealand. From EU and EEA countries, personal information can be transferred to countries that have “adequate protection,” namely all other EU and EEA member states plus Switzerland, Canada, Argentina, and Israel. Germany's Data Protection Act, §11, introduced in Section 8.3.2, states that “where other bodies are commissioned to collect, process or use personal data, the responsibility for compliance within the provisions of this Act and with other data protection provisions shall rest with the principal.” The implication of this Act is that users must know the exact location of their data and their cloud providers' court of jurisdiction, and export or movement of data is not possible without prior notification of the customer.

Transborder dataflow is the transfer of computerized data across national borders. Restricting it is a form of prevention necessary to protect the personal data of the citizens of a country, but this has indirectly limited cooperation between countries and affected economic development. (Manap and Rouhani 2014) reported the outcome of a study they conducted. They described the fundamental concepts of free dataflow and how it can help the growth of a country's economic power, and proved that unrestricted dataflow allows for more competition in business activities and, therefore, more growth. Internationally dealing with businesses such as research and development, design, protection, sales, and support services, companies gain profit from transborder dataflow because they can receive the best services from the best suppliers. Furthermore, transborder dataflow can promote not only the economic growth of a country, but also its political and social development.

Cyberspace technology, including cloud computing, has changed the way data flows. Through global communication, the distribution of information permits efficient management of businesses by allowing a better way of using resources and services. However, growing public concern about the misuse of personal data has led to the introduction of privacy laws in various countries, with the disadvantage that cross‐border restrictions can negatively impact the choice and quality of products and services offered to the consumers around the world. In addition to the ban of transborder dataflow, another factor that can adversely affect a country's economy is the presence of different and incompatible standards and regulations of privacy protection from one country to another. Unstandardized data protection has created challenges in dealing with transfer of data worldwide. Moreover, complicated regulations on cross‐border restrictions lead to lost business prospects, with dramatic effects particularly in developing countries. It is suggested that the protection of law is important to enforce the proper use of data and assure the rights and concern of individuals. In order to alleviate the challenges introduced by strict laws, the mixture of procedures and cross‐border privacy protections can be minimized through the development of standard privacy regulations. These standard regulations should be adopted by corporate sectors in delivering business transactions over the internet. Adopting consistent corporate privacy rules would resolve several difficulties caused by the various personal‐data laws enforced in different countries. In addition, such adoption would allow the administration of data in a more unified and consistent way throughout organizations, independent of where the data may be exported.

8.4.6 The Need for Compliance

Because of the many cloud challenges mentioned so far, cloud customers must understand the terms and conditions of a CSP and which of these terms have to be spelled out in SLAs to maintain cloud compliance. Unfortunately, not all CSPs are in favor of providing detailed security assurances to customers. As Donna Scott, a Gartner Vice President (VP), said during an interview in 2013, “Gartner customers have logged many complaints about weak SLAs lacking the necessary guarantees when it comes to security, confidentiality and transparency.” It is essential, as John Morency, a Gartner Research VP, stated: “The devil is in the details,” meaning that these contracts should be written very clearly, with no ambiguity, and should explicitly indicate which party is responsible for which security operation. Jay Heiser, another Gartner Research VP, said that enterprises very often struggle to explain, when asked, what security controls a CSP can provide to its cloud customers. Many CSPs answer with very vague responses or unclear documentation. Heiser said that it is necessary for industries to have to define common certifications, so enterprises can offer their services more easily: for example, by using the US government's FedRAMP initiative, which is one of the closest to a global standard. On June 26, 2014 the EC published the Cloud Service Level Agreement Standardization (Cloud SLAS) Guideline. This document is a crucial component of the contractual relationship between a cloud service customer and a CSP. Based on the global nature of the Cloud, SLAs usually cover many jurisdictions, often with varying applicable legal requirements, in particular with respect to the protection of personal data hosted in the cloud service. In addition, different cloud services and deployment models require different approaches to SLAs, which add to the complexity of the SLAs. Moreover, SLA vocabulary, as of today, very often varies from one CSP to another, which makes it even more difficult for customers to compare cloud services. In fact, standardizing characteristics of SLAs improves clarity and increases the understanding of SLAs for cloud services in the market, in particular by highlighting and providing information on the concepts usually covered by SLAs. These are some of the reasons the Cloud Computing Strategy company is working on the development of standardization guidelines for cloud‐computing SLAs for contracts between cloud service providers and cloud service customers.

In February 2013, the EC Directorate General for Communications Networks, Content, and Technology (DG Connect) set up the Cloud Select Industry Group, Subgroup on Service Level Agreement (C‐SIG‐SLA) to work on these issues. The C‐SIG‐SLA, an industry consortium assisted by DG Connect, has created the Cloud SLA Standardization Guidelines document to provide a set of SLA standardization guidelines for CSPs and professional cloud service customers, while still ensuring that the specific necessities of the European cloud market and industry are taken into account. This initial standardization effort has will have the highest impact if standardization of SLAs is done at an international level, rather than at a national or regional level. Taking this into consideration, the C‐SIG‐SLA set up a connection with the ISO Cloud Computing Working Group to provide concrete input and present the European position at the international level.

In 2013, PCI DSS was updated to the next major revision, PCI DSS V3.0 (https://www.pcisecuritystandards.org/minisite/en/pci‐padss‐supporting‐docs‐v31.php). In January 2015, the first set of requirement changes rolled out for PCI‐DSS. While there are a number of areas where new requirements could affect the compliance plans of suppliers and service providers, both cloud‐based and traditional ones, the part that may potentially have the greatest impact related to the Cloud is in situations where the cardholder data environment (CDE) deals with cloud technologies. Suppliers are aware that PCI DSS compliance in the Cloud is a challenging topic, and for this reason the PCI Security Standards Council published a document intentionally for those providers using the Cloud in a PCI context. Three new requirements in PCI DSS V3.0 are particularly relevant to the use of cloud technologies in a PCI‐regulated infrastructure:

1. Requirement 2.4: “Maintain an inventory of system components that are in scope for PCI DSS.”
2. Requirement 1.1.3: “[Maintain a] current diagram that shows all cardholder dataflows across systems and networks.”
3. Requirement 12.8.5: “Maintain information about which PCI DSS requirements are managed by each service provider.”

Today, enterprises have to deal with increasing challenges meeting the different compliance and regulatory requirements applicable to them. The challenge significantly increases if the enterprise offers services in various geographies or across different verticals. For instance, an e‐commerce supplier with international customers needs to comply with data‐privacy and ‐disclosure mandates such as the EU data‐protection directive, California privacy laws, and PCI‐DSS. In addition, as a response to the recent economic losses experienced by enterprises as a result of e‐commerce hacking, more regulatory and compliance orders are expected as governments and regulatory organizations design requirements in an attempt to prevent future incidents.

Instead of handling each compliance requirement as an independent effort, an organization can obtain significant reductions in effort and cost by adopting standards such as ISO 27001 and 27002 (introduced earlier in Section 8.4) as their base security guidance. In addition, a governance, risk management, and compliance (GRC) framework allows an organization to avoid conflict, reduce overlap and gaps, and obtain better executive visibility to the risks faced. It also enables the organization to be proactive in addressing risk and compliance issues. There is no doubt that meeting compliance requirements by moving to a cloud‐based solution offers significant benefits.

Cloud compliance, or rather the lack of it, is an obstacle to cloud adoption. A lack of compliance for services in the Cloud makes the elasticity of the Cloud difficult for operations that must meet compliance mandates. However, a cloud‐based solution that includes compliance services opens up the general benefits of the Cloud to many more applications. Based on the type of business process (for example, handling credit‐card transactions) or industry (such as healthcare), the majority of organizations are subject to different compliance and regulatory authorizations. For instance, openly traded financial services institutions situated in the US, running their own data centers where both customer and corporate data and applications are located, must comply with the following standards:

• Gramm‐Leach‐Bliley Act (GLBA) guidelines
• PCI
• SAS 70 type II
• Various state and federal privacy‐ and data‐breach disclosure requirements

Companies depend heavily on reporting and auditing frameworks that include requests to detail controls and policies that have been implemented in order to ensure compliance. In general, a cloud solution must prove that it possesses identical types of safeguards and controls that are otherwise implemented privately. A CSP must also be able to provide evidence of compliance by showing regulation‐specific reports and audits, such as SAS 70 type II audit reports. In 2014, Citrix Systems Inc. published a document entitled Citrix Cloud Solution for Compliance describing the Citrix‐specific solutions and products that Citrix offers in helping cloud services comply with various regulations (https://www.citrix.com/content/dam/citrix/en_us/documents/products‐solutions/citrix‐cloud‐solution‐for‐compliance.pdf). The IEEE Cloud Computing group, which is part of the IEEE organization, is also working on standards for cloud computing (https://cloudcomputing.ieee.org/standards).

References

1. Alomari, E.A. and Monowar, M.M. (2014). Towards data confidentiality and portability in cloud storage. Design, user experience, and usability. User experience design for diverse interaction platforms and environments. Lecture Notes in Computer Science 8518: 38–49.
2. Arockiam, L. and Monikandan, S. (2014). Efficient Cloud Storage Confidentiality to Ensure Data Security. 2014 International Conference on Computer Communication and Informatics (ICCCI), Coimbatore, India.
3. Aslam, M., Gehrmann, C., and Björkman, M. (2013). Continuous security evaluation and auditing of remote platforms by combining trusted computing and security automation techniques. The 6th International Conference on Security of Information and Networks, Aksaray, Turkey.
4. Badger, L., Bernstein, D., Bohn, R., et al. (2011). National Institute of Standards and Technology Special Publication 500–293 (Draft). US Government Cloud Computing Technology Roadmap Volume I Release 1.0 (Draft). http://www.nist.gov/itl/cloud/upload/SP_500_293_volumeI‐2.pdf (accessed 16 December 2014).
5. Bishop, M. (2002). Design principles. In: Computer Security: Art and Science. Professional: Addison‐Wesley.
6. Bowers, K.D., Juels, A., and Oprea, A. (2009a). Proofs of retrievability: theory and implementation. In: Proceedings of the 2009 ACM Workshop on Cloud Computing Security, 43–54. ACM.
7. Bowers, K.D., Juels, A., and Oprea, A. (2009b). HAIL: a high‐availability and integrity layer for cloud storage. In: Proceedings of the 16th ACM Conference on Computer and Communications Security, 187–198. ACM.
8. Cabot, T. and Pizette, L. (2014). Leveraging public clouds to ensure data availability. White paper. http://www.mitre.org/sites/default/files/pdf/12_0230.pdf (accessed 10 January 2015).
9. Chen, Y., Paxon, V., and Katz, R.H. (2010). What's new about cloud computing security? EECS Department, University of California, Berkeley. Technical report UCB/EECS‐2010‐5
10. Djebaili, B., Kiennert, C., Leneutre, J. et al. (2014). Data integrity and availability verification game in untrusted cloud storage. Decision and game theory for security. Lecture Notes in Computer Science 8840: 287–306.
11. Doelitzscher, F., Reich, C., Knah, M. et al. (2012). An agent based business aware incident detection system for cloud environments. Journal of Cloud Computing: Advances, Systems and Applications 1: 9.
12. German Parliament (2009). Federal Data Protection Act. Federal Law Gazette I p. 66, as most recently amended by Article 1 of the Act of 14 August 2009 (Federal Law Gazette I p. 2814). http://www.gesetze‐im‐internet.de/englisch_bdsg/englisch_bdsg.html (accessed 13 December 2014).
13. Gondree, M. and Peterson, Z.N.J. (2013). Geolocation of data in the cloud. In: Proceedings of the Third ACM Conference on Data and Application Security and Privacy, 24–36. ACM.
14. Goodin, Dan. (2014a). Crypto attack that hijacked Windows Update goes mainstream in Amazon Cloud. http://arstechnica.com/security/2014/11/crypto‐attack‐that‐hijacked‐windows‐update‐goes‐mainstream‐in‐amazon‐cloud (accessed 12 December).
15. Goodin, Dan. (2014b). AWS console breach leads to demise of service with “proven” backup plan. http://arstechnica.com/security/2014/06/aws‐console‐breach‐leads‐to‐demise‐of‐service‐with‐proven‐backup‐plan (accessed 20 December 2014).
16. Hamza, Y.A. and Omar, M.D. (2013). Cloud computing security: abuse and nefarious use of cloud computing. International Journal of Computational Engineering Research 3 (6): 22–27.
17. Han, K.J., Choi, B.Y., and Song, S. (2014). High Performance Cloud Auditing and Applications, 6–7. Heidelberg, Dordrecht, London: Springer New York.
18. Kortchorski, B. and Rutkowska, S. (2009). Cloudburst. Technical paper. http://www.blackhat.com/presentations/bh‐usa‐09/KORTCHINSKY/BHUSA09‐Kortchinsky‐Cloudburst‐PAPER.pdf (accessed 15 December 2014).
19. Loutas, N., Kamateri, E., and Tarabanis, K. (2013). Cloud computing interoperability: the state of play. Centre for Research and Technology Hellas, Thessaloniki, Greece; Information Systems Lab, University of Macedonia, Thessaloniki, Greece.
20. Manap, N.A. and Rouhani, A. (2014). Issue of transborder data flows in cloud computing the impact on economic growth. International Business Management 8: 113–117.
21. Meng, S., Iyengar, A.K., Rouvellou, I.M. et al. (2012). Reliable state monitoring in cloud datacenter. Proceedings of the IEEE Fifth International Conference on Cloud Computing, 951–958. IEEE.
22. Mell, P. and Grance, T. (2009). The NIST definition of cloud computing. http://csrc.nist.gov/publications/nistpubs/800‐145/SP800‐145.pdf (accessed 28 January 2014).
23. Paladi, N., Aslam, M., and Gehrmann, C. (2014). Trusted geolocation‐aware data placement in infrastructure. The 13th IEEE International Conference on Trust, Security and Privacy in Computing and Communications IEEE (TrustCom), Beijing, China.
24. Patel, P., Ranabah, A., and Sheth, A. (2009). Service Level Agreement in Cloud Computing. Fairborn, Ohio: Wright State University CORE Scholar.
25. Puttaswamy, K.P.N., Kruegel, C., and Zhao, B.Y. (2011). Silverline: toward data confidentiality in storage‐intensive cloud applications. In: Proceedings of the 2nd ACM Symposium on Cloud Computing, article 10. ACM.
26. Ries, T., Fusenig, V., Vilbois, C. et al. (2011). Verification of data location in cloud networking. In: Proceedings of the Fourth IEEE International Conference on Utility and Cloud Computing (UCC), 439–444. IEEE.
27. Rutkowski, M. (2013). An Introduction to DMTF cloud auditing using the CADF event model and taxonomies. https://wiki.openstack.org/w/images/e/e1/Introduction_to_Cloud_Auditing_using_CADF_Event_Model_and_Taxonomy_2013‐10‐22.pdf (accessed 27 December 2014).
28. Sathiskumar, R. and Retnaraj, J. (2014). Secure privacy preserving public auditing for cloud storage. International Conference on Engineering Technology and Science (ICETS'14), Tamilnadu, India.
29. Saxena, R. and Dey, S. (2014). Collaborative approach for data integrity. Verification in cloud computing. Recent Trends in Computer Networks and Distributed Systems Security. Communications in Computer and Information Science 420: 1–15.
30. Schaub, Hillary. (2014). FedRAMP to monitor cloud service providers. https://www.brookings.edu/blog/techtank/2014/06/05/fedramp‐to‐monitor‐cloud‐service‐providers (accessed 22 January 2015).
31. Schiffman, J., Moyer, T., Vijayakumar, H. et al. (2010). Seeding clouds with trust anchors. In: Proceedings of the ACM Workshop on Cloud Computing Security, 43–46. ACM.
32. Sotto, L.J., Treacy, B.C., and McLellan, M.L. (2010). Privacy and data security risks and cloud computing. Electron. Comm. Law Rep.
33. Sunagar, S., Patil, U., and Sheshgiri (2014). Dynamic auditing protocol for data storage and authentication forwarding in cloud computing. IJRIT International Journal of Research in Information Technology 2 (4): 429–436.
34. Tiwana, B., Balakrishnan, M., Aguilera, M. et al. (2010). Location, location, location!: modeling data proximity. In: Proceedings of Hotnets, the 9th ACM Workshop on Hot Topics. ACM.
35. Vaish, A., Kushwaha, A., Das, R. et al. (2013). Data location verification in cloud computing. International Journal of Computer Applications 68 (12): (0975–8887).
36. Venkatesa, K.V. and Poornima, G. (2012). Ensuring data integrity in cloud computing. Journal of Computer Applications 5 (EICA2012‐4): ISSN: 0974–1925.
37. Waqas, A., Yusof, Z.M., and Shah, A. (2013). A security‐based survey and classification of cloud architectures, state of art and future directions. In: Proceedings of the Advanced Computer Science Applications and Technologies (ACSAT), International Conference, 284–289.
38. Weise, E. (2014). Apple's iCloud network under attack. http://www.usatoday.com/story/tech/2014/10/21/apple‐icloud‐‐attack‐network/17669603 (accessed 21 December 2014).
39. Wright, Rob. (2014). CSA to closely monitor enterprise cloud data privacy issues in 2015. http://searchcloudsecurity.techtarget.com/news/2240237429/CSA‐to‐closely‐monitor‐enterprise‐cloud‐data‐privacy‐issues‐in‐2015 (accessed 11 January 2015).
40. Yu, Y., Niu, L., Yang, G. et al. (2014). On the security of auditing mechanisms for secure cloud storage. University of Wollongong Research Online, Faculty of Engineering and Information Sciences. University of Wollongong, New South Wales, Australia and University of Electronic Science and Technology of China.

Further Reading

1. Arockiam, L. and Monikandan, S. (2013). Data security and privacy in cloud storage using hybrid symmetric encryption algorithm. International Journal of Advanced Research in Computer and Communication Engineering 2 (8): 3064–3070.
2. Arora, R. and Parashar, A. (2013). Secure user data in cloud computing using encryption algorithms. International Journal of Engineering Research and Applications 3 (4): 1922–1926.
3. Bugiel, S., Nurnberger, S., Poppelmann, T. et al. (2011). Amazonia: when elasticity snaps back. In: Proceedings of the 18th ACM Conference on Computer and Communications Security, 389–400. ACM.
4. Doelitzscher, F., Reich, C., Knahl, M. et al. (2012). An agent based business aware incident detection system for cloud environments. Journal of Cloud Computing: Advances, Systems and Applications 1 (1): 9.
5. Kamara, S. and Lauter, K. (2010). Cryptographic Cloud Storage. In: Financial Cryptography and Data Security (FC 2010) Workshops, RLCPS, WECSR, and WLC 2010, Revised Selected Papers, 136–149. Springer‐Verlag.
6. Karbe, T. (2013). Design and development of an audit policy language for cloud computing environments. Cloud Research Lab; University of Applied Sciences Furtwangen, technical report. http://wolke.hs‐furtwangen.de/publications/theses (accessed 10 January 2015).
7. Khorshed, M.T., Ali, A.B.M., and Wasimi, S.A. (2012). A survey on gaps, threat remediation challenges and some thoughts for proactive attack detection in cloud computing. Future Generation Computer Systems 28 (6): 833–851.
8. Krutz, R.L. and Vines, R.D. (2010). Cloud Security A Comprehensive Guide to Secure Cloud Computing. Wiley Publishing, Inc.
9. Mather, T., Kumaraswamy, S., and Latif, S. (2009). Cloud Security and Privacy: An Enterprise Perspective on Risks and Compliance. O' Reilly Media.
10. Morsy, A. and Faheem, H. (2009). A new standard security policy language. IEEE Potentials 28 (2): 19–26.
11. Ristenpart, T., Tromer, E., Shacham, H. et al. (2009). Hey, you, get off of my cloud: exploring information leakage in third‐party compute clouds. In: Proceedings of the 16th ACM Conference on Computer and Communication Security (CCS'09), 199–212. ACM.
12. Vieira, K., Schulter, A., and Westphall, C.B. (2010). Intrusion detection techniques for grid and cloud computing environment. IT Professional, IEEE Computer Society 12 (4): 38–43.
13. Wang, B., Li, B., and Li, H. (2012). Oruta: privacy preserving public auditing for shared data in the Cloud. In: Proceedings of the IEEE International Conference on Cloud Computing, 293–302. IEEE.
14. Wang, C., Wang, Q., Ren, K. et al. (2010). Privacy preserving public auditing for data storage security in cloud computing. In: Proceedings of INFOCOM, 525–533. IEEE.
15. Xiao, Z. and Xiao, Y. (2013). Security and privacy in cloud computing. IEEE Communications Surveys & Tutorials 15 (2): 843–859.
16. Yang, K. and Jia, X. (2012). Data storage auditing service in cloud computing: challenges, methods and opportunities. World Wide Web 15 (4): 409–428.
17. Zhang, L.J. and Zhou, Q. (2009). CCOA: cloud computing open architecture. In: IEEE International Conference on Web Services, 607–616. IEEE.
18. Zhang, X., Wuwong, N., Li, H. et al. (2010). Information security risk management framework for the cloud computing environments. In: Proceedings of 10th IEEE International Conference on Computer and Information Technology, 1328–1334. IEEE.
19. Zhu, Y., Hu, H., Ahn, G. et al. (2012). Cooperative provable data possession for integrity verification in multi‐cloud storage. IEEE Transactions on Parallel and Distributed Systems 23 (12): 2231–2244.