Cloud Storage: Privacy Enhancement Challenges
Recent tendencies in digital technology are associated with the outsourcing of data and granting storage and processing to remote systems. Experts in the IT industry and various businesses recognize that this method opens wide perspectives for any operations that involve data processing and storage. For instance, Jadeha and Modi (2012) claim that cloud computing and storage allows outsourcing the data away from desktop and portable computers, which opens new perspectives for their use. Among the most essential benefits of this technology, scholars name the absence of the need for payment for the infrastructure, its installation and human resources necessary for granting its maintenance (Jadeja & Modi 2012). One of the most beneficial platforms in this sense is portable computers and smartphones, the computing and storage capacities of which are significantly improved by cloud computing and storage. As a result, observing the benefits of the new technology, individuals and organizations demonstrate a tendency towards rapid shifting to its use (Sanaei et al 2014). Such tendencies lead to the fact that the complexity of cloud system networks is increasing at a great speed, which is why some of its non-solved issues require analysis and mitigation. Sanaei et al. (2014) identify the following problems: the need for cross-platform applications, which allow accessing cloud storage and computing services regardless of the device of the end user. However, they believe that the gravest issue connected with cloud storage technologies is data security and integrity.
The literature review devoted to the investigation of the problem of data integrity in cloud storage systems reveals that although current cloud networks possess specific data integrity tools, they are far from being efficient. One of the reasons for this is that modern businesses that engage with cloud storage companies fail to realize that date outsourcing leads to the fact that organizations do not own the data physically (Wang et al. 2013). Such misjudgment can be explained by the fact that multiple businesses are not aware of the taxonomy of cloud networks, presuming that their data integrity patterns are similar to those in local and metropolitan area networks (Wang et al. 2013). However, data outsourcing means that the users’ sensitive data can be exposed to other users in multiple ways, including network hacking, data interception, direct retrieval of information by accessing physical part of the cloud storage and other. Due to the fact that these vulnerabilities are diverse, various experts attempt investigating the ways of creating IT solutions that would suit the users in terms of granting rapid access to data and ensuring its integrity. The major issue associated with this aim is that outdated methods of encryption lead to delays in data access, and restrain the process of the search of information while failing to improve data security and integrity. For instance, the drawback of the typical cloud privacy-preserving logic is the creation of a large index size, which drastically increases the complexity of the search and restrains information access capacities within a cloud (Wang et al. 2014). In this respect, search index security within a cloud is one of the most critical factors because unauthorized users can retrieve privacy information from encrypted data by using indexing methods (Zhang et al. 2013). As a consequence, search index requires special protection in the cloud that would not lead to the restraint of the computation and search capacities of the network. In order to provide a balanced solution, which would allow combining data integrity and rapid access to the cloud, Wang et al. (2014) propose using multi-keyword fuzzy search mechanics, which exploit the locality-sensitive hash method. However, the drawback of this research is that it tends to bypass the need for enhancing the protection of such a cloud-based search engine. It is possible to mitigate this problem by turning to the use of traditional end-to-end encryption and data utilization services as proposed by Li et al. (2013). However, one may presume that the use of this encryption pattern would neutralize the rapid access results achieved by using the locality-sensitive hash method. The cause for this neutralization might be the excessive demand for computing resources for the end users used for decryption, which would bring no benefits from the overall idea of data outsourcing.
Among the recent investigations of data integrity aspects in the cloud storage and computing systems, the most perspective studies are associated with the usage of the methods of auditing the cloud. The two main branches of research that attempt to solve this issue are the ones that develop special cloud auditing protocols, advanced data encryption tools and methods of enhanced public auditability of cloud servers. For example, Yang and Jia (2013) propose creating an effective and secure dynamic protocol used in cloud networks for data storage. Their research is impelled by the need for mitigating the vulnerability, which is the result of the fact that third parties store and transmit the users’ information. In this sense, scholars suggest that such networks should migrate to enhanced protocols designed especially for cloud networks, which would grant data integrity and multiple user data security within a cloud (Yang & Jia 2013). A solution proposed by the researchers is that cloud storage systems have to establish an autonomous auditing network on the basis of batch auditing systems designed for multiple clouds (Yang & Jia 2013). The benefit of this method is that this auditing protocol would grant data integrity disregarding the taxonomy and complexity of the clouds involved in data processing and storage. Other scholars suggest that data integrity checking protocol should not be the part of the cloud, but should remain remote instead, which would increase its security. By using this method, Yu et al. (2014) propose establishing a “zero-knowledge privacy” cloud auditing system that would ensure “privacy against third party verifiers.” An important issue raised by this research is that data auditing systems also require the patterns of data protection because in case they are hacked, they fail to approve the data integrity within a cloud. One of the best solutions in this respect is modern encryption systems designed especially for cloud network use. Among these, Zhang et al. (2013) propose establishing a k-anonymity technique for the attribute indexes to prevent users from having their data hacked. Another promising method of establishing secure cloud storage systems is a homographic encryption scheme that is based on the method of the Elliptic curve cryptography developed by Chakraborty et al. (2013). Currently, this method of encryption is one of the most beneficial as it allows bridging the flexibility of data operation together with granting its secure storage and integrity. Therefore, the homographic encryption method is one of the most efficient because it combines several means of supporting data integrity within a cloud. One of them is the use of a third-party auditor, a rational model of encryption the efficacy of which has been approved experimentally (Chakraborty et al. 2013). Consequently, one may suggest that using a combination of this model of encryption together with multi-keyword fuzzy search mechanics would make modern cloud storage services fast, flexible, and secured. At the same time, recent studies of the search engines demonstrate that the competitive methods of protected search, such as ked keyword search and structured data search, might also be efficient without a significant increase of the self-cost of such systems (Li et al. 2012). Therefore, although cloud systems have diverse flaws in terms of sensitive data integrity and security, modern IT technologies have up-to-date means for their potential mitigation.