PAS-Cloud : Toward a Framework With privacy preserving public Auditing Strategies for Heterogeneous Data in cloud storage

Abstract As enterprises cannot maintain local copy of outsourced data due to local storage issues, it is essential to have secured and privacy preserving means of outsourcing. Otherwise the data owners do not trust cloud service providers. Efficient auditing techniques are important to ensure privacy, security and data integrity. Many public auditing schemes came into existence. Recently, proposed a novel public auditing scheme based on Dynamic Hash Table. Third Party Auditor (TPA) in their scheme cannot view the data while performing auditing. However, most of the existing public auditing schemes do not have mechanisms differently for different type of data outsourced to public cloud. Structure, unstructured and semistructured data can exist in the outsourced cloud storage. One size does not fit all as no single audit method can satisfy all types of cloud data. Therefore it is the new research trend to design a comprehensive framework for public auditing strategies for structured and unstructured outsourced cloud data. In this paper, we propose such framework which supports multiple public auditing strategies to cater to heterogeneous types of data. It is named as PASCloud. It framework consists Dynamic Hash Table (DHT) based public auditing for unstructured data and a verifiable auditing scheme (VAS) for outsourced structured data or relational database. The former can detect data integrity issues caused by CSP either intentionally or unintentionally while the latter can withstand intentional misbehaviour of CSP when a query is made to outsourced relational database. Amazon AWS cloud platform is used for experiments. The experimental results revealed the utility of the framework in rendering auditing services for user's unstructured outsourced data while the implementation for structured data is deferred for our future work.