Deep Belief Network-based Decentralized Secure, Lightweight Data Auditing Scheme for Streaming Outsourcing Data

A cloud storage environment is an emerging paradigm that provides storage services to various user categories to outsource their data. The multi-tenancy nature of cloud data storage leads to multiple security concerns like data breaches and unauthorized data access. The cloud user must outsource their private data to the cloud server. To mitigate those challenges and enhance data integrity in the multi-tenancy cloud environment, various researchers have designed and implemented many secure auditing schemes to ensure the data integrity of their outsourced data. However, these architectures often struggle with computation overhead and managing the data integrity verification among the streaming outsourcing data at runtime. They are also less capable of block-based audits due to improper data auditing structures. In this context, the proposed Decentralized lightweight data integrity verifying model for secure auditing of the streaming of outsourced data, implemented using a deep belief network, offers significant benefits. It secures the data on various layers of the neural network, constructs the feature map to compute the duplicate of the data, and enhances the attacker's data integrity and identity traceability to attain user revocation. The model's hidden layer efficiently processes the dynamic cloud data on insert, delete and update operations with less computation cost and time. It can eliminate duplication of data extracted from cloud databases. These features of our model significantly enhance data integrity and security, providing a robust solution for cloud data storage.