Elucidation of Adaptive Long Short-Term Memory for Data Deduplication and Data Security Enhancement by Hashing Algorithm over Cloud Sector

Abstract

The advanced big data technology and cloud computing methodology change the method of the user and efficiency while processing the data, where efficient storage and scalable computing are provided to the user by the cloud servers anytime and anywhere. Many cloud service providers have been attracted to the data deduplication approach that highly decreased storage costs. The bandwidth requirement of users has been reduced and data redundancies in cloud storage have been removed by using data deduplication techniques. Most of the general data deduplication models are affected by many privacy and security problems because of the outsourced data transmission techniques of cloud storage. Therefore, deduplication approaches have been implemented to handle specific privacy and security issues that leads to a wide range of trade-offs and solutions for cloud data. Hence, a new approach is introduced based on the adaptive network and hashing for secure data deduplication. Input attributes are collected initially using the standard dataset. The collected attributes are given to the Adaptive Long Short-Term Memory (ALSTM) model for data deduplication. Here the attributes in the implemented ALSTM are tuned by the Bald Eagle Search (BES) strategy. Further, the hashing function is included to encode the deduplication data to improve security. Finally, the attained result from the implemented framework is compared with the standard data deduplication method for analyzing the deduplication efficiency and security of the proposed model.