IABHE: Identity and Attribute-based Honey Encryption Algorithm and Deep Learning-based Key Generation for Securing Big Data in Hadoop Framework

Abstract

Nowadays, Hadoop is considered promising to store and reliably process big data in the cloud computing platform, and it also provides low-cost and flexible services for large amounts of data. However, the possibility of malicious attacks on stored or processed data has increased due to the absence of inherent security mechanisms in Hadoop. In this approach, an Identity and Attribute-Based Honey Encryption Algorithm (IABHE) is developed for the encryption of data to offer security to the big data in the Hadoop framework. Here, the system model of the cloud is considered initially. Then, acquired input big data is applied to the MapReduce framework, where data encryption is performed by IABHE, which incorporates identity and attribute-based encryption with honey encryption. Moreover, a deep learning model, PyramidNet is utilized for the generation of secret keys used in encryption. Moreover, the encrypted data obtained in the mapper phase is aggregated by performing polynomial interpolation in the reducer phase. Finally, encrypted data is securely saved in the cloud. Here, the efficiency of IABHE is evaluated utilizing various performance metrics and it attained better values of encryption time at 17.756sec, decryption time at 8.767sec, and key complexity at 0.368.