Distributed authentication framework for Hadoop based bigdata environment

• Published in: Journal of ambient intelligence and humanized computing Vol. 13; no. 9; pp. 4397 - 4414
• Main Authors: Hena, M., Jeyanthi, N.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-09-2022; Springer Nature B.V
• Subjects: Access control; Algorithms; Artificial Intelligence; Authentication; Authentication protocols; Big Data; Blockchain; Comparative studies; Computational Intelligence; Computer systems; Computing time; Cost analysis; Cryptography; Cybersecurity; Data systems; Datasets; Digital signatures; Distributed processing; Engineering; Original Research; Passwords; River beds; Robotics and Automation; Time synchronization; Traffic delay; User Interfaces and Human Computer Interaction; Digital signature; Kerberos; Big data; Apache Hadoop; Threshold ElGamal cryptosystem; Blockchain
• ISSN: 1868-5137; 1868-5145
• DOI: 10.1007/s12652-021-03522-0

Big data, the upcoming technology in the field of computing, refers to a large complex dataset. It deals with large complex datasets and yields great valued information when analysed properly. Data Security has become the greatest challenge in the minds of cyber experts and researchers in this scenario. Apache Hadoop frameworks that let distributed processing of these large datasets rely on Kerberos Authentication for mutual authentication and verification. The protocol comes with inherent challenges like Single point of failure, Dictionary Attacks, Replay Attacks, and Time Synchronization problems. This paper puts forward a one-off approach based on recent technologies like Blockchain Networks, Digital Signatures, and Elliptic ElGamal and Threshold Cryptosystem. The proposed scheme aims to mainly deal with the Single Point of Failure problem. Riverbed Modeller (AE) simulation is performed to do the comparative study of the proposed scheme with existing systems that use traditional encryption standards like RSA cryptosystems. Analysis of the simulation results proves that the proposed scheme is more efficient in terms of time and memory without compromising the level of security offered. The response time, network delay and traffic rates of the proposed system are compared with the existing RSA based system and the results strengthen the claims of this work. Lastly, the results from comparative analysis of security features and computational time cost indicate that the proposed method heightens the security level offered for big data systems with a nominal effect on performance.