Decentralized Systems and Distributed Computing

14Merging Blockchain and Deep Learning for Authentication and Security Architectures

Abhay Kumar Yadav^1* and Satya Bhushan Verma²

¹Guru Gobind Singh Indraprastha University, New Delhi, India

²Shri Ramswaroop Memorial University, Barabanki, India

Abstract

Deep learning (DL) has attracted a lot of attention recently due to its ability to make informed decisions. Features such as dependability, operational transparency, security, reliable data provenance, and traceability are missing in centralized servers based on DL systems. This chapter highlights the need to merge DL and blockchain technology for eliminating the drawbacks of centralized servers. Generally, DL algorithms are employed to authenticate, spot anomalies, and offer system security in cryptography and biometric systems. Blockchain, on the other hand, provides decentralized and secured storage of data. It is important for DL algorithms and the blockchain-based security system to establish balance between confidentiality and efficacy as network sensors are energy-constrained devices. This paper also highlights security models presented for providing authentication and security features using Blockchain, DL, or both based on different systems based on clouding type, models, consensus protocols, applications, services, and deployment goals.

Keywords: DL, blockchain, multi-layered security architecture, healthcare, IoT, SCM

14.1 Introduction

Deep learning (DL) idea has been present in nearly every industrial sector. ...