Title: Creation of an iterative process for blockchain networks protecting privacy combining secure routing based on homomorphic encryption with dynamic routing in deep Q-networks

Authors: Hiralal Solunke; Pawan Bhaladhare; Amol Potgantwar

Addresses: School of Computer Science and Engineering, Sandip University Nashik, Maharashtra, India ' School of Computer Science and Engineering, Sandip University Nashik, Maharashtra, India ' School of Computer Science and Engineering, Sandip University Nashik, Maharashtra, India

Abstract: Technology is changing rapidly every day. The current blockchain routing protocols frequently fail to adequately handle all privacy concerns and provide less-than-ideal compromises between privacy protection and network throughput. To circumvent these constraints, this work presents a revolutionary privacy-preserving blockchain system that combines several state-of-the-art technologies from data science, machine learning, artificial intelligence, and encryption. The model that has been suggested includes five novel techniques. Firstly, dynamic routing with deep Q-networks keeps strong privacy levels while reducing average latency by a significant 20%. Second, autonomous privacy-preserving routing system. Third, by analysing and optimising network data using graph analytics combined with differential privacy, privacy-preserving graph analytics for blockchain networks reduces privacy leakage by 30%. Fourth, Homomorphic Encryption-based Secure Routing achieves almost negligible information leakage by using homomorphic encryption to enable calculations on encrypted data. Last, Bayesian privacy-preserving routing models increase privacy levels by 35% while having little effect on routing efficiency.

Keywords: blockchain optimisation; privacy-preserving technologies; deep Q-learning; homomorphic encryption; differential privacy.

DOI: 10.1504/IJBC.2024.142286

International Journal of Blockchains and Cryptocurrencies, 2024 Vol.5 No.2, pp.136 - 160

Received: 20 May 2024
Accepted: 30 Aug 2024
Published online: 16 Oct 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article