Energy-Efficient Secure Architecture For Personalization E-Commerce WSN

Abstract

A crucial challenge within e-commerce Wireless Sensor Networks (EWSNs) is the subtle equilibrium between personalised user experiences, transaction security, and real-time data processing. A comprehensive framework is introduced to enhance energy efficiency and security in EWSNs through the integration of Federated Learning (FL), edge computing, and blockchain technology. The key challenges, such as user privacy preservation, energy efficiency, and transaction trust, are addressed. The transaction trust and transparency are ensured by blockchain, contributing to a 30% reduction in transaction-related security breaches. The data privacy in the cloud layer is maintained through homomorphic encryption, resulting in a 27% decrease in privacy breaches. The effectiveness of the framework is quantitatively validated by experimental results, showing improvements of approximately 15% in privacy preservation, convergence speed, throughput, latency, and communication overhead. The security analyses include the resistance of the Proof-of-Energy (PoE) consensus mechanism against Sybil and Sinkhole attacks, with a success rate of 95% in preventing such attacks. Additionally, space and time complexity analyses, performance comparisons, and security theorems are presented, showcasing improvements of approximately 21% across various metrics.