Lightweight Federated Homomorphic Encryption for Secure and Energy-Efficient Data Transmission in 6G IoT-WSNs

Abstract

The 6G-enabled Internet of Things (IoT) applications that use Wireless Sensor Networks (WSNs) produce large amounts of sensitive information that need to be transmitted safely and with minimal power consumption. Traditional cryptography methods ensure data security during transmission but needs processing without decryption hence exposing confidential data and consuming more computational power. The fully homomorphic encryption (FHE) allows processing encrypted data, and its huge complexity and resource usage make it inapplicable in the case of limited IoT-WSN nodes. This leaves a severe research gap in the development of lightweight privacy-preserving encryption mechanisms that enable the use of secure computation with minimal energy use. The present paper presents a Lightweight Federated Homomorphic Encryption Framework (LFHEF) a framework that facilitates the implementation of encrypted data analytics without decryption by integrating optimized partial homomorphic encryption with federated secure aggregation. The trust-based node evaluation model is integrated into the framework to address the malicious contributions and minimize the communication overhead. Python simulations of real-time environment IoT datasets, network scenarios simulated with NS-3, are evaluated experimentally. The areas of performance examined include encryption time, energy usage, latency, throughput and privacy leakage. Findings illustrate that the proposed framework simultaneously decreases encryptions overhead by 47, energy usage by 39, and end-to-end latency by 34 relative to the current frameworks based on HE and ensures a high level of privacy is preserved.