Secure and Efficient Data Fusion in IoT Systems Using Homomorphic Encryption and Machine Learning

This research examines the secure and effective combination of information in Internet of Things (IoT) frameworks utilizing homomorphic encryption and machine learning strategies. The study proposes and assesses four calculations: Secure Aggregation utilizing Homomorphic Encryption, Unified Learning with Secure Model Aggregation, Privacy-Preserving Ensemble Learning, and Secure Multiparty Computation for Collaborative Information Combination. Through tests conducted on manufactured IoT datasets, we survey the security, productivity, and precision of the proposed calculations. Results illustrate high levels of security, with secrecy and judgment protected in all calculations. The computational overhead and communication costs are found to be direct, guaranteeing commonsense possibility in real-world IoT organizations. Furthermore, the precision of the melded information remains competitive, with cruel squared error (MSE) values extending from 0.018 to 0.027 over the calculations. By and large, our investigation contributes to progressing the state-of-the-art in secure and privacy-preserving IoT information combination, empowering educated decision-making and prescient analytics in differing application spaces.