Privacy-preserving federated learning across decentralized edge devices in iot networks

The rapid growth of Internet of Things (IoT) technologies has significantly increased the generation of sensitive user data through healthcare wearables, industrial sensors, smart transportation systems, and intelligent monitoring devices. Conventional centralized machine learning architectures require transfer of raw datasets to cloud servers, thereby increasing privacy risks and cybersecurity vulnerabilities. Federated Learning (FL) has emerged as a decentralized machine learning paradigm that enables collaborative model training across distributed edge devices without exposing sensitive local data. The present study evaluated a privacy-preserving federated learning architecture integrated with secure aggregation and differential privacy techniques in decentralized IoT environments. Experimental analysis demonstrated progressive improvement in model accuracy, communication efficiency, and cybersecurity performance while minimizing privacy exposure risks. The findings indicate that federated learning can significantly contribute toward the development of secure, ethical, and scalable artificial intelligence systems for healthcare and smart IoT ecosystems.