Edge AI and Federated Learning for Privacy-Preserving IoT Systems: System Architectures, Non-IID Optimization, and Security Bounds
Abstract—The proliferation of Internet of Things (IoT) devices has created an unprecedented surge in decentralized telemetry, sensor data, and localized computational workloads. Conventional cloud-centric machine learning architectures require centralized pooling of raw edge data, which induces prohibitive wide-area ne...
Yogesh Kumar, Manas Pareek, Dr. Vishal Shrivastava et al.
· Zenodo (CERN European Organi... · 0 citations