Fault-Tolerant Framework with Federated Learning for Reliable and Robust Distributed System

Data intensive computing system like distributed systems suffer a potential problem of data storage, load balancing and privacy of confidential data. The distributed data needs to be analyzed and protected properly for ensuring data privacy and improving the fault tolerance. Data privacy and security have become prominent issues for data-oriented applications. Fault-tolerant must be deployed for smooth functioning of the distributed computing system. A Federated learning (FL) model can be deployed on multiple edge devices. The Federated learning model is deployed on both horizontal and vertical scopes. The feature of distributed computing is that it is growing tremendously towards horizontal and vertical directions. FL deployed with distributed deep learning could identify, recognize, and resolve the faults at a large scale. Federated learning is a multimodal deep learning system that captures training data across various distributed and decentralized edge devices.