FedEdge: Federated Learning with Docker and Kubernetes for Scalable and Efficient Edge Computing

This paper introduces a novel framework that integrates Docker and Kubernetes to address well-known challenges in federated learning. Federated learning has gained significant attention as a privacy-preserving and scalable machine learning paradigm. However, existing frameworks often lack portability, scalability, resource efficiency, fault tolerance, standardization, and ecosystem integration. To overcome these limitations, we propose a conceptual framework that combines Tensorflow FL with Docker’s containerization capabilities and Kubernetes’ orchestration capabilities. Our approach fills all the gaps in existing FL frameworks. By leveraging Docker containers, our model achieves efficient resource allocation, maximizing computing resources while maintaining portability and scalability. Kubernetes further enhances resource allocation by orchestrating the deployment of these containers, minimizing resource consumption. Our proposed framework provides opportunities for largescale distributed machine learning applications, enabling the widespread use of federated learning methodologies.