Machine learning on telecommunication networks with cloud implementation

• Grant number: 21/00253-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2021
• End date: January 31, 2022
• Field of knowledge: Engineering - Electrical Engineering - Telecommunications
• Agreement: MCTI/MC
• Principal Investigator: Rodrigo Caiado de Lamare
• Grantee: Leonardo Farina de Souza de Botton
• Host Institution: Centro Técnico Científico (CTC). Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)
• Associated research grant: 15/24499-0 - Cloud radio access networks for 5th generation wireless communications: new concepts, algorithms and applications, AP.TEM

Abstract

This project aims to implement a network of devices based on Arduino, using machine learning in a new way, referred to as Federated Learning. In this method, learning is distributed so that little information is exchanged between devices, avoiding congestion of the communications network. The cloud stores the model to be trained, which undergoes minor changes on each device, creating a system with non-centralized training data, and faster processing speed. We then seek to develop a model, on a small scale, relatively efficient and versatile, illustrating the advances and possibilities of using this technique. (AU)