This project is being developed as part of a cooperation between Uppsala University and Ericsson research for the course Project CS in autumn 2023. It aims to evaluate the performance of federated learning and split learning for user equipment (UE) mobility prediction. Both models provide data privacy by enabling training directly where data is located, preventing the necessity of sending and collecting potentially sensitive data at a centralized location.
Contained in the bloom
folder is the main project including running instructions, requirements and the code. lit_research
includes summaries of papers relevant to the project.