Federated Learning in Intelligent Traffic Management System

The data generated by the intelligent traffic management system (ITMS) is large containing videos and images of the traffic. The machine learning (ML) based ITMS requires the transmission of all the generated data to the centralized server which involves the transmission cost and time. The drawbacks of these kinds of systems are expensive communication, heterogeneous devices and location, and privacy user. Federated Learning (FL) is the technique that provides the solution to deal with these challenges. Focusing on the use of FL in ITMS, this paper includes a brief overview of FL. This also includes the discussion on the ITMS application for which FL can provide an efficient and better solution. A brief discussion on the benefits of FL in ITMS is included followed by the conclusion.