Distributed Machine Learning for Large-Scale IoT Systems with MEC

Machine learning in large-scale Internet-of-Things (IoT) systems can bring overwhelming demand on the limited resource of IoT devices, e.g., simple cameras and ambient sensors. However, it can substantially benefit from the additional computation and communication resources provided by a new wireless computing framework called Mobile/multiaccess Edge Computing (MEC). In this project, we study machine learning within the multi-level hierarchy of IoT devices and MEC servers, with an aim to develop new methods and techniques for data collection, task scheduling, and system optimization. This project is expected to generate theoretical insights and engineering guidelines to serve as foundation for innovative design in wide ranging products and services by our industry partner. It will also create significant market impact in a strategic sector of the Canadian economy.

Faculty Supervisor:

Ben Liang

Student:

Partner:

Ericsson Canada Inc (Quebec)

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

University of Toronto

Program: