Machine Learning Paradigms for Massive IoT

Massive Internet of Things (mIoT) is deemed to connect billions of miscellaneous mobile devices or IoT devices that empowers individuals and industries to achieve their full potential. A plethora of new applications, such as autonomous driving, remote health care, smart-homes, smart-grids etc., are being innovated via mIoT, in which ubiquitous connectivity among massive IoT devices are fully automated without human intervention. To achieve this trend, modern wireless networks need to satisfy the increasing demand of quality of services (QoS) with massive numbers of IoT devices, meanwhile face to the limit spectrum, expensive resource, green communication and security.

Machine learning is a field that concerns mainly with a fundamental question of how to learn generative models that would best explain the observed data. The key idea behind all the developed learning methods is to discover structures or latent variables that would i) most compactly represent the data and ii) retain most generalisation ability. Therefore, there is a great acquired the application of machine learning in the study and design of key functionalities of mIoT.

Objectives:

Although some research attempts have been made towards understanding theoretical and practical potential of these disruptive machine learning solutions, current studies to IoT are not complete. The challenging and breakthrough objective of this research is to leverage the potential of using machine learning technique to mIoT. The main objectives are to use advanced machine learning and artificial intelligence techniques (deep learning) to:

1) Make sense of massive amounts of data that might be generated from wireless channel measurements and sensor readings (massive IoT)
2) Analyse how physical layer functions (coding and modulation) perform and operate at both transmitter and receiver levels in order to come up with new operation and design definitions;
3) Learn end-user behaviour models to help adapting new wireless networks to the human users such that the overall quality-of-experience of the users is maximised.

Academic Requirements:

A minimum 2.1 honours degree or equivalent in Electrical and Electronic Engineering or relevant degree is required. International English Language Testing System (IELTS) 6.0 with a minimum of 5.5 in all four elements of the test or equivalent. A strong knowledge in optimisation, wireless communications, and machine learning is desired.

General Information

This 3 year International PhD studentship, potentially funded by the School of EEECS, commencing as soon as possible (intended start date 1 October 2018), covers tuition fees and a maintenance grant (approximately £14,000 per annum). The deadline for submission of applications is 15 July 2018 at the latest.

Applicants should apply electronically through the Queen’s online application portal at: https://dap.qub.ac.uk/portal/

Further information available at: http://www.qub.ac.uk/schools/eeecs/StudyattheSchool/PhDProgrammes

Supervisors Names:

Dr. Trung Q. Duong (Reader in Signal Processing Communications) Tel: 028 9097 1766,
QUB Address: Room 01/53 ECIT, Queen’s University Belfast NI Science Park Queen’s Road Queen’s Island Belfast BT3 9DT Web: https://sites.google.com/site/trungqduong/

Dr. Vien Ngo (Lecturer in Machine Learning) Phone: +44 (0)28 9097 1824

Deadline for submission of applications is 15th July 2016

For further information on Research Area click on link below:

http://pure.qub.ac.uk/portal/en/persons/trung-q-duong(00ca0df3-5113-489d-a43b-1c3527a10a1d).html