Distributed deep learning for cognitive radio networks
Project Leader: Tansu Alpcan
Staff: Chris Leckie (CIS), Sarah Erfani (CIS)
Sponsors: Northrop Grumman Corporation, USA
Primary Contact: Tansu Alpcan (email@example.com)
Keywords: communications and networks; complex and intelligent systems; defence; machine learning; optimisation
Disciplines: Computing and Information Systems,Electrical & Electronic Engineering
The project aims to develop a distributed cognitive network based on Software-Defined or Cognitive Radio (SDR/CR) nodes, which will detect various patterns on the electromagnetic spectrum using machine and deep learning algorithms. The undelying goal is to provide defence, law enforcement, and emergency response agencies the much needed situational awareness on the radio spectrum. The research challenges include finding the segments of the spectrum that are used by adversaries and distinguishing adversaries' communication from background traffic while combining inputs from multiple small-scale SDR/CR receivers in an online, distributed, and adaptive fashion. An important related challenge is optimum allocation of distributed resources available to the Cognitive Radio Network. The methodology includes machine, adversarial, and deep learning techniques in addition to system, game, and optimisation theories.
The interested students should have a solid background in mathematics, engineering, or computing. The project will balance experimental/computational analysis with developing theoretical insights. The candidates with Australian citizenships will be preferred. Successful applicants will benefit from a generous top-up scholarship and future employment opportunities.