Currently Available Research Projects (April 2018)

Please send all inquires to a.austin [at] auckland.ac.nz

Millimetre-Wave Communication Systems Engineering

This project will investigate the use of millimetre-waves for extremely high data-rate wireless communication systems. Of particular interest are millimetre-wave systems operating in the 60 GHz frequency band where considerable bandwidth is available, compared to the lower microwave bands (1-6 GHz) that are currently used for most wireless systems. However, the step to millimetre-wave wireless systems entails a number of research challenges.

Objectives: The first stage of the project will involve the design and hardware implementation of a millimetre-wave hardware testbed operating in the 60 GHz band. In the second stage, the project will focus on using the hardware testbed to experimentally investigate aspects of millimetre-wave systems, e.g., receiver architectures, modulation schemes, and/or beam-steering algorithms.

Reflectarray Antennas for Indoor Millimetre-Wave Systems

The considerable bandwidth available at mmWave frequencies has the potential to deliver wireless systems with transmission rates comparable to wired systems, and enables new applications. However, moving from the conventional sub-6 GHz to mmWave frequency-bands introduces a number of significant challenges for antenna design. To overcome the increased propagation loss, future mmWave systems require highly directional 'steerable' antennas. This project will investigate reflectarray antennas, which offer the ability to create very directional beams that can be electronically steered to achieve area coverage.

Objectives:
  • Investigate the design and experimental verification of reconfigurable reflectarray antennas for indoor mmWave systems; and

  • Examine novel approaches to increase the coverage area of mmWave systems by using additional reflectarray surfaces to 'bounce' the energy into the shadowed regions.

True Full-Duplex Wireless Communications

The limited availability of suitable radio frequency spectrum has renewed interest in physical-layer technologies to improve spectral efficiency. One of the most promising techniques is 'true' full-duplex communication, where transmission and reception can occur simultaneously in the same frequency band. The practical realisation of full-duplex systems requires the strong self-interference signal (arising from the closely coupled transmitter and receiver circuits) to be suppressed, ideally to the noise-floor.

Objectives:
  • Investigation and proposals for new SI cancellation algorithms and corresponding full-duplex architectures;

  • Development of a hardware testbed; and

  • Implementation and testing in real RF hardware.

Uncertainty Quantification in Computational Electromagnetics

Computational techniques are widely used to analyze and design electromagnetic structures and devices. However, uncertainties in the description of the problem (e.g., fabrication tolerances, temperature, and material properties) are difficult to include in most computational methods. Currently, Monte-Carlo techniques are used to estimate the impact of input uncertainties, however, these are slow to converge and require multiple simulation runs.

Objectives: This project will investigate new and novel approaches to more efficiently include uncertainties in computational electromagnetic techniques using Polynomial Chaos.

Inverse Scattering Problems in the Presence of Uncertainty

Inverse scattering problems arise when electromagnetic waves are used to image 'hidden' objects, e.g., buried landmines, underground oil reserves, or tumours in the body. Inverse problems in electromagnetics are considerably more difficult to solve than the corresponding forward problems, as wave scattering leads to non-linear partial differential equations for the inverse operators. Existing analytical inversion techniques applied to electromagnetic imaging often fail for 'ill-posed' scenarios, particularly when parameters of the problem are uncertain.

Objectives: The aim of this project is to investigate computationally efficient techniques to explore the multi-dimensional parameter space created by the uncertainties in inverse scattering problems. In particular, the specific objectives are to develop new numerical methods for computational electromagnetics where the inputs are probabilistic measures, instead of exact numerical values.