Research Projects

The following list represents a selection of projects belonging to our members.

RECRUITING

Indicates projects currently recruiting research students.
Visit the Melbourne School of Engineering: Study for program and application information.


RECRUITING A novel Deep Brain Stimulation system for Parkinson’s disease therapy

Team: David Grayden, Anthony Burkitt

This translational research project will develop a new implantable system to enable closed-loop Deep Brain Stimulation (DBS) therapy. Current DBS technology was approved for treating Parkinson’s disease in 2002 and uses pacemaker-like stimulation to treat the symptoms caused by the loss of dopaminergic neurons. We are developing a next-generation system, capable of recording neural signatures from the brain electrodes, in order to provide adaptive stimulation. With the enhanced electrode array and lead system, we also aim to reduce the incidence of undesired side-effects (30%) and medical complications (14 – 27%) related to limitations of the technology.

This project involves technology development and conducting preclinical evaluation of the electrode array and lead, to enable a first-in-human clinical trial. The in vivo efficacy of the electrode array for recording neural activity will be assessed, as well and the tissue response to the new lead system. The project is focused on developing a simpler, minimal surgical approach where all the components are implanted in the head and a single-step electrode insertion is performed. The system will improve DBS therapy by improving stimulation selectivity, which will translate into better symptom relief with fewer side-effects. This disease contributes to a major quality-of-life impairment and burden for caregivers. There is a critical need to address the efficacy issues as DBS is considered the last resort for refractive Parkinson’s disease patients.

Read more

View description


Adaptive learning visual sensor networks for crowd modelling

Team: Jayavardhana Gubbi Lakshminarasimha, Marimuthu Palaniswami, Slaven Marusic

The prevalence of camera networks for surveillance, together with the decreasing cost of infrastructure, has produced a significant demand for robust monitoring systems. Current systems offer limited functionality, particularly in their reliance on centralised processing of gathered information. This project addresses end-to-end system challenges of wireless visual sensor networks. Integrating developments across the spatial, spatio-temporal and decision domains, the project will incorporate distributed sensor network technology with intelligent fusion of information, to deliver unique long-term behaviour analysis capabilities for efficient planning in highly crowded environments.

Read more

View description


Advanced epileptic seizure warning methods

Team: David Grayden, Anthony Burkitt, Levin Kuhlmann

This project will develop epileptic seizure prediction methods, warning patients of the likelihood of an impending seizure, so that precautionary measures can be taken. Seizure prediction will be of great clinical significance as it will improve the lives of 33% of epileptic patients who have drug-resistant epilepsy, by warning of impending seizures and potentially allowing acute therapies to prevent seizures, such as electrical stimulation of the brain or targeted drug delivery.

We have a unique opportunity as a result of the first and only clinical trial of an implantable intracranial EEG monitoring system and seizure predictor developed by NeuroVista Corporation. This trial was completed in Melbourne with CI Cook as a lead investigator. Patients were implanted with the system for periods of up to 2.5 years, providing unprecedented amounts of EEG data per individual. The large amount of data means we can robustly evaluate new and existing seizure prediction methods, and quantify the effects of the diurnal cycle and EEG statistics on seizure prediction. This has not been done before.

The NeuroVista trials showed that seizure prediction is possible and useful, but further research is needed for it to work for everyone.
Our goals are:
1. Evaluate established seizure prediction methods using the unique NeuroVista dataset.
2. Develop new features to better discriminate patient-specific seizure-related time periods.
3. Use the diurnal cycle to improve seizure prediction algorithms.
4. Use iEEG statistics to better quantify long-term epilepsy changes relevant to seizure prediction.

This project will develop robust seizure prediction methods, which will have a resounding impact on the management of drug-refractory epilepsy. Our prediction methods will provide seizure warnings that will greatly reduce the anxiety and stress linked to the uncertainty of when a seizure will occur and greatly enhance the quality of life of patients.

Read more

View description


Analysis of time-varying and uncertain interconnected dynamical systems

Team: Michael Cantoni

The project aims to develop mathematical tools for the analysis of time-varying and uncertain interconnected dynamical systems.

Topics of particular interest include:

  • Closed-loop stability verification for distributed-parameter systems under sampled-data / networked feedback control; 
  • Input-output (i.e. operator theoretic) approaches to robust stability analysis for feedback interconnections of time-varying linear systems in a discrete-time setting;
  • Computational tools for the analysis of time-periodic systems, especially for the evaluation of gap metrics, which quantify the difference between open-loop systems from the perspective of closed-loop performance and;        
  • Scalable stability verification for networks of dynamical systems using tools such as integral-quadratic constraints.

Read more

View description


Anticipation of epileptic seizures using electrical probing of the cortex

Team: David Grayden, Dean Freestone, Anthony Burkitt, Alan Lai

Epilepsy is a chronic disease of the brain that affects 1-2% of people. Its defining characteristic is recurrent seizures, carrying a risk of injury, brain damage or death. Anti-epileptic drugs are the mainstay of treatment; despite this 1/3 of focal epilepsy patients have uncontrolled seizures. For these people, seizure onset is unpredictable, severely impairing quality of life.

Although seizure occurrence appears to be random, there is evidence that cortical excitability increases prior to attacks. Tracking cortical excitability is challenging. The field of seizure prediction has developed considerably over the last 30 years, yet this important problem is unsolved. We have developed an active strategy to monitor cortical excitability by measuring responses to low-intensity electrical stimulation. This approach is a paradigm shift from conventional passive techniques with the potential to lead to clinically relevant outcomes such as an implantable therapeutic device.

Data is obtained from St. Vincent’s hospitals. The standard preparation process for epilepsy related surgery involves electrode implantation directly on the brain to localise epileptic and functional tissue. These electrodes present a rare opportunity to stimulate and record from the human brain.

Our novel approach overcomes problems with the existing theory for seizure prediction. By actively determining input-output relationships, we have a direct measure of the brain’s excitability compared to passive observations. This provides the opportunity to track excitability.

Read more

View description


Bionic Vision Australia

Team: Anthony Burkitt, Stan Skafidas, David Grayden, Tatiania Kameneva, Bahman Tahayori, Craig Savage, Emily O'Brien, Shun (Leo) Bai, Hosing Chin, Colin Hales, Sam (Yuhua) He, Omid Kavahei, Vijay Muktamath, Nhan Tran, Jiawei (Jeff) Yang, Yuanyuan Yang

Bionic Vision Australia (BVA) is a national consortium of researchers working together to develop a bionic eye. The aim of this technology is to restore the sense of vision to people with vision impairment due to retinal degenerative conditions such as retinitis pigmentosa and age-related macular degeneration.

The Australian bionic eye project brings together a cross-disciplinary group of world-leading experts in the fields of:

  • Ophthalmology
  • Biomedical engineering
  • Electrical engineering and materials science
  • Neuroscience
  • Vision science
  • Psychophysics
  • Wireless integrated-circuit design
  • Surgical, preclinical and clinical practice.

In December 2009, the Australian Federal Government awarded a $42 million ARC grant to Bionic Vision Australia to develop bionic vision technology.

Read more

View description


Codes over rings

Team: Margreta Kuijper

This project focuses on algebraic codes over finite rings of the type Z_{p^r}. In recent years, novel linear algebraic tools were developed that overcome the difficulties of the presence of zero divisors in such rings.

Current research is focused on further application of these fundamental ideas. We consider the design of efficient algorithms for decoding of Reed-Solomon codes over rings; network coding over finite rings; shortest recurrence algorithms for sequences over rings; non-Hamming metric decoding over finite rings.

Read more

View description


RECRUITING Communication and computation over erroneous networks

Team: Girish Nair

Communication over networks of unidirectional channels is a well-studied problem in information theory, and is known to be difficult. The (ordinary) capacity regions for networks with multiple information sources are not generally known, and capacity-achieving codes are not guaranteed to be linear. In most of the literature on this topic, the objective is to reconstruct source signals with negligible error probability, or with bounds on expected distortion. Shannon's Information Theory is the main tool for finding capacity regions.

In this project, such networks will be studied when the objective is to obtain exactly zero errors, or when the distortion at each sink (i.e. destination) must satisfy worst-case bounds. The techniques will be based on nonstochastic information theory. [see Nair, IEEE Trans, Automatic Control, 2013; Nair, IEEE Conf. Decision and Control (CDC), Osaka, 2015]

An alternative goal may be for the sinks to evaluate a specified function of the source signals, such as their sum or median, to some specified tolerance. It is anticipated that if the network is a directed acyclic graph then structural routability conditions similar to [Nair, 2014] may be useful. However, if there are feedback links between network nodes, then a different angle is needed, based on “directed nonstochastic information”. [Nair, IEEE CDC 2012; Nair, IEEE CDC 2015]

Read more

View description


Computational neural modelling of bottom-up information and top-down attention in auditory perception

Team: David Grayden

The primary aim of this project is to advance our understanding of how the brain processes auditory information and how it can make sense of acoustic signals that are often mixed with other sounds in frequency and time, depending on the current behavioural or perceptual needs. In particular, we aim to investigate the processing strategies used in the auditory cortex to enhance the perception of relevant sounds in the presence of background noise and distractors. We will develop neuronal network models that will be used to elucidate the mechanisms by which attention and plasticity modify neuronal responses in a task-dependent fashion. The models will be constrained by electrophysiological and behavioural data recorded some of the world’s top experimentalists. This understanding is likely to be relevant for other sensory modalities.

A better understanding of the impact of top-down processes on perception through a computational approach will be a significant step forward in neuroscience. The outcomes of this research will yield new insights into information processing in the brain, which is of interest to neuroscience research in general and to those working on brain-inspired computation. In addition, the procedures used will offer valuable results concerning the development, optimisation, simulation and analysis of large-scale spiking neural network models.

Read more

View description


Computer aided diagnosis of melanoma

Team: Mohammad Aldeen

Malignant melanoma is the deadliest form of skin cancer. In Australia, melanoma is the most common cancer in people aged 15–44 years. It represents10% of all cancers and its per-capita incidence is four times higher than in Canada, the UK and the US, with more than 10,000 cases diagnosed and around 1250 deaths annually. The worldwide steady increase in incidence of melanoma in recent years, its high mortality rate and the massive respective medical cost has made its early diagnosis a continuing priority for public health. Early diagnosis of melanoma is particularly important for two reasons. First, the prognosis of melanoma patients highly depends on tumour thickness. If melanoma is detected at early stages, it is highly curable, with a 10-year survival rate between 90 and 97%.  Due to the enhancements in skin imaging technology and image processing techniques in the recent decades, there has been a significant increase in interest in the computer-aided diagnosis of melanoma. The aim is to remove subjectivity and uncertainty from the diagnostic process, provide a reliable second opinion to dermatologists and overcome the low reproducibility found in clinical diagnosis. Through digital representation of dermoscopy images and in-depth mathematical and statistical analysis of colour and intensity of the lesion, it is expected that the computer algorithm (software) recognises features that are not detectable by human eyes, and therefore improves the diagnostic accuracy of even well trained and experienced clinicians.

Vitamin D Digital Meter: This project, which is a joint venture with the Royal Melbourne Hospital, aims at designing and constructing a device that measures the amount of vitamin D a human body absorbs when exposed to the Sun. The device reads and records the total exposure time, time of day, sun intensity level, and UV profile. All this information is processed by a micro-processor to calculate the amount of vitamin D absorbed based on clinical data provided by the Department of Dermatology at the Royal Melbourne Hospital.

Psoriasis Index Valuator: Psoriasis is a disease that affects the human skin causing red scaly spots. The most widely used measure of severity used by dermatologists is a score called PASI (Psoriasis Area Severity Index), which is a subjective measure of colour, texture and depth of affected areas. This index is assessed through visual examination and thus relies heavily on the expertise of the physician. To remove subjectivity, signal and image processing techniques are being investigated to provide consistent and highly accurate alternative to visual examination by dermatologists.

3D Mapping of Human Body: This project involves the mapping of human body onto a 3-dimenstional template so that any lesion on the skin is accurately mapped and its statistics recorded for future reference by dermatologists. Some of these lesions may not be detectable by human eyes, and therefore this technology improves the diagnostic accuracy of even well trained and experienced clinicians.

Read more

View description


RECRUITING Continuous wave magnetic resonance

Team: Leigh Johnston, Peter Farrell, Iven Mareels, Stephen Moore

The project aims to augment the capabilities of Magnetic Resonance Imaging (MRI) systems using continuous
wave (CW) transmission and signal reception, in order to image objects using very low excitation power. Whether
articulated mathematically or not, it is the objective of any given MRI sequence to solve an inverse problem,
involving estimation of some subset of the hidden states and parameters of the system, given the observed data.
Exploiting the transient and steady-state CW magnetisation dynamics to solve inverse problems in Magnetic
Resonance spectroscopy and imaging will potentially lead to technological advances in control of the magnetic
resonance signal, and will open new avenues of research in the mature domain of MRI.

Read more

View description


RECRUITING Coordination of linear multi-agent systems over digital networks

Team: Girish Nair

In this project, a  distributed coordination problem will be studied, where multiple agents cooperate over an errorless, digital network to keep an underlying linear dynamical system in a desired target region of operation, despite bounded dynamical and measurement noise.

What makes this topic challenging is that the agents’ data sets may not be “well-aligned”; consequently, some agents may have to take actions that sacrifice performance, in order to indirectly signal information to other agents through the global dynamics.This leads to a dual effect that generally complicates the problem.

This problem will be studied from the point of view of a recent nonstochastic theory of information for worst-case problems with bounded noise [Nair, IEEE Trans, Automatic Control, 2013; Nair, IEEE Conf. Decision and Control, Osaka, 2015]. By studying the nonstochastic information flows in these multi-agent systems,the aim will be to derive  criteria for control/coordination, possibly by extending the  “structural routability” conditions of [Nair, 2014].

Read more

View description


Design of nonovershooting tracking controllers

Team: Robert Schmid, Denny Oetomo

This project involves developing new state-space techniques for the improvement of transient performance in tracking control. Tracking control problems are as old as control theory, and have applications in nearly all areas of engineering, including servo problems, airplans, robotics.

The aim of this project is to develop new mathematical algorithms for the improvement of the transient response in the tracking of step reference signals, in order to guarantee the avoidance of overshoot and undershoot, without compromising the rate of convergence. The unified design method developed within this project will be applicable to linear and non-linear, continuous and discrete, SISO and MIMO, square and non-square, minimum and nonminimum phase, and also strictly proper and biproper systems.The effectiveness of the proposed algorithms will be demonstrated by their application to controlling the movement of a
robotic arm.

Read more

View description


Dielectric metasurfaces

Team: Kenneth Crozier

It is an exciting time to be working in optics. Previous generations of researchers have been largely limited to the use of naturally-occurring materials such as glass, etc.  Recent advances in fabrication made by the integrated circuit industry have dramatically improved our ability to produce nanostructured materials. Recent years have also seen an explosion in computing power. These two developments have made it possible to design and fabricate nanostructured materials ("metamaterials") that have optical properties not present in natural materials. These have been used in the demonstration of several, previously unimaginable, phenomena, e.g. materials with negative refractive indices and "invisibility cloaks". We previously demonstrated a very thin metamaterial, termed a "metasurface", consisting of silicon nanostructures on a glass substrate (Nature Communications). This had the unusual property of deflecting infrared light into different directions, depending on its polarization. PhD projects are available on developing dielectric metasurfaces with new and technologically-useful properties. Projects will be primarily experimental, but also with theory/simulation components.

Read more

View description


Easing the squeeze: dynamic and distributed resource allocation with cognitive radio

Team: Tansu Alpcan, Subhrakanti Dey

The radio spectrum is a scarce and valuable natural resource which is being squeezed by the rapid growth in wireless communications. Cognitive radios make efficient use of radio spectrum by dynamically reusing frequencies. This requires cognitive radios to sense the local environment and to control the interference caused to existing users of the spectrum. In this project we will design novel dynamic and distributed resource allocation algorithms for cognitive radios in order to significantly improve their performance. We will do so using techniques from extreme value theory, game theory and mechanism design and large random
matrix theory.

Read more

View description


Energy systems modelling, analysis, and control

Team: Mohammad Aldeen

This project aims to develop models for all components of conventional power systems, as well as renewable energy systems, such as solar and wind power systems. The modelling of such systems includes the control mechanisms involved, such as exciters, governors, stabilisers and associated power electronics devices. The modelling is developed so that any faults in any components of  the systems are accounted for, that is the derived models are generic enough to accurately represent the dynamics of such systems with or without faults

Power systems are complex dynamic systems and in most part exhibit nonlinear behaviour. They are, by their operational and physical nature, prone to frequent faults, such as short circuit caused by lightening or objects (plants or animals) coming into contact with high voltage distribution lines. Also they operate under highly elaborate control and protection schemes, which comprise a very large numbers of sensors, actuators, circuit breakers, and relays. These devices, as well as the communication channels associated with them, often experience faults and malfunctions. Any fault or device failure has the undesirable effect of changing the dynamic behaviour of these systems, which may in some cases lead to instability or loss of generation with detrimental financial impact on suppliers and consumers alike.

The fault detection and identification project aims at using state estimation and observer theory to design fault diagnosis schemes for large-scale interconnected power systems. The schemes are required to detect and identify faults in any part of the power systems, such transmission and distribution lines, generators, exciters, governors and associated controls, in real time.  Linear and nonlinear fault detection schemes are considered. Both model-based and knowledge-based approaches are investigated.

This project contains a few sub-projects each dealing with a different control design problem as outlined below

Design of Power System Stabilisers:  This project involves the design of a controller that injects supplementary stabilising signal into the exciter to damp out intra-machine mechanical oscillations. Both conventional as well as modern designs are considered. The emphasis in this project is placed on stabilisation of multi-machine systems using locally available information only.

Voltage Stability: In the daily operation of large power systems, voltage stability is paramount. It is a measure of the ability to transfer reactive power from generation sources to loads during steady operating conditions. Under “abnormal” operating conditions, such as outages of generation or large loads, voltage stability may not be retained if the new equilibrium voltages post-outages are below acceptable level. This may lead to either partial or total collapse or blackout. The aim of this project is to use dynamical modelling to analyse the voltage dynamics of power systems undergoing large disturbances or load changes. An estimation of when the voltage will reach the level of collapse will then be determined using estimation and fault detection techniques.

Transient Stability: Transient stability is associated with large disturbances that cause the load angle of generators to grow in time to a level where synchronism, and thus stability of the entire system, is lost. This project aims at identifying such load angle excursions and to trigger control signals that would adjust governors’ actions and in extreme cases activate protection systems to initiate load shedding and tripping of circuit breakers.

Read more

View description


Fault detection in engineering systems

Team: Mohammad Aldeen, Rahul Sharma

This project aims to use state estimation and observer theory to design fault diagnosis schemes for engineering systems. The schemes are required to detect and identify faults in any part of such systems in real time.  Linear and nonlinear systems are considered using model-based and knowledge-based techniques. Large-scale interconnected systems (such as power systems) are of special interest in this project.

Read more

View description


Finite sample issues in system identification, change detection and filtering

Team: Erik Weyer

Models of dynamical systems are commonly used in many fields of science and engineering. In order to make proper use of a model, it is important to know the uncertainties associated with it. An evaluation of the model quality must necessarily be based on a finite number of data points from the true system. In this project we have developed methodologies: Leave-out Sign dominant Correlation Regions (LSCR) and Sign Perturbed Sums (SPS), which stand on a solid theoretical footing in a finite sample context and which deliver useful evaluations of the model uncertainties. In particular, it delivers a probabilistically guaranteed confidence set for the true system parameters for any finite number of data points under very weak assumptions about the noise processes affecting the system. The LSCR and SPS principles are very powerful and they also find applications in filtering and fault detection, and filtering and detection algorithms with guaranteed statistical properties for any finite number of data points that have been developed.

Current research is focused on further developments of the methods and their properties for different types of linear and nonlinear models and identification algorithms (e.g. least squares, prediction error methods, instrumental variables), efficient numerical implementation of the methods and applications of the principles in different areas.

Read more

View description


Gigabit wireless access using millimeter-wave over optical fiber systems

Team: Masud Bakaul, Thas Nirmalathas, Christina Lim, Stan Skafidas

Millimeter-wave radio-over-fiber (RoF) systems are widely considered as a disruptive technology for Gigabit/s wireless communications. In these systems, the benefits of optical fiber and mm-wave radio technologies are combined to provide an alternative approach for high-speed wireless access to customers. An optical fiber feeder network is used to interconnect a large number of remote antenna base stations (BSs) to the local Exchange (central office, CO), where most of the switching and signal processing equipment are installed. Usual distances between the CO and the BSs and the BSs and the customers are 5-50 km and 10 -1000 meters respectively. This project explores various system technologies and architectures in simplification of optically modulated millimeter-wave generations, transports and detections that enable Gigabit wireless access potentially at low-costs.

Read more

View description


Harnessing the capacity of optical fibres via few-mode transmission

Team: William Shieh, Abdullah Al Amin, An Tran, Trevor Anderson

This project aims to explore novel approaches to leapfrog the capacity limit of a conventional single-mode fibre using concept of few-mode transmission. The proposed research entails few-mode fibre design, few-mode channel modelling, few-mode transmission simulation and digital signal processing for few-mode transmitter and receiver. The successful execution of the project can potentially bring about breakthrough technologies that could find applications in a wide range of places from optical interconnect to long-haul back-bone networks.

Read more

View description


Internet of Things for Creating Smart Cities: Designing an urban information architecture

Team: Jayavardhana Gubbi Lakshminarasimha, Marimuthu Palaniswami, Jayavardhana Gubbi, Slaven Marusic

The project aims to create a Smart City capability through seamless urban environment monitoring via large-scale sensing, data analytics and information representation. Interconnection of sensing and actuating devices as an ‘internet of things’ addresses the ability to share information across platforms through a unified framework, developing a common operating picture for city management. The interpretation of events and visualisation of information for end-users will ensure sustainability and higher quality of life in the urban environment. Major outcomes will include energy-efficient sensing, network quality of service, cloud computing for sensor networks and high level analytics to detect and interpret events for decision making.

Read more

View description


Large scale multiple antennas for energy-efficient heterogeneous wireless networks

Team: Phee Lep Yeoh, Brian Krongold

This project investigates new network architectures for future wireless broadband inspired by recent advances in large scale multiple antenna technology and heterogeneous networks. The aim is to support flexible and scalable wireless services across diverse network regions with energy-efficient management of radio spectrum and interference. Targeted applications include smart energy metering, intelligent transport systems, mobile health monitoring and green data centres. Outcomes of the research will be new wireless protocols and algorithms drawing upon the foundations of random matrix theory, game theory, and large system analysis, which will offer fundamental insights into large scale multiple antennas for heterogeneous wireless networks.

Read more

View description


Modelling the human nervous system with human pluripotent stem cells.

Team: Mirella Dottori

The human nervous system is one of the most complex structures evolved to date. In order to understand how it functions, and dysfunctions in a diseased state, it is fundamental to decipher how it develops to generate various neuronal populations that form this elaborate network. Human stem cells provide a valuable source to study such processes. Our aim is to use human stem cells to study how early progenitor cell types that structure the nervous system are generated and how their neuronal derivatives form connectivity and functional synapses. The outcome of these studies is that we will establish a cellular model of human neurogenesis that can be utilized to study developmental disease processes. 

Read more

View description


Nanowire photodetectors

Team: Kenneth Crozier

Semiconductors such as silicon and germanium absorb photons with energies greater than their bandgaps. For many applications, however, it would be desirable to achieve control over the absorption spectrum (absorption vs photon energy).  We recently demonstrated this with silicon nanowires (Si NWs). Bulk silicon appears silver or grey, but we showed that vertical Si NWs take a range of colours, depending on their radii (Nano Letters 11, 1851 (2011)). We showed that arrays of Si NWs could be used as filters for colour and multispectral imaging (Applied Physics Letters 101, 193107 (2012) and Scientific Reports 3, 2460 (2013)). We incorporated p-i-n junctions into the NWs, and showed that they could be used as photodetectors for colour imaging (Nano Letters 14, 1804 (2014)). PhD projects are available on Si and Ge NW photodetectors for multispectral (visible to infared) imaging. These will be primarily experimental, but will also involve numerical modelling.

Read more

View description


Network Timing System

Team: Darryl Veitch, Julien Ridoux

This project aims to develop the next generation system for computer clock synchronisation for the Internet.  Our goal is to replace the incumbent, the `NTP' system, currently used by (almost) every computer on earth.

All computers incorporate a software clock, essential to software applications. An inexpensive and convenient way to synchronise such clocks is over a network, however the approach the Internet currently depends upon is unreliable, inflexible, and ignores major sources of error. This project will define and solve the key research problems underpinning an optimal and maintainable system for network timekeeping, and implement and test the outcomes under realistic conditions over the Internet. The result will be software clocks of high accuracy and reliability supporting applications like cloud computing, tele-medicine, smart grids, and network measurement, well positioned to become the next generation timekeeping system for the Internet.

This project builds on the existing testbed and RADclock synchronisation client developed under the SyncLab project.  Open Source software is available for download at http://www.synclab.org/radclock/.  Support for RADclock has been accepted into FreeBSD, the operating system of choice for servers.

Current directions:

  • Virtualised OSs     ( timing architecture for virtual machines )

               Virtualisation-friendly timing including live migration;  testing on Xen, VMware Player

  • Asymmetry measurement and mitigation   ( large error, largely ignored )

               Tightening bounds via spatial diversity; OS calibration; server recommendation service 

  • Server heath monitor   ( remote assessment of time-server quality )

               Exploiting known RADclock performance to detect server anomalies

  • Timing system health   ( evaluating systemic anomalies and performance )

               Mapping and evaluating public stratum-1 servers; assessing vulnerabilities

  • IEEE-1588 (PTP) support   ( exploiting 1588 masters for software clocks )

               Benchmarking 1588-capable RADclock against alternatives; contributions to standard

  • Unidirectional feed-forward algorithms   ( currently RADclock is bidirectional )

               Currently only feedback algorithms exist, unstable to noise (latency variability)

  • RADclock performance   ( formal analysis and performance enhancement )

               Statistical analysis of algorithms (for optimisation, calibration..);  LAN-specific enhancements

  • Adoption   ( building momentum towards replacing the NTP system )

              Accepted into FreeBSD 10.0; visibility with Linux stakeholders; adoption by CAIDA’s Ark monitors

Read more

View description


New frontiers in ultra-wideband electro-optic measurement technologies

Team: Christina Lim, Ampalavanapillai Nirmalathas, Alan Ka-Lun Lee

Emerging applications such as wireless communications, medical bionics and medical imaging are being developed to exploit higher frequencies of the electromagnetic spectrum, with non-invasive methods for precise measurement of electric fields becoming paramount for engineers and designers of such systems. With recent advances in photonics, it is possible to conceive novel systems with significantly improved sensitivities over the presently available methods. The primary aim of this project is to develop novel high resolution non-invasive electro-optic techniques capable of measuring relatively weak electric fields with an ultra-wideband of frequencies with greater spatial resolution and enhanced accuracy.

Read more

View description


RECRUITING Next generation brain-machine interface: Minimally-Invasive electrode array for robotic limb control

Team: David Grayden, Anthony Burkitt, Sam John

Loss of limb function occurs with a variety of conditions leading to significant disability. Stroke, spinal cord injury and traumatic limb amputation can all lead to limb failure. A majority of these patients retain a normal functioning motor cortex. This brain tissue holds capacity for complex motor control, but has no means of neural signal transmission. The current gold standard brain machine interface comprises an invasive electrode array, inserted directly into the motor cortex via craniotomy. A human clinical trial has demonstrated capacity to reach and grasp with imagined thought, using a robotic limb. However, implantation requires major brain surgery, and significant signal degradation occurs at six months due to various factors.

We hypothesise that a novel, minimally-invasive neural interface can be fabricated that is capable of recording and transmitting meaningful neural signal from the motor cortex, without chronic signal deterioration.

Aims:
1. Establish methods that allow access to motor cortex.
2. Fabricate a robust electrode array.
3. Preclinical Trials: Demonstrate biocompatibility and safety of device.
4. Preclinical Trials: Record and stimulate motor cortex.
5. Human Pilot Clinical Trial: Implant electrode array and record motor cortical activity.

This pioneering, world-first project collaborates across nine departments of the University of Melbourne including Medicine, Electrical Engineering, Materials Engineering, Anatomy, Neurology, Radiology, Neurosurgery and Florey Institute of Neuroscience. The significance of demonstration of proof of principle may represent a paradigm shift in neural interfacing, and a new therapy in a previously untreatable group of patients.

Read more

View description


Optical nanotweezers

Team: Kenneth Crozier

Optical tweezers use the forces exerted by focused laser beams to trap and manipulate particles. Conventional optical tweezers employ lenses to focus laser beams. Due to the diffraction limit, these can only focus light to spots no smaller than roughly half the wavelength. This sets a limit to the force that can be exerted by conventional optical tweezers on a particle of a given size, with a given refractive index, and a given laser power. This makes it challenging to trap very small particles. We have previously overcome this limitation via using nanostructures, rather than just lenses, to focus light. These have included gold nanostructures (e.g. Nature Communications 2, 469 (2011)), silicon photonic crystals (.e.g. Nano Letters 13, 559 (2013)) and silicon micro-ring resonators (e.g. Nano Letters 10, 2408 (2010)). PhD projects are available on developing new types of nanostructures for optical tweezers, and their integration with microfluidic chips. Projects will have both experimental and theory/simulation components. 

Read more

View description


RECRUITING Optimal design of large scale mechatronic systems

Team: Chris Manzie, Marcus Brazil Doreen Thomas Hyam Rubenstein

This project is an industry based PhD opportunity focusing on developing methodologies for the optimal design of mechatronic systems. This will involve the use of optimisation techniques in conjunction with computational mechanics software to develop improved structures that meet specified design requirements. To demonstrate the methodology, the student will work closely with the industry mentor, Australian Turntable Company in Bendigo, on application to next generation vehicle turntables. These designs are expected to have international deployment opportunities in sectors including mining and construction.

Read more

View description


Optimisation of signal processing and electrical stimulation algorithms for the abatement of epileptic seizures

Team: David Grayden, Dragan Nesic, Dean Freestone, Anthony Burkitt

This research will address the complex problem of the interactions between electrical stimuli and excitable neural tissue. Electrical stimulation is finding increasing acceptance as a novel means of providing therapeutic benefit across a wide range of neurological disorders. Despite this increase in application, the understanding of the ways in which electrical stimulation may convey therapeutic benefit is limited. Aspects of computational neuroscience will be employed in order to better characterise the effect that electrical stimuli have when applied to computational models of neural circuits. A greater understanding of these interactions may aid the future development of implant technologies across a broad range of diseases.

Read more

View description


Perceptual continuity

Team: Tatiana Kameneva, Tania Kameneva, David Grayden

We take it for granted that the visual world is stable, yet we move our eyes almost continuously. Moreover, we only have high-resolution vision in a small region at the centre of our visual field, yet we believe that we see the entire visual image in high resolution. How does the brain give us the impression of visual stability and uniform high spatial definition? Answering these questions has been one of the most fundamental tasks for vision scientists and psychologists for over a century.

This project conducts experiments both at its Carlton site and in its satellite laboratory at Monash University that investigate these fundamental issues. We work closely with our collaborators at Monash University, Professor Marcello Rosa and Dr Nicholas Price. We have established that visual processing undergoes very fundamental changes at the time of saccadic eye movements. Saccades are eye movements that shift our gaze from one location in the visual scene to another and occur at a rate of three times per second. We have shown that certain areas of the brain are attenuated during these eye movements, which assists in removing the blurred signals of a moving world that would otherwise occur during saccades, which can move the eyes at speeds of 500 degrees per second. Perhaps more exciting is the observation that immediately after saccades those same brain areas become hyperactive and respond more strongly to visual stimulation than normal. Our observations suggest that attenuating visual sensitivity during saccades and improving visual sensitivity after saccades are integral parts of maintaining visual stability in the face of continuous gaze changes. Therefore, we get the best of both worlds: we can change our gaze direction to maximize our capacity for high-resolution vision across the entire visual field, while not losing our sense of visual stability. The other exciting conclusion to be drawn from this work is that the motor system is influencing how the visual system takes in information: we are not passive observers – rather we actively search each visual scene for what we want and tune our visual systems to see what we need.

Read more

View description


Precision Timekeeping Infrastructure: bridging the hardware/software divide

Team: Julien Ridoux, Darryl Veitch

Accurate time is essential for critical services from telecommunications to banking, and increasingly, must be performed with software clocks within computers, using hardware clocks accessed over the Internet. This project with Symmetricom Inc. will bridge the hardware/software divide to deliver reliable and cost effective access to precise timing. 

The project builds on the existing testbed and RADclock capabilities with the broader SyncLab project.  Its focus is on working to brings software based timekeeping on a LAN (Local Area Network) down to the 1 microsecond level.

 

Read more

View description


RECRUITING Quantised causal inference based on directed information

Team: Girish Nair, Erik Weyer

In the field of causal inference, the aim is to determine the causality relationships between 2 or more observed time series, under suitable assumptions on the underlying generating model classes. Causality here is defined in terms of conditional probabilities, based on Granger’s ideas [e.g.  Amblard and Michel, Entropy, 2013].

All current techniques for causal inference are based on full-resolution measurements of the time series of interest. However, in many remote sensing applications only quantised, low-resolution measurements may be available.

This introduces several major theoretical challenges not addressed in the existing literature.Firstly, existing causality inference/system identification techniques cannot be applied directly. Secondly, optimal quantiser design becomes a critical issue. This project will explore these challenges, by using the Marko-Massey directed information and/or the nonstochastic directed information [Nair, IEEE CDC 2012; Nair, IEEE CDC 2015] as measures of causality.

Read more

View description


RECRUITING Robust filtering and estimation

Team: Girish Nair

A basic goal in stochastic filtering/estimation problems is to form an estimate Xest of a parameteror state X from noisy past data Y (time indices omitted). A common distortion criterion is the mean-square error (MSE) under a linearity constraint, which is often unsuitable in nonlinear or discrete-valued settings. An alternative approach in the literature is to maximise the Shannon information I[X; Xest ]. However, this can involve an infinite-dimensional optimisation over distributions.

The aim of this project is to explore estimators  that maximise instead the *nonstochastic*  information between X and Xest [Nair, IEEE Trans, Automatic Control, 2013; Nair, IEEE Conf. Decision and Control, Osaka, 2015]. This has the potential to yield filters with reduced computational cost for situations with bounded disturbances that are nonstochastic or have poorly known distributions.

A complementary approach is to finding an estimator that satisfies a nonstochastic “unrelatedness” principle, whereby the estimation error  is “unrelated” to the data Y in the sense of [Nair, IEEE Trans, Automatic Control, 2013; Nair, IEEE Conf. Decision and Control, Osaka, 2015]. Solving these problems is potentially much simpler than the corresponding probabilistic versions.

Read more

View description


Smart devices for ubiquitous health monitoring

Team: Jayavardhana Gubbi Lakshminarasimha, Marimuthu Palaniswami, Jayavardhana Gubbi

Provision of healthcare services remains a critical challenge across all levels of society, with the cost burden and constrained resources limiting the accessibility to healthcare. Advances in hardware development have made available efficient, low-cost, low-power miniature devices for use in remote sensing applications. This iconic project will drive the development of smart devices for low-cost monitoring, analysis and treatment of a number of medical conditions addressing continuous monitoring, aged care, and rehabilitation. The project consolidates a number of biomedical engineering initiatives into a single interdisciplinary project.

Read more

View description


The impact of the mass adoption of electric cars on the Australian electricity grid

Team: Iven Mareels, Doreen Thomas, Marcus Brazil, Kevin Prendergast, Tansu Alpcan, Julian de Hoog

This project will study the impact of the mass adoption of electric vehicles (EVs) on the electricity grid.  The success of the uptake of EVs promises significant greenhouse gas reduction but depends on making it convenient and affordable for motorists to move away from the use of fossil-fuelled vehicles.  The electricity grid infrastructure required to realise the full potential of EVs will be quantified, both from a traditional as well as a smart grid perspective.  The impact of this fleet on the power management of the distribution network and power quality will be analysed.

Specific research goals include:

  1. Identification of limitations in the current distribution network with respect to EV charging
  2. Development of an optimal charging policy for demand aggregators
  3. Evaluation of the benefits of the charging policy with respect to variable (green) power sources

Read more

View description


The role of information in game-theoretic decisions on distributed systems

Team: Tansu Alpcan, Girish Nair, Rob Evans

Game theory is an important instrument for analysis and design of resource allocation algorithms on distributed systems. In many real-world problems, information available to agents is incomplete in contrast to the perfect information assumption often made. This project will investigate game-theoretic decisions and quantify information analytically using concepts from information theory. A better understanding and quantitative modelling of information will lead to distributed systems that are optimal and robust with respect
to communication constraints. Project outcomes will be applicable to electrical power grids, renewable energy generation and storage, water irrigation networks, and communication systems.

Read more

View description


RECRUITING Topological feedback entropy for nonlinear control

Team: Girish Nair

The topological feedback entropy (TFE) of a fully observed, deterministic, continuous, nonlinear plant measures the lowest rate at which it must generate information in order to be able to keep state trajectories within a specified compact set [ Nair et al, IEEE Trans Automatic Control, 2004]. It is an adaptation of the classical concept of topological entropy from dynamical systems theory. In this project, the aim will be to extend TFE to handle partial observations, discontinuous dynamics, and/or bounded disturbances, and to then explore the use of TFE in analysing networked coordination problems with multiple nonlinear agents.

Some other relevant literature is [Colonius and Kawan, MCSS, 2011; Hagihara and Nair, 2013; Kawan and Delvenne, IEEE Trans. Control of Netw. Syst, 2015].

Read more

View description


Ultrahigh-speed optical transport for sustaining the internet growth

Team: William Shieh, Abdullah Al Amin, An Tran, Trevor Anderson

The project aims to investigate various transmission issues for the emerging Terabit optical transport networks through mixture of theoretical analysis, computer simulation and experimental demonstration. The scope of the research encompasses study of ultra high-speed signal nonlinear propagation in dispersive fibre optic channel, coherent optical single-carrier and multi-carrier (or OFDM) performance comparison and advanced optical network functionalities.

Read more

View description


Unified digital networking for wireless and optical access

Team: Christina Lim, Yizhuo Yang, Ampalavanapillai Nirmalathas

Wireless and wired access are traditionally developed on two separate backbone infrastructures with minimal interaction which leads to inefficient operation with high maintenance and upgrade cost in the long run. This project aims to provide cost-effective solutions to unify the backbone infrastructure for both services. An innovative approach based on digitisation of wireless channels and using high performance and bandwidth digital optical transceivers will be investigated which improves the system performance and overcomes impairments inherent to analog optical links. Key outcomes of the project will lay a foundation for future broadband wireless and wired network integration.

Read more

View description


Wireless sensor networks

Team: Doreen Thomas, Marcus Brazil, Charl Ras

Wireless sensor networks (WSNs) are an exciting new technology that lies at the intersection of communication networks and smart monitoring. It is easy to imagine the benefits of a system consisting of hundreds or even thousands of sensors distributed over a region or an object and able to measure any of a multitude of aspects of their environment. In general the sensors communicate their results to other sensors via multi-hop paths and also pass their data to central nodes for intelligent processing. The applications are countless: bushfire early detection, smart-home systems, battlefield surveillance, animal population tracking, and environmental pollution monitoring – to name just a few.

Although WSNs are already extensively utilised in the real word, their full potential remains largely untapped. In contrast to other types of communication networks, sensor network technology is severely limited by battery efficiency. This is because WSNs are required to be autonomous for as long as possible, as they are often deployed in remote, harsh, or even hostile environments.

The topology of the network (i.e., the overall pattern formed by the communication links between pairs of sensors) and the relative locations of the sensors play a large role in the efficiency of the network. This project will therefore focus on the optimisation of any aspect of the topological design phase of WSN deployment. A PhD student can choose to focus on aspects of any of the following design stages: the development of mathematical tools to model and understand optimal WSN deployments, the application of these tools to the construction of new algorithms, or the performance analysis of algorithms by mathematical means and/or by coding the algorithms and performing simulations.

Read more

View description