New innovations for the "massive access" required for the Internet of Things are being developed right here in Australia.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
At the University of Newcastle, in NSW's Hunter Valley, Professor Sarah Johnson is leading numerous research projects, devising ways to cope with exponential growth in demand for data.
She and her team are designing digital technologies "that will sit out of sight, inside the smart devices".
"These algorithms define how the smart devices will communicate with a base station to share their sensor data with the wider world via the internet," she said.
The need to improve network capacity is an ever-expanding purpose, as the Internet of Things is a network of connected devices that will revolutionise technology in the near future.
Billions of ordinary objects will be connected to the web, communicating with each other through a more advanced information highway.
This revolution will connect many things like appliances, home security systems and wearable health monitors. Other connections to this grid will be things like biometric cybersecurity scanners, smart glasses, bridges, dams, farm, factory and safety equipment and ingestible electronic devices to monitor disease.
"The Internet of Things is already here, but it has a long way to go," Professor Johnson said.
"We have barely scratched the surface of how many smart devices and sensors we will be trying to connect to the internet in future years."
IN OTHER NEWS:
Many of these devices will need connection through mobile phone networks.
"Handling all these new devices calls for major changes to data-transport infrastructure and the ways in which these devices connect and communicate," she said.
Professor Johnson, of the School of Electrical Engineering and Computing, devises ways to deal with potential bottlenecks.
Newcastle-based research is being done with academics at the University of Sydney and in the UK and Germany.
They're designing new communication strategies better suited to Internet of Things traffic.
In particular, this involves enabling multiple devices to transmit on the same frequency at the same time, without a base station directing traffic.
"This is a really interesting and challenging problem. If we can come up with a novel solution, it has the potential to find use on a very large scale," she said.
Digital cities
The Internet of Things [known by the acronym IoT] is like any new technology.
"Some people and some industries are first movers when it comes to new technology," Professor Johnson said.
"Already, on their smartphone, some people answer their doorbell, monitor their window sensors and set their thermostat from anywhere in the world via IoT devices.
"Meanwhile some people don't have a smartphone yet."
It's the same for commercial applications.
"In the Hunter, we have some innovative IoT deployments already," she said.
"The Newcastle City Council smart cities trial is based on IoT technology. Things like bin sensors, street lighting and parking sensors all communicate via IoT.
"Hunter Water is trialling IoT-linked sensors to detect water leaks on their mains infrastructure."
She said it would be interesting in the future to see how the data collected from countless devices is used.
"How can that be even more useful to us? For example, Hunter Water intends to see if their sensor data can be used to predict future leaks."
Machine Learning
One of Professor Johnson's research projects is titled, "Novel Techniques for Uncoordinated Massive Access in the Internet of Things".
Due to start in July, the project will explore ways to "re-design the receivers for the IoT signals".
This is a collaboration with Associate Professor Stephan Chalup and Associate Professor Kaushik Mahata, both of the University of Newcastle.
They are experts in machine learning and signal processing.
"The receivers will be trained using machine learning - artificial intelligence technology - to identify transmitting devices, recover their data and resolve any collisions," Professor Johnson said.
Associate Professor Stephan Chalup said "artificial intelligence techniques such as deep neural networks will be central enablers for the IoT".
"In recent years, there were surprising successes of these techniques in computer vision and board games such as Chess and Go," Dr Chalup said.
Now these methods are being applied in other domains.
"In our research at the university, we learned that by developing some relevant mathematics and fine-tuning the deep-learning algorithms in computer experiments, they often can be made simpler and faster which makes them ideal for the IoT."
He and his colleagues were "very excited about these developments and the opportunities that come with this new form of AI".
TV Character
We told Professor Johnson that she could be a character in the hit TV series, Silicon Valley, which involves the invention of a revolutionary data-compression algorithm.
"I love working in tech. It's a constantly changing and challenging field. If a particular problem gets old, there is always a new one to solve. I also get a kick out of seeing technologies I helped create working in real-world devices," she said.
In recent years, she has worked with an Australian company called Quintessence Labs on "quantum key distribution technology".
This involves the use of quantum physics for encryption and secure communication. Professor Johnson works on "the digital information processing part" of this. "To see products out there running algorithms I helped design is super cool."
She also designs new technology to "improve people's health".