U of T investigating possibilities of self-powered large scale networks

Imagine a network of wireless sensors scattered across thousands of hectares of forest watching for signs of dry conditions that could set off fires.

Or spread across the 49th parallel watching for people trying to cross our border with the U.S. Or in a warehouse sniffing for biohazards.

A project to look into the feasibility of building such large-scale self-powered wireless networks has been given a boost with a $3 million grant announced Tuesday from the Ontario government to a University of Toronto team.

Headed by Dr. Dimitrios Hatzinakos of the university’s department of electrical and computer engineering, the team is partnering with Motorola in a five-year quest to create sensor hardware, software and a power system in network a that can run by itself yet span great distances.

The five-year project has an estimated cost of just over $9 million, although much of that will be contributed by Motorola and the university in services.

Hatzinakos said the system has many challenges. It needs a wireless system that’s scalable enough to cover a large area yet easily software upgradable; an efficient power system; and, if sensitive personal or corporate information is gathered, a way to ensure security of any data transmitted.

It’s a project that excites Dan Gamota director of printed electronics, physical and digital realization research at Motorla Labs, who is the company’s principal investigator on the project.

“This has elements that surpass efforts at other universities,” he said Wednesday. “What I liked about this program is it’s a system level architecture that’s being developed, and the fact is you have four technolgies [nanosensors, wireless, energy saving and data mining] that really are at the leading edge.”

“The prime objective is energy,” Hatzinakos said in an interview in an accent that betrays his birth in Greece, “which has other implications as well, because what we learn from this technology can be applied to other areas.” Batteries might be first thought of as a power source, but he pointed out if they need to be changed that could cut a node out of the network. So a logical alternative is solar power, as long as a storage system can be devised to conserve energy on overcast days.

But what if the sensor is in a building? Some sort of thermal source of energy could be used, or what Hatzinakos calls “scavenging of energy” – a solar-powered sensor near a window might transfer energy via infra-red technology to a sensor that isn’t near a light.

“After all,” he smiles, “sunlight is the wireless transfer of energy.”

A number of optical and wireless technologies already exist that the team can take advantage of, Hatzinakos said, while Motorola has its own ad hoc network technology that could be modified.

The challenge, he said, will be to integrate what’s known at what will be discovered by the team into something powerful and sustainable.

One scenario the U of T team imagines is a system using what are called quantum dot (QD) devices watching for the scent of specific chemicals and biohazards. They would be sensed by a wireless reader, which is also linked to other non-QD based sensors such as accelerometers, chemFETs, photovoltaic and pH/colorimetry devices. All will be linked by a low-power wireless mesh network, forwarding data to a central co-ordinator which then transmits it to managers using standard cellular technology.

Such systems could be used on farms to watch nitrogen levels in the soil, or by municipal electrical utilities looking to balance electrical loads to homes and businesses.

“The approaches the professors at the U of Toronto are undertaking are novel and really haven’t been to my knowledge been commercialized,” said Motorola’s Gamota. “So it’s really an exciting program from the standpoint of being leading edge.”

Motorola’s contribution will largely be advising the Toronto team on the communications industry, he said. For its part, the company hopes to learn about advanced nano, wireless energy-scavenging and data mining techologies the university it working on. It’s also an opportunity for Motorola to evaluate the students working on the project it might like to hire after graduation.

Hatzinakos has been working oncommunications problems for some time. After receiving an undergraduate degree in electrical engineering from Aristotle University in Greece in 1983, he won a scholarship to study satellite communications at the University of Ottawa, where he earned a master’s degree.

He moved to Northeastern University in Boston, where he got his doctorate, to work on signal processing. In 1990 he was hired by U of T as an assistant professor, where for the he continued his work on signal processing. He is also the chair of management committee of the newly established Identity, Privacy and Security Initiative (IPSI).

Lately he’s been working on ad hoc networks and solving security problems in multimedia applications. Since November 2004, he ha held the Bell Canada Chair in Mutimedia at the university.

Other U of T faculty members on Hatzinakos’ team include Konstantinos Plataniotis, who is on the steering committee of an IEEE’s mobile computing group; Keryn Lian, who holds patents in areas of electrochemical energy storage materials; Ted Sargent, a specialist in quantum dot devices; Renee Miller, a data management specialist; David Lie, an expert on computing infrastructure security; and Shana Kelley, whose work has focused on biological and chemical detection.

Related Download
The Fast Path to Software-Defined Networks Sponsor: F5 Networks
The Fast Path to Software-Defined Networks
Download this white paper to learn how new partnerships are pioneering ways to ensure that they can transfer knowledge to enterprise IT staff.
Register Now