4.10.18 – Wi-Fi NOW
New research shows that discarded data from Wi-Fi systems can be used for indoor surveillance systems.
While X-ray vision seems like something that belongs only in the comic books, research shows that it is indeed possible to use Wi-Fi to see through walls.
Researchers in Denmark have discovered a way to track people in an area covered by a wireless network by using Wi-Fi’s “discarded data”. The method identifies changes in a physical environment – like when a burglar is wandering around the living room while you are on vacation.
The beauty of discarded data
As with standard radar systems, a Wi-Fi imaging system measures how waves hit and bounce off various objects, and then forms an image based on that data. But unlike radar systems, a Wi-Fi-based system doesn’t need specialised hardware. It uses Wi-Fi, which is already available almost anywhere.
“Today’s standard home security systems are based on Passive Infrared Sensors (PIR), which use heat to detect changes in a room,” says John Aasted Sørensen, an associate professor at The Technical University of Denmark (DTU) “These sensors come at a price, and most of them are communicating through a Wi-Fi server anyways.”
The secret to Wi-Fi-based surveillance systems is the use of so-called “discarded data” from Wi-Fi. This is the data that a Wi-Fi system uses to detect how radio waves change or ripple around a room. The data can identify objects that might get in the way of a signal – like your sofa or TV. Thanks to this new research, it can also detect unwanted visitors.
Usually, discarded data is deleted automatically. But Professor Sørensen says it’s valuable.
With this data, a visual ‘heat map’ can be created and sent to homeowners’ phones when motion is detected – and all of this intelligence will run on your existing Wi-Fi system. That means no extra sensors, no alarm units, and no installation.
“It’s about simply removing the PIR and using Wi-Fi as the sensing system as well as a unit communicating with the Wi-Fi. It’s cheaper and simpler than traditional surveillance systems,” he says.
A fan, a person, and a pet
A few Wi-Fi based surveillance products are already on the market today. Aura, made by Canada-based Cognitive Systems, is set to start shipping out this month. The product consists of a small white base station and companion units that are placed at opposite ends of a home.
It uses a cognitive radio chip that can pick up a vast range of signals, from 680 MHz to 4 GHz. Origin Wireless, in the US state of Maryland, is also using the same principles to develop indoor positioning systems via Wi-Fi and LTE
Still, the mass-market deployment of Wi-Fi surveillance systems is some ways off. Cognitive Systems’ website says that Aura is “learning to tell the difference between a fan, a person, and a pet.” Accuracy, Sørensen says, will be a challenge. Due to their complexity and dynamic nature, the recognition of Wi-Fi-generated images seems an obvious use case for AI and machine learning algorithms.
“There is still a lot of work to do within product development,” he says. “Decisions will need to be made for which thresholds to set to prevent false alarms from a dog or fan.”