(Note: This is a fictionalized account of how a journalist may have covered the paper, “Feasibility and Limits of Wi-Fi Imaging” by Donny Huang, Rajalakshmi Nandakumar and Shyamnath Gollakota. None of the quotations below are true and must be interpreted as fiction only.)
University of Washington
A team at the University of Washington has developed Wision, a wi-fi based imaging approach that allows the researchers to see through walls like Superman and detect the presence of objects such as a couch or a laptop.
“Imaging systems based on radar technology have been around for a few decades,” says the lead author of the study. “However, all of these systems require specialized hardware that generate radar signals with a particular structure. While these signals are great from an imaging perspective, they cannot be used in wireless communication protocols, like wi-fi, because of their large frequency requirements. We have tried to break this inherent mismatch between imaging and wireless communication by developing Wision, a system that uses common narrowband wi-fi signals for imaging. To our knowledge, this is the first such prototype ever built.”
Wision performs imaging on multi-antenna wi-fi access points by analyzing transmitted wi-fi signals that bounce off the environment before being received. Unlike most radar imaging systems, these wi-fi transmissions need not have originated at the access points where the imaging is performed. “Unlike active imaging systems that require the antennas being used for imaging to both send and receive wireless signals, our system works as long as some wi-fi transmissions are occuring in the environment,” adds the lead author. “However, due to the inherent randomness in the locations of access points and the nature of wave propagation, the received signals consist of a combination of those signals reflected off all possible objects in the environment as well as direct signals that did not undergo any reflections in the path from the transmitter to the receiver. As these direct path signals typically have the largest power among all the received signals, Wision first removes them by nulling the largest received responses.”
To form images of the objects that may have reflected the remaining received signals, the system takes insight from optical cameras.
“Optical imaging systems use lenses to separate all incoming light by the directions from which they originate. Similarly, we use mathematical tools from Fourier analysis and the availability of multiple antennas to separate every received signal by the angles at which they arrive at the receive antenna array. Computing the intensity of these signals at every such angle gives us an image of the environment.”
To also allow the system to differentiate between objects located in the same direction, the researchers change the signal transmissions to selectively focus on each such object at once via a process known as beamforming. This allows the system to exactly localize objects in the formed image.
The team has tested the system in a variety of scenarios with different objects, such as laptops and leather couches. “The main limitation of using wi-fi for imaging is the poor resolution of the formed images. However, this is expected because of the small frequency bandwidths used in wi-fi, and is not a defect of our system,” says the senior principal investigator. “In addition, due to the larger wavelengths of wi-fi radiation compared to light and GHz radar signals, objects smaller than 10 cm in size cannot be imaged very well by our system. Thus, Wision cannot be used for applications such as gesture sensing. However, Wision shows the feasibility of using wi-fi in applications such as non-destructive evaluation and concealed weapon detection.”
“This is an exciting time for wireless communication,” says the lead author. “We hope to keep enabling newer capabilities of the wireless spectrum besides just communication, and foresee imaging to be an integral service provided by wireless networks of the future.”
###
