RF Sensing with Robots, X-Ray Vision with WiFi, Through-Wall Imaging

RF Sensing and Through-Wall Imaging with Unmanned Vehicles and WiFi

Chitra Karanam, Belal Korany, and Yasamin Mostofi, UCSB

Video (from 2017) on first demonstration of 3D Imaging through walls with WiFi and drones [paper][project page] Video (from 2014) on two ground vehicles imaging through walls in 2D with only WiFi [project page][most recent paper] Two copters imaging through walls in 3D with only WiFi [paper][project page]

In the News (June 2017): BBC Interview, Engadget, TechCrunch, Mashable, TechRadar, Sputnik News, New York Post, PC Mag, Digital Trends, IFLScience, World News, Science Daily, Tecmundo, The Register, Phys.org, New Atlas, UCSB press release, IEEE Spectrum, ACM News, Yahoo News, MSN News, and other outlets

In the News (Aug. 2014) : BBC Interview, Engadget, Gizmag, Daily Mail, Gizmodo, IDG (PC World, IT World, Computer World), International Business Times (Yahoo News), Headline and Global News, I-Programmer, The Verge, Ubergizmo, Outer Places, UCSB press release, SD Times, Investors Business Daily, and other outlets

Research Summary

WiFi signals are everywhere these days. Unmanned aerial vehicles are expected to become a part of our near-future society. In this research, we are interested in the possibilities created at this intersection of robotics and communications. For instance, imagine unmanned vehicles arriving behind thick concrete walls. They have no prior knowledge of the area behind these walls. But they are able to see every square inch of the invisible area through the walls, fully discovering what is on the other side with a high accuracy. The objects on the other side do not even have to move to be detected. Now, imagine robots doing all these with only WiFi RSSI signals and no other sensors.

Our lab has been working on this problem of imaging through walls with WiFi signals since 2008. Here are some highlights of our work in this area. In IEEE ACC 2009, we proposed our initial approach for imaging with WiFi received power measurements (RSSI) and unmanned vehicles, and showed the first experimental demonstration of imaging with WiFi in IEEE Milcom 2010. The approach is based on devising robotic paths (TX/RX antenna positioning) that are most informative for robotic imaging, approximated wave modeling, and sparse signal processing. It is noteworthy that our approach only uses WiFi RSSI measurements and further does not rely on making prior measurements in the area of interest. In IEEE TMC 2012 (DOI: 10.1109/TMC.2012.32 for 2012 online publication) we then showed the first demonstration of imaging through walls with WiFi and unmanned vehicles. The 2012 Ph.D. Thesis of A. Gonzalez-Ruiz from our lab also contains several key findings and experimental results for see-through imaging. In IEEE Sensors Journal 2013, we then showed the tradeoffs between using different robotic paths and imaging quality. The IEEE Sensors Journal 2014 further shows how WiFi and laser scanners can be integrated to image more extensive areas through walls, using binary compressive sensing. The 2015 IEEE TVT paper has an extended method that can image more complex areas, based on using Rytov wave models. For a summary of robotic through-wall imaging, see our 2017 IEEE Antenna and Propagation Magazine paper. Our project page for 2D imaging through walls also has more details on the methodologies and results.

More recently, we are excited about our new approach that has enabled the first demonstration of 3D imaging through walls with only WiFi RSSI and drones in IPSN 2017 (See the video and the project page). There are four tightly-integrated key components to our proposed approach to enable 3D through-wall imaging. First, we have proposed robotic paths that can capture the spatial variations in all the three dimensions as much as possible while maintaining the efficiency of the operation. Second, we have modeled the 3D unknown area of interest as a Markov Random Field and utilized Loopy Belief Propagation to update the imaging decision of each voxel. In order to approximate the interaction of the transmitted wave with the area of interest, we have used the WKB linear wave model. Finally, we have also taken advantage of the compressibility of the information content to image the area with a very small number of WiFi measurements (less than 4%), using sparse signal processing. See the video and the project page for more details.

Back to top


Back to top


Dept. of ECE and College of Engineering at UCSB for facilitating the experiments.