Off-air DTV viewers using indoor antennas know that moving around in the room can cause reception problems. Newer receivers work better than old ones in this regard, but the effect on the signal is the same—a human body moving about in a room changes the multipath environment and the signal at the indoor antenna.
Now imagine a receiver aimed at the room looking at the multipath caused by a moving human. It could indicate that someone is moving, but more analysis would be needed to determine where the person is at any given point in time.
Joey Wilson and Neal Patwari in the Sensing and Processing Across Networks (SPAN) Lab at the University of Utah in Salt Lake City have developed techniques that not only identify someone moving in a room using multipath, but can also track their path.
Wilson and Patwari's method takes advantage of the motion-induced variance of received signal strength measurements. Instead of broadcast stations, the researchers used a Zigbee wireless network with 30 nodes operating at 2.4 GHz. The researchers' paper, Through-wall Motion Tracking Using Variance-Based Radio Tomography Networks [PDF], includes plots of motion in a room showing the difference between mean-based radio tomography and variance-based radio tomography. It also describes the formulas used to create the images. Kalman filter tracking is used to track and coordinate motion estimates.
It seems to me that ATSC DTV signals, if they come from different transmitter sites, might be able to provide some indication of what's going on in the environment. When I was working at KSCI in Los Angeles more than 25 years ago, I received a call from a company that wanted information on our channel 18 facility. They were working on a system that used TV stations to illuminate objects (such as airplanes or missiles) that they could then track without the need for a radar transmitter. Today, with the pseudo-random noise sequence in the ATSC signal, such a system should be able to operate with much greater accuracy, provided signals from transmitters at different sites was available.
Although Through-wall Motion Tracking Using Variance-Based Radio Tomography Networks uses complex mathematics to explain how the SPAN system functions, there is sufficient information outside the formulas to get an idea of how the system works and its potential real-world applications. The paper does not discuss the use of signal sources other than the Zigbee network with many transmitters arranged around the room to be monitored.
