If you’ve ever tried to use GPS indoors, you know that the signals aren’t easy to acquire in any sort of structure. Now imagine trying to get a satellite fix underwater. Researchers at MIT have a new technique, underwater backscatter localization or UBL, that promises to provide a low-power localization system tailored for the subsea environment.
Like other existing solutions, UBL uses sound waves, but it avoids some of the common problems with using sonic beacons in that environment. A typical system has a fixed beacon constrained by the availability of power or battery-operated beacons that require replacement or recharging. Since the beacon acts as a transponder — it receives a signal and then replies — it requires either constant power or time to wake up from the external stimulus and that time typically varies with the environment. That variable startup time interferes with computing the round-trip time of the signal, which is crucial for estimating position.
A UBL reader pings for nodes and, when a node replies, the pair change frequencies in a hopping scheme that allows measurement at many different frequencies. This allows an estimate of the true propagation time that is more accurate in the presence of reflections of the signal on the seabed, the water’s surface, and other elements of the environment. The speed of the communications between the node and the reader can also alter to tradeoff between multipath rejection and better ability to locate objects in motion.
Like traditional navigation or even GPS, knowing how much delay there is between you and a fixed position allows you to draw a circle around that position and you know you are somewhere on that circle. Find two fixed positions and you must be at one of the points where the two circles overlap. With three fixed positions that are far enough apart, you’ll have only one point where all three circles intersect. Finding fixed positions could use a star, a GPS satellite, or, in this case, an acoustic signal. The principle is the same.
We’ve looked at finding submarines before which is a similar problem. Another solution to the variable delay time is to not look at the transit time, but the difference between arrival times at two different points.