I've been following the news about the recently Iranian hijacked US drone with some fascination. I'm a bit surprised by the lack of coverage -- or maybe not. Perhaps there are sensitivities that haven't been revealed. So maybe keeping it all under wraps is extremely important, even though the information flow is a bit leaky. Many parallels have been drawn between this incident and the downing of a U2 plane in 1962.
Of particular personal interest to me, is the navigation (or guidance) system used to guide and direct the drone. It appears that the drone relied on GPS (Global Positioning System) for navigation. Of course, GPS is the same technology used for your car navigation system and all those cool location based services you use on your IPhone. Reports indicate the drone was compromised through an attack vector on the GPS navigation system. (Side note: I've watched with equal fascination how easy it is to compromise your automobile through really unsophisticated attacks.) From the Christian Science Monitor Report, Iranian specialists reconfigured the drone's GPS coordinates to make it land in Iran at what the drone thought was its actual home base in Afghanistan.
Certainly, if this is true, the security system that allowed access to the navigation system was compromised. So, there is some fault in the design/implementation of that system. The obvious fault lies here.
However, stepping back, maybe there's a more strategic design flaw?
Serious design considerations for how to build "modern" navigation/guidance system were contemplated probably as early as Sputnik -- the Beep Heard Round the World in 1957. And, as guidance systems for Minuteman, SRAM, the B1, and all commercial airline systems were designed and deployed, an approach that did not rely on GPS was developed. The design goal was to build a guidance system that was completely self contained and thus not reliant on any external systems. In particular, any system relying on GPS was rejected. The fear was that external systems, such as GPS, could be compromised, especially through EMI -- Electromagnetic Interference (resulting from enemy jamming or emissions from a nuclear blast). Such compromises would render the navigation systems useless.
So, what was developed instead of GPS for navigation?
Basically, math and a spinning top -- a gyroscope. Using a Kalman filter, a gyroscope (to make measurements in the physical world), and a pretty unsophisticated computer (that had less than few hundred bytes of memory) inertial guidance systems were built. Missiles and airplanes were completely autonomous and self contained from external systems, hence insusceptible to compromises to GPS failures or communication attacks on such systems.
[I don't quite remember the math involved. It had something to do with an optimization allowed by the Kalman filter so that instead of doing an NxN matrix inversion in O(n^3), the problem was reduced to inverting N*N 1X1 matrices in O(n^2).]
So, the take away for me today is that sometimes you need to take a step back from the problem you are solving. Look at the big picture. Maybe we're overly reliant on technology. I romantically believe that math is the answer to many problems.
Google+
