GPS, Geo-Positioning System is providing a variety of services so seamlessly that it has almost disappeared from our perception. It is not just about the "big things" like airplane navigation or military use, it is also about small apps that we may use often, like Everytrail, the navigation system in our car, the tagging of photos in many digital cameras...
Sometimes the GPS cannot be used, like when you are in a garage or in a tunnel: it requires a direct line of sight from the receiver to at least three orbiting satellite. This is why it may fail in urban areas with skyscrapers: they block the line of sight.
Now DARPA is looking at alternative ways to pinpoint a Position, assist Navigation and provide a precise Time (PNT) to overcome the weak points of the GPS (among this the fact that GPS signals may be ruined by interference like the ones created by solar flares or malicious one created by an enemy attack).
In particular DARPA has launched investigation in four areas that eventually may cause the retirement of the "old" GPS (and of the European Galileo as well..).
The first one is looking at the development of new algorithms and architecture to rapidly deploy and use sensors that are basically incredibly accurate inertial systems. We know that if you spin a wheel that wheel will resist any change in the direction of its axes keeping its movement. In reality, depending on various factors, after a while (think about the top you played with many years ago) the axe will change its direction and can no longer be used as a reference system. At DARPA with project ANS, Adaptable Navigation System, they are looking at spinning a cloud of atoms and then measuring any change in the spin to infer change in position (direction and acceleration). With project Micro-PNT they are looking at extreme miniaturisation of sensors to create micro-gyroscopes usable as inertial system.
A second area is looking at making extremely precise clocks portable. With project QuASAR they aim at reaching the same precision of optical atomic clocks (of the order of 1 second shift in a 5 billion years) in portable clocks that can be used as reference beacons.
The third area is about making synchronisation possibile, something that gets extremely complicated once you are dealing with the kind of precision provided by atomic clocks (you might remember the ruckus two years ago when some physicists said that they found a glitch in the Einstein relativity theory: there was no glitch in Einstein theory, it was, as it turned out, in the synchronisation of clocks measuring the departure and arrival of photons). With projects PULSE they aim at providing much more accurate synchronisation systems.
The fourth area is actually aiming at going beyond the GPS making PNT possibile in absence of a GPS signal. The project, STOIC, Spatial Temporal and Orientation Information in a Contested Environment, aims at providing a variety of systems to support extremely reliable PNT information among multiple users.
It is interesting to see that in a very concrete way DARPA is studying ways to leverage extremely sophisticated technology, like the ones involved in the first three areas AND to exploit a variety of systems that because of their variety can provide PNT information in a very robust way. You are no longer coding reliability into a single component, rather you expect reliability as an emerging property of a very dense set of independent entities.