DME, from Wikipedia...
"An airplane’s DME interrogator uses frequencies from 1025 to 1150 MHz. DME transponders transmit on a channel in the 962 to 1150 MHz range and receive on a corresponding channel between 962 to 1213 MHz. The band is divided into 126 channels for interrogation and 126 channels for reply. The interrogation and reply frequencies always differ by 63 MHz. The spacing of all channels is 1 MHz with a signal spectrum width of 100 kHz."
The DME channels are "paired" with VHF counterparts so when you switch your Nav radio, the DME follows along.
There's usually a cable between the two for "remote" setting of the DME. My DME also has a switch to manually set the DME frequency but you set the VHF Navaid frequency you want the DME to base itself off of, and it internally selects the correct DME frequency, since DME frequencies aren't published on charts.
We also have a switch that selects whether the Nav 1 receiver or Nav 2 receiver remote is active, so you can flip it to have the DME reading off of either tuned Nav source.
Transponder receives primary radar sweep at 1030 MHz and transmits back at 1090 MHZ. Transponder is right in the middle of the DME band. Works out good for swapping antennas.
http://en.wikipedia.org/wiki/Secondary_surveillance_radar
I was Googling for a channel chart to show the gap in the DME channel assignments around the radar transponder, but can't find it right now.
If you think about it, DME is just "backwards" secondary surveillance radar without any azimuth information. The DME ground station is just a transponder. Signal from the airborne DME is received on one frequency and bent-piped back out on another. Time measured and distance known.
Add a little math to the DME receiver and it could provide a ground track. Add more ground DME stations and a frequency hopping DME systems, and...using this knowledge, it's possible (easily) with today's high speed CPUs and Nav data information storage that a navigation device could be built that could range to multiple DME stations and determine the aircraft's location as long as more than one were in range.
It'd need two stations for finding two possible locations on a map (the distance curves cross at two points) and a third to remove the ambiguity and determine which of those two crossing points the aircraft is at. Cheap and easy. How's that for cool? (And the exact same principal as GPS, but GPS uses a clocked signal for range instead of the aircraft measuring round-trip time to the birds.)
Nifty huh? All sorts of neat Nav stuff possible with all those ground based systems that folks haven't built. DME-RNAV. Could be done. You'd probably want the box to do auto-Ident. Indenting multiple DME stations would be a bugger.