If you are the BTS, hopping is optional in GSM. And you should be able to pull enough BW, even over USB2, to hop in software.

As for the cost, the BOM for the USRP is only about $250 if anyone wants to make their own. It is an open source design, after all.

We are mostly rich-world people with pretty toys. I count myself in that group, but when I pull my head out of my backside and get a little perspective I see this: GSM will be with us always. Even unto the end of the Earth.

(1) As GSM systems get decommissioned in the US and Europe they show up in places like South America, Africa and the high steppes of central Asia. Used 2.5G networks will still be a big upgrade for a lot of people.

(2) There are at least 3 billion people out there who are poor and live in areas with population densities too low to allow WCDMA networks to be profitable. They don't need mobile YouTube. They need a cheap, reliable telephone with modest power requirements. GSM still does that better than any other air interface technology out there.

(3) The point of UMTS was to allow carriers to upgrade their air interfaces while keeping most of their core networks intact. The problem is that the typical site backhaul is a DS1, so if a lot of subscribers actually try to *use* that air interface bandwidth, the core network is going to choke. AT&T learned that the hard way when the iPhone started to catch on.

(4) Don't even mention wi-max for mobile telephony outside of major urban centers. You just make yourself look silly.

In summary: Just because GSM isn't *your* future, doesn't mean it's not *a* future.