O2 and Vodafone are in talks about sharing network sites and infrastructure in the UK, the Financial Times reports. This would be interesting if the companies hadn't already done this for decades. According to the FT a deal between the two companies would save costs and, if spectrum sharing is included, could solve the 900MHz …
I'd settle for T-mobile and 3 just having 900MHz for rural coverage.
While things work fine in _most_ urban environments their coverage is abysmal near buildings in the countryside - to the point where i have a Vodafone 900Mhz prepay phone just to allow for those spots.
The original Nordic analog system (which became 800MHz AMPS) ran at 450Mhz for similar reasons. I'm willing to bet that once the TV bands are freed up these lower frequency blocks will come on the market for mobile coverage.
AMPS & 3G
(Actually, AMPS was not based on NMT, which is the Nordic system that ran at 450mhz... I think they were developed independently.)
In most areas, I don't know if the 450mhz band is used for *anything*... I've read in eastern Europe providers have began running CDMA and EVDO in this band. (In most cases the 450mhz spectrum available is not large enough to run 1 channel of WCDMA, this needs a 5mhz up and 5mhz down block, while CDMA and EVDO need 1.25mhz up and 1.25mhz down.)
Auctioning off TV band should certainly help -- here in the US they have auctioned off 700mhz spectrum. It's not being used yet but the plan among most owners is to put LTE in this band. (LTE is the successor to WCDMA, and Verizon's chosen upgrade path from CDMA+EVDO too.. )
Getting off on a tangent, when LTE is rolled out it means finally here in the states we'll be back to the AMPS situation where an LTE phone should pick up service wherever avaiable. As opposed to now, where there's far more CDMA coverage than GSM in the US, but some few areas have GSM only and no CDMA. (So if you have the wrong technology phone, there's service but your phone can't use it.)
Note, the correct solution was NOT to just go GSM as some may claim -- UMTS solves these problems, but "regular" GSM was unsuitable for rural areas here due to the 35km cell site radius limit (yes, there are areas out west where that drop off is a problem), and it was not as suitable for some cities due to it's much higher spectrum use per user compared to CDMA.. The GSM providers spent big in cities just to avoid network collapse (more spectrum, more cell sites, etc.) and have not rolled out much UMTS (maybe 40% coverage, if that), while the CDMA providers (Verizon and Alltel at least) have nearly 100% of their networks overlaid with EVDO coverage (3.1mbit/sec peak).
It doesn't appear LTE is very closely related to GSM *or* CDMA, but has provisions to be a reasonably smooth upgrade from both, which should be VERY nice. Ironically, considering some consider LTE a GSM tech, Verizon (the big CDMA provider here) is planning a much more agressive LTE upgrade than AT&T (the big GSM provider) -- Verizon's nearly done with EVDO upgrades, and has their current networks in good working order, while AT&T's spending enough getting their current GSM and UMTS network in working order that they don't have as much to spend on LTE upgrades.