The problem is a) and b) are often difficult to have simultaneously, as you will naturally have more people contending the limited set of available bandwidth slices when you improve signal strength/reach of a single mast. I suppose the best technical solution would be to have many lower powered nodes so that each mast has a smaller number of people contending for bandwidth. Would be rather expensive though, I suspect.
That's essentially happening today with 2100mhz 3G - on the networks with good capacity (and O2 is not one) and not overloaded in an area, the 3G/HSPA technology works extremely well.
However LTE does provide a massive improvement as it forgets voice traffic and turns everything into a data packet. Unlike 2G and 3G which handle voice separately to data today.
Ofcom seem to be waiting for the final Freeview/DigitalTerrestrialTV switchover next year to release the frequencies for LTE to all the networks. These should be low enough frequency to have quite a large coverage area from a single mast. Similar to todays 2G footprint from Voda/O2. And hopefully with sufficient in-building penetration.
Everything Everywhere has 1800mhz, and put a lot of money into getting good in building penetration by having double the sites to Voda or O2, but that still hasn't helped their 3G coverage.
3 has amazingly good 3G performance; but is pretty useless for a mobile phone due to the frequency anything on the edges of coverage ends up a blackspot indoors.
James -
be* pro - on THFB - sync about 17.2mbps -
BQM