[sclitheroe]

Originally Posted by HotHotHeat

Here's one more article. Seems to be something that's accepted as a problem going forward, as opposed to something that doesn't exist because they don't know what they're talking about.

The problem is mostly inefficient allocation of spectrum and under-utilization by those that obtain spectrum. It's estimated that about 5% of all useful commercially licensed spectrum is in use.

Furthermore, all of this talk of spectrum is ignoring the fact that improved signal processing gets you more mileage out of existing spectrum. LTE, for example, is vastly more spectrum efficient than HSPA out of the box.

Down the road past LTE, with greater and greater on-device signal processing capability, we can drive spectrum density higher, as well as operate in spectral areas that are less than ideal compared to the spectrum cell carriers are currently using, because we can apply more sophisticated modulation and error correcting algorithms against the signal.

In total, U.S. operators have licenses for about 538MHz of wireless spectrum. Only about 192MHz of that spectrum is currently being used. And of the spectrum that is being used, 90 percent of it has been allocated to existing 2G, 3G, and 3.5G wireless services by larger wireless carriers, such as AT&T, Verizon Wireless, Sprint Nextel, and T-Mobile USA.

The move to 4G is very important for these operators because it offers them a much more efficient way to deliver service. According to the Citigroup report, 4G LTE uses the available spectrum roughly 700 percent more efficiently than the 3G wireless technology EV-DO.

Source: http://news.cnet.com/8301-30686_3-20...trum-shortage/

I'm not an expert on RF and signal processing by any stretch, but when we look at every single physical/electrical/signalling constraint in IT, it's always overcome with more processing power allowing us to squeeze more data out of existing contraints (eg. CAT5 ethernet went from 10 megabits to 1 gigabit, hard drives continue to increase in density despite having to deal with quantum physics level effects, we're pushing hundreds of megabits throughput on regular coax (DOCSIS3), etc). There's no reason to think it won't apply here too.

The biggest issue being faced is economics and logistics - it's really really expensive to continually build out, grow density, and refresh existing tower networks. It's also expensive to transition millions of subscribers to newer technology as well. And both take a long time to make happen.

I'm willing to wager that the simple economics and logistics will stifle growth in the mobile space long, long before spectral capacity will.