For the target of a corporate acquisition, getting ready to be swallowed by another company is lot of work, sapping countless man-hours that could be well spent elsewhere were the merger not taking place. It's enough work, in fact, that AT&T has agreed to kick out $3 billion in cash to T-Mobile -- on top of a new roaming agreement and some AWS spectrum -- in the event that its planned acquisition doesn't meet regulatory muster. A new rumor popping up on Reuters yesterday pegs the total value of the package at $6 billion: roughly $1 billion for the roaming agreement and $2 billion worth of spectrum in addition to the previously-disclosed $3 billion in cash.

T-Mobile already uses AWS for its HSPA network, so if the merger fell through, those new licenses would be a nice fit. But just how much AWS spectrum does $2 billion buy you these days? And more importantly, is it enough for the nation's fourth-largest carrier to deploy LTE?

In talking to a consultant with Aircom International, Reuters postulates that $2 billion would cover roughly 10MHz worth of bandwidth (and though they don't spell it out, I'm assuming they're suggesting that it would cover major markets nationwide). That would line up roughly with the results of the FCC's Auction 66 held in 2006, where T-Mobile spent about $4.2 billion for some 120 AWS markets, mostly in 20MHz blocks.

AT&T currently holds a total of 158 AWS licenses, some covering major markets like Boston, Miami, Atlanta, and Jacksonville. Since 700MHz spectrum is more appealing in some respects -- it penetrates buildings more effectively and allows cell sites to be placed further apart -- AT&T looks to be concentrating its phase-one LTE deployment in its 700MHz holdings, which lines up with its statements that the AWS license transfers to T-Mobile would not be critical to the rollout.

So would it be enough to finally give T-Mobile an effective next-generation network strategy? Amazingly, yes, it probably would. Even accounting for spectrum appreciation in the five years since Auction 66, the company could probably find a way to combine its new holdings with its existing AWS spread to deploy somewhere between 5 and 10MHz of LTE bandwidth across major markets (the standard is designed to effectively scale from 1.4MHz up to 20MHz). And if the deployment would be too speed- or capacity-restrained, we're sure they've got Clearwire and LightSquared on speed-dial for a wholesale deal; $3 billion in newfound pocket change certainly doesn't hurt when you're sitting down at the bargaining table.

But the real story here isn't necessarily the crazy failed merger package, it's AT&T's supreme confidence that the merger will succeed. Never mind the roaming deal or the $3 billion in cash -- no executive board in its right mind would approve a plan to relieve itself of any amount high-quality, field-tested spectrum right now unless it was operating with near-certainty that its doomsday scenario won't play out. The FCC will be the first to tell you that it takes years, sometimes decades, to free up new spectrum for broadband wireless -- and with as little spectrum as we have to go around right now, the last thing AT&T wants to do is turn T-Mobile into a viable 4G contender at its own expense.