AT&T and XM/Sirius to Get More Spectrum Online? And What Does It Mean for LTE In The U.S.

Every now and then, spectrum licensees do something so sensible it actually seems possible that Coase was right about all that rational actor stuff.   So it is with AT&T’s recent joint submission with XM Sirius on how to resolve the long-standing interference fight between the Satellite Radio licensees (“SDARS,” for reasons we won’t get into now), and the “wireless communications Service” (WCS) licensees. You can read a general summary over here and the AT&T/XMSirius letter here.

Briefly, the deal protects XM Sirius from possible interference by sacrificing some spectrum right next to it. This will mostly shaft a company called Nextwave. But AT&T is taking enough of a haircut on its theoretical spectrum rights to deflect the argument that it is trying to benefit itself solely at the expense of others. And while Comcast (also a licensee here) might try to slow things up as a favor to its new BFF VZ Wireless, that seems unlikely. Since this plan appears the only path forward for getting this spectrum into productive use, I expect the FCC will approve it. Hopefully the FCC will approve it quickly, because it will take a lot of work to develop equipment for the band.

AT&T deserves credit for working with XMSirius on this and coming up with something positive and pragmatic rather than continuing the endless cycle of blame between satellite radio and WCS licenses. On the other hand, like the current AT&T effort to refarm its 2G spectrum for 4G, AT&T had to be pushed into it kicking and screaming by denying it the right to buy out T-Mo. This should also emphasize the next direction for the wireless industry: more efficient multiband radios. The Future of LTE is a multiband future (with the exception of Verizon Wireless). Robust technologies like the TV White Spaces will, as predicted by TVWS advocates, play an increasingly important role in licensed technologies as well.

More below . . . .

The WCS band is one of those bands that everyone keeps hoping will turn into something good but nothing ever seems to work out.   The FCC auctioned it in 1997 in the immediate rush after the success of the first PCS auctions and the FCC was auctioning just about any block of spectrum it could lay its hands on without much concern for who would use it or how. The rationale being that if you just dumped spectrum into the market and people bid on it, market forces would ensure that it got used efficiently and stuff. Located at approximately 2.3 GHz, the FCC divided into four bands: an unpaired 5 MHz C and D, and a 10 MHz A & B auctioned as paired 5 MHz blocks. At the time, a 5 MHz block was considered adequate for voice and most other functions. The FCC also wanted to move away from the traditional pairing of an “up” and “down” band of equal size (a consequence of technological limitations of the 80s) to allow for more flexible and efficient spectrum use. So while this band plan looks rather bizarre today, it made sense back in the early days of spectrum auctions when the world seemed full of possibilities if the FCC would only get out of the way and let the market work everything out.

From the get go, however, WCS faced a bunch of problems. The biggest problem, as illustrated by this band plan diagram is that the Satellite Digital Audio Radio Service (SDARS), aka satellite radio, sits smack dab in the middle of the WCS band. As a result, WCS had to protect satellite radio signals. Satellite radio signals are fairly weak, and the consumer devices used to receive these signals are mobile. Add to this that consumer devices always try to minimize cost by skimping on shielding, and the problem of protecting satellite becoes extremely complicated and costly. To this add the additional complication that the 2.3 GHz band has good-but-not-great penetration characteristics, making it much less interesting for mobile services (or even fixed data services) than a lot of other bands that were available back then. Finally, as data became increasingly important, the 5 MHz channel size proved less attractive than the larger blocks in places like the AWS bands.

So when the FCC did auction the band, it went for practically nothing (the WCS Auction netted approximately $13.5 million — keep that in mind everyone who thinks that spectrum auctions automatically mean billions in revenue without rasing taxes). The WCS band sat idle, while the satellite radio guys in the neighboring band built out their business. With the licenses due for renewal in 2007 and no one building anything out, the FCC gave everyone a 3 year extension until 2010. But that didn’t address the problems that made the band pretty worthless. So in 2010, as part of the National Broadband Plan’s “lets win the future by solving the spectrum crisis” initiative, the FCC issued a new set of rules to try to address some of the problems while still protecting SDARS receivers.

Unfortunately, none of the WCS licensees found the changes particularly useful for deployment, and the band continued to sit there. A bunch of licensees filed Petitions for Reconsideration. The FCC staff, having already done its best to arbitrate between XMSirius and the WCS licensees, encouraged the parties to come up with something else that represented a better way forward. While I am normally critical of the FCC for ducking responsibility and asking private parties to do its job for it, I have sympathy for them in this case. They made an effort to arbitrate and impose a solution on all sides. Unless the licensees could agree on a different approach, why should staff think another effort to impose a solution on the licensees would turn out any better?

The AT&T/XMSirius Proposal

AT&T holds the largest number of licenses in the WCS band (as measured by the population coverage). After the DoJ and FCC rejected AT&T’s efforts to buy T-Mo, and with no new auctions likely any time soon, AT&T became highly motivated to find new sources of spectrum for LTE. So making the WCS band work went from “nice if we can manage it” to “reasonably important and something we should focus on doing.” XMSirius, for its part, would like to get this resolved since there is always a chance that the demand for new spectrum will result in the FCC siding more aggressively with the WCS licensees. So with both parties motivated and the negotiation reduced to two sophisticated parties, the market did indeed work its magic and a deal emerged.

Lest the neo-Cosians and worshipers of the Gods of the Marketplace become unduly smug, this still requires the heavy hand of government regulation to impose the solution on the other licensees. If you insist on a property analogy, consider this analogous to when the government exercises “eminent domain” to take private property for public purposes. I stress, however, that since there is no property interest in spectrum licenses those parties losing spectrum rights get no compensation. In classic “scarcity rationale” regulation, the FCC regulates the licenses because in the absence of regulation no one can use the spectrum. Anyone who thinks the “scarcity rationale” is dead, take note. Scarcity is not just Red Lion and media ownership. It goes to the heart of wireless regulation, and overturning the “scarcity rationale” would have broad unforeseen consequences. But I digress . . .

The deal proposed by AT&T and XMSirius would further reduce the utility of the C & D blocks, the two unpaired blocks directly adjacent to the SDARS band. This has the effect of creating a 5 MHz guard band around SDARS, permitting more robust mobile operation in the A&B block pairs. As a result, the A&B 5 MHz pairs can be used for LTE. AT&T has some C&D block licenses, so it takes a hit on this. it has A Block licenses in some markets, B block licenses in other markets, and combined A&B in a few markets. So this will ultimately give them some useful deployment. Sprint benefits potentially as well, since it has a bunch of A &B licenses in the deep south. This benefit is somewhat limited, however, because they don’t cover a lot of population. Some companies no one ever heard of in the Northeast also have A&B licenses, as do Comcast and Nextwave. I expect AT&T will acquire these over time.

In the C&D blocks, the biggest single holder is AT&T, so they have pretty good moral authority to press for approval. They are taking a significant hit. The next highest holder is Nextwave, followed by a company called Stratos.

What Happens Next?

Procedurally, this is all part of a pending Petition for Recon, so the FCC can grant it at any time (subject to a vote by the full Commission). If the Bureau wants to be cautious, it can put the proposal out on public notice to eliminate any future objection on procedural grounds. Given the way the general practices of the Genachowski FCC, I would bet on a short public notice followed by fairly quick approval unless something turns up in the public notice that acts as a major surprise.

Why do I think it moves forward? Because this represents the best chance of getting the band into productive use and no one else has a better path forward. I do not expect the FCC to be moved by the complaints of C & D block licensees, especially when there is no evidence they could ever do anything with an unpaired 5 MHz block. As noted above, AT&T is taking a major hit on its C&D licenses, so this isn’t an unjust windfall on the backs of one licensee over another. Comcast might be tempted to make trouble because it and Verizon are now total BFFs and Verizon has no reason to want to see AT&T get more spectrum into productive use. But Comcast has already publicly stated it has no intention of getting into the wireless business, so its motives in objecting would be pretty transparent. As a general rule of thumb, with the potential gain of messing with AT&T so low and the potential success of messing with this fairly low, Comcast is unlikely to want to spend a political chip and piss off AT&T just to lend VZ a helping hand here.

Is it Good Policy?

From what I can tell at this stage, I think the FCC ought to approve the deal. Mind you, stuff could come out in the public comment process to change my mind. But at the moment this looks like a pretty reasonable and positive thing to do.

It may surprise folks to see me so chipper about this proposal given my usual concern about the “spectrum gap,” i.e. that AT&T and Verizon will hold so much more spectrum than their competitors that no one else can really compete. But there is a huge difference between carriers using spectrum they already have v. consolidating the spectrum further. AT&T has the spectrum already, so no one else can possibly use it. We want carriers like AT&T to use what they have to meet their needs, rather than suck up more spectrum to warehouse until needed. AT&T has not used WCS because of this rather unfortunate interference fight. Getting this spectrum into productive use allows AT&T to expand capacity in a way that leaves any other available spectrum for competitors. A win for spectrum efficiency without any significant loss on the competition side.

In addition, to make this band work, AT&T will need to do some serious investment. I’m all for AT&T, or any other company for that matter, reaping rewards for investing in their network and new technology. I just want them to work for a living like the rest of us rather than enjoy monopoly rents. If AT&T can take a band that is going nowhere and make it something productive, I’m happy to see them enjoy the benefit from that. Rah rah capitalism, markets and stuff!

Impact on AT&T and The Future of LTE

Some folks are talking about this as a potential “game changer” for AT&T. I wouldn’t go that far. No equipment exists for this band, so AT&T will need to go to the standards bodies, get a standard, get manufacturers to build stuff, and then deploy. Fortunately, AT&T is big enough to make this happen. Still, it will be a few years, at least, before we see AT&T deploying LTE on WCS spectrum.

What this highlights, however, is that the future of LTE is multiband, and we will need to dramatically improve the current multiband technology. AT&T will need to make this work with its Lower 700 MHz footprint and its refarmed PCS spectrum. Other carriers are in similar straights. Sprint needs to cobble together the BRS and EBRS spectrum at 2.5 to work with its broadband pcs spectrum. T-Mo and other carriers that are even more spectrum constrained will need to depend on multiple roaming partners on a variety of bands, as well as tricks such as Wi-Fi offload.

The only exception to this is Verizon. Verizon will have a fairly neat and efficient architecture with its upper 700 MHz C Block spectrum and big AWS paired 10 MHz blocks. But everyone else is going to need to manage a patchwork of access rights and licensed bands to meet the growing demand for mobile data traffic.

Happily, as AT&T has just demonstrated, the wireless carriers can manage this challenge when forced to do so. Sure, everyone would rather have more spectrum to keep doing what they have always done. R&D for new multiband radios is expensive, and competitors hate to work together on things like roaming and interoperability if they think they can win the game on their own. But when the alternative is learn to work together and innovate or die, people take a second look at their options and decide cooperation and investment aren’t so bad after all.

And to tie this back to my other spectrum obsession, the technology leading the way to the future of mutliband is the TV white spaces. It is the next generation of cognitive radio. As new bands outside the TV bands become accessible, the technology will evolve exactly the sort of multiband cognitive radio handling complex spectrum environments that carriers will need in the future. Instead of multiple chip sets with a separate radio and separate shielding for each spectrum band, we will see a much smaller set of chips each capable of handling a larger range of frequencies and rules.


Happily, we need not prognosticate so far for the FCC to determine whether the A&T/XMSirius proposal is a good thing. “Sufficient unto the day are the evils thereof,” and whether I am right in my long term prediction does not alter the fact that the proposal to get WCS into production looks like a good thing, and the FCC ought to move it along with all due speed.

Stay tuned . . . .

One Comment

  1. What in the world makes you think they will actually build this out? They have not built out what they already have at all in many areas and not efficiently (small cells). Is this not another strategy to prolong the actual use and keep it warehoused? It has been 15 years for this band to sit fallow…..

Comments are closed