Even FCC Commissioners Unsure!Wednesday, August 01, 2007
By a four to one vote, the FCC yesterday agreed to the rules for the upcoming 700-MHz spectrum auction. The rules won't make anyone completely happy and may, in fact, devalue some of the spectrum, but the open access camp can feel as though it won at least a partial victory (although it is evident they don't have a clue what this type of open access means) and the first responder community received additional spectrum to be used on a shared but priority basis. The real question of the day is whether the shared public safety/commercial spectrum will have value to commercial bidders.
Even the Commissioners, in their comments, are not sure of this with at least two saying that perhaps they won't get any bids for this spectrum and that it would then have to rethink the plans. Further, one Commissioner was right on the money when he said this shared spectrum would be more expensive to build out than a strictly commercial network. That is because the FCC wants the network to be public safety grade, meaning extensive coverage, which requires many more cell sites, back-up systems at every cell site, in-building coverage and coverage even in areas where there may not be a population base to support such a network.
If you look at the new band plan carefully, you can see the differences. The bands in yellow have already been auctioned, but some of the guard bands are being relocated in order to make this all work. There are some firsts here. For starters, the shared 10-MHz D band will be auctioned as a single nationwide license. Second, open access will be required for the C block (22 MHz, 11X11).
Let's look at that more closely. In the C block, there will be twelve licenses awarded on what is called a REAG or Regional Economic Area Grouping. So, the first question here is whether this spectrum will be worth as much as the spectrum that does not carry open access requirements. The second question is whether open access proponents understand that this spectrum could be bid on and won by twelve separate companies and that there is nothing I see in the rules that requires these spectrum holders to employ the same technology, so we could end up with twelve separate companies using up to four different technologies. Let see, I will need a 700-MHz device with four different technologies built in, and all the companies that own the spectrum will probably have to figure out how to handle roaming agreements across these four technologies. Does anyone really think we will have cheap, open access devices in this situation?
The FCC is trying to make it easier for one bidder or perhaps only a few to walk away with this spectrum. In the auction procedures that were also part of this Report & Order, the FCC will use package bidding, which means companies can bid on one, five, seven or all twelve licenses as a single package.
If one bidder were to win all twelve licenses, it would not matter. But reality is somewhere in between, say, four different companies and three different technologies. And for the record, the press I talked to today and yesterday all seem to be coming to the same wrong conclusion! Open access is NOT free access. You still have to have an approved device and pay for service. There is no free ride with open access.
All All of the 700-MHz spectrum is prime real estate, but after the FCC rules, some portions have more value than others. The A block (12 MHz, 6X6) will be broken into 176 licenses based on EA or Economic Areas, while the B block (12 MHz, 6X6) will have 734 licenses based on CMA or Cellular Market Areas (this spectrum fits with the existing Cellular and PCS license areas). These two segments of the spectrum are the most valuable because they are not encumbered with any restrictions, open access requirements or co-ownership.
The C band (22 MHz, 11X11), which is the largest of the bands, is where open access rules apply. On my list, this makes it the third most valuable portion of spectrum. The D band (10 MHz, 5X5), for which there is one license available, brings up the rear.
There is also some unpaired spectrum (some has already been auctioned), the D band (UHF Channel 55, which Qualcomm already won at auction) and the next UHF TV channel up, 722-728 MHz, that will be up for grabs in EAs, resulting in 176 potential licenses.
The bidding will be anonymous, which means no one will know who is bidding for what at what price until the auction closes, and the Wireless Telecommunications Bureau has been ordered to establish reserve (minimum) prices for the entire spectrum allocation.
First, let's consider what technologies will be deployable when these networks are built in 2010. We will have UMTS but with HSPA+ and CDMA2000 1xEV-DO Rev B or C and WiMAX. Three new technologies that should also be available in that time frame are LTE or Long-Term Evolution from the GSM/UMTS world, UMB or Ultra-Mobile Broadband coming from the CDMA community and perhaps WiMAX M, the next generation of WiMAX. The newer technologies will be faster with more capacity and, just as important; they will be able to be deployed into the segments set out by the FCC. Today's UMTS, for example, takes 5 MHz of spectrum per carrier, LTE can operate within almost any type of band from 1.25 MHz up to 20 MHz, as will UMB and, I assume, WiMAX M.
If the timing is right and the stars are properly aligned, these new networks could all use the newer, faster, more efficient technologies. If not, the older technologies can be deployed and the newer ones used in major metro areas where there is a greater need for capacity.
Now let's look at the spectrum. A little refresher will help here. Most systems today, with the exception of the current version of WiMAX, use one segment to transmit from the cell site to the device and the other segment for the device to transmit back to the cell site. This makes it easier to deal with interference.
First is the A block, which is 12 MHz (2, 6X6). There are 176 licenses for the entire nation so it is unlikely any one company will win (or even bid on) all of them. Since 12 MHz of spectrum is not enough for a new operator to build a new network, these licenses will probably go mostly to incumbents or perhaps to companies that want to use licenses as currency in dealing with the incumbents.
The B block, also 12 MHz (2, 6X6), has 734 licenses that are also basically aligned with existing wireless network licenses. These licenses will be most valuable to the incumbents, probably on an area-by-area basis as extensions to their existing 850/1900-MHz networks where they need additional capacity. Once again, it would be difficult for a new entrant to buy up a nationwide footprint, and 6 MHz is simply not enough spectrum. Some regional or local companies might want to enter the wireless business and some licenses in these two blocks might be of interest to them, but I believe most will go to incumbents.
This leaves the biggest block: The 22-MHz C block that requires open access and for which only twelve licenses will be auctioned, and the D block, which will be a shared network with the first responder community (10 MHz, 5X5). Let's take the D block first where there is one license holder sharing this spectrum with the public safety community. While it is one license and the winner will have a nationwide footprint, this is not enough spectrum on which to build a new network. Even if a company were willing, I'm not sure how customers would feel about subscribing to a network they might not have access to when the system is being used by first responders who have priority. Since this network has to be built to public safety standards, I believe it would be the most expensive commercial network ever built.
If this spectrum is bid on, I believe it will have to be by an incumbent or someone who believes an anchor tenant such as public safety will provide enough income to offset the disadvantages of commercial customers being forced off the network from time to time. Another factor is that the bidder will have to match the technology chosen by the public safety community for its own 10 MHz of broadband spectrum. If this is not the case, then the public safety community will already have a problem trying to support two different technologies on adjacent bands―one that is its primary band and one band on a shared basis.
It is possible that no one will bid on this spectrum, which would be a shame. I believe if a network is built correctly, includes Rural America and is used by a commercial operator with other capabilities on other portions of the spectrum, this system has a good chance of working, and I think there are many opportunities for funding this system. It remains to be seen whether commercial operators will see value in spectrum they have to build out but where their customers are secondary to public safety.
This leaves the C block (22 MHz, 11X11), the largest segment of the spectrum and where winners must adhere to open access requirements. I have already talked about the fact that there are twelve licenses and you might be able to make a case for building a reasonable nationwide network in 22 MHz of spectrum. So perhaps a company such as Google will step up and lay its money down.
A single company with sufficiently deep pockets could win all twelve licenses and end up with a nationwide system. Open access could rule the day, and there would be more competition. However, Rural America still won't have wireless broadband access. Why, you ask? Because the FCC build-out requirements for this segment of spectrum requires the winner to build out coverage to only 40% of the population within four years and 75% of the population by the end of the license term. That leaves 25% of the population in the United States still underserved.
Apple will still be able to sign an exclusive deal with the network of its choice, the networks will still be able to make exclusive deals with handset vendors and there will probably be some really cool open access devices―but they won't be cheap! My bet is they will have to include at least three different technologies and will ONLY operate on the one 700-MHz portion of the spectrum. In my estimation, if a device includes other bands for a network it won't qualify as an open access device.
Andrew M. Seybold