As we enter the last 24 hours before the critical and transformative November 4 vote (no, not this one, the FCC vote!), a last battleground has emerged. While the broadcasters and wireless microphone guys have generally not generated any traction, a final possible hitch has shown up on the question of higher power for rural providers. While I applaud the sentiment, this has become the last ditch effort to sneak a “poison pill” into the Order by keeping alive the hope/fear of exclusive licensing in the band.
As I have long warned, the potential last-minute threat to unlicensed in the band would not come from broadcasters, whose interference claims have been discredited and who have stooped to rather ridiculous smear tactics, or even from wireless microphone manufacturers and their vast horde of politically powerful pirate users. No, I have always believed that at the last minute, the real flank attack against the public interest would come from the licensed wireless guys pushing for licensed backhaul.
Which is why I am unsurprised to find the last potential stumbling block toward the finish line, after five years of unprecedented testing and investment, comes from a push for some kind of exclusive licensing scheme, either as an immediate set aside in the existing order or as part of a further proceeding.
I call this the “Part 101 Poison Pill.” Part 101 of the FCC’s rules governs high-power point-to-point transmission links of the sort used by telecommunications companies for transmitting significant distances. Part 101 is different from cellular licensing, in that it can accommodate multiple users on a “first in time, first in right” basis. Whoever comes in later must protect everyone who comes in earlier, which essentially makes it a very high-cost game of “king of the mountain.”
What makes exclusive licensing, even the relatively more open licensing such as Part 101, such a poison pill for unlicensed?
See below . . . .
As noted above, the licensed wireless carriers have always opposed making access to the spectrum any easier. Mind you, they have a real problem finding back haul, and more spectrum for commercial use generally since AT&T and Verizon have done such an excellent job out-consolidating and out-bidding everyone. So unlike the broadcasters, who are essentially just making up scary stuff, the folks pushing wireless have a serious problem that generates a fair amount sympathy from folks at the Commission. Heck, I’m sympathetic to it myself. But my sympathy for the competitive and rural licensed carriers does not blind me to the fact that their proposals would kill unlicensed access to the white spaces dead, DEAD, DEAD in everything but name. And, as I have written ad naseum over the last five years, unlicensed access to the white spaces has far too much power to transform our lives for the better to have it killed at the last minute for the benefit of a handful of licensed carriers.
To understand how licensing is so utterly inimical to development of unlicensed, let us delve into the two proposals for licensing pressed on the FCC in the last two days. Despite their disagreements on details, the licensed carriers have been unanimous in trying to derail unlicensed use. It also explains why, despite very serious disagreements with Wireless ISP Association and their “licensed-lite” proposal, the Public Interest spectrum Coalition (PISC) and
Exclusive Cellular Licensing — the Wireless Carrier Plan
CTIA and Qualcomm want to keep “half” the available broadcast channels for exclusive licensing and require unlicensed to “prove itself” before being “phased in” (I put “half” in quotes because CTIA wants the channels 38-51 for licensed use, which is where the bulk of the unassigned channels (aka “the white space) actually is, so it amounts to most of the available space even if it is only half the channels). CTIA wants exclusive licensing of the cellular variety. Basically, CTIA (and Qualcomm, which holds a fair number of 700 MHz licenses) would like another 700 MHz-type auction, ideally for its members who got squeezed out of the original 700 MHz auction.
This plan is obvious death to the unlicensed use of white spaces by prohibiting any kind of shared operation, effectively excluding unlicensed devices from any major market by eliminating any available white space (remember, especially after the digital transition, and after subtracting channels set aside for wireless microphones, the bulk of available white space will be in the channels above 25). Unlicensed devices would also have to protect the new licensed white space services. No one would want to waste their time or money trying to ”prove“ that this would work, especially after five long years and many millions of dollars already spent proving this will work. If the FCC adopts this proposal, it might as well not bother pretending to adopt unlicensed at all.
The Rural Licensed Carriers and Part 101 Exclusive Licensing
The much smaller rural carriers, combined with Sprint-Nextel (which is already taken care of for wide area service with the New Clearwire deal) don’t want another 700 MHz auction, wherein they will get screwed. They want high-power service for backhaul. This leads them to Part 101 of the Commission’s rules. Part 101 is non-exclusive enough so that there is no requirement to auction the licenses. However, it is exclusive in that you need to get a site license. It favors first movers because it has ”first in time, first in right“ rules that mean anyone who comes in must protect everyone else who has been there before. As noted above, this creates a ”king of the mountain“ mentality that encourages people to throw up towers and start blasting just to reserve their primacy. Worse, from the standpoint of those of us who worry about non-commercial operations, these higher-power licensed towers are expensive (relatively speaking) to build, power and operate.
So they have shot out this last minute proposal. On the surface, it looks moderate. Reserve up to 6 Channels on third adjacent in rural areas for Part 101-type high-power licensed backhaul. Lots of empty spaces out in rural areas, yes? So why not throw the rural guys a few backhaul channels while leaving the urban and suburban markets for unlicensed?
There are a number of reasons why Part 101 licensing turns into a total poison pill for unlicensed. First, the Part 101 power levels are completely incompatible with operation of unlicensed devices based on sensing. Even limited to just six channels, each one a ”third adjacent“ from an active broadcast channel (that means there must be at least two empty broadcast channels on either side of the Part 101 channel) the out-of-band-emissions (OOBE in spectrum speak) from these high-power point-to-point backhaul providers will convince any device using spectrum sensing that all channels are occupied. Use of the white spaces on an unlicensed basis in rural areas, including the higher-power fixed unlicensed nodes intended for rural broadband, will never develop. The whole promise of rural broadband from unlicensed use of the white spaces will die stillborn from the accidental leakage from a handful of rural backhaul systems. This does not really make sense as a viable trade.
More importantly, the introduction of a new, high-powered, licensed-and-therefore protected unknown element into the band will have the same impact as the CTIA proposal on deployment and development. No one is going to put money or research into this if every time it looks like the rules are settled the FCC changes the game again and issues a general invitation to those who want licensed to keep coming back and trying again.
Case in point, we got the rules for ”licensing lite“ in the 3.65 GHz band back in 2005. But development of equipment for the band was delayed by 2 years because WiMax supporters filed Petitions for Recon after Michael Powell left the FCC and no one knew what Kevin Martin would do. Given the regulatory uncertainty, no one tried to get devices certified until the FCC resolved the Petitions for Recon in 2007.
If the FCC sends a strong signal that it backs unlicensed in the band, hardware developers like Motorola, and Phillips, as well as software developers like Microsoft and Google, stand poised to invest big bucks into unlicensed. But if the FCC starts another round of rulemaking on a possible high-power licensed solution, these players will take their money and go home. The fact that players such as Microsoft and Motorola have stuck with this for five years has been amazing by the scope of industry standards (Intel, an early supporter, dropped out after a couple of years to focus on WiMax). But change the rules on them at the last minute again (after putting up with an addition Further Notice back in ’06), based on no engineering evidence and introducing maximum uncertainty, and they will find it very difficult to maintain enthusiasm.
Reject The Final Push By Licensed Carriers To Keep Unlicensed Out of The Picture.
It is noteworthy that while CTIA and the other licensed carriers, despite their disagreements on the flavor of licensed, agree entirely on screwing this up completely for unlicensed. In other words, the licensed carriers will gladly delay this whole thing and divide the corpse at their leisure. Their sole goal at this point is to — once again! — drag out the rulemaking, frighten off investment for unlicensed, and ward off any competition from the unlicensed providers (such as my friends out at the Mountain Area Information Network. It is also noteworthy that the Wireless ISPA Association (WISPA), although opposed to pure unlicensed in favor of a ”licensed-lite“ regime modeled on the rules developed for the 3.65 GHz band, has vigorously opposed both CTIA and FiberTower.
In the ecosystem of rural broadband, you can find a clear trend. The bigger you are, the more exclusive you like your license. CTIA and Qualcomm want exclusive licenses. The rural licensed carriers, who can’t compete with CTIA’s bigger members in an auction, want Part 101 licensing. WISPA, whose members are tiny businesses even when compared to ”small” licensed rural carriers, want the non-exclusive licensed-lite. And the community wireless networks, who are provisioning themselves and relying on volunteers and donated equipment, want true unlicensed.
Exclusive licensing is a poison pill, and as long as a the specter of exclusive licensing hangs over the broadcast white spaces, it will never realize its potential. Yes, getting rural backhaul and more spectrum for competing wireless carriers is important. But you can’t solve those problems here today, and the effort to do so — however well meaning — sacrifices too much. If the FCC really wants to encourage innovation AND make real progress on rural broadband deployment, it will vote the Order tomorrow as proposed and lay the licensed ghost to rest once and for all. Come back in another proceeding to examine the possibility of licensed-lite for rural areas, as well as safe ways to boost power for unlicensed devices. I like licensed-lite a lot, and think it holds real promise as a middle ground between unlicensed and licensed for a higher-power service in rural areas compatible with unlicensed. But that is a proceeding for another day, when the licensed carriers aren’t ruthlessly exploiting every opportunity to stunt unlicensed and keep it from breading new competitors. Tomorrow, vote the Order and settle the exclusive licensed v. unlicensed in the broadcast white spaces question once and for all.
Stay tuned . . . .
“. . . interference claims have been discredited . . .”
Only in terms of propaganda, not engineering; read the data and substance of the OET reports, not just the executive summaries.
“First, the Part 101 power levels are completely incompatible with operation of unlicensed devices based on sensing. Even limited to just six channels, each one a ”third adjacent“ from an active broadcast channel (that means there must be at least two empty broadcast channels on either side of the Part 101 channel) the out-of-band-emissions (OOBE in spectrum speak) from these high-power point-to-point backhaul providers will convince any device using spectrum sensing that all channels are occupied.”
Not necessarily. With proper bandpass filtering (which would be required anyway) there’s no reason adjacent channel out of band emissions can’t be kept at 80dB down (and over 100dB for 2nd adjacent channel), thus providing a power level *at the transmit antenna* of -23dBm, given a +63dBm (2KW) ERP. Given a distance of let’s say 1000 feet from the antenna, which provides about 78dB free space path loss at 650MHz provided a completely unobstructed fresnel zone, and a vertical beamwidth loss of 8dB, the WSD would be see about -109dBm. Although this is within the proposed spectrum sensing sensitivity of -114 to -120dBm, it is low enough power that the WSD could A) rely on the geo-location database and B) a sort of reverse beacon; the backhaul data stream could include a few bits indicating which channel it’s actually using, thus permitting the spectrum sensing algorithm to allow use of the channel(s) receiving only the OOBE. Since the backhaul signal will be broadband, there’ll be no confusion with Part 74 devices.
Since spectrum sensing will be fully functional, it’s only a matter of a few lines of code – certainly no problem for the likes of Microsoft, Philips, Motorola, Dell, Google and HP.
Further, as the people in most desperate need of high speed internet are those in the rural areas abandoned by the telcos and cable providers (and not the city and suburban folks already getting broadband via cable, teleco, EVDO, HSDPA and Starbucks), isn’t this a more socially responsible use of the spectrum? If you’re concern is the licensing, simply argue to make this an unlicensed service and with lower power levels.
Mr. Cohen, you are like some demented energizer bunny.
Seriously.