Depending on whom you ask, the Progeny Waiver will either (a) totally wipe out the smart grid industry, annihilate wireless ISP service in urban areas, do untold millions of dollars of damage to the oil and gas industry, and wipe out hundreds of millions (possibly billions) of dollars in wireless products from baby monitors to garage door openers; (b) save thousands of lives annually by providing enhanced 9-1-1 geolocation so that EMTs and other first responders can find people inside apartment buildings and office complexes; (c) screw up EZ-Pass and other automatic toll readers, which use neighboring licensed spectrum; or (d) some combination of all of the above.
That’s not bad for a proceeding you probably never heard about.
For me, the Progeny Waiver is a microcosm of why it has become so damn hard to repurpose spectrum for new uses. The added twist here is that this time it is largely the unlicensed spectrum users acting like incumbents and saying that it will be the end of the universe if Progeny lights up its system (although the licensed neighbors say the same thing, pretty much), and Progeny, the licensee, arguing that everything will be JUST FINE, really, and if it isn’t too damn bad because we are licensed and they are unlicensed so there!
You might ask, “if this Progeny thingie is so gosh darn important, why have I never heard of it?” Well that’s why you read this blog, you clever reader you. This amazing little proceeding is still so deep in the bowels of the FCC that only the true spectrum wonks have noticed. But action now appears imminent, so consider this a sneak preview of this season’s favorite telecom reality show, Spectrum Wars.
What raises the stakes on this too damn high, however, is the implications for the future of unlicensed generally and the implications for the credibility of the FCC as an agency able to actually do the technical job of managing an increasingly complex spectrum world. Fairly or unfairly, everyone is going to compare this to Lightsquared (waiver, followed by worries about interference, arguments that the FCC failed to follow its own rules and procedures, blah blah). Let us add to this House Republicans who would love to call the FCC on the carpet for mismanaging spectrum – especially around unlicensed. Add to that the car manufacturers in the 5 GHz band and the federal users generally wanting to show that the FCC can’t adequately manage the stuff it has and you have a pack of circling sharks just waiting for the FCC to screw this one up and commence the feeding frenzy. So no pressure.
Happily, I have, if not a solution, at least a better way for the FCC to cover it’s rear-end and contain the damage, below . . . .
As always, the backstory takes awhile, so bear with me.
It All Started Almost 20 Years Ago . . .
Back in the mid-1990s, the FCC had a proceeding to set aside a slice of spectrum for the much anticipated “Intelligent Highways Initiative.” As part of this, the FCC set aside some spectrum for the “Multilateration Location and Monitoring Service” (M-LMS) and a non-multilateration version, LMS. The new M-LMS service would help track cars and trucks on the new intelligent highway by using broadcast beacons that would triangulate with units in the cars and trucks. Unfortunately, things being rather crowded, the FCC selected the 902-928 MHz band for the new M-LMS and LMS service.
For those of you who do not have the table of allocations conveniently posted on your wall as if it were some prized Jackson Pollack print, the 902-928 MHz band is the one significant chunk of spectrum below 1 GHz open for unlicensed use at reasonable power. Spectrum below 1 GHz (as regularly readers have heard me say countless times) has propagation characteristics that make it extremely good for mobile data – particularly in urban areas, because the spectrum will more easily penetrate objects like walls. As cars also drive through urban areas, the same properties that made it attractive for unlicensed use made it very attractive for M-LMS.
Mind you, this was still in the days before WiFi. Most of the inhabitants of the band were consumer devices, although some folks (including a long-forgotten dotcom darling called Ricochet) were using the band for Internet access even then. Normally (and especially back in those days), the FCC resolves these conflicts by telling unlicensed users to suck it up because licensed users are important and unlicensed is about as valuable as stuff scraped off your shoes. The vast proliferation of unlicensed devices in the 900 MHz band (and the intense lobbying of their manufacturers), however, prompted the FCC to take the unusual step of actually protecting unlicensed use in the band from the future licensees (at least modestly).
Under the rules adopted by the FCC, an M-LMS licensee would need to show that operation of its system would not cause “unacceptable levels of interference” to unlicensed use in the band before commencing operation. Of course, the FCC did not define what “unacceptable levels of interference” actually meant, how it compared to the usual (and equally unclear) “protection from harmful interference” usually given to licensed services, or how it intended to certify that future M-LMS licensees had met their burden. In fairness to the FCC, defining what this sort of thing actually means – especially when predicting the nature of future systems that don’t even exist yet – is hard. The FCC also likes to have flexibility to deal with changes in technology and so forth. Less charitably, this is a good way to kick the can down the road and hope things work out.
The FCC finally auctioned off a bunch of the LMS and M-LMS licenses in 1999. The LMS service ultimately morphed into those automatic toll collection thingies like EZ-Pass. But M-LMS went nowhere. For one thing, the intelligent highways initiative never really took off. For another, a much better and cheaper mobile location system called “GPS” came along at about the same time and pretty much eliminated the primary reason for the M-LMS service.
Fast forward to 2011. A company called Progeny LMS (Progeny) has a whole bunch of M-LMS licenses and a new technology for repurposing the band. Progeny has a system that it claims will provide much more accurate geolocation for mobile phones than standard GPS. If someone is in a multi-story apartment of office complex, GPS does not tell you what floor or apartment they are in, it just gets you to the right building. Progeny’s modified M-LMS system will, according to Progeny, provide you with an accurate location so that EMTs don’t have to keep knocking on doors asking “excuse me, are you suffering a stroke and are non-responsive so you couldn’t tell the 9-1-1 dispatcher your address?” Then breaking down every door with no response.
To implement its system, Progeny needed a waiver of some of the old rules designed for the mid-1990s, particularly the one limiting the use of service to cars and trucks, but also some of the other network rules as well. Progeny applied, and the Wireless Bureau put the waiver application out on Public Notice. Perhaps unsurprisingly, very few people noticed the public notice seeking comment on “Request by Progeny LMS, LLC for waiver of certain Multilateration Location and Monitoring Service rules,” or would have understood the significance even if they had, so the FCC got very few responses.
I will pause here to say that the FCC’s actions here were entirely appropriate. The FCC is supposed to publish applications for waiver. This request involves a single licensee trying to implement a new technology. That’s a situation waiver is for – does this licensee need a limited exception from the general rule for some very specific purpose. So there was nothing sneaky or underhanded about Progeny applying for a waiver, or the way the FCC simply tossed it out on routine public notice with the dozens of other items you find every day in “FCC Daily Digest” (official motto: “No one reads this unless they get paid to”).
One of the companies that did notice the Notice, however, was Itron. Itron makes smart meters that use the 900 MHz spectrum. They also are one of the companies doing smart grid using 900 MHz. (Most companies doing smart grid use unlicensed, in fact.) Itron objected to the waiver, largely on the grounds of “we use this these frequencies so we don’t want anyone to start mucking about and causing interference and stuff.” The Wireless Bureau granted the waiver over Itron’s objection, noting that Progeny would need to demonstrate that it did not cause “unacceptable levels of interference (whatever that means)” for unlicensed users in the 902-928 MHz band.
And Then Things Start To Go Horribly Wrong
Regular viewers of Spectrum Wars will recognize that testing is where things start to go wrong, because the definitions are fairly nebulous. Here, Progeny and the Wireless Bureau needed to address the fact that it had no definition for “unacceptable levels of interference.” This makes designing a test to show that you won’t cause “unacceptable levels of interference” a shade difficult, as I shall illustrate in this hypothetical conversation between Progeny and the Wireless Telecommunications Bureau (WTB).
WTB: Now you need to go and test whether turning on your service will create unacceptable level of interference.
Progeny: O.K., what do I need to do.
WTB: We’re not sure actually, no one has ever tried to turn on a M-LMS system. What do you think you ought to do?
Progeny: Lets start with the most important question. What does “unacceptable levels of interference” mean? I can’t design a test to show I don’t cause it without knowing what it means.
WTB: Well, it’s level of interference that are . . . ummmmm . . . unacceptable.
Progeny: Right. [pause] Lets try this the other way. What would be an acceptable level of interference?
WTB: Less than unacceptable.
WTB: Look, just go do some testing, O.K.?
So Progeny went off and did some testing. They came back later and were pleased to report that – on the test they made up themselves based on a definition they made up themselves – the TOTALLY ACED IT. WTB then put the test results out for public comment. Itron, now joined by the Wireless ISP Association (WISPA) and a few others (including yr hmbl obdn’t blogger), objected in a conversation that basically went like this.
Itron, et al.: That test proved nothing!
Progeny: It proved we don’t cause unacceptable levels of interference.
Itron, et al: It did not!
Progeny: Oh yes it did! Prove it didn’t.
Itron, et al: You did no outside testing. You just grabbed a couple of Part 15 900 MHz devices from radio shack!
Progeny: And your point is?
Itron: There’s no way that this shows that you won’t cause unacceptable interference!
Progeny: Point to me in the rules where it says I need to do outdoor testing. Hmmmm……
Itron: Ummm….Show me where you don’t have to.
Lets Do More Testing!
Regular viewers of Spectrum Wars! will not be surprised that, after a couple of rounds of this, the WTB sent Progeny off with Itron and WISPA to jointly come up with some tests. WTB still cherishes the belief that engineers can somehow magically forget that their companies futures ride on the outcome of these things and that, locked in a room with shiny lab equipment, they will revert back to being engineers. Sadly, this just about never happens.
Instead, what happened is exactly the same thing that happens whenever the FCC tells parties opposed to each other to work out their own testing. They squabbled about everything, including the results. Progeny reported that, after rigorous testing, it was utterly and absolutely obvious that operating Progeny’s systems would not cause unacceptable levels of interference and can we please turn our system on before our investors give up and go home. Itron and WISPA argued that the testing had in no way been adequate to determine any such thing, and the data from the tests that were done shows that Progeny’s systems will cause unacceptable levels of interference so Progeny should no way be allowed to turn on its system.
Those who harbor the cherished belief that engineering is a science that provides clear answers to any problem may wonder how this is possible and surely someone must be lying somewhere. But engineering is not a straight up or down answer. Engineering is applying the laws of physics to the real world, which makes it very messy. Most importantly, engineering cannot answer questions without some clear direction. Without knowing what the heck “unacceptable levels of interference” means, how do you know if a test proves or does not prove that a system will cause it?
More Parties Notice Progeny, Which Does Not Make Things Easier.
Meanwhile, as the process dragged on, more parties took notice of Progeny and their intent to activate a licensed system in this set of frequencies. This set off alarm bells with a surprising and diverse set of companies and interests that either manufacture devices that use this band of unlicensed frequencies, or which use the devices fairly extensively. This turns out to include WISPs, oil and gas companies (who use Part 15 devices in this band for a variety of purposes), other smart grid companies, other smart meter companies, and a whole bunch of others. So when the FCC put the second set of test results out on public notice, it got a *lot* more comments from a lot of very upset people (including the licensed LMS users) saying “Please don’t destroy our business by letting Progeny go live!”
Why would so many companies invest so many millions in technology without interference protection? Because unlicensed means cheap and available, whereas getting ahold of licensed stuff in the right frequencies is impossible without striking a deal with one of the few licensees that has flexible spectrum in the right bands – which is so expensive and hard to do that it is practically impossible for most purposes. As a result, a lot of people have increasingly loaded up important functions on this band because they don’t really have an economically viable choice. Sure, they understood the environment in which they worked meant no interference protection. And yeah, it’s been on the table since 1999 that someone in M-LMS land might turn on a system and they would need to deal with the whole “unacceptable levels of interference” thing. But most people never thought about it. They just went and bought devices because if they wanted to have a business in the first place, they needed to buy devices.
And, of course, lots of consumers bought devices. That’s why the FCC adopted the whole “no unacceptable levels of interference” in the first place.
And just to make things more interesting. The licensed LMS guys are also worried. After all, whenever you are a licensed user, you worry about what happens when a neighbor suddenly cranks up the volume. LMS licensees have interference protection, of course, but the FCC’s process for resolving interference complaints assumes either that one of the users will have a superior right (primary v. secondary) or that one of the co-equal users is causing interference by operating in violation of its rules. Confronted with the paradox of two co-primary services operating within their rules causing interference to each other, the FCC tends to act like those computers on Star Trek confronted with an insolvable dilemma — shouting ‘it does not compute!’ — and shut down.
What’s at Stake and What Happens Next?
At this point, the Wireless Bureau has three options: a) let it die by failing to act; b) act on the record as it exists now; or, c) something else.
Before deciding, lets pause to consider what’s at stake. Part of the problem is that once Progeny goes live, no individual user of unlicensed devices has any protection from increased interference. That’s the trade off for unlicensed. On the plus side, it’s cheap, ubiquitous, and completely flexible. On the downside, you get no protection from interference.
Worst case, if it turns out the FCC is wrong and Progeny does cause “unacceptable levels of interference (whatever that means),” the unlicensed users have no recourse. Progeny going live could wipe out a huge chunk of the investment in smart grid (and smartmeters generally), kill a couple of hundred million dollars in investment by the oil and gas industry, drive WISPs out of urban areas, and kill a whole bunch of consumer product lines. That’s pretty bad. Even if it seems unlikely to happen, it does give one pause.
Even without worst case, a bunch of folks who are providing service or using the 900 MHz unlicensed devices may find themselves experiencing significant declines in throughput or usefulness. Mind you, the same operators might have experienced similar problems at some point from the crowding of the band. Still, even a non-worst case jump in interference could be pretty bad for a bunch of folks providing useful services via unlicensed.
On the flip side, doing nothing effectively kills the M-LMS service, since it’s not like things are going to become less crowded. Also, enhanced geolocation for Next Gen 9-1-1 is a rather useful little service. Mind you, the fact that Progeny has received only tepid support from public safety community to date (one half-hearted endorsement from NENA) raises a legitimate question as to just how valuable this service would be. But this is hardly conclusive.
Progeny’s most powerful argument, IMO, is that it played nice and followed all the rules. What exactly could Progeny have done differently to get its service online. As far as I can tell, nothing. That bespeaks a real problem, because the rules encourage companies like Progeny to invest in new technologies that follow the rules. As I noted last year with regard to Lightsquared, no one will want to invest in developing new services if it is impossible to repurpose spectrum ever.
As if actual consequences of betting wrong weren’t enough, the Bureau needs to worry about political consequences. Here, I confess I worry much more about the negative consequences for unlicensed if the Bureau allows Progeny to light up its current system on the basis of the existing record.
Lets start with the Hill. The American Petroleum Institute, oddly enough, knows Republicans on the House Energy and Commerce Committee. These guys already hate the FCC. Worse, Committee Chair Rep. Fred Upton (R-MI), and Rep. Greg Walden (R-OR), Chair of the Telecom Subcommittee, have launched a crusade to kill any attempt by the FCC to permit unlicensed in the reclaimed broadcast spectrum set aside he Incentive Auctions. I can just hear these guys bellowing at the next oversight hearing: “Why on Earth do you want to set aside 12 MHz for guard bands in the 600 MHz incentive auctions, costing taxpayers $19 Billion dollars, when you are trashing the existing allocation of 25 MHz below 1 GHz? Shouldn’t you do a better job managing the unlicensed spectrum that already exists rather than looking to take spectrum designated for auction off the table?”
Sure, it’s a rather silly argument. But it has just taken me 3000 words to explain why that is a silly argument (with 2000 more to go to get to the answer). On the surface, if you are not a regular viewer of Spectrum Wars, this argument makes sense. Given the ridiculous arguments that the House Rs have made so far against allowing any unlicensed in the reclaimed broadcast spectrum, I see no reason why they wouldn’t make this one. Heck, given the significant number of parties arguing for more testing in the record, it’s a very easy argument to make.
The counter argument is that if unlicensed users can stop Progeny from lighting up its system, it’s going to re-ignite resistance by fed users to sharing spectrum. “Sure,” say Fed users. “We let you guys start sharing and then you start taking over.” While I acknowledge that’s a concern, I am somewhat less worried about that one because the Fed users don’t want to share. So either way, the FCC is going to be facing a Progeny argument. If the FCC doesn’t lets Progeny light up, then the feds will say: “See, there is no sharing. As son as you guys get a toe-hold in the band, you want to take it all for yourselves.” But if the FCC allows Progeny to light up, the Feds will say: “why are you coming after our spectrum? You should take better care of the stuff you already have.” The car manufacturers, who don’t want to see unlicensed expanded in the 5 GHz band, will argue that they are part of the same intelligent highways as M-LMS, and therefore deserve to be privileged as against unlicensed users.
So What Should the FCC Do? Surprise! I Have an Opinion.
So the FCC, as is so often the case, finds itself in a no-win situation. On the one hand, if they act, they get accused of putting smart grid out of business. If they don’t act, they are killing innovation and investment in new licensed services. The best outcome is that the Wireless Bureau allows Progeny to light up under some set of conditions that minimize interference, interference is minimized, and everything works fine. In which case, no one will notice give the FCC credit for anything.
So what is a poor Wireless Bureau to do?
As you might imagine, I do not ask the question rhetorically. I actually have some recommendations for the FCC.
Let me start by restating my own prejudices and perspectives. As regular readers know, I’m a big fan of unlicensed. At the same time, I recognize that licensed stuff has an important role in the wireless ecosystem (at least for the foreseeable future). For years, I’ve advocated that we need to move to a “share the pain” approach where all users of the spectrum have to learn to work cooperatively and regard interference mitigation as a shared responsibility rather than the “I’ve got mine, Jack, and I’ll fight to the death to keep you out.”
The problem of Progeny is basically the problem we always have repurposing spectrum. No one using the spectrum now wants anyone new to come in and potentially mess things up. Since this is people’s actual livelihoods on the line, they are not inclined to take a dispassionate view. If the spectrum they use gets messed up, they go out of business, or incur heavy costs. The best thing that can happen from an incumbent user’s perspective is that nothing bad happens. So the rational spectrum incumbent user doesn’t take the policy wonk view, the rational spectrum incumbent fights for its life.
The FCC needs to recognize this and stop hoping that somehow engineers can rise above that and come to an agreement based on engineering realities. Engineers are human beings like anyone else. The existing user community will always want conservative safeguards that the new entrant will always regard as ridiculously expensive. We just saw this with DISH v. Sprint in the AWS-4 v. H Block fight. We saw this with Lightsquared. We saw this with TV White Spaces. We saw this with M2Z.
Engineers are not special. They are human beings like everyone else. Get over it.
Notice and Comment Decisonmaking Is Your Friend
The FCC has to embrace that it will make hard decisions and that parties will not come together on the engineering issues. The way to protect yourself and make progress is to give up on the strategy of pushing the sides to work together and admit we are in an adversarial proceeding where the FCC needs to make the call.
What I think the FCC needs to do, as I have repeatedly said in this proceeding and elsewhere, is to figure out what it means by “unacceptable levels of interference” and what test (or set of tests) would determine whether Progeny will or won’t cause unacceptable interference. You put that out on notice, take input, decide what changes need to get made, issue a final decision then run the tests. The new entrant either passes the test, or doesn’t.
That’s it. You don’t argue about the test after you run it, because the entire point is to construct a set of tests that give you reason to believe you can move forward. Without this certainty, the debate always becomes about the test itself and the results of the test. Engineering is probabilistic, and the real world is hideously complicated. As a result, parties will always find things to argue about. Doing notice and comment proceedings that clearly define the definitions and the goals makes the process transparent. That won’t make the parties happier, but it does mean the FCC can protect itself from allegations that it is futzing around with needless delay or cutting corners. In a no win situation like this, the only way to protect yourself is to use rigorous process as a shield.
No one likes this, because notice and comment proceedings take a long time and requires the agency to make tough decisions. In fact, Progeny has made it clear they totally hate this suggestion and that they ant to move forward already. Notice and comment processes also mean having less flexibility. Parties will still do their best to politicize the process. And, at the end of the day, there is still no guarantee the FCC won’t guess wrong. So what’s the advantage?
Following notice and comment consistently provides the agency with protection by enhancing its legitimacy and warding off criticism that it makes decisions based on politics or prejudices rather than engineering. To go back to our hypothetical attack by Rep. Walden that the FCC “contaminated” the 900 MHz unlicensed band so it shouldn’t allow unlicensed in the guard bands in the 600 MHz Incentive Auctions, the FCC can say in response “we did not contaminate the 900 MHz band. We acted cautiously and transparently to find ways for multiple users to co-exist. We understand some people are unhappy, but we have a set of transparent procedures that make sure this is decided on the basis of engineering data.”
Planning for when Things Go Wrong
The other way for the FCC to protect itself (and manage the spectrum responsibly) it to have some safety nets in their to address what happens if things go wrong. Because the reality is that even with the best engineering experiments in the lab, and the best open and transparent process in the world, actual deployment in the real world can still create lots of surprises. In the TV white spaces proceeding, we talked a lot about how the use of the database created a “kill switch,” so that if we were wrong and use of the TVWS did cause interference the FCC could shut it down.
In the case of Progeny, I’ve proposed two precautions. First, the FCC ought to require Progeny to do a phased-in roll out. Choose a market, go live, watch for awhile to see what happens. If it works out, light up the next market. Have clearly defined criteria for what circumstances would prompt the FCC to either halt the roll out or shut down Progeny until the problems get resolved.
Second, the FCC ought to consider what happens if Progeny interferes with the existing LMS systems. In theory, M-LMS is not supposed to cause interference with existing LMS systems. I have three words: Sprint Nextel Rebanding. Stuff can go wrong, and the FCC’s processes for resolving interference issues always breaks down when you have co-primaries operating within the scope of their service rules causing interference to each other. A little forethought now can solve big problems in the future.
None of this is easy, and no one likes delay. New entrants fume at yet-more-delay and the costs, with good reason. Incumbent users worry about interference screwing up their livelihood, also with good reason. That’s why the road map out of this is to have a process and stick to it, accompanied by safeguards that let you pull things back if the reality on the ground evolves in a very different direction than it does in the lab.
It’s a pain in the ass, but there are no short cuts. That applies here to the Progeny Waiver, and it applies for every effort to repurpose spectrum.
Otherwise, we will keep having new seasons of Spectrum Wars, with the same plot but different players.
Stay tuned . . . .