Last time on Spectrum Game of Thrones (hereinafter “SGoT”) I spent 6500 words discussing the first two edges of the sword. The Wi-Fi dependent side has strong reason to suspect the LTE-U crowd of either reckless indifference or actual malice toward deployment of Wi-Fi based streaming services in the newly refurbished U-NII-1 band up in 5 GHz. Even if the Wi-Fi Dependents could trust the motives of the LTE-U crowd, what happens if everyone is wrong about the ability of the two technologies to co-exist? Under the current structure, the Wi-Fi dependents would be screwed, and they could do nothing about it. So the rational Wi-Fi Dependent must fight tooth and nail against deployment of LTE-U.
It doesn’t help that the Wi-Fi Dependents know that this is an utterly impractical solution for the long term. Unless there is a way to answer the two questions central to the survival of Wi-Fi streaming in U-NII-1 in the face of LTE-U (what happens if something goes wrong, what happens if somebody deliberately does something bad post-deployment), rational Wi-Fi dependents have no choice but to fight deployment.
The LTE-U crowd, for its part, has good reason to want to deploy LTE-U and has a legitimate gripe that Wi-Fi Dependents cannot keep saying no without defining the conditions for yes. If we admit the possibility that we can deploy LTE-U consistent with reasonable use of Wi-Fi (which everyone does), then there has to be some way to actually deploy it. And while I savor the fine irony of seeing licensees in the same position I have been in countless times, it is still crappy policy. Also, unlike me and other would-be new entrants, the wireless guys and Qualcomm have enough political muscle to make the current stalemate untenable. Eventually, they will get to deploy something.
Which brings us to the third edge of the Vorlon sword of understanding – the FCC. As I shall explain below, government actually is the solution here. Not by imposing a standard or a rule, but by providing both sides with a process for resolving the problem. As a happy side effect, this will also help resolve the general class of problems that keeps coming up on how to manage more and more intense use of the airwaves. Just like we all learned in high school math, and most of us forgot about 30 seconds after the exam, you solve an intractable problem by trying to break it up and simplify it into solvable problems.
The only problem is, and I know most people are not going to believe me, the FCC actually hates asserting and clarifying its authority. Yes. Really. Which gives rise to the question of whether the FCC actually has the willingness to do what needs to be done and create a general solution, or if they will continue to try to do the minimum possible, what I call the “Snow Goons Are Bad News” approach immortalized in this classic Calvin and Hobbes strip.
So, as we get to SGoT 2: Storm of Spectrum Swords, we come to another dramatic turning point. Will the Wi-Fi Dependents and the LTE-U Dependents see the wisdom of allowing the FCC assert authority over the land of Spectrumos? Can the FCC be persuaded to fulfill its destiny and its duty? And will the anti-Regulatory Zombies from beyond the Wall crash the party and devour both Wi-Fi and LTE-U because of their hatred of the FCC?
More below . . .
It will not surprise regular readers that I will once again refer folks to George A. Akerlof’s seminal paper “The Market For Lemons.” Go ahead. Look those up. I’ll wait.
So How Did The FCC Get Involved In All This?
As I mentioned last time, the LTE-U fight brewed pretty much all through 2014. By the time we got to the start of 2015, the WiFi Dependents (including yr hmbl obdn’t blogger and other public interest folks who have long supported unlicensed spectrum access) had reached a point of franticness. Qualcomm, T-Mo and VZ made it clear they planned to deploy their version of LTE-U as quickly as possible, without even waiting for the 3GPP standard – despite the fact that Qualcomm controlled the 3GPP process and had pretty much shut out IEEE and Wi-Fi Forum at that point.
So that meant going to the FCC, but that involves one of the classic Catch-22s of the regulatory world. Even though the possibility of bad stuff happening was pretty clear, no one was actually doing anything yet and nothing bad had actually happened. The FCC doesn’t generally get involved with industry hypothetical stuff (with good reason most of the time), and generally avoids getting involved in standards issues — lest it be accused of trying to “intrude on industry self-regulation” and “expand its authority.” So it was too soon to get the FCC involved, because nothing bad was happening yet. But at the same time, if one of the carriers went and began to deploy, it would create “facts on the ground” and the FCC would not want to act in a way that “disrupted settled business expectations” and “did not violate any rules.”
In other words, once anything actually started happening that would justify FCC action, it would be too late for the FCC to act. This is fairly common in D.C. It is always either too soon or too late to regulate anything – no matter how bad the consequences of inaction might be. Apparently, 5:27 a.m. on Thursday last, when the stars were properly aligned and Mars was the dominating influence in the House of Congress (aka “the Bullshit Bearer”), was the one time you could regulate. Now the feng shui is wrong and regulating would disrupt the industry chi and create bad karma.
Happily, the FCC was getting ready to wrap up the latest phase of the 3.5 GHz band. Some of you may remember this from last January’s Spectrum Wars Season Preview. The 3.5 GHz band is supposed to be the target for the next stage of evolution for spectrum sharing and the implementation of the PCAST Report recommendation on sharing spectrum between federal users and commercial users. Existing Fed users have priority over anyone else. The remaining 150 MHz of space is divided between General Access Licensees (GALs) and Priority Access Licensees (PAL). PALs pay money for short-term exclusive use (3 year license). GALs are basically like TV white spaces unlicensed, but are technically “licensed by rule” under Section 307(e) rather than authorized by Part 15 as pure “unlicensed” space. While “licensed by rule” and “unlicensed” are functionally indistinguishable to the lay person, this makes a significant legal difference.
If that makes no sense, remember this is SGoT. Law is like magic. It works by its own funky rules that even the special class of learned geezers who practice it don’t fully understand or agree on. So just take it from me that this difference matters.
The 3.5 GHz band was a likely target for QCOMM and Friends to deploy their flavor of LTE-U. Not only was it going to become available, but it fit QCOMM’s general goal of screwing up the PCAST approach. Ever since the FCC opened the 3.5 GHz proceeding in 2012, QCOMM and friends have pushed to make it as much like regular licensed and as little like anything actually new or potentially disruptive to the Established Order as possible. But suddenly, at the end of 2014/beginning of 2015, QCOMM and Friends were all interested in getting 3.5 GHz moving and capable of supporting LTE-U.
So Public Knowledge and friends asked he FCC to clarify that it would not allow LTE-U in 3.5 GHz or, if it did, it would not be linked to bands outside the 3.5 GHz band or interfere with GAL. You can see our original filing on this here.
Wait, Didn’t You Say in Part I That Banning LTE-U Would Be Bad?
I said banning LTE-U from unlicensed would be bad. But the 3.5 GHz band wasn’t for unlicensed. General Access Licensees are licensed by rule. The FCC sets interoperability standards and interference protection rules for licensed services all the time. This made it much easier for the FCC to do something in 3.5 GHz.
You Realize That Makes No Sense Whatsoever.
Did it make sense in Game of Thrones that ancient mage geezers could make 5 times as much explode-ey liquid as usual because, on another continent an ocean away, some blond chick is wandering around with baby dragons?
Point. But Do Not Take This SGOT Things Too Far. No “Red Open Commission Meeting!” What happened next?
As soon as we Public Interest Spectrum Coalition (PISC) folks broke the ice, all heck started to break loose on this issue in the 3.5 GHz proceeding. Wile this debate was going on, and the FCC staff and even the Commissioners were getting educated on what the heck this “LTE-U” thing was, the spectrum hit the fan big time at the Mobile World Congress in Barcelona Spain in March of ’15.
MWC is a huge deal, with mobile carriers and vendors from all over the globe. It also attracts regulators from around the globe, like FCC Chairman Wheeler. With coordination between the IEEE and 3GPP at an all time low, and tension about it among the WiFi Dependents running at an all-time high, VZ and T-Mo announced that they planned to deploy QCOMM’s pre-standard LTE-U/LAA flavor within the next few months as part of overall network deployment upgrade.
The Wi-Fi Dependents saw this as the final declaration of war and a complete rejection of even a pretense of interest in working together to protect Wi-Fi. They reacted pretty much as you would might expect – by throwing a total collective shit fit. And it wasn’t just one company like Google or one vendor like Cisco up in arms. It was pretty much every single Wi-Fi Dependent announcing as loud as possible that QCOMM and LTE-U/LAA were trying to kill Wi-Fi to squash the emerging Wi-Fi first competition. You can see good coverage of the whole mess by Dave Burstein over at 5G News.
Needless to say, none of this was lost on attending Chairman Wheeler, who politely asked why the entire mobile world decided to dissolve into a collective freak out, complete with apocalyptic predictions and hysterical finger-pointing. In the babble of conflicting answers, accusations, counter-accusations and denials that followed, Chairman Wheeler realized that – just maybe – this LTE-U stuff was important and the FCC needed to give it a closer look.
As an aside, I actually take it as a sign of T-Mobile’s (and VZ’s) innocent intent that they apparently had no idea that their announcement was going to trigger this kind of reaction. They appeared utterly flabbergasted that the Wifi Dependents saw this as a life or death struggle to save themselves from the spectrum equivalent of an invasive species like Frankenfish in the Potomac or kudzu in the American south.
So With the FCC All Alerted And All Active And Stuff, What Did the FCC Do?
The FCC actually analyzed this in two different ways. First, what were the potential impacts on the 3.5 GHz? After all. 3.5 GHz was the actual proceeding in which this got raised. But of course, the FCC was also thinking about the issue of LTE-U/LAA generally.
From the FCC’s perspective, this looked a lot like the interoperability problem that made it so hard for competing carriers with 700 MHz A block spectrum to get equipment and deploy competing services. Commissioner Clyburn, who as acting Chair brokered the deal that resolved the 700 MHz interoperability problem, gives a good summary in her concurring statement on the 3.5 GHz Order of the issue and similarities in the 3.5 GHz context.
For those to whom the above was gibberish, here is the short version. The 700 MHz band has a bunch of different blocks of licenses (which is fairly common). The most useful were the lower 700 MHz A block, B block, and C block. VZ got C Block. AT&T got a big chunk of B block and pushed in post auction sales to get a national 700 MHz B block footprint. Smaller providers had mostly A Block. 3GPP developed a standard that would work for A&B block, but they also developed a standard for just B block and another that would work for just C block. VZ ordered equipment that used the C Block only standard, despite owning some A block licenses (which they later sold). AT&T ordered equipment that was just the just B block standard.
This meant that no one bothered to mass produce equipment that would work on A block, because without AT&T and VZ the market was too small to interest any major manufacturer or get economies of scale. That made it effectively impossible for the A Block licensees to deploy and compete. So the A Block licensees petitioned the FCC to require that any handset that works on any 700 MHz frequency band must work on all 70 MHz frequency bands. That would require AT&T and VZ to switch to the 3GPP standard that worked for A&B block, thus creating usable/affordable equipment for the A Block licensees. The A Block licensees also claimed that AT&T and VZ were using their power in 3GPP to manipulate the standards process in an anti-competitive fashion. Needless to say, AT&t and VZ denied these accusations and opposed any interoperability mandate.
While the FCC issued a notice of proposed rulemaking on 700 MHz interop, it never actually issued a rule. Instead, it sent an observer to 3GPP to make sure everyone knew the FCC was watching and would not tolerate any funny business, and pressured AT&T and VZ to cut a ‘voluntary’ deal.
For the FCC in 3.5 GHz, therefore, the big danger was that the carriers would deploy LTE-U/LAA dependent on having CMRS licenses before anyone could deploy technology that would work in 3.5 GHz on a stand-alone basis. One the carriers achieved dominance – especially with QCOMM LTE-U’s reputation as a “Wi-Fi killer,” it would dissuade an rival equipment manufacturer from trying to develop a rival technology. So the FCC imposed an obligation that all 3.5 GHz equipment had to be capable of operating on a “stand alone” basis in 3.5 GHz. i.e. any equipment that works in 3.5 GHz has to be at least capable of transmitting and receiving in 3.5 GHz alone. While you can put capacity to work on other frequencies in the device, you can’t develop a chip or a standard (like LTE-U/LAA) that won’t work in 3.5 GHz unless you also have a license in some other band.
So you can still use LTE in 3.5 GHz. You just can’t use LTE-U/LAA.
How Could Just Having the Wrong LTE in 3.5 GHz Keep Anyone Else From Developing and Alternate Technology, Like Wi-Fi For 3.5 GHz?
Those who don’t know the industry economics and history scoff at the theory that dominance of a standard and equipment can have such dramatic effect. “Surely someone would develop a technology for Wi-Fi in 3.5 GHz is they thought it would work!” They then point to things like how Microsoft was once dominant in desktop and then technology changed and blah blah blah, without pausing to consider possible differences between the examples cited and the issues here.
What this misses is that chip manufacturing at the scale needed to make mobile phones and other consumer electronic equipment requires HUGE upfront investment. That makes manufacturers capable of making these chips very nervous about going up against an established standard or equipment market. Look at what happened to anyone who invested in the capacity to make WiMAX chips. Anyone who spent that money lost big time in terms of sunk cost and market share when the market ultimately consolidated into LTE. So if the 3.5 GHz space became quickly dominated by LTE-U/LAA, which was the likely outcome given the presence of QCOMM and Ericson willing to make the equipment and T-Mo and VZ eager to deploy, it would dissuade any other major manufacturer from even considering investing in a rival technology. Why bother to spend the money and take the risk, after all, when there are other more lucrative markets?
So, given the experience in the 700 MHz band, the FCC figured it had enough evidence of how things could go down badly with LTE-U/LAA to take a precautionary measure in 3.5 GHz. But the FCC did not feel it had enough evidence to ban LTE-U from being used in 3.5 GHz on a stand alone basis. (They also declined to adopt our suggestion of a FRAND condition. *pout*)
What About Outside the 3.5 GHz? After All, You Said The Real Action Is In The 5 GHz U-NII-1 Band. What Happened There (Or Anywhere Else).
To understand the FCC’s reaction, you need to understand the institutional mentality of the FCC. Not the caricature version of the FCC as either a captured tool of industry or as a power mad regulator, but the actual way FCC staff and Commissioners tend to think about things.
The FCC doesn’t play any formal role in standards setting. It is, however, charged by Congress to regulate all use of “radio” (meaning spectrum) in the U.S. So the FCC needs to keep informed about everything that’s going on, because when major spectrum stuff starts to go down that people don’t like, they run to the FCC.
The thing is, contrary to the constant bitching and moaning from Libertarian think tanks, the Republican Commissioners, and industry folks who dislike whatever the FCC happens to be doing at the moment, the FCC generally does not like to get involved with stuff. The FCC is underfunded and already has more than enough to do without inventing new problems for itself. So when people bring exciting new problems to the FCC, the FCC is always interested in learning about them — but is extremely anxious about actually doing anything about them. The FCC would always, always always prefer that industry folks work issues out among themselves.
Mind you, that’s not a bad instinct generally. You don’t want the FCC to become a “moral hazard” where folks paralyze industry standard making by threatening to call in the FCC every time there is a disagreement. But at the same time, it means that when you have real problems that would benefit from the FCC taking action, the FCC is extremely reluctant to do anything. Since everyone in the industry knows the FCC does not want to act, and that it can be easy to paralyze the FCC by accusing it of “seeking to expand its authority,” this encourages bad actors that can leverage power in industry standards bodies to do so without fear of FCC consequences that you would think would temper bad behavior.
Libertarians and Republicans who have never gotten over the initial insight of public choice theory utterly refuse to believe this. They respond with the catechism that federal agencies always want to expand their jurisdiction because this invariably leads to bigger budgets, more prestige, and enhances the value of agency staff when they leave the agency to cash in, because only they will know how to manipulate the vast overarching regulatory apparatus they have erected. This, of course, makes it so easy for bad actors to manipulate them into making the accusations that paralyze the FCC.
If you repeat the above to any actual FCC career staffers, you will get the sort of hysterical laughter/sobbing combination you get when you ask abuse victims why they don’t just leave their abuser. FCC staffers know whatever they do – or decline to do — will get people mad. The FCC can’t even run a voluntary survey without Republicans trying to score political points by claiming it’s the return of the Fairness Doctrine. Rather than getting rewarded, agencies that actually do anything get punished big time. As the current Chair of the House Appropriations Committee told Chairman Wheeler directly, House appropriations keeps cutting the FCC’s budget to punish it for regulating and to get it “to do less with less.” Further, as a “reward” for Title II reclassification, the House Appropriations Committee has slashed the FCC’s budget again, along with a bunch of other riders to slap them around.
So while the basic insight of public choice doctrine – that what decisionmakers at agencies perceive as in their best personal and institutional interests influences their decisonmaking – remains true, Libertarians and Republicans generally have utterly failed to actually apply this insight to the incentives to decisionmakers as they exist today. Rational agency employees never want to expand their authority – or do anything controversial. It is not so much about reward, but about avoiding punishment. To understand how the FCC agency staff think, imagine all those experiments you read about in high school where people administer electric shocks to mice. FCC staff pretty much work in a constant state of trying to do their job while expecting at any minute to get punished for dong their job, while being endlessly told what powerful awful monsters they are for either acting or not acting.
The reason the agency actually ever does anything is that most people who go into government service do so because they want to do good things, and because they believe in the mission of the agency. So you will generally see FCC staff trying to do their job and trying to address the problems people bring to them, but wanting to do it in a way that involves minimum action, minimum controversy, and as much industry buy in as possible.
Yo! Monologue About Updating Public Choice Theory Later! Focus on the LTE-U Stuff.
By the time the FCC got to the 3.5 GHz Order in April, between the spectrum hitting the fan at MWC2015 and the record in the 3.5 GHz proceeding, the FCC knew that something looked really, really funky about QCOMM and the LTE-U Gang. At the same time, it wasn’t really clear what the FCC should do about it. As I noted at great length last time, they don’t want to ban LTE-U, since the FCC does not regulate the unlicensed space (and, in any event, that is the wrong policy). Besides, the fact that things looked all suspicious did not mean things were actually as bad as folks were saying.
So, consistent with seeing this as being like the 700 MHz interoperability issue, the FCC treated it the same way. Overall, there strategy has been to slow things down and nudge the standards process back toward cooperation between 3GPP, IEEE and Wi-Fi Forum, while pressing the various industry participants to “work it out.”
Specifically, the first thing the FCC did was open a Notice of Inquiry (NOI) to “gather information” on LTE-U/LAA. The NOI itself was vanilla-neutral in terms of not taking any sides, but the opening of the NOI itself sends a message to everyone familiar with how this works. Furthermore, the questions in the NOI point out the chief areas of potential FCC concern. Notably, the FCC asked about the state of coordination between the Wi-Fi standards bodies and 3GPP on LTE/LAA, exactly what flavors of LTE and LTE-U/LAA did the industry look to deploy, how parties were determining that LTE-U of any flavor would play with Wi-Fi, and how industry intended to deploy any version of LTE-U – particularly any “pre-standard” version.
Finally, the FCC started sending an observer again to 3GPP. The FCC (and other national and global regulators) are always free to send observers to 3GPP, and the FCC has done so in the past. Notably, the FCC did this back during the 700 MHz interoperability fight to assess whether AT&T and VZ were leveraging their power at 3GPP to prevent development of an interoperability standard.
To us FCC watchers, this combination of signals from the agency sends a message that the agency is worried something funny is going on and wants it to stop. It is a Level 1 nudge to the industry to get back on track and work things out without regulatory intervention. It very deliberately does not take a side, but does warn those insisting that all is well or refusing to engage with others that the FCC is not buying the happy story so stop being jerks and start playing nicer – or else we might need to do something at some point.
But Things Seem To Have Escalated, Rather Than Calming Down.
As I mentioned in Part I, Qualcomm and friends at the LTE-U Forum displayed an absolute genius for inflaming the situation by acting in exactly the way you expect people with something to hide to act. First, they rushed to release a stand-alone version of LTE-U called LTE-U MuLTEfire. MuLTEfire had all the hallmarks of a rush job intended to defang an issue (the fact that LTE-U/LAA restricted LTE-U to mobile carriers) with no serious effort at support or intent to deploy (and, as an added bonus, proprietary to Qualcomm). Qualcomm and friends filed comments assuring everyone that everything was just wonderful, LTE-U was the product of lots of standards body coordination, and generally trying to pretend that the entire history I spent 4000 words on in Part I never happened. The LTE Dependents also explained how they had done lots of testing to show that LTE-U/LAA would totally play nice with Wi-Fi.
In response, IEEE and Wi-Fi Forum filed comments that the LTE Dependents were totally lying, there was no effort at coordination. The Wi-Fi Dependents slammed LTE-U Forum for its secretive processes and efforts to control LTE-U development. The LTE-U Forum responded they were shocked and outraged with the Wi-Fi Dependents, that the Wi-Fi Dependents ignored all the stuff LTE-U Forum published, and that the Wi-Fi Dependents were just trying to stifle competition. The Wi-Fi Dependents responded by releasing their own testing data that showed LTE-U would totally destroy Wi-Fi if deployed in the 5 GHz U-NII-1 band. You can find all the relevant comments in the FCC’s Electronic Comment Filing System (ECFS) here. For a good summary in the replies from someone you’ve never heard from, see these reply comments from by Leigh Chinitz. (I have no idea who Leigh Chinitz is, but these reply comments do a good job of summing things up in a fairly neutral way.)
All of this is pretty standard in FCC Land. It generally produces a stalemate. But Qualcomm also did two things that made the FCC worried enough to go to Level 2 Nudge – aka “serious pushback.”
First, in April 2015, after the FCC made it clear it intended to get interested in LTE-U/LAA, Qualcomm started a pushback campaign called “Why Wait.” The “Why Wait” campaign promised unlimited awesomeness if nasty bad-bad bureaucrats and tech-stifling rivals would just get out of the way.
Second, Qualcomm and friends began to say very loudly that since this was the unlicensed space they could, under the rules, totally deploy LTE-U of any flavor any time they wanted and the fact that they were still being all nice and sharing information and testing showed what wonderful guys they were.
I can think of few ways Qualcomm could have antagonized anyone more. They were effectively saying to the Wi-Fi Dependents (and the FCC): “we don’t have to play nice if we don’t want to, we can totally deploy whenever we want, and we don’t think the FCC has the means or the balls to stop us. So take the co-existence mechanisms we’re offering or f— off, because there ain’t nothing you can do about it.”
How Did The FCC Respond?
The FCC and the Wi-Fi Dependents took it to Level 2. Notably, a bunch of Senators sent a letter with the FCC calling for greater oversight by the FCC of the LTE-U/LAA process to ensure that it didn’t mess things up. A week later, the FCC Office of Engineering and Technology sent a follow up letter to the LTE-U Forum with a bunch of follow up questions. Specifically, after summarizing the situation in highly technical terms, noting that a bunch of different flavors existed, OET asked for more information over how “key parameters” in the primary coexistence mechanism touted by LTE Dependents, “carrier-sensing adaptive transmission” (“CSAT”). You can read more on CSAT in the old Qualcomm white paper describing the “harmonious existence” between LTE-U and Wi-Fi here. Cablelabs has a blog post on why they don’t trust CSAT here.
Most pointedly, the OET letter asked why Qualcomm and friends were in such a huge rush to get a pre-standard version out, given that everyone in the carrier world usually stresses the critical importance of waiting to deploy a standardized version because of the importance of the global market, economies of scale, and everything else I mentioned in the “standards bodies” section of Part I.
What OET basically was telling Qualcomm and the other LTE-U dependents, through the polite question language, was “the FCC does not believe the happy happy joy joy story of harmonious cooperation and mutual respect. Now go back to the standards process and start behaving or you will get in big trouble!”
But Who Cares What The FCC Thinks? Isn’t Qualcomm Right That Under The Rules, They Can Do Whatever They Want and Screw The FCC and the Wi-Fi Dependents?
Not quite. As the OET August 5 letter reminded Qualcomm and the LTE Dependents, the FCC does have fairly broad authority when it wants to use it. In this case, before carriers can deploy any new Part 15 equipment – such as for any flavor of LTE-U – the equipment must be certified as complying with all necessary FCC regulations. As the letter pointed out, Qualcomm cannot manufacture, import or sell any LTE-U equipment until it gets certified by the FCC as compliant with all rules set forth in 47 C.F.R. §15.
As the OET August 5 letter also reminded Qualcomm & Friends, “Telecommunications Certifications Bodies must consult with the FCC Lab before certifying such equipment. The Lab will request a full technical description of how the device will operate and we plan to require submittal of sample devices for testing. This is our standard process when dealing with new technologies.” While no one could possibly characterize this polite reminder as a threat or demand, no one in tune with the subtleties of FCC approval would miss the implication of this. The FCC was saying “yes, we can stop you from deploying this if you decide to try to bulldoze this through, so wouldn’t it be better for everyone if you stopped behaving like total jerkwads and decided to genuinely cooperate with the rest of the community?”
Did the LTE Dependents Get The Message?
Pretty loud and clear. The carriers pretty much told Qualcomm to shut the $#@! Up and stop pissing everyone off (a good start). They filed a substantive (rather than bullsh**) response to the August 5 OET letter on September 9. If you click through, you will see the tone is positively collaborative. The carriers (and Qualcomm) also replaced the “Why Wait” campaign with something called “Evolve”. You can see the EvolveMobile website here.
Two things leap out right away about EvolveMobile. First, it’s not just Qualcomm or even just LTE-U Forum. It involves all the major carriers, the competing carriers, and their equipment manufacturers. The carriers have clearly understood that Qualcomm has NOT BEEN HELPFUL in reassuring people about the happiness that is LTE-U.
Second, EvolveMobile is much less about evil government regulation holding back the wonderfulness and much more about the wonderfulness. Evolve is a “love bomb” type campaign, with the major carriers reassuring everyone how much they totally want to work with everyone to make sure that unlicensed spectrum continues to grow and prosper and remain innovative. No bad mouthing Wi-Fi or the FCC. Just reasonable willingness to move forward and why can’t we all get along.
But It’s An Insincere Trap And We Should Totally Ignore It, Right?
Ummm . . . . not that simple.
But Evolve Is just About the LTE Guys Getting What They Want!! They Don’t Really Mean All That Love and Co-Existence Stuff, Do They?
As I always say, pandering in Washington is how you show respect to the other side. I am not interested in dating the carriers and I don’t need them to be sincere. The fact that the LTE Dependents feel compelled to frame this fight as about who loves unlicensed spectrum more (and we will ignore everything that happened before 2014) shows that they recognize a need to bargain, not just steamroll over the Wi-Fi Dependents.
Mind you, it doesn’t tell you how far they will go in bargaining. It could easily be a trap. But it’s the LTE Dependents that need to show their reasonableness and move the ball at this point, for now. By going all reasonable and no longer demonizing the FCC and Wi-Fi, the LTE Dependents force the Wi-Fi Dependents to demonstrate their reasonableness or risk looking like they are simply stalling for anti-competitive purposes and losing sympathy.
This is not lost on the Wi-Fi Dependents, who have – in response – laid out their demand that LTE-U adopt “listen before talk” (LBT) as the means to ensure that Wi-Fi and LTE-U coexist, and that 3GPP ultimately work with IEEE to adopt LBT in the upcoming 3GPP standard rather than CSAT, as currently proposed by LTE-U Forum.
Ummm . . . . You Lost Me There.
Listen-before-talk (LBT) means that before a transmitter starts sending signal on an open channel, it will listen to see if the channel is available. If no channel is available, then the transmitter waits some random period of time, then checks again to see if there is an open channel. The device will keep doing this until it finally sends a signal.
Wi-Fi uses LBT. This permits lots of Wi-Fi devices to operate simultaneously. As everyone has known since the early days of networking, most traffic is “bursty.” It works in spurts of data coming down the pipe – even if to human beings it looks like one continuous flow. This allows networks to share capacity and accommodate multiple users at once. For Wi-Fi, LBT allows lots and lots of devices to communicate at roughly the same time. Eventually, of course, you hit a limit, which is why we need to open up more spectrum for unlicensed use. But that’s a digression for another time.
As discussed above, LTE-U does not use LBT. It uses something called CSAT, which is proprietary to the LTE-U Forum specification. (Remember, LTE evolved for licensed use, so it never had to share a channel with anyone else before, so those favoring LTE over unlicensed needed to come up with a way to share the space with other users.) Normally, LTE-U (like LTE) just keeps talking over the available channel. But the CSAT introduces breaks in the transmission so that another signal can get a chance. This concept is called a “duty cycle.” LTE-U, using the CSAT algorithm, shuts up for some period of time to allow Wi-Fi or other entity to operate. Then the LTE-U transmission starts up again, potentially drowning out Wi-Fi until the next break in the duty cycle.
This is where the distrust lies, and why the Wi-Fi Dependents look at LTE-U as an invasive species capable of wiping them out. As long as control of the duty cycle remains a black box in a proprietary CSAT, and as long as the standards body capable of altering the CSAT standard (LTE-U Forum) remains a closed membership of Qualcomm & friends, there is simply no way for the Wi-Fi Dependents to trust LTE-U of any flavor.
But The LTE Dependents and the Wi-Fi Dependents Are Actually Negotiating Now, Right? The FCC Can Just Keep Pushing Them To Play Nice Together Like They Did With 700 MHz Interoperability, Can’t They? Just Stall Things Until 3GPP Works With IEEE and Resolves This With A Good Standard.
That is definitely what the FCC wants. Unfortunately, it totally won’t work.
Why Not? 3GPP Is Supposed To Get It’s Work Done On LTE-U/LAA by the end of March 2016. That’s Not So Far Away.
First, the LTE Dependents are not going to consent to an indefinite delay and moving goal posts. They have a trust problem as well. LTE Dependents have to worry that the Wi-Fi dependents will keep changing the rules, and will try to leverage their current upper hand to give them a competitive advantage. Just as the LTE Dependents cannot prove to a rational Wi-Fi Dependent that they will always play fair, since the Wi-Fi Dependents would have no rights if the LTE Dependents decided later to screw up Wi-Fi (or if it just screwed up on it’s own), the Wi-Fi Dependents cannot prove to the LTE Dependents that they are not trying to leverage the regulatory process for their own competitive advantage.
Keep in mind, I’m not saying that is what is happening. I’m saying that the rational LTE Dependent has to act as if the rational Wi-Fi Dependent can never be satisfied, so you never reach a deal on a standard with the rational Wi-Fi Dependent. Because even if you can agree on a standard, you still end up with the two unanswered questions:
- What happens if we guessed wrong on the standard and things go to Hell after the standard gets selected?
- What happens if, sometime down the line, the LTE Dependents change the standard to mess up Wi-Fi deliberately?
Until you can answer those two questions, the rational Wi-Fi Dependent cannot trust the rational LTE Dependent, therefore the rational LTE Dependent cannot trust the rational Wi-Fi Dependent.
This Is That Whole Akerlof “Market For Lemons Thing” Isn’t It?
Exactly. To refresh for those who don’t remember, George Akerlof’s “The Market For Lemons” explained why, in the 1960s and early 1970s, you could never find a good used car. Used cars are divided into two categories: good cars (called “cream puffs”), and cars with problems (called “lemons”). Because a buyer cannot determine if a car is a cream puff or a lemon, the rational buyer must assume the car is a lemon. Therefore, the rational buyer will only pay as much as the value of a lemon – even if the car is a cream puff. Because no one will pay what a cream puff is worth, no one who owns a good used car will put one on the market. So, despite the fact that everyone wants to buy cream puffs, the only cars available are lemons.
Same dynamic occurs here. Because the Wi-Fi dependents cannot tell if the proposed LTE-U is a “creampuff” (a technology that coexists with Wi-Fi) or a “lemon” (a technology that will cause destructive interference with Wi-Fi, or wipe out latency sensitive applications), the Wi-Fi Dependents must assume that LTE-U will be a dangerous invasive species and destroy the Wi-Fi ecosystem. The only thing the Wi-Fi Dependents know for certain is that LBT works for Wi-Fi, so they want it incorporated into LTE-U. On the other side, because the LTE Dependents know that the Wi-Fi Dependents will always fight them, they do not trust that even adopting LBT will get them what they want.
Which brings us to the current stalemate. The Wi-Fi Dependents are extremely unlikely to get the FCC to require LBT. First, the FCC does not like to impose standards. While the FCC will force a tech standard if they have no choice (for example, as they did with DTV), they really, really really hate doing that for all the reasons I described above. Additionally, as I said at the end of SGoT Part I, the FCC does not want to impose a standard on unlicensed. Yes, there are many differentiating factors in this specific case. I can (and have) argued that because licensees have licensed spectrum available to them and have incentive to and capability to undermine Wi-Fi first services, the Commission should view this as regulating licensees not regulating the unlicensed space. But while that is a fine argument for a lawyer, it does not sit well with the engineers at the FCC.
On the other hand, the LTE Dependents need to get their devices through testing in order to deploy. While they can force the issue, and then take the FCC to court if the FCC denies them certification for no good reason, that is very time consuming and expensive and something the LTE Dependents want to avoid. Additionally, given the deference showed to the FCC on spectrum matters, it’s not clear that the LTE Dependents will prevail.
At the moment, the stalemate favors the Wi-Fi Dependents. But the stalemate will not persist. Evolve is definitely gaining traction, and the longer this sits out there, the harder it becomes for the FCC to delay without requiring a clear path forward. Nor is it clear what happens if 3GPP adopts a standard in March 2016 but IEEE and Wi-Fi Forum remain unhappy. After all, there is no legal requirement for standards bodies to agree with each other. If 3GPP adopts a standard, says that hey believe it coexists with Wi-Fi and other unlicensed users, and submits devices for certification, what does the FCC do?
The upshot is we now have a stalemate where no one is happy and no one has a good way forward. It is possible that either the Wi-Fi Dependents or the LTE Dependents will ultimately come out ahead, but neither party has an advantage over the other at the moment and everyone has an incentive to see this damn thing resolved.
More importantly, even if Wi-Fi Dependents and LTE Dependents agree on a standard now, it leaves the two vital questions unanswered: What happens if something goes wrong when this ramps up; and what happens if, after a co-existence standard for LTE-U is approved, someone alters the standard?
So How Do We Get Out of This Mess?
Akerlof proposed a simple mechanism for resolving the problem in Market for Lemons. He did not try to define what a “creampuff” was or what a “lemon” was, try to force some kind of standard, or even prevent people from selling lemons. Instead, he proposed a mechanism by which if a consumer discovered she had bought a lemon, she could return it to the dealer get her money back. This became the basis for the various “Lemon Laws.”
The existence of a clear, easily executed remedy solved the problem – enough. No one would imagine that the used car market works perfectly today. But, on the whole, the existence of a remedy provides enough trust in the market for buyers to risk believing that a car is a creampuff and not a lemon, which means those trying to sell creampuffs can actually sell for the proper value.
How Does That Translate Into FCC Terms?
Well, I see we’ve come to more than 7,000 words, so it’s time to end Part II.
OH NO YOU DON’T!! I DID NOT JUST PLOUGH THROUGH 14000 WORDS FOR YOU TO LEAVE US AT ANOTHER CLIFFHANGER SO YOU CAN WRITE ANOTHER SEQUEL!!!
OK, I will give you the short version/preview.
Section 333 of the Communications Act states: “No person shall willfully or maliciously interfere with or cause interference to any radio communications of any station licensed or otherwise authorized under this Act.” The FCC has already said that Section 333 applies to unlicensed spectrum by prohibiting Wi-Fi jamming.
The FCC needs to issue an Order clarifying that the words “knowingly” and “cause interference” also apply where someone knows that deployment of their technology/equipment would cause destructive interference to the unlicensed ecosystem. Obviously, that means more than just interference with a single device, or mere degradation of service generally associated with the deployment of existing Part 15 unlicensed devices. In addition, the FCC should clarify that it has the power under Section 324 (use of minimum power), Section 302a (equipment certification and marketing), various subsection of Section 303, and probably a bunch of other sections, to issue whatever orders are necessary to remediate the situation in the event that deployment of a new technology causes “destructive interference” (a term stolen from the [Progeny] rules) in the unlicensed ecosystem.
LTE Dependents would still have to show enough evidence before deploying LTE-U that there is good reason to believe this isn’t going to blow everything up, but they would not have to remain trapped in permanent testing Hell or adopt a specific co-existence protocol like LBT. Nor would this be a safe harbor. In the event an unforeseen disaster starts to occur, the FCC can force LTE-U to back off and play nice, even if that means reducing power or altering the duty cycle. But LTE Dependents could go ahead, subject to preliminary testing that their coexistence claims work out. If they alter the coexistence mechanism later – especially if the alteration appears designed to provide an anti-competitive advantage over the Wi-Fi dependents – the FCC can act immediately to require the LTE Dependents to stop mucking with the co-existence mechanism. Since LTE Dependents know that they muck with the coexistence standard at their own risk, they cannot complain about being treated unfairly.
Nor does the rule only apply to LTE-U. It applies to any new technology that may come along in the future. Nor does it guarantee the success of any specific protocol. If LTE-U (or something) is actually better than Wi-Fi in the marketplace, then LTE-U will displace Wi-Fi the way Wi-Fi displaced Bluetooth and other technologies as the default for wireless LANs.
That Sounds Good In Theory, But It Raises A Lot of Questions.
I know, but we’re at over 7500 words here for Part II. So we will have to take up how the FCC could actually do this in a practical way in Spectrum Game of Thrones Part III – A Song of Licensed and Unlicensed.
This Is Going To 5 Books, Isn’t It?
Stay tuned . . .