My Insanely Long Field Guide to Lightsquared v. The GPS Guys

For some time now, I’ve been rooting for Lightsquared. Despite the fact that it faces tough odds trying to build out an expensive wireless network, a wireless network built from ground up for wholesale only could totally change the wireless market (which is entirely different from the mobile cellphone (aka the “commercial mobile radio service” or “CMRS”)  market, but that’s a rant for another time). But now, I just love the fight between Lightsquared and the GPS industry because it manages to contain everything that makes spectrum policy in this country like running a marathon with concrete blocks on your feet: bad neighbors operating critical systems so they can get away with being prima donnas, hostility from other federal agencies, unanticipated interference issues that crop up on deployment, and efforts to politicize the FCC’s technical process.

And, as always, a special guest appearance by a very tired looking Julie Knapp.

For a spectrum wonk such as myself, it simply does not get better than this. I also get one more real world example where I say to all the “property is the answer to everything” guys: “Ha! You think property is so hot? The rights are clearly defined here. Where’s your precious Coasian solution now, smart guys?” Which usually sends them back muttering that it’s not their fault no one in the real world follows the models that explain how it’s all supposed to work out in the world of rational actors and no transaction costs where unicorns frolic in the golden sunshine.

So, in the latest installment of my occasional “Insanely Long Field Guide” series, I take a lengthy look at Lightsquared, how we got here, and what I think will happen. Short version, ignore all the pseudo-Whitewater nonsense flogged by the conservative conspiracy theorists and complaints that the FCC bypassed their own process. So far, and I do not say this often so please pay attention, the FCC has behaved entirely appropriately, even intelligently. (Yeah, yeah, don’t let it go to your heads.) What matters is that the FCC is about to receive a report that confirms that, yes, when Lightsquared operates it system, it creates interference for existing deployed GPS systems. As a result, only the following things matter:

1. The Lightsquared folks are right about how the GPS guys knew this day would come and conveniently chose to do nothing. But in the short term it doesn’t matter, because the FCC will not allow anything to happen to GPS.

2. OTOH, if the GPS guys get their way, it means taking another 40 MHz of prime spectrum and rendering it useless forever. That also isn’t going to happen. That suggests a phase in/compromise.

3. Whether Lightsquared actually survives the compromise as a viable service will depend on a lot of things. The dimensions of any such compromise will depend on the interference tests. So while it is pretty clear from what’s been leaked that Lightsquared’s system as proposed causes interference with GPS systems, a lot of questions remain about what ought to happen to make it so that GPS and Lightsquared can live together in harmony.

At this point (from my wonkish perspective), the precedent of how to deal with annoying neighbors is almost more important than what actually happens to Lightsquared. If the GPS guys get their “sit on your rear-end veto,” then we can pretty much kiss off spectrum reform in the most useful spectrum bands. Every potentially useful band has neighbors that built systems on the assumption that nothing would ever change. So the FCC either finds a way to balance the interest of incumbents with fostering the expanded use we need for our expanding wireless  demand, or we forget about “spectrum flexibility” and resign ourselves to the current state of the universe pumped up by the occasional auction.

More below . . . .

Mike Marcus has written up a pretty good (pro-Lightsquared) history of the Lightsquared GPS controversy here. You can see the GPS argument on why Lightsquared will inevitably cause interference here. Finally, just about anything written on this by Stacey Higginbotham at GigaOm (like this piece here) is well worth reading.  But because understanding the story is important, especially given the increasing politicization of the process, so I spell it all out below.

Where Did All This Spectrum Come From? I Thought We Were Fresh Out.

We have bands allocated for Fixed Satellite Service (FSS) and Mobile Satellite Service (MSS). These bands have very nice physical properties that make them  extremely useful for modern wireless broadband networks. These same properties make them useful for devices trying to receive relatively weak signals from space, which is why satellite systems got there first. For years that was cool. Satellite serves a lot of useful purposes, especially when you need something global that reaches areas where no business can make a case to build out. Then came the crunch for wireless broadband, both fixed and mobile.

The satellite industry, and occasionally the FCC, like to pretend that satellites can do broadband. This is like pretending a Clydesdale can win the Triple Crown.  Satellite faces too many physical obstacles, from speed of light lag to the expense of upgrading technology, to offer what most folks would consider even vaguely adequate broadband. (Assume usually angry protestations by satellite industry inserted here.) Plus its expensive to provide as a niche service – hence the steady stream of bankruptcies in the satellite sector. So the FCC has been looking for over 10 years now to use the spectrum allocated to satellite more efficiently for wireless broadband and mobile services.

And thus was born into the world “Ancillary Terrestrial Component,” or “ATC.” ATC works on a simple principle. Satellite services communicate up to the sky and down to Earth. That, in theory, leaves space for communication systems on the ground to use the same spectrum without interfering with the existing satellite service. Imagine I could draw a diagonal dotted line from a stick figure going up to a satellite. Now imagine I am drawing different, wavy lines underneath the diagonal line.  That’s the basic idea.

Reality, of course, is far more complicated. Of particular relevance here, the energy from wireless communication does not stay neatly confined to the right frequency bands. Invariably, some leaks out. We call these “out of band emissions” (OOBE). OOBE creates interference with neighboring services. Change the nature of the service, from a satellite service to one with a significant terrestrial component, and you increase the OOBE, which requires your neighbors to rejigger their systems or may make it impossible for your neighbors to operate at all. To employ another of those bad analogies we use in spectrum policy, imagine you want to convert a residential area to a more densely packed commercial area. That may have lots of benefits for the community as a whole, but neighbors are likely to hate all the extra noise and traffic.

One of the satellite bands, called the “L Band” (don’t worry about why, it is totally arbitrary), makes a great candidate for ATC. It sits pretty close to one of the bands used for mobile voice and broadband, so you can pretty much plug in the right technology to work with existing systems.  If you believe in the “spectrum crisis,” the L Band represents about 40 MHz of prime spectrum going fallow. From that perspective, repurposing the L Band to provide much more intense use through ATC represents a gigantic improvement over the status quo.

Unfortunately, the L Band has a fairly important neighbor, the global positioning system, aka GPS. Inconveniently enough, GPS sits smack dab in the middle of L Band where the FCC and the satellite industry wanted to put all this relatively high-power terrestrial ATC. Needless to say, the GPS folks kicked up a ruckus about possible interference. In 2003, the FCC adopted an order (affirmed in 2005) setting rules for ATC in the L-Band. The rules set a limit on OOBE to protect GPS, which the GPS guys considered totally inadequate but were told “tough noogies, this is the trade off because we need to get more spectrum out for broadband.” Meanwhile, in an unrelated part of the order, the FCC also establishes a “gating criteria” that requires a satellite provider offering broadband through its ancillary terrestrial component to integrate its satellite service into the system. “This will make sure satellite actually survives,” said the FCC, “and the ATC won’t swallow up the actual service.”

Time passed. Various companies tried to make the ATC work. More satellite companies went bankrupt. Various companies asked for and got various waivers, but nothing seemed to really work.

Enter Lightsquared

Then came Philip Falcone, head of the Harbinger Hedge Fund. Falcone bought up a number of ATC rights and satellite companies to pull together a new venture now renamed Lightsquared. These transfers required FCC approval. The FCC approved the transfers in March 2010 subject to a bunch of conditions (Order here). Notably, Lightsquared (then called Skyterra/Harbinger) agreed to build out a wholesale LTE network. Lightsquared also agreed to a bunch of provisions that limit how much capacity it can lease to “the top two” cellular providers, aka AT&T and Verizon. Needless to say, Verizon and AT&T were less than thrilled with this last part and filed Petitions for Reconsideration (ATT, VZ). OTOH, Sprint and other competing wireless carriers looking for new backhaul and roaming partners were quite happy with the limitations and filed in opposition to the Recon (here and here) and in support of Lightsquared (as did my employer Public Knowledge and a number of other public interest groups loosely branded as “The Public Interest Spectrum Coalition” (PISC) (here)).

A few months later, Lightsquared came back asking for a new favor to get its network up and running. (Because of the way that proceeding is set up, it is not easily available through ECFS, so no links. Sorry.) Lightsquared wanted the FCC to approve its “dual mode” combined satellite/terrestrial wireless handset and plan to provide network services that were primarily terrestrial with some token satellite service thrown in to satisfy the “gating” requirement.  AT&T and Verizon, as expected, showed up to oppose anything that might enhance competition in the wireless space. Unexpectedly, however, a bunch of GPS guys showed up to complain that the proposed Lightsquared system would overload neighboring GPS systems.

Lightsquared, for its part, admitted that, yeah, when they had turned on the switch for some test systems, some unanticipated problems emerged, but the FCC should not trouble its pretty little head about it because Lightsquared had it totally under control. GPS guys argued that Lightsquared ought to be shut down immediately or planes would start dropping out of the sky and cars would all start going crazy and drive off bridges and into deserts. Lightsquared countered that the GPS guys had known for 8 years this day would come and if the FCC held up deployment, Lightsquare’s funding would disappear, lots of spectrum would go to waste, rural America would never get broadband, and the mobile wireless market would remain an uncompetitive backwater where T-Mobile and Sprint would wither away and die (insert ATT-TMO joke here).

The FCC, to its credit, actually made (in my opinion) a smart set of choices in this much maligned Order here. First, the Bureau rejected Lightsquared’s fig leaf of “dual mode,” recognizing that it might as well ditch the whole gating requirement if it accepted what Lightsquared proposed as satisfactory. On the other hand, the Bureau recognized that Lightsquared potentially provided something critical to the wireless market – a wholesale only provider built from the ground up with no intention of ever providing retail services. One has only to look at the number of companies such as Best Buy willing to sign deals with Lightsquared even before it’s built to understand the pent up demand for wireless broadband that is not inextricably tied up with one of the major carriers. So the Bureau decided to grant Lightsquared a waiver of the gating requirement, subject to Lightsquared sticking to all the good stuff and build out requirements to bring the service out to rural places.

Then came the GPS interference stuff. The FCC decided to allow Lightsquared to go ahead with the build out, but would not allow Lightsquared to actually begin operations until it satisfied the FCC that Lightsquared would not cause GPS problems. The FCC reasoned that it will take a while to build out the system, and that if everything comes to a full stop while wrangling over the interference stuff, deployment will be seriously delayed and funding may well go away forever. On the other hand, it will take enough time for Lightsquared to build out that the FCC can determine whether the GPS problems are real or not, and whether Lightsquared and/or the GPS guys can fix them. The FCC told Lightsquared to set up a committee that included representatives from the federal agencies and GPS receiver folks to go study the problem and come back with a report (due today) on interference testing that shows whether there is really a problem and, if so, to make recommendations for fixing it.

All of which brings us up to the present. On the engineering side, we have Lightsquared and the GPS guys doing testing, with the usual fights between incumbents and new entrants on whether the tests are a total failure showing that switching on the system means instant death or whether the tests are a total success showing the GPS guys are full of baloney. On the other hand, we have the current political and procedural circus, with GPS guys whipping up Republicans and claiming that the laws of physics just make it gosh darn impossible to do anything other than what the GPS guys want and that anyone who says otherwise is a crazy ignorant terrorist who wants airplanes to fall out of the sky.

Meanwhile, Lightsquared is telling everyone that there are no problems, or not terribly serious problems, and moreover, even if there are interference problems, it is all the fault of the GPS guys for sitting on their rear ends for 9 years and building crappy GPS units. Lightsquared is also running a whole bunch of advertisements in various places around the DC policy world extolling the virtues of its future system as the cure for all our mobile broadband blahs.

Blowing Past The Procedural Smoke

As I said back above the fold, arguments that the FCC exceeded its authority in this case, misbehaved procedurally, or – even more ludicrous – are part of some bizarre conspiracy, are just sideshow. Ignore them. The FCC has granted waivers before, including waivers for ATC, and nothing they did here is out of line with what the agency does all the time. Sure, this waiver is different from previous waivers because they permanently waived the gating criteria, rather than a 16 month temporary waiver like they did with Globalstar blah blah. But all waivers are different. That’s the point of waivers – tailoring specific relief to the specific facts of a case. I certainly do not blame the GPS guys for making the argument. Lord knows I have certainly bitched about the FCC abusing its waiver authority or doing “rulemaking by waiver” in other cases. I just expect the argument against the practice to be as effective for them as it has been for me, i.e., not very. While opponents of the Lightsquared waiver may not like it, the idea that the FCC overstepped its authority or did anything out of the ordinary here is a non-starter. If this went before a court, it would get affirmed. Also, for what it’s worth, I actually think the International Bureau was right on the merits here. (As I said above FCC, don’t let it go to your heads.)

The conspiracy theory stuff is even more ridiculous for anyone who knows the FCC. I’m sorry, but this is not an agency where you get licenses by providing staff with unlimited supplies of hookers and blow. Yes, the agency has its general issues with incumbents, particularly the revolving door problem (and the occasional Cupcakegate). But the idea of Falcone walking around buying influence doesn’t pass the laugh test.

So what prompted the FCC to push through Lightsquared’s application?  Lightsquared solves a problem for the FCC. The FCC recognizes the lack of competition in the wireless world, as well as the lack of available spectrum for broadband. Lightsquared potentially solves that with the wholesale model. It potentially lets the FCC take a pass on more structural regulation (which, despite accusations from industry and assorted cheerleaders, the FCC is loath/terrified to do) while simultaneously promoting competition. It was the FCC, after all, who inserted the pro-competitive condition on the transfer of spectrum rights to Harbinger that limited the ability of Harbinger/Lightsquared to lease capacity to AT&T and Verizon.

Which brings us to the only question that really matters: does Lightsquared’s proposed system cause interference to GPS? And, if so, what happens then?

The FCC Is Not Going To Let Anything Happen To GPS –But That Doesn’t Mean The GPS Guys Will Be Happy.

Here is really the most important thing. The FCC’s engineers on spectrum issues are extremely conservative. Indeed, I have often argued they are too conservative. This is not because they are in the pay of the incumbents, but because they recognize that making predictions about possible interference is not nearly the precise science that people like to think it is.  So the FCC’s engineers tend to err well on the side of caution when setting interference limits. From an engineering standpoint, it is easier to loosen interference restrictions later if you were too conservative than try to mitigate interference if you were too optimistic. But from an economic standpoint, this approach can take a service that was economically feasible and make it into something simply not worth doing. The FCC generally tries to balance these competing interests, taking into consideration things such as how important (economically and politically) is the existing service and how useful (from the FCC’s perspective) would the new service be.

Let us apply this to the specifics of Lightsquared and GPS.  Press reports indicate that the tests mandated by the FCC show that, if Lightsquared turned on its network tomorrow, it would have fairly widespread impact on existing GPS systems.  True technical tests often yield complex data subject to lots of dispute by all interested parties, and each side has a tendency to loudly promote the interpretation of the data most favorable to its side. Nevertheless, unless the press reports are wildly at variance from the actual results, it looks like the GPS guys stand on solid ground when they say flipping the switch to activate Lightsquared under the existing rules will cause significant interference to existing deployed GPS systems.

So is the show over for Lightsquared? Not necessarily. Even if the tests show interference, Lightsquared can try to demonstrate to the FCC (and the GPS guys, but they seem fairly dug in on “no new service in the L Band ever”) that it can take “mitigation measures” to bring the risk of harmful interference down to safe levels.

“What’s a ‘mitigation measure?’” I hear you ask. “Isn’t interference a straightforward physics question? Isn’t it either there or not?” Actually, no. The dirty little secret of wireless is that “interference” (and, even more importantly, “harmful interference”) is a function of many different factors, including how much you are willing to spend on the receiving device.  Interference happens when unwanted energy hits my receiver and my receiver gets all confused, spitting out garbage (or getting totally overloaded and blanking out entirely) instead of information. You can solve the interference problem by reducing the amount of unwanted energy from the transmitter, building a better receiver that can screen out or ignore the unwanted energy, or a combination of both.

So when someone loudly asserts that something is absolutely true as a matter of physics in spectrum engineering, they are usually blowing smoke.  I confess this is one of the things that annoys me about the GPS guys. All this “it is just a law of physics that you a can’t have ANY operation next to GPS receivers because the GPS signal is a measly 25 watts” is just baloney.  In white spaces, my dinky little receiver is picking up a 40 miliwatt signal while a goddamn 50,000 watt transmitter is sitting on top of the neighboring channel.

On the other hand, the statement that “current equipment can’t handle the OOBE from Lightsquared’s system, even when it operates in the parameters approved by the FCC back in 2003” may very well be true.  Mind you, it may turn out that any proposed solution means spending too much money or may render existing equipment obsolete or may prove impractical (at least in the short term) for some other reason. But if we really face a “spectrum crisis,” then we do not want to walk away from 40 MHz of prime spectrum for mobile broadband without at least trying to solve the problem.

Of course, this raises the question of who pays for these “mitigation measures.” Again, I start being rather unsympathetic to the GPS guys because, like it or not, these are the rules the FCC established back in 2003 and the GPS guys chose to build devices based on the idea that no one would ever want to take advantage of those rules rather than develop devices that could handle the FCC approved out-of-band emission limits. The GPS guys usually say “Oh, but that was with gating criteria, and we expected much less intense use of the terrestrial component.”  But that statement contradicts all the (much louder) statements about how the laws of physics make it impossible to have any operation in the neighboring L Band. If the GPS guys are right about the physics, the gating criteria acting to limit deployment doesn’t matter because ANY terrestrial use in the L Band will create interference with the existing GPS receivers.

But the GPS guys, although they would never admit it because it would mean admitting that they could build more interference resistant receivers, actually have a good argument for why they didn’t start developing and deploying better receivers in the 8 years since the FCC set the current OOBE limits. Consumer electronics equipment, the category for most GPS receivers, is an insanely competitive and price sensitive business. Once devices move from the high-end premium market to the mass consumer market, manufacturers face incredible pressure to cut costs because a large segment of the consumer market is extremely cost-conscious. CE manufacturers have incredibly small margins per device, making their millions by selling huge numbers of units (and, in the case of GPS services, things like map updates and other subscription or enhanced features).

Developing more interference resistant devices is expensive, and the typical solutions from the device side (adding more filters to block out of band energy or putting more “intelligence” in the device so it can ignore unwanted signals) drive up the cost of the device overall. In addition, building interference solutions that improve the overall robustness of the device is harder when you don’t know the details of the new interference challenge. Yes, the GPS guys knew the new OOBE limits. But they could not realistically tell until Lightsquared came along just how much new interference risk was likely – or whether anyone would ever even use the new “ancillary terrestrial component” at all.

So faced with the choice between spending money to improve their devices against a possible worst case interference threat that might never happen or keeping the price as low as possible, the GPS guys chose not to spend the money. It’s understandable, like a healthy person not wanting to pay for health insurance, but it’s also a risk. If you end up getting hit by a bus or needing major medical treatment, you are out of luck no matter how rational your decision to avoid the cost of insurance was. Arguably, the GPS guys face the same dilemma here. They made a reasonable bet. Then came Lightsquared wanting to develop the spectrum, creating the interference equivalent of getting hit by a bus and having no health insurance to pay for the hospital stay.

On the other hand, some folks will point the finger at Lightsquared. After all, Lightsquared knew the GPS guys were there, and that they had done diddly to prepare for a change in the use of the neighboring L Band. Why the heck did Lightsquared spend billions of dollars without doing GPS testing first?

And this is where I get to laugh my patootie off at all those economists and others in the “property school” who keep insisting that “what we really need is a well defined system of property rights and the free market will take care of everything.” Lightsquared’s problem was that it believed it could operate at the power levels and for the purposes approved by the FCC in rules the GPS industry never challenged in court. True, Lightsquared needed some waivers, but – as I discussed above – these waivers don’t have anything to do with the interference question.

Lightsquared’s “sin” was that it believed it could use the spectrum in a manner consistent with the FCC rules, for a purpose (mobile wireless broadband) the FCC identified as a critical national priority. Lightsquared didn’t think it needed to do interference testing for GPS because the entire point of the FCC rulemaking is to create rules that allow licensees to develop the spectrum. But then reality got in the way. Because, as I keep saying, there is simply no way in Hell that the FCC is going to tolerate serious interference risk to GPS receivers. Like it or not, GPS is part of our critical infrastructure and too many lives and too much economic activity relies on the deployed technology. Or, put another way, the current deployed GPS technology is “too big to fail.” No matter what should happen under any reading of the existing FCC rules or “definition of property rights,” that is the defining reality.

To take a bad analogy, imagine if you have land zoned for mixed commercial and residential use. A developer buys some of that land, builds some nice residential housing developments, and hundreds of happy people eager to enjoy the new suburban splendor quickly buy up the available lots. A few years later, a developer buys neighboring land to develop into a strip mall and applies for a construction permit. The residents protest about noise and traffic and security issues associated with all the strangers coming into the neighborhood. The strip mall developer asserts that they own the land, their proposed use as a strip mall meets the zoning requirements, the new use will create jobs and bring other consumer benefits, and that the residents knew or should have known that the land was zoned for commercial use as well as residential use. While folks may be more or less sympathetic to the strip mall developer (depending on your views about property rights and strip malls), anyone who has ever been in development fights before knows that – at a minimum – the strip mall developers will need to spend money to address the concerns of local residents to get a construction permit. At the same time, however, the residents are unlikely to block the project entirely.

So too with Lightsquared and GPS. As you will have gathered by now, my sympathies lie with Lightsquared for a number of reasons: we need to bring the spectrum into production, Lightsquared’s business model has the potential to change the economics of the wireless industry (particularly for non-CMRS uses), and licensees (and everyone else) trying to bring new services to market should not face an “incumbent veto” because incumbents made the convenient assumption that the world would never change. At the same time, I recognize a reality when I see it.

What Mitigation Measures Seem Likely, And Who Pays?

So now we get to the real meat of the matter that will occupy time and consideration at the FCC: what kind of mitigation measures will the FCC require, and who pays for them? The GPS guys will of course argue that there is simply no way on this Earth or in space above that any mitigation measures can possibly provide adequate protection against dangerous interference and if by some miracle there are such measure Lightsquared ought to pay for them (in fact, a representative has already testified to that effect to Congress).  But setting that aside for the moment, here are the standard kinds of interference mitigation measures in these sorts of situations.

Require upgrades for the existing systems. Lightsquared has already indicated it will pay to protect “critical” systems related to national security, public safety, and extremely expensive precision applications. This probably also applies to other wireless providers collocated in the same area whose towers rely on the GPS for precision timing rather than the location function. While a good start, this approach alone cannot possibly work for millions of already deployed consumer GPS systems, and will probably be resisted by the GPS coalition as underinclusive. Which brings us to the next popular choice.

Internalized Guard Band: The problem of energy leaking out of the band allocated to one service into another service is an old one, and the traditional solution – when receivers were more finicky and transmitters were more noisy – as to require a “guard band” between the services. Lightsquared can be required to limit the actual operation in its system to something far smaller than its actual 40 MHz. For example, Lightsquared might limit its operation to 20 MHz, using the other 20 MHz as a guard band to protect neighboring GPS. The idea is that by increasing the separation between the Lightsquared operations and the frequencies GPS units receive, you attenuate the strength of the energy that interferes. Remember, GPS units are already shielded to some degree, so activity further away from the GPS band should pose less of an interference risk.

Phase in over time to replace deployed systems. A third approach is a gradual phase in on the theory that consumer GPS units, the most numerous and with the worst shielding, will be replaced over time with better units designed to shield against the increased energy generated by Lightsquared’s operations. The Commission can also require that, as Lightsquared ramps up operation, it must work with providers to notify consumers about the possible need to upgrade their GPS units.

None of these is mutually exclusive, and a combination play seems the most likely – if the FCC does anything at all. Because in addition to sorting through conflicting claims by everybody about what the test results mean, the FCC has to weigh the following additional factors.

Does a solution leave Lightsquared financially viable? The most likely problem for Lightsquared is if it has to sacrifice too much capacity for an internal guard band. Lightsquared’s business model depended on access to 40 MHz of spectrum, giving it tremendous capacity for LTE. If Lightsquared must take too much spectrum out of action, or if it needs to eliminate the large contiguous blocks that make LTE more efficient, it may jeopardize its ability to sell enough wholesale capacity to give adequate rate of return. Even if Lightsquared is ultimately permitted to bring more of its spectrum online as better shielded GPS units come on the market, the FCC will need to consider whether Lightsquared can actually handle the combination of reduced capacity and increased cost.

Chilling investment and innovation in wireless broadband deployment. In theory, this proceeding involves one licensee and its direct neighbors. But the reality is that what happens here will dramatically impact whether investors try to innovate and repurpose spectrum for wireless broadband generally. Lightsquared has spent well over a billion dollars already. If the FCC indicates it will pull the plug on that kind of investment because of political pressure, only the most well-connected incumbents will take any kind of risk, and then only for the most conventional kinds of investment. Proceedings designed to enhance flexibility in both the licensed and the unlicensed space will remain largely theoretical, because you can never anticipate all of the things that might possibly go wrong in the wireless neighborhood when you start to deploy.

Consequences of guessing wrong are potentially disastrous. On the GPS side of the ledger, the FCC has to take very seriously what happens if they guess wrong and the mitigation measures they apply are not enough.  Because all you need is one plane missing a runaway and landing in the ocean or one first responder team getting lost on the way to a fire to set up a public clamor denouncing the FCC as reckless villains who destroyed GPS so that rich kids could have better eBook readers. And while such claims are made so frequently by incumbents that they are almost laughable, it is an actual real concern here.

Where does the consumer benefit lie? The FCC will weigh two competing claims of competing benefits. On Lightsquared’s side, the benefit is a massive infusion of mobile broadband capacity with a business model that could enable lots of new services and products. On the GPS side, cheap available GPS clearly provides lots of utility to the millions of consumers using it every day. What’s the right trade off on cost? If something makes Lightsquared more likely to succeed, but drives up the cost of consumer GPS units, is it worth it? For what set of trade offs?

“Interference” v. “harmful interference.” As with nearly every other fight about new wireless services or increased flexibility, things will turn on the difference between protecting against “interference” and against “harmful interference.” All consumers expect a certain level of interference with their wireless devices, including GPS units. Such interference can happen from weather, background noise, or any of a hundred other things. The question for the FCC is whether a new service creates harmful interference. No good definition of “harmful” interference exists, and this is where the tradeoffs happen. Do you have to be certain of protecting every unit everywhere? Against every possible risk? Back in the TV white spaces proceeding, we spent a lot of time arguing about the possibility that someone’s TV at the edge of receiving range might experience interference from a TV white space device (the so called “hidden node” problem), and to what extent the FCC needed to limit the utility (and increase the cost) of TV white space devices as a precaution against fairly rare circumstances. Was something likely to effect only one in a 100 million people harmful? Or was it enough that it might be “harmful” to that particular individual?

There is no clear answer, which creates much mischief and posturing. But even when parties are not simply talking smack about each other, the lack of any clear definition of “harmful interference” means that knowledgeable and skilled engineers can have real arguments about what test results actually mean and what constitutes acceptable changes in the radio frequency environment. As a result, every time the FCC faces a question about trade offs between services, the FCC has to figure out the answer to this question all over again. And no matter what answer the FCC comes up with, the incumbents will complain that the FCC has been shamelessly reckless and put their service at risk, while new entrants will complain that the FCC has loaded them down with needless conditions that drive up the cost of deployment and threaten innovation. And each side is completely sincere.

So How’s It Going To Turn Out In The End?

At the end of the day, it seems likely that the FCC will want to do something to prod GPS manufacturers to upgrade their shielding so that the L Band spectrum can be used efficiently for mobile broadband. Whether Lightsquared actually survives until that happens, however, remains to be seen. It would be cold comfort for Lightsquared and its investors to achieve vindication post-Bankruptcy, but sometimes that’s how it goes.

And for those of us in spectrum land looking for a bit of distraction from the non-stop debate around incentive auctions on the one hand and ATT/T-Mobile on the other, Lightsquared promises to be the sleeper hit of the summer. Pretty much every spectrum reform issue debated over the last ten years around flexibility, defined rights, and balancing responsibilities between new entrants and incumbents is on display. And while I am rooting for Lightsquared finds a way to navigate through the political pitfalls and engineering difficulties to get to actual deployment, I can’t make any real predictions at this point about what happens next.

Stay tuned . . . .

Be Sociable, Share!
This entry was posted in Life In The Sausage Factory, Spectrum, Tales of the Sausage Factory and tagged , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.

6 Comments

  1. Brett Bobley says:

    Thanks — very well done. I learned quite a lot.

  2. Harold
    Very well done and on target. One thing that isn’t clear is whether the GPS issue is Lightsquared out of band or inband radiation. One engineer I respect says that Lightsquared can do a good job controlling out of band, but that the typical GPS receiver is vulnerable to signals that are properly in Lightsquared’s band. Which makes GPS claims even more untenable on the law, but that, as you note, is not the only issue.

    Incidentally, 20 MHz is plenty for the Lightsquared network for at least several years. That’s what Verizon is using for its LTE network, for example. It’s effectively 30-50 megabits shared and Verizon is showing that reliably delivers 5 megabits to each user. Doing so requires a low cap, however, somewhere between 25 and 75 gigabytes a month or equivalent. That limits how much TV or other high bandwidth use is practical in 20 MHz.

    In fact, LTE (as opposed to the coming LTE advanced) isn’t designed to use more than 20 MHz, I believe. Which doesn’t mean tying up 20 MHz in a guard band forever is a good idea, but it’s viable for a while.

    Separately, LTE Advanced – early deployments 2013-2014, standard pretty much set – is designed to thrive on 40-100 MHz of bandwidth. That becomes ideal for rural broadband, where there is massive amounts of unused spectrum nearly all the time.

    To make that practical, we need to find ways of sharing spectrum across licensees. My choice is “use it or lose it” forfeiture rules, but I’m sure you can come up with more politically acceptable measures. Whatever the best way, it’s time to move ahead. db

  3. Tim Farrar says:

    A few comments on the history here:

    Actually the GPS industry urged the FCC to approve expeditiously (what was then) MSV’s application to provide ATC, because they thought that the agreed OOBE emissions limit would fully protect them from any problems. You might ask why the GPS industry thought all would be fine at that point in time and why apparently no-one on either side did any tests to check this one way or the other. All the FCC did was say “if the GPS industry is happy then we’re happy”.

    Also worth noting that the FCC’s intention at that time was primarily to resuscitate the viability of the MSS industry not to “get more spectrum out for broadband”. The cellular industry got their piece of the pie when the FCC reassigned 30MHz of the original 70MHz in the 2GHz MSS band for terrestrial use (which is why Sprint ended up clearing part of this spectrum).

    This gets it backwards. LightSquared (then MSV) agreed the limits with the GPS Industry Council, which were ratified by the FCC. You could argue that one or both sides weren’t smart enough to realize that there was another (overload) problem. However, what is very clear is that back then MSV simply wanted to get the ATC license and sell the spectrum before they got into operation (remember the rumors of potential deals with DirecTV and DISH back in 2006 before the AWS-1 auction?). Thus there was no incentive for MSV to try and surface any potential problem that might have prevented that. Even as Harbinger got ever more heavily invested in 2008-2009, there was still an expectation that the spectrum would eventually be sold to a wireless carrier, who would be left with the problem of how to implement a real system (remember Harbinger’s talk about “vast global demand” for satellite phones, which would make the company profitable until someone needed the spectrum for a terrestrial buildout – basically the disastrous strategy that TerreStar went to market with last year).

    However, we ended up in early 2010 with no third party buyer, Harbinger a couple of billion dollars in the hole, and no other option to move forward other than to do what would appeal to the FCC and actually offer to build a wholesale wireless network. That’s when a potential issue that had simply been buried for many years actually became a real problem. You might feel sorry for Harbinger at this point, though you might think that they should have done better due diligence before they invested, but simply to say that its purely the GPS industry’s fault for not objecting until now is just not a convincing explanation of what happened.

  4. Tim Farrar says:

    Sorry for the duplication – the previous version cut out a couple of quotes.
    A few comments on the history here:

    “In 2003, the FCC…set a limit on OOBE to protect GPS, which the GPS guys considered totally inadequate but were told ‘tough noogies, this is the trade off because we need to get more spectrum out for broadband.'”

    Actually the GPS industry urged the FCC to approve expeditiously (what was then) MSV’s application to provide ATC, because they thought that the agreed OOBE emissions limit would fully protect them from any problems. You might ask why the GPS industry thought all would be fine at that point in time and why apparently no-one on either side did any tests to check this one way or the other. All the FCC did was say “if the GPS industry is happy then we’re happy”.

    Also worth noting that the FCC’s intention at that time was primarily to resuscitate the viability of the MSS industry not to “get more spectrum out for broadband”. The cellular industry got their piece of the pie when the FCC reassigned 30MHz of the original 70MHz in the 2GHz MSS band for terrestrial use (which is why Sprint ended up clearing part of this spectrum).

    “Lightsquared didn’t think it needed to do interference testing for GPS because the entire point of the FCC rulemaking is to create rules that allow licensees to develop the spectrum.”

    This gets it backwards. LightSquared (then MSV) agreed the limits with the GPS Industry Council, which were ratified by the FCC. You could argue that one or both sides weren’t smart enough to realize that there was another (overload) problem. However, what is very clear is that back then MSV simply wanted to get the ATC license and sell the spectrum before they got into operation (remember the rumors of potential deals with DirecTV and DISH back in 2006 before the AWS-1 auction?). Thus there was no incentive for MSV to try and surface any potential problem that might have prevented that. Even as Harbinger got ever more heavily invested in 2008-2009, there was still an expectation that the spectrum would eventually be sold to a wireless carrier, who would be left with the problem of how to implement a real system (remember Harbinger’s talk about “vast global demand” for satellite phones, which would make the company profitable until someone needed the spectrum for a terrestrial buildout – basically the disastrous strategy that TerreStar went to market with last year).

    However, we ended up in early 2010 with no third party buyer, Harbinger a couple of billion dollars in the hole, and no other option to move forward other than to do what would appeal to the FCC and actually offer to build a wholesale wireless network. That’s when a potential issue that had simply been buried for many years actually became a real problem. You might feel sorry for Harbinger at this point, though you might think that they should have done better due diligence before they invested, but simply to say that its purely the GPS industry’s fault for not objecting until now is just not a convincing explanation of what happened.

  5. christopher hoover40 says:

    the gps signal (L1) seen by a receiver is no where near 40 mW = 16 dBm. it is around
    -130 dBm. that’s 100e-18 W if I did the math correctly.

  6. Steve Stroh says:

    You make the leap that all satellite systems are created equal. Nope. What you state applies to satellites in geosynchronous (Clarke) orbit. That’s because it’s really, REALLY expensive to get them all the way up there, they’re there for a decade or so, the orbit slots are very precious, and from that vantage point they “paint” literally half the planet (absent very expensive, fussy, failure-prone spot antennas)

    Low Earth Orbit (LEO) satellites do not suffer from the speed of light lag because they’re much, much closer, it’s cheaper to launch them into LEO, there’s no dedicated orbital slots (but everything in orbit has to be coordinated), antennas can be smaller and cheaper because you don’t have half-planet footprint, etc.

    Harold sez:
    The satellite industry, and occasionally the FCC, like to pretend that satellites can do broadband. This is like pretending a Clydesdale can win the Triple Crown. Satellite faces too many physical obstacles, from speed of light lag to the expense of upgrading technology, to offer what most folks would consider even vaguely adequate broadband. (Assume usually angry protestations by satellite industry inserted here.)

  • Connect With Us

    Follow Wetmachine on Twitter!

Username
Password

If you do not have an account: Register