Blog Post

Welcome to a New Era of Spectrum Speculation

Reportedly, LightSquared, the company trying to create a wholesale, fourth-generation, wireless network, is thinking about an initial public offering. While part of me loves the idea of reading the details behind the network operator that doesn’t yet have a network in the S-1 filing, there’s a huge part of me that says the company is planning to take investors for a ride using the current spectrum crisis as cover for a questionable business plan.

Because, at the end of the day, what LightSquared has right now is 53 MHz of spectrum, an interference problem with GPS in some of that spectrum, and lots of plans. Yes, back in the days of the bubble, IPOs were built on less, but I think LightSquared may be milking the spectrum crisis. The rallying cry of that crisis claims the U.S. will run out of spectrum by 2013 and the demand for mobile broadband will overtake the nation’s cellular networks and grind them to a halt.

Some people view this coming crisis as making holding spectrum akin to finding an oil well or a platinum mine in your back yard, but that’s not exactly true. Because all spectrum is not created equal, it’s more like finding a meteorite that may contain some really valuable metal, or it may contain something that’s so fragmented or worthless that getting it out of the rock is an economically unfeasible proposition.

That being said, if LightSquared can raise the money that so far has eluded it, placate the GPS industry and then make good on some of its build-out plan, it may actually build a working network and business. To its credit, LightSquared has no shortage of the type of boldness required to build a wireless network. You have to be in it to win it in this business, and Phil Falcone, the owner of Harbinger Equity Partners, the private equity firm backing LightSquared, is most certainly in it — to the tune of billions.

Meanwhile, the spectrum crisis has brought us another questionable deal, AT&T’s $39 billion bid for T-Mobile. Using the spectrum crisis as a rationale for taking out the No. 4 player in the market and hurting the No. 3 player is pretty ballsy, but brilliant. We’ll have to see if AT&T (s t) gets away with it. As long as the FCC continues pitching a crisis, though, I’m wondering who else might take advantage of this airwave apocalypse to boost their bottom lines.

21 Responses to “Welcome to a New Era of Spectrum Speculation”

  1. Dish Network’s Charlie Ergen’s is another big player to watch. He is getting his hands on Terrestar and DBSD’s 2 GHz MSS S-Band spectrum and he has 700 MHz spectrum through Echostar. Maybe he flips his spectrum to Lightsquared, who desperately needs it or builds out his own network on a wholesale basis. I wouldn’t be surprised to see him cut a deal with Sprint to use their new network.

    Another player who is well positioned to take advantage of the spectrum crisis rhetoric is cable companies. They have spectrum (AWS) and a high capacity fixed infrastructure. They have been successful in building out WiFi networks.

    As for a spectrum crisis, I agree with Steve Crowley. The spectrum crisis is overstated. More spectrum is always better, but it is only one input into building a solid network. Smaller cell sizes and higher capacity backhaul are two things carriers will need to invest in to handle future traffic demands.

      • I agree. It will be interesting to see what you uncover on both. Cox has been making progress in their build-out though the 3G network they are building does seem to limit their ability to offer advanced services. I think the Huawei equipment they deployed will allow for an easy upgrade to LTE. Maybe we have to wait until after AT&T brings down the equipment costs from its planned deployment of LTE in AWS to see the true play for Cox and other MSOs. I’m guessing some type of mobile video delivery service where they can take the exemption wireless networks received under the FCC’s net neutrality order to favor delivery of its content.

  2. Steve Crowley

    Before new spectrum becomes available, I believe the mobile broadband infrastructure industry will respond to any shortfall with innovative products and services. In the next year, for example, we’ll see a comparative flood of picocell and Distributed Antenna System products that will reduce the time and cost to increase capacity. I think rate plans will become more rational, with users increasingly charged for the resources the use. The Wi-Fi Alliance recently predicted mobile broadband data offloading to Wi-Fi to be considerably higher than assumed by the FCC in its spectrum demand estimate.

    Much spectrum anxiety would be reduced if the FCC and trade associations would wean themselves from Cisco’s famous mobile data demand forecasts. It can be shown that there is overlap between the people who prepare the forecast and the people responsible for marketing Cisco’s line of core-network hardware to service providers — hardware intended to address forecast increases in traffic. The forecast is used by Cisco to help market that hardware. For business purposes, that’s fine. As for spectrum policy, the FCC is taking Cisco’s claims as-is, and not considering how the company’s interest in selling hardware conflicts with its interest in preparing the forecast, let alone critically and openly examining the assumptions that go into the forecast.

    The Commission should rely more heavily on two forecasts in its possession that, unlike Cisco’s, are independently prepared. One is by Yankee Group and the other is by Coda – both independent research firms. These two forecasts are cited, along with Cisco’s, in the FCC’s October 2010 Technical Paper, Mobile Broadband: The Benefits Of Additional Spectrum.

    • You’re inconsistent about the numbers you accept on the basis of disinterest. The Wi-Fi Alliance is a trade organization of companies with vested interests in pumping up the forecast for Wi-Fi usage, every bit as much as Cisco has vested interests if not more so.

      The problem that you run into with the Yankee Group and people of that ilk is general cluelessness; Yankee has written some things about how the Internet handles congestion that aren’t even close to accurate, but they rely on their own analysis to forecast demand for bandwidth.

      In a perfect world, you could meet the growing demand for mobile broadband with smaller cells, but in this one each cell needs a permit and a backhaul, and they’re not always easy to come by. In San Francisco and New York, for example, the bribes that have to be paid to clear a tower permit are quite formidable.

      • Steve Crowley

        I put more faith in the Yankee Group forecast because it was based on a survey of people in the mobile broadband business. Cisco has practically no experience in mobile broadband radio access networks (RANs), and that is where most of the congestion is today. (It does have experience in core networks, in part through its acquisition of Starent a year and a half ago.)

        I agree adding sites is a challenge. But if we accept traffic is doubling every year — or whatever — we can double or triple mobile broadband spectrum and we’re still doomed. The solution will be a combination of spectrum, infrastructure expansion, innovative technologies, and business solutions such as traffic shaping that will incent users to offload more traffic than is now predicted by some.

        Regarding other countries’ spectrum repurposing, I think it’s unfortunate that Congress has not yet ordered an inventory of federal and non-federal spectrum authorizations and utilization, as was first proposed several years ago. If it had, we’d be much further along with repurposing spectrum today.

      • It seems odd to discount Cisco for not understanding radio networks and then to endorse a consulting group that’s notorious for its lack of technical clue. I still don’t get that. Cisco has a long history with telco switches and it certainly has reputable businesses in the backbone space such as Webex and in the Wi-Fi space such as Linksys and Airespace.

        Spectrum is like memory, you can assume that users and applications will find a way to use whatever’s there. Would you forecast that demand for memory by applications will increase at some fixed muliplier?

        And yes, the spectrum inventory is something that we’ve been asking the FCC to complete for some time now, but we know the answer: The biggest spectrum hog in America is the federal government, and government allocations are the reason the US lags the rest of the world. The second biggest hog is TV, and we know how to deal with that. GPS has an enormous allocation as well, and dealing with that is proving to be a problem.

      • Steve Crowley

        “It seems odd to discount Cisco for not understanding radio networks and then to endorse a consulting group that’s notorious for its lack of technical clue. I still don’t get that.”

        I didn’t endorse the group.

        “Cisco has a long history with telco switches and it certainly has reputable businesses in the backbone space such as Webex and in the Wi-Fi space such as Linksys and Airespace.”

        I agree will all that. Great company. Unfortunately, that experience has almost nothing to do with mobile broadband RANs. If it had such experience, I think it’s forecast would be different in several respects.

        “Spectrum is like memory, you can assume that users and applications will find a way to use whatever’s there. Would you forecast that demand for memory by applications will increase at some fixed muliplier?”

        I may not understand this point. I would not forecast that the demand for memory by applications would increase by a fixed multiplier.

      • How do people forecast the demand for memory? They pretty much don’t anymore, because they accepted long ago that applications would grow to demand whatever amount of memory technology enabled devices to contain. In the old days, prognosticators put out forecasts to the effect that application demand for memory would double every 18 months, but they were just restating Moore’s Law.

        There are two ways to forecast spectrum use, from the supply side according to what’s in the pipe or the demand side as a multiplier of what we already have, but a forecast is a forecast, which is to say it’s only a guess.

        One thing we do know is that Wi-Fi is a nomadic technology but not mobile one, so it’s not a substitute for mobile spectrum.

        Another thing we know is that mobile broadband increases from more spectrum are cheaper than those from denser cells, and the economics have a lot to do with making more bandwidth available to apps.

      • Steve Crowley

        What happens sometimes is that mobile broadband is used in situations where Wi-Fi is available — you might say it’s a mobile technology sometimes used in a nomadic manor. At home, at work. Maybe half the time total. Under flat rate plans, that hasn’t mattered. If operators start to disincent users from streaming video (the bulk of traffic Cisco predicts long term), and I think some may because of that traffic’s inordinate demand on the RAN, I’d suggest Wi-Fi can become a substitute in those limited circumstances.

        Furthermore, forgetting Wi-Fi, under some evolving schemes you might download a movie into the phone overnight using the 4G network for playback later. The operator could give you a discount for doing that instead of streaming in peak times. Thus, total traffic on the 4G network increases, but peak traffic does not, with implications for spectrum requirements.

      • I can find a way to use the 500 MHz of new mobile spectrum the National Broadband Plan asked for; just look at the increases in fixed line broadband since 2000. I can also find a way to get by pretty well with no increased spectrum, but I’d prefer to have the spectrum available.

        The difference is whether we burn our innovation investment on increasing spectral efficiency or on applications that make life better. I’d rather do the latter, and I don’t see unlicensed as contributing to the efficiency problem.

      • Steve Crowley

        I have a correction. It’s not the Wi-Fi Alliance predicting greater offloading. It’s Juniper Research. In a recent press release it forecasts “the majority of traffic (63%) generated by Smartphones, Tablets and Feature Phones will transfer onto the fixed network via Wi-Fi and Femtocells by 2015.”

      • We still have the same problem: Some forecasts come from companies with a lot of technical knowledge but vested interests in the answer, and others come from market researchers who use the same methods for beans and broadband.

        A forecast is only a guess, no matter who offers it.

      • Steve Crowley

        I agree. What I then object to is policymakers cherry-picking worst-case projections from among the three suboptimal forecasts in communications with the public, instead of numbers from the average of forecasts as was done in the October 2010 report. It would be as if the head of the FDA rummaged through clinical trial data on a new drug, picked the best results, and mounted a public campaign to fight the disease the drug treats citing the best-case data and the drug company by name. Sort of.

        I agree with your suggestion elsewhere that the market can help sort this out. I’m not opposed to the incentive auctions as proposed.

      • Steve Crowley

        Though I agree with the forecast quandary generally, I think it does not apply here as Cisco isn’t a mobile radio company except in the area of core networks. The congestion problem is in the 3G/4G air interface, and that’s not Cisco’s focus, as it will admit. If one digs into the forecast methodology, one finds techniques similar to those used by market research firms. It’s my understanding it’s produced by a marketing department within Cisco. Technical input does not seem to apply much anyway. The forecast seems to be a market research product analogous to those of the independent market research firms.

  3. Steve Crowley

    I would say we need more spectrum for mobile broadband but the urgency is overstated by the FCC. What’s happening now is like in 2000 when there were estimates of internet traffic doubling every 100 days. The FCC even told Congress that internet traffic was doubling every 100 days. Those estimates fed the dot-com boom and later the bust when it was found internet traffic was not growing nearly as fast. As with the internet estimates of 2000, there’s considerable innumeracy, gullibility, and mania associated with mobile broadband spectrum discussion today — ingredients of a bubble.

      • I do think the FCC is overstating the problem. It’s report on the issue ignored femtocells, Wi-Fi offload and even uses of MIMO. Yes, it takes a while to get spectrum auctions set, but falsely creating a 2013 date for the end of our current spectrum causes problems.

      • If the FCC is overstating the problem, then so is every other national regulator in the world. There are massive releases of new spectrum taking place in the EU, Asia, and Australia/New Zealand that leave the US in last place in terms of spectrum available for 4G.