Is WiMax like the tragic Barbaro?

(Editor’s note: The wireless world is an important one, yet a jargon-filled one, so we asked investor Paul Grim to give us a lay of the land. We also asked him to spell out the acronyms, so that the rest of us can keep up. He wrote this several days ago, before the latest comments by a high-level Qualcomm exec, also critical of WiMax technology. Investors are plowing billions of dollars into WiMax.)

In addition to the demise of the MENS Club, there were several other interesting insights to be found at the Orlando CTIA conference a few weeks back:

Mobile Broadband is now a three-horse race, but one of them (mobile WiMax) may end up like Barbaro –- a visibly thrilling candidate that tragically crashes and burns in the final round.

Broadly, there are three acronym-laden contenders for next-generation mobile broadband rollouts:

Ultra Mobile Broadband, or UMB: This is Qualcomm’s upgrade path from their current Code-Division Multiple Access (CDMA) Evolution, Data-Only (EV-DO) networks. Interestingly, UMB will not be CDMA, but rather Orthogonal Frequency Division Multiplexing (OFDM), based in part on Qualcomm’s recent acquisition of OFDM player Flarion.

High-Speed Packet Access, or HSPA: This is the upgrade path favored by the GSM crowd. When they moved to 3G networks, they adopted something called Wideband-CDMA, which was different enough from Qualcomm’s CDMA to lower the royalty bill a little. They are continuing this track with HSPA; current flavors are HSDPA (D for Download), then HSUPA (U for Upload), HSPA, and of course then “HSPA+”. If that isn’t confusing enough, the next step is something called Long-Term Evolution or LTE, which like UMB is OFDM as well.

Mobile WiMax: This is the third major contender, also OFDM, but pushed most strongly by Intel, Clearwire and lots of startups. It began life as a number in the Electrical Engineering IEEE standards body (802.16), and like WiFi continued to add letters as it went (802.16a, d, e, and now m). As all the mobile operators would rather do software upgrades of their network (see options 1 and 2 above), this is a hard sell for a lot of them. Nevertheless, some operators insist they can go this route more cheaply, including Sprint Nextel.

So in summary, you have an OFDM upgrade supported by the, uhh, CDMA Development Group, a W-CDMA upgrade supported by the, uhh, GSM Association, and a relatively green version of OFDM supported by the PC guys and a lot of nervous VCs.

In Orlando, GSMA Chairman Craig Ehrlich noted that the GSM crowd still control about 80% of the mobile market, and that this guarantees that HSPA will dominate (he helpfully suggested that investors “think really, really hard before backing alternative technologies” – thanks Craig, consider us warned). Qualcomm will ensure that UMB becomes viable, no matter what. WiMax has been getting the most hype, but it’s not clear whether the technology is better, the spectrum available, and the market big enough. Incumbents will typically take the path of least resistance (ie, software upgrade to current infrastructure). Even with Intel and crew backing it, WiMax won’t get the economies of scale on equipment cost that the GSM or even CDMA operators do. That’s why WiMax may stumble at the end of the day.

While the hardware guys at CTIA worry about upgrades, the content guys have taken over the place.

Five years ago, virtually all the booths were Original Equipment Manufacturers (OEMs), carriers, and resellers. This year, the balance has shifted toward content and applications: The Google, Yahoo and Microsoft booths were bigger than some OEMs, and the Viacom/EMI keynotes made perfectly clear that content people want wireless to live up to its potential as a distribution channel. The biggest challenge is no longer technical, but structural: the carriers are used to charging in minutes (and megabytes), not sharing revenue with content partners, or giving it away in an ad-supported model.

Even with ubiquitous broadband, you still need a device with a UI (user interface) that doesn’t suck.

The iPhone may indeed become the holy grail of decent UI, but the OEMs could do a lot to catch up. Some are starting to, but none (including Apple) have yet to overcome the inherent challenges of small displays and inefficient text input. The Microsoft spinout ZenZui is a nice attempt, and there will be many others, but I think some combination of breakthroughs in micro-displays and speech-driven interfaces will be needed to take things to the next level.

The more functionality loaded onto these devices, and the more complex the wireless protocols become, the worse battery performance will get.

There do not yet appear to be any real breakthroughs in power management or battery technology to support the growing demands on cell phones. The newer standards use multiple antennas to boost performance, but adding antennas drains power. I didn’t see anything that suggests the power gap has been addressed yet.

Cell phones are evolving into personal media/communications devices, but evolution stops at my wallet.

VISA CEO John Coghlan presented a plan for mobile payments, and even showed a survey that 57% of users would use their phone to pay for stuff. Sorry, but I don’t buy it just yet (in the US). Aside from the massive behavioral change required, this won’t happen here until biometric access is on every handset. It’s bad enough when I lose my phone; it would be infinitely worse if whoever found it could empty my bank account. There’s also the non-negligible challenge of getting retailers to invest in new POS equipment. This could be a classic case of “I’m not the market” but I think a big chunk of consumers aren’t the market either.

Next Story: The incubator bubblet: YEurope, Techstars, HitForge…
Previous Story: BoomTown launches….again

Bookmark and Share
Photo of Paul Grim

About the Author, Paul Grim

Paul Grim is a General Partner at SunBridge Partners, the US affiliate of Japan-based SunBridge Corporation; previous investments include Alien Technology, Flarion (acquired by QCOM), Eclipse Aviation, and Salesforce.com (NYSE: CRM).Prior to co-founding SunBridge Partners, Paul was a GP at Equitek Capital, and spent ten years in Europe at Gemini Consulting and IBM. Paul holds an MBA from MIT Sloan and a BS in Mechanical Engineering from MIT.

  • They're spending so much time on technology and content without focusing on UI. As you said, the iPhone is the first device I've seen that really spent time on getting the user experience right. And you know what? That might be enough to make me buy it (not to mention stay with Cingular).

    I wish companies would realize that user experience matters. That means good UI and good customer service. I don't care how much content, speed or bandwidth you have if you don't get those two things right.

    GJ
    http://www.60in3.com
  • The UWB and LTE upgrades are not going to be a software upgrade. The Base stations will need to be forklift upgrades, and it's possible that new base station location's will be required. That's why Picocell/Femtocell's are interesting to those guys.

    From a networks/billing point of view, UMB and LTE have an advantage, since they will migrate the current billing approaches.

    I agree with GJ, that a bigger issue is customer service is crucial. Everytime I have to deal with Verizon Wireless, I feel the veins in my forehead starting to pop out....

    Dave
    http://www.keystoneradio.com
  • Paul Grim
    Dave,

    Fair point about UMB and LTE that it's not purely about software (HSDPA mostly was, as was EV-DO Rev. A). But vendors have a good idea of future upgrades and generally try to future-proof base stations to avoid wholesale "Rip-and-Replace" (a blade server here, a modem there, etc.). I think the bigger issues in the network for all of these are about backhaul demands as speeds head north, and moving towards pure IP in the core NW.

    In both cases though, switching to WiMax still feels like a leap of faith in comparison. Greenfield operators maybe, but incumbents not so much.

    Paul
  • Paul:

    I agree that the network backhaul is going to be interesting. The incumbents certainly have an advantage there.


    Base station locating will also be a challenge. To get the high data rates that are being discussed, you need a strong signal, stronger than you need for a voice call. The carriers may need to re-architect their networks for better coverage.

    It's still not obvious what the use case for the added BW is going to be though. It's my understanding that the uptake rate on anything other than text messaging is pretty small. I'm sure that the UI has a lot to do with it.

    WiMax on a nationwide network is a real leap of faith. I can see the WISP market jumping on it if they can get spectrum.

    Dave
  • rk
    Paul,

    There is some good coming out of WiMax world mainly around open standards for Baseband to DFE links such as OBSAI, and CPRI which can enable carriers to deploy different configuration of Macro Cell, Pico Cell type base stations , by using Fiber in between, other standards can leverage lot this work as well. enabling them to optimize on deployment cost. On spectral efficency and performance all UMB, HSPA and WIMAX is upto carriers which way to go. Mobile WiMax has power eff issue on the handset side, jury is out if any one can solve that issue and add more features than other two standards. Also proposed data rate of 22Mbits by WiMax is shared medium in that spectral area.

    Raj
  • Paul Grim
    Raj,

    I agree that the mere existence of WiMax might help the efforts to move towards open standards for the mobile operator base stations, and certainly WiMax itself will follow the WiFi route of commoditization thanks to Intel et al (at least on the device side).

    But given that LTE, UMB and WiMax are all heading toward the same bits/Hz/sector on spectral efficiency, I'm guessing the overall cost to roll out for incumbents will still be lower to stick with existing upgrade paths.

    Despite my skepticism, I still hope that WiMax at least provides enough competition to keep the mobile folks on their toes. With a Centrino WiMax chipset in every laptop, and Clearwire and Sprint in 2.5GHz, then Verizon, T-Mobile and AT&T will have a harder time maintaining a walled-garden world with crazy charges for content and blocking traffic they don't make money on (ie, VoIP, etc.).

    Paul
  • RK
    Hi Paul,

    I agree, LTE & UMB has better chance in existing carriers market, WiMAX will probably get rolled out by Cable operators, DSL,Satellite, and WISPs! I feel some of this crowd will go with LTE & UMB as well. some reason, I am not also sold on the use case Intel presents about Laptop yet. As I see UMTS cards in the market by carriers, I have not seen traction.

    Raj