7 Comments

Summary:

The enthusiasm in Silicon Valley over the growth of mobile broadband and mobile applications is palpable these days, but there’s one thing missing: an understanding of how the underlying network affects both the physical hardware and the way applications run. What we need is the return […]

iStock_000007457826SmallThe enthusiasm in Silicon Valley over the growth of mobile broadband and mobile applications is palpable these days, but there’s one thing missing: an understanding of how the underlying network affects both the physical hardware and the way applications run. What we need is the return of the Radio Wars, those loud and acrimonious battles over exactly how we’re going to get our gigabytes of mobile data from Point A to Point B.

In the mid-’90s the debate was over CDMA vs. GSM; 2G vs. 3G wireless arguments came later in the decade and then in the last few years, shifted to 3G vs. WiMAX. So why have the “Wars” gone away? Put simply, with WiMAX finding its niche and most of the world’s major wireless operators moving to LTE (see PDF), there’s no longer a conflict. But the relative silence in 2009 — the “Kumbaya” of LTE-solves-all — has led to a reduction in critical thinking around wireless network evolution.

Looking back, the Radio Wars served a serious commercial function: They identified the challenges inherent in increasingly complex, multistandard, multiband radio networks. At the same time, they made clear through public debate and white papers the depth of complexity faced by network operators and their vendors. And they were material to the wireless world, as the lines drawn surfaced the underlying reality that wireless broadband is different than wired broadband.

Today it seems that LTE as a foregone conclusion has convinced people that there’s no longer a need to have debates around radio link budgets and “bits per hertz,” as if by ignoring technical realities somehow makes those realities disappear. Worse, discussion about network design, latency, and consumer experience are often drowned out by companies — big and small — that, coming from a PC or “Web 2.0” perspective, have a blatant disregard for the complexity of making all this stuff work. Witness AT&T’s very public challenges when it comes to supporting iPhone traffic. As vendors increasingly speak about “one Internet,” whether wired or wireless, and companies big and small increasingly pitch high-bandwidth services such as streaming video to mobiles — and users increasingly adopt them — the issues around the iPhone may end up seeming like the canary in the coal mine.

And that’s the problem. Those that believe in “one Internet” aren’t paying attention to the fact that use cases for wireless broadband users hitting their 5GB-per-month limit over their cellular laptop card and home desktop PC users coming up against their cable modem’s newly imposed 250GB-per-month cap are not the same, that ever-richer media (including all manner of video) cannot continually be pumped onto mobile networks without consequence.

Expanding 3G coverage and adding more RF channels will help, as will the addition of WiMAX networks and LTE, but none will prove a panacea. Take LTE: There are more than 15 LTE spectrum bands globally; the mere thought of designing functional handsets that support even a subset of those bands (never mind legacy bands, plus Wi-Fi, GPS, Bluetooth, DVB-H, FLO, etc.) makes my head hurt.

Jeff Belk Headshot Mar 08In order to realize true mobile broadband, over the next decade we will need networks that can support tens of gigabytes of traffic per user, per month. Multimode 3G and 4G networks won’t cut it. There will be more radios in our devices, enabling the offloading of traffic to other networks such as Wi-Fi, or the downloading of content via so-called unicast networks such as DVB-H or MediaFLO. New spectrum and network topologies, including femto/picocells, will help, as will better handset performance, new handset and base station antenna technologies, and new ways of structuring applications and shifting traffic.

But such changes will not come from those that are blissfully ignorant of the current network reality, and education is hard to come by absent a debate over radio technologies. Without such a debate, we are potentially setting ourselves up for another cycle of hype and disappointment when it comes to the evolution of networks, devices and services. In other words, it’s time to bring back the Radio Wars.

Jeffrey Belk is Managing Director of ICT168 Capital, which advises and invests in emerging and growing companies in the information, communications and technology space.

  1. Wirelessly Tethered Sunday, September 27, 2009

    There will always be different Radio networks serving needs based power, range and application requirements. That however does not make them camps at war. For the Cellular networks indeed there were two camps – GSM/UMTS and CDMA. They have reached a truce with a common OFDMA technology for LTE and a common architectural evolution path. As for WiFi, WiMax they will deliver their value suited to the right economy based on coverage area or spectrum. Unlike the common perception, they will not compete with cellular.

    Share
  2. Jeff, you are absolutely correct. One of the problems is that the radio engineers from the days of the wars are no longer in the industry — replaced by exuberant engineers and marketeers who do not have historical memory. And operators such as Verizon and AT&T aren’t going to throw away their billion dollar fiber networks nor will cable operators. Laws of physics and all the points you eloquently raised will define what mobile broadband is or is not. And of course the regulators who decide how to allocate spectrum around the globe only complicates matters…and least but not last net neutrality is rearing its ugly head. And the poor consumer has high expectations which will be hard to realize for a long time.
    Keep the historical memory alive…for without it we will have chaos as we have never seen before.

    Share
  3. [...] Bring Back the Radio Wars [...]

    Share
  4. Mishan Kontroll Monday, September 28, 2009

    Do mesh networks provide the answer? Is the allocation of spectrum like property an anachronism? Isn’t wireless bandwidth potentially unlimited if smart radios communicate with only enough power to reach adjacent parts of the mesh?

    Share
  5. I think it’s a waste for those carriers using W-CDMA to go to LTE. Use HSPA+ to get the throughput and relative immunity to multipath. The next step for all will be OFDM with each “carrier” being a narrow-band cdma channel. Immunity to multipath, again. LTE/WiMax are too susceptible to multi-path.

    For rural areas allocate the lowest frequencies for HSPA+/EV-DO rev C.
    For urban areas use higher frequencies, but not high enough to be affected by precip. and other atmospheric attenuation.

    Beef up the error correction for higher throughput services.

    In some cases, meshes might help, but not generally necessarily.

    Share
    1. Beefing up error correction for higher throughput can involve many improvements.

      Higher frequency stability and lower drift of the BTS by using rubidium references or GPS Rb, lower distorion and phase noise using the latest GPS and GPS disciplined oscillators, either OCXO or Rubidium, but with -110dBc/Hz@1Hz offset phase noise and 8 x 10-13/s short term stability (this has a great impact on jitter etc.

      These devices are now available from specialist manufacturers and at “normal” Rb costs.

      Using a kalman filtering 1PPS timing module can transform accuracy, jitter and stability for €1 cost on the PIC processor….IPR come free.

      All the above from Quartzlock.com

      Share
  6. [...] are starting to emerge, such as the role that smartphones play in distracted driving, or the effect some apps can have on battery life or the cellular network. A recently launched startup — Waze — exemplifies some of those issues, while also [...]

    Share

Comments have been disabled for this post