Yes, a lot of the marketing spin on WiMAX leads people to believe that a lot fewer BTS (cell sites/towers) will be needed to cover the same area as an incumbent cellular (2G/3G) network. But that’s mostly not so. First off, look at the operating frequency that WiMAX is intended (2300 MHz upwards) to work in – and compare that to 800 MHz (or 1900 MHz/2100 MHz) for cellular. I’d say to provide competitive coverage, you’d need close to 1.5x the number of 2G/3G BTS – and that’s just for moderate urban environments. Dense urban? I don’t know, but it sure looks like a lot more!

Mobility in WiMAX is the next big question: cellular voice networks seem to be a lot better suited for handover but not WiMAX. What happens when an ongoing TCP/UDP session is timed out? IMHO, a far greater number of BTS (as compared to GSM/CDMA) will be needed in order to ensure “seamless” handover in mobile WiMAX networks.

(And oh, BTW, 802.20 is intended to address exactly that shortcoming in WiMAX: handoffs in mobile data networks. So QC may not be completely irrelevant especially if they own signifcant IPR.

These two factors alone point to the fact that CapEx considerations for WiMAX are at least as much as existing voice networks today, if not more. OpEx will vary from operator to operator and country to country, I daresay.

Then there are other factors such as in-building losses (first-wall penetration) that still remain to be seen in real-world environments – but those arguments are subjective and, to be fair, there’s nothing to suggest that pico- and femto-cells cannot be used equally well for cellular voice and WiMAX networks.

WiMAX was never intended to replace cellular voice and it is not ready to do so today, anyway. The kind of performance that cellular voice networks deliver today cannot be matched by WiMAX for at least the next couple of years. And I’m sticking my neck out on this one.

What WiMAX really delivers today is latent data capacity. This is far in excess of what 3G or EV-DO offer for a comparable network investment. Really, really much greater. But the question that faces most operators is: do I really have that volume of subs to consume this capacity. And in most cases, I think the answer will be “No”.

The operator will really need to grow his subscriber base by about 5x or 7x, if it is to use all it’s capacity effectively and therefore get RoI. Or hope to get it back faster, anyway. MIMO and OFDMA will be adapted to fit UMTS/LTE over the next couple of years and when those are mature, that’s when there will be a real apples-to-apples comparison about feeds and speeds in WiMAX and cellular (voice) technologies. The efficacy of MIMO in a number of environments is yet to be substantially proven and perhaps the rollout of WiMAX networks over the next couple of years will show some results – either way!

Up until then, it’s just a matter of using WiMAX as a filler or for pure greenfield ISPs. WiMAX will be good for ISPs, there’s no doubting that. For the next two years, I’d put my money on using WiMAX in fixed environments only – whether you use 802.16-2004 or 802.16e-2005.

DSL replacement or wireless DSL. Call it what you will.

“The 802.20 group points to its focus on full mobility as a differentiator. The standard will be designed to support high-speed connectivity at client speeds as fast as 200 miles per hour, says Ronny Haraldsvik, senior director of marketing for Flarion. ‘That’s the key differentiator with 802.16,’ she says.” Looks like QCOM might be able to get the market for the high speed trains, and German autobahns. For the rest of the time, WiMAX will be cheaper. I find it hard to believe that WiMAX will be similar cost to 3G to deploy. Is a WiMAX router going to be the same cost as a 3G switch?? We are talking about Cisco equipment vs Ericsson equipment. I don’t know of many cases where their price lines overlap!

MIMO is about attentas, not an air interface. Nevertheless I think QCOM did pick up some IP in this domain, but thats not germane to my point.

I didn’t mean crush as in the marketplace, just in terms of raw air-interface performance. Which I thought would be more obvious since my main point was that Carriers and Vendors were trying their best to avoid QCOM. And which is the only way to compare with WiMAX at the moment.

What will be fun to watch is Verizon and maybe Alltel.

VZW and ATT are having their cake and eating Sprint’s lunch. If Sprint cant manage two technologies (CDMA and iDEN), what makes anybody think that a third will improve the situation? WiMAX is no silver bullet for wireless, much less a tonic for crappy leadership.

I’m not saying QCOM is perfect and they have certainly been out played by Nokia for the last year plus, but the benefits of WiMAX will evaporate and then be crushed by future CDMA/OFDM versions. Check the last QCOM CC.

While the cost of the network is not trival, the cost of marketing, churn and CC are much more important especially in the beginning.

If a carrier can not compete today, changing technologies wont necessarialy help, other than to buy the executives more time to bank some badass bonuses before the board calls bullshit on the whole operation.

Which largely explains WiMAX in america.