There is a huge, and I mean HUGE, demand (as of yet, vastly unrealized) for wireless broadband anywhere you walk, at a reasonable price. I work for a realty MLS, and there are a good number of listing agents (500+ at this MLS, I would say, out of 17,000ish total) that use their PDA's to access the listing database online via wireless at places like Starbucks & Barnes & Noble, because it saves a ton of driving time. This is but one example....personally, I'd love to go to the beach and play online games with a great view in front of me :)

802.16 is a highly asymmetric protocol. To SSs (SS = Subcriber Station) cannot talk together directly. They communicate with a BS (Base Station).

There is a mesh version in the standard, but it is incomplete and insecure.

WiMax is to 802.16 as the Wifi Alliance is to 802.11. It determines interoperability criteria for 802.16 systems.

802.16 is an effort to standardize an existing market in MMDS and LMDS systems. There are many manufacturers that have been making and selling this stuff for a long time. What is new is that there is a standardized MMDS/LMDS protocol coming out of the IEEE.

WiMax can serve eiher big carriers, small carriers or private users. The standard is flexible in this respect. It can work in licensed or unlicensed spectrum. It can be fixed or mobile. It can be point to point or point to multipoint.

This looks nothing in the slightest like Iridium.

The 802.16 standard standard stack terminates at one of
A) An ATM convergence sublayer
B) An IP Packet convergence sublayer
C) An 802.3 Packet convergence sublayer
D) An 802.1Q CS - Ignore, this one is braindead

So there is no compulsion to spit L2 user traffic out of the BS.

I have seen a variety of implementations, from the IP routing being right in the base station, next to the radio, through to L2 traffic being routed over a closed IP network back to an aggregation point elsewhere in the network and varieties that lay somewhere in between.

SS1 -- 802.16 BS -- SS2 is feasible and real.

There is nothing in 802.16 that demands you work hard at L2. Although some people clearly think there is a reason to try, hence 802.1AB provider bridging.

Once again the media get it wrong. ClearWire is NOT using WiMax. There is no WiMax gear available which uses the U.S. spectrum, and there won't be such gear for another (probably) another 18 months.

What McCaw is doing is using the equipment from NexNet (which he also purchased) to make everything work. NexNet builds MMDS (Multichannel Multipoint Distribution System) equipment. Transitioning that equipment to WiMax may not be too difficult, but, again, there is no WiMax equipment currently on the market in the U.S.

Here in Jacksonville, FL, broadband is typically 45/mo, more if your not already a customer to the local comcast. On top of all that, this city has a very large landmass for its population, making things like cable and DSL available only to the areas with highest demand.

If I had to guess, i would say nearly 40% of the city is still without broadband of some kind. Not from lack of desire for it, but simply not available in your zip code. Nobody in Jacksonville will be doubling their ISP bills by paying $50/mo, not unless your a member of AOL.

I think that should answer your question, the price is reasonable, and do to the large physical size of jacksonville, hi-speed by hardline is limited. Last time I searched for a wireless provider that offered similiar speeds, they wanted $100/mo, $700 in equipment, and they wanted me to pay for labor to profesionally install an antenna.

http://www.nextelbroadband.com/ [nextelbroadband.com] is using Mobile-Fi (IEEE 802.20). This technology is superior to WiMax in many ways. First of all, Mobile-Fi actually provides mobility today, while 802.16e will probably never provide realisitic mobility. And Mobile-Fi is very low-latency when compared to WiMax, WiFi, and 3G/3.5G/4G networks.

The primary benefit of WiMax is in the architecture. It lends itself to be very flexible. The person who mentioned it as a replacement for LMDS/MMDS and other wireless technologies is correct. The people making comparisons to ATM and Iridium are mostly incorrect.

If WiMax components become cheap, mass-marketed, and ubiquitous -- that is a good thing for everyone. Since Intel, Alcatel, and Siemens are behind the WiMax movement, there stands a good chance of this. Nokia got out of the WiMax alliance, so maybe they know something that the others do not (and maybe it's Mobile-Fi or 4G).

The WiMax POP architecture is where the true power is. Being able to mix/match licensed and unlicensed spectrum via antennas, while using the same "Access Point" electronic components for cost reasons makes complete sense. A WISP could easily build a survivable backhaul wireless network across a city, while providing the best-effort CPE/customer networks a few miles here and there on the same device.

>Ok, seeing as you wrote it, explain to those who might be too lazy to RTFA, why 802.16 is better than 802.11[a|b|g].

It's not better; it's different.

5 mile radius cells = metro area service = MAN = 802.16 = lower bps/unit_area
100m radius cells = local area service = LAN = 802.11 = higher bps/unit_area

The IEEE hasn't even started writing the 802.20 spec [ieee.org], so Nextel can't be using 802.20.