Michael van Noorden, Managing Director of Sevida, a consulting firm in the UK, provides Wi-Fi ecadamy with interesting observations on the battle between mobilized 802.16e (using both licensed and unlicensed bands) and 802.20 (using licensed bands below 3.5 GHz) - for vehicular-speed handoff.

In March, the standards body ratified 802.16a, the non-line of sight, fixed wireless version of WiMAX, and kicked off the 802.20 process. At this stage, Intel and Nokia had only just decided to put their full weight behind WiMAX and both technologies were still obscure. They were positioned, quite realistically at the time, as complementary.

WiMAX had come from a background of addressing last mile requirements using fixed wireless, while 802.20 was seeking to standardize various efforts to provide a fully mobile broadband solution using IP. Even the mobile variant of WiMAX, 802.16e, was still widely seen as an extension to a fixed wireless standard rather than a fully mobile standard in its own right.

Speaking to 802.20 activists last week, it was clear that they no longer regard e as a distant cousin. The role of WiMAX has been enhanced considerably, driven by the new prominence of public Wi-Fi.

But once the potential of WiMAX to dramatically enhance the potential of Wi-Fi by backhauling hotspots and providing wireless networking over 30 miles and more, it became far hotter property. For Intel, it was a way to make wireless notebooks and devices even more attractive by increasing their capabilities through a more powerful technology than the limited Wi-Fi. For Nokia, even more significantly, what had started as a potential new string to its ailing base station unit, suddenly became a means of creating a whole new handset business, with the company promising WiMAX cellphones by 2005.

This encroached well into 802.20 territory. This effort was initially driven mainly by the pioneers of fourth generation wireless IP technologies, notably Flarion and Navini Networks, who refused to have anything to do with 802.16e, claiming theirs is a purer IP solution.

But with the politics in the background, seems that 802.20 will fade into insignificance, or will be redirected into a niche application, notably one area where it excels mobile communications in fast moving vehicles (the standard specifically supports vehicular mobility, at speeds up to 250km per hour).

The best result will be if both camps come together, with the 802.20 rump adopting WiMAX specifications into their products and the best of 802.20 finding its way into 802.16e. Ironically, this has a far better chance of happening once Motorola accepts defeat and seeks out another weapon to push forward its strategies, and the vendor stand-off collapses with the victory of WiMAX. It is an indictment of how the standards processes are hijacked by the suppliers politics that such a coming together could not have been achieved with the blessing of the big names, rather than as a result of a battle that was over almost before it started.

Navini prefers CDMA modulation rather than the COFDM modulation used in 802.16a/e, while Flarion’s Flash OFDM, is a leading contender for 802.20. Navini says CDMA penetrates better. Either way, they argue, their everything over IP approach will lower costs by a factor of ten over “3G”.

The 802.16a standard can deliver fixed broadband using licensed MMDS frequencies (2.5-2.6 GHz), 3.5 GHz or the unlicensed 5.8 GHz band. Aperto’s fixed wireless system typically delivers 7-20 mile range and should resemble 802.16a coverage. The 802.16e standard (still being formulated) may be backward compatible with 802.16a and provides “walk-about” mobility and handoff. The 802.20 standard only works with licensed frequencies below 3.5 GHz and provides hand-off at vehicular speeds.

Unlicensed (2.4 & 5.8 GHz) mobile users might use 802.16e while MMDS users could use either 802.16e or 802.20.

Still, T-Mobile has invested in Flarion and bought Worldcom’s MMDS. There seems to be a movement to licensed 802.20 on the part of cellular operators who might trade 802.16e’s bit rates for easy installation, and fast handoff (duplex) channels which may also handle voice more simply than (simplex-based) Navini or Arraycomm.

Or maybe not.

Posted by Sam Churchill on Sunday, July 27th, 2003 at 11:46 pm.