Carriers, standards bodies and equipment manufacturers like to draw battle lines among the varied flavors of broadband wireless. That became clear when Intel propped up the moribund 802.16a by dubbing the fixed broadband wireless standard "WiMax" and urging that its mission be broadened to include mobile client platforms. While Intel was hoping for a Centrino/Wi-Fi model, pundits were quick to point out that WiMax would never displace 3G services. It's hard to find evidence that Intel had ever suggested it would. But analysts assumed Intel would only be interested in a standard if it applied to millions of consumers--and that such ubiquity would be predicated on WiMax's mowing down every cellular standard in its way, including high-speed packet access in both downlink and uplink forms (HSDPA/HSUPA). This view also assumed that an expanded 802.16 would supplant 802.20, the original mobile broadband wireless spec.
The "fight to the finish" model is one way to interpret the IEEE's recent suspension of 802.20 activity, though seeing Intel's late entry into the working group as part of a plan to destroy a competing standard would overlook earlier efforts by 802.20 participants such as Qualcomm and Flarion to skew the voting process.
In the operator realm, however, there's a far more important battle being waged. Whether you believe in 802.16, 802.20 or the 3G Partnership Project's (3GPP) Long-Term Evolution (LTE) program, they all are based on Internet Protocol, advanced quality-of-service and, probably, orthogonal frequency-division multiplexing modulation. Given the vast number of frequency bands available for 802.16, there are likely to be many packet-based broadband wireless services offered in different regions--many certifiably WiMax, others based on 802.20, if it survives, some coming from LTE standards.
But real battles have little to do with the physical and media-access layers in which the IEEE largely dwells. Instead, one should look to the network layer. Each succeeding 2.5G and 3G standard has moved closer to a native packet format, though even HSDPA/HSUPA relies on a connection-oriented model in the network, because most cellular nets still are circuit-switched.
Carriers know that future networks should be optimized for IP. The 3GPP recognized this three years ago, developing the IP Multimedia System standards so that wireless operators could preserve TDM traffic where necessary but still move to IP backbones.
At the broadband wireless physical layer, there's no reason to worry about a mix of WiMax, 802.20 and LTE environments. It can accommodate them all. At the network layer, smart wireless operators know that adding broadband provides an impetus for moving to voice-over-IP sooner rather than later. The dinosaurs will see circuit switching as their last means of differentiation and will only discover a few years hence what wireline local incumbents realize today: If they don't move away from circuit switching, their revenue base for voice will vanish. WiMax, LTE and 802.20 are merely wrinkles on the elephant, and any hesitancy to move to IP is beginning to look downright silly.