Wireless USB revealed (Score 1)

Wireless USB is a 128-carrier, OFDM system that instantaneously occupies 528MHz of spectrum. Each carrier is separated by about 4MHz, and in general uses simple modulation schemes to keep the complexity of the FFT processing to a minimum. While the theoretical speed is advertised at 480Mbps, practical demonstrations to date in real world environments, especially with more than one radio link running simultaneously in the same physical space, show more like 30-50Mbps. However, there's few real-world applications that truly use a significant portion of USB2.0's available bandwidth, so no one (except the external disk drive manufacturers) seem to be too worried about the bottleneck. WUSB was born from the FCC's ruling back in 2002 to open up 7.5GHz of spectrum, from 3.1 to 10.6GHz, to license-free, ultra-low-power, ultra wideband transmissions. The salient definition of UWB to the FCC was instantaneous occupancy of at least 500MHz of contiguous spectrum, thus the 128 carriers and 528MHz bandwidth. Otherwise, WUSB could be called an "802.11a on steroids", since it uses the same modulation scheme but with many more carriers. WUSB's promoters divided that spectrum up into 528MHz "channels", (see Figure 2 at http://www.deviceforge.com/articles/AT8171287040.h tml) and have a scheme whereby the link between two radios hops between channels in a psuedo-random fashion. Current technology radios use only the bottom three channels (3.1 - 4.8GHz) as current inexpensive CMOS technology doesn't perform well enough at higher frequencies. A significant challenge for UWB systems under the FCC ruling is the maximum power output. The FCC specifies a level of -41.3dBm per MHz, or about 0.007 mW per MHz. So, for a 528MHz wide signal, the maximum output power is still less than 10% of your basic Bluetooth headset, thus the extremely short range. And the regulatory climate in the rest of the world is not good and getting worse for WUSB, with the regulatory authorities outside the US only willing to allow use on the frequencies above about 6 or 7GHz, making the propagation issues even worse. Challenges from multipath and absorption by materials in the path create a system that may practically have only a meter or two maximum range. Right now, even though there are a number of companies that claim to have working, shipping, useful silicon, the real story is that it's a long way off from commercial utility, the price is very high, as is the power consumption. This is not a $5 chipset, nor should it be used in battery-starved devices. The Bluetooth SIG adopted the use of a form of the WiMedia WUSB radio for its upcoming Bluetooth 3.0 release, originally scheduled for the beginning of 2007, but now slipped out to (at least) October 2007, with real products not expected until 2009. The mandate from the BT SIG is that only UWB radios that operate in the >6GHz spectrum will be allowed, and the industry is years off from being able to meet that requirement. 802.11n is much closer to reality, its data rates are already comparable to the practical instances of WUSB, and it is not constrained by the exceedingly severe regulatory power limitations. It has a strong ability to trump anything that WUSB could eventually roll out, and crafted for very low power operation, be just as power efficient as WUSB.