All Ultra-Wideband (UWB) systems are not created equal

To serve these applications, the UWB radio architecture must be able to scale in such a way that the ultra-high-speed needs of the handheld can be met with an extremely simple (lowest cost, lowest power) implementation. In other words, an architecture must be selected that minimizes complexity and power drain so handheld devices achieve high performance and full interoperability wih AC line powered devices.

You may have heard that "UWB doesn't work well," based on UWB product tests published over the last year. Both industry organizations and manufacturers claim file transfer rates "up to" 480 Mbps for Certified Wireless USB devices. However, independent testing has revealed rates that are more than 10 times slower. (See Comprehensive UWB tests give video a green light but caution on wireless USB.)

Manufacturers blame these results on defective test procedures, or early hardware and software that doesn't truly represent the capability of their systems.

What is a consumer to believe—the claim of a trusted manufacturer, or the verifiable results of testing performed by respected independent laboratories? Is the problem simply "immature software drivers," or is there something else going on?

This article will examine two UWB radio architectures, short-pulse (<2ns) UWB and long-pulse (>250ns) Pulsed Multi-Band OFDM (MB-OFDM) based UWB, along with the underlying physics upon which the two radically different UWB signaling schemes are based.

UWB basics
The term "UWB," as defined by the FCC's regulations, can be applied to vastly different radio communication architectures. Consequently, as you might expect, all UWB systems are not created equal. Some UWB systems will perform better than others, and to find the winner you must look deeper than just the name "UWB."

Design choices fall into two categories, architectural and implementation. A simple example of an architectural choice is picking AM versus FM for a radio system with a goal to eliminate static in a received signal.

While implementation choices can make a difference, the architectural choice sets the fundamental performance capabilities, or intrinsic system flaws, of the design. Intrinsic architectural-based flaws cannot be fixed by clever implementation choices.

In the case of short-pulse-UWB versus long-pulse MB-OFDM UWB, we are discussing two very divergent architectural approaches that do, indeed, establish drastically different fundamental performance capabilities.

Why? In a nutshell, there are drastic differences between how much multipath fading occurs for the different waveforms used. A long-pulse MB-OFDM based radio and a short-pulse based radio experience drastically different fading. This fading difference strongly affects how well the two architectures perform in real-world multipath environments—like your and my office or living room.

When a radio wave is sent from one antenna to another, along the way it reflects off of a multitude of objects, like walls, floors, chairs, desks, stereo systems, etc. The transmit antenna transmits in all directions, the receive antenna receives from all directions, and the reflections from the multitude of objects in the vicinity arrive at the receive antenna with random time shifts. At some frequencies, these reflection components subtract (or combine out of phase) causing a multipath fade.

At other frequencies the components add creating an enhanced signal. Given the random delays and strengths of these multipath reflections, the result is a random distribution of fades that radios must deal with. The wireless link, or path "between" the antennas, with all its fades and enhancements over the bandwidth of the system, is called a "channel."

Page 2: next page
Page 1 | 2 | 3 | 4