MANHASSET, N.Y. — The fundamental characteristics of ultrawideband (UWB) radio and its principal of operation help explain the excitement and hyperbole the technology has generated of late.
UWB begins with the generation of a narrow pulse of electromagnetic radiation, on the order of 0.5 nanoseconds. At the important half-power point of -3 dB, the range is a little over +/- 1 GHz, providing 2 GHz of bandwidth, hence ultrawideband.
The first UWB systems used a series of constant-phase pulses, with the position of these pulses relative to each other being modulated according to information contained in the baseband, creating a carrier-less train of tightly controlled, coherent pulses.
Unmodulated, this train creates a comb of power spectra with regular power peaks that can severely impact narrowband signals. However, pulse-position modulation (PPM) disperses this power around a given carrier frequency, thus alleviating this interference. This distribution is further enhanced through the use of temporal coding using a pseudo-random number (PN) which disperses the signals even more, pushing the instantaneous power below the noise floor.
Reception of the widely spread signal requires the use of a time-gated correlator. A correlator multiplies the received RF signal with a stored template waveform, and integrates the result to get a dc output. Because of the PN coding, the correlators must be time-gated according to the PN sequence so it samples the incoming signal at exactly the right time, otherwise the signal would be lost in the noise.
An alternate modulation scheme that has found favor recently is called binary phase-shift keying (BPSK), or bi-phase modulation. This uses the baseband information to modulate the phase of the signal (0° or 180°), instead of modulating the position. Pre-processing of the signals eliminates the spectral lines, or comb effect, thereby allowing bi-phase systems to meet regulatory requirements without having to reduce the total transmit power levels. Bi-phase proponents argue that it holds a 3-dB efficiency improvement over PPM, and reduces jitter requirements as well, as signal recovery isn't as position-dependent.
Whatever the chosen scheme, UWB has reveled in controversy and the hyperbole of pundits, some plainly misinformed. The propagandist claim that UWB can transmit information across all possible frequencies at all times probably has its basis in the fact that the PN codes can be as large as a designer wants, hence the dithering range is theoretically infinite. Also, with power levels so low (under 75 nW, typical), who cares if the signal crosses into other bands?
In fact, the Federal Aviation Administration cares. The military cares. And so does any carrier that's overspent on valuable spectrum. The FAA is worried about flight safety, while the military is concerned about interference in both its communications and GPS bands. Given the current political climate, UWB proponents would be wise not to go against their concerns. Many have already openly sided with the U.S. Department of Defense.
As for cellular operators, they've spent billions on spectrum and are unlikely to willingly share it with anyone competing for their customers.
The end result is that UWB will be confined to a very narrow band, if the Federal Communications Commission allows it at all. The regulations will also complicate circuit design and add cost. However, the homodyne UWB transceivers still hold the advantage in cost over their venerable heterodyne cousins.
UWB also has inherently high security features, thanks to the tight control over power limits, which can be adjusted according to range requirements in real-time, as well as the coherent nature of the pulses. The coherency of the impulse allows the receiver to resolve the signal even in the face of severe multipath, while also allowing for pulse addition at the receiver. This allows for lower transmission power levels, which can slightly impact on data throughput rates.
The PN coding gives UWB two of its most exciting, yet overblown capabilities — processing gain and channelization.
Processing gain is defined as the ratio of instantaneous bandwidth to information bandwidth. A figure of merit that can indicate a signal's resistance to jamming runs anywhere from 50 dB to 100 dB for UWB.
But because UWB is unlike direct-sequence, spread-spectrum (DSSS) signals in that it has no compression at the receive end, the processing gain is effectively much lower than 50 dB.
Channelization refers to the number of individual communications channels that can operate over a given frequency range, thanks to the PN coding. UWB's proponents have declared this can number in the thousands, due to the PN coding. But studies have shown that in a practical cellular-like application with a distance of 1 mile, the number drops to under 100 channels.
Regulatory controls for UWB also affect its operating range. If passed by the FCC, UWB will most likely operate above 4 GHz, where its propagation capabilities are severely limited for a given power level, which will also be controlled. That pretty much does away with long-distance communications, but allows for personal area networks.
Many current arguments involving UWB are largely matters of market positioning. UWB proponents claim that GPS is lobbying against UWB solely for fear of competition, due to UWB's position-location capabilities. Yet the NTIA and independent groups have presented ample evidence that shows lower-cost GPS systems are indeed susceptible to interference. Even if UWB were passed, in a region of the spectrum and at power levels that would prevent interference, GPS vendors have nothing to fear, as UWB is being promoted primarily as an indoor, local-positioning technology, which would complement GPS.
