A comparison of UWB radios

Virtually all ultrawideband systems considered before federal UWB regulations were enacted in February 2002 were based on the concept of impulse radio, where information is encoded in a pulse either in position, amplitude or polarity. Since that time, the concept of multiband orthogonal frequency-division multiplexing (MB-OFDM) has developed, based on the principle of sequencing through different frequency bands, each of them 528 MHz wide.

Since multiband and MB-OFDM were new concepts that the Federal Communications Commission had not examined before the ruling, there has been some misunderstanding about how to measure power emission and compare it to that in an impulse radio system. That led the Multiband OFDM Alliance to ask the FCC for a clarification in the form of a waiver.

The purpose of this article is to analyze power emissions and compare them to those in impulse radio systems. To that end, it is useful to understand the concept of power emission when applied to all UWB systems, impulse and MB-OFDM alike. This analysis is particularly important since MB-OFDM systems are expected to ship commercially soon in applications like wireless USB.

Inside UWB systems

All UWB systems use noncontinuous waveforms. As such, their transmit power is a function of measurement bandwidth and measurement time. A comparison of a three-band MB-OFDM system with four impulse radio systems of different pulse repetition frequencies (PRFs) follows.

All systems are set to have the same power and average power-spectral density (PSD), where Pave = -41.25 dBm/MHz. Consequently, they emit the same power and their average PSD is within the value specified by the FCC for the average PSD measurement. The PRFs of the impulse-UWB radios were set at 100 kHz, 1 MHz, 10 MHz and 100 MHz. Their bandwidth is 528 MHz for easy power comparison to the MB-OFDM radio, because that level doesn't require further scaling.

The peak power of the impulse radios using more than 528 MHz would be higher, proportionally to their bandwidth (that is, twice as large for a 1,056-MHz system). The waveforms were then plotted for each radio.

The chart for the 100-kHz PRF system was chopped because of the larger signal amplitude for the 100-kHz PRF system, and the MB-OFDM system looked constant because of its long symbol duration. However, it was obvious from this plot that the impulse radios' peak power, if measured only in the transmit time, decreases with increasing PRF because all systems are assumed to have the same average PSD. In that time the MB-OFDM transmitters always had lower peak power than the impulse radios considered here.

All the impulse radios' spectra were assumed to be flat for ease of comparison. This can be realized with a sync envelope in the time domain and a spreading code to remove spectral lines. In reality it is very difficult to achieve, and this assumption made the impulse radio power profile optimistic.

The next step was to calculate the peak PSD and to compare it among the various systems for different measurement times. Peak power and peak PSD change as a function of the measurement time, given that all UWB systems considered here are not continuous. The peak PSD is constant when more than one pulse is contained in the measurement time; the peak PSD increases when the measurement time gets smaller than the period between pulses.

In the case of the MB-OFDM system, the peak PSD doesn't vary when the measurement time becomes shorter than the pulse. In addition, for MB-OFDM systems, the power stays constant when the measurement time is smaller than the length of the MB-OFDM pulse because the MB-OFDM signal appears continuous in that time frame. This would happen for impulse radios only when the measurement time is smaller than the pulse length-2 nanoseconds, in this case.

This demonstrates that all impulse radios are not continuous signals and that measurement time plays a key role in determining their peak power or peak PSD. It also shows that the transmit power from MB-OFDM transmitters is equal to or lower than that of impulse radios when the measurement time exceeds 1 microsecond, and is always lower than UWB radios with PRF below 1 MHz. For measurement times lower than 1 microseconds, MB-OFDM transmit power is sometimes higher and sometimes lower than other UWB systems, depending on PRFs and measurement times.

As stated in the initial assumptions, all UWB systems in this article have the same power or average PSD by definition. It is now possible to determine how all these systems compared for the peak PSD used by the FCC's peak-power measurement.

When the UWB systems' peak PSD values were calculated, the peak PSD above 0 dBm for a PRF of 100 kHz and 1, 10 and 100 MHz, as well as for the MB-OFDM waveform, was found to be 2.79, -7.27, -17.27, -24.26 and -19.50, respectively. So, the 100-kHz PRF system exceeds the peak PSD value set by the FCC, while all other systems transmit less power.

In summary, MB-OFDM transmit power in normal operation is lower than the average and peak PSD allowed by the FCC measurement limits, while impulse radios that comply with the FCC rules transmit more power than MB-OFDM transmitters under some conditions and less under others.

Roberto Aiello (roberto.aiello@staccatocommunications.com), chief technology officer at Staccato Communications (San Diego).