Page 72 - Complete Wireless Design

Modulation

Modulation 71

10. Thus OQPSK does not allow the carrier to be snuffed, and will have carri-
er amplitude variations of, at most, only 3 dB. As stated above, this permits the
use of more efficient nonlinear amplifiers, with higher output powers possible,
and with less spectral regrowth. Another similar QPSK modulation format,
referred to as /4 DQPSK, will also not let the carrier be completely snuffed
out, while also allowing for easier clock recovery during demodulation. Other
modulation schemes have a modulation envelope that has a completely con-
stant envelope, and can thus employ very efficient nonlinear Class C amplifiers
without any spectral regrowth. Two of these modulations are minimum shift
keying (MSK) and its derivative gaussian minimum shift keying (GMSK).
Digital power measurement. The peak amplitude of a digital signal is complete-
ly unpredictable, and will vary dramatically over time, because of its noiselike
nature. The average power output of an analog amplitude-modulated trans-
mitter will vary depending on the baseband waveform, and its peak envelope
power can easily be measured over a single cycle of the waveform (and need not
be averaged over time) by commonly available, low-cost test equipment. The
situation is not quite the same with digitally modulated signals. To find the
peak amplitude of these erratic digital signals, the power measurement must
be taken over time to obtain a statistical peak amplitude, since there is only a
statistical chance that a much higher peak will come along at any time. This
peak reading is then compared to the digital signal’s average power (the aver-
age power is the same power that a DC signal would require to heat up a resis-
tance element to the same temperature as the RF signal). With this type of
average power measurement, any wave shape can be gauged.
If an RF signal did not vary its power over time (as with DC signals, with
P DC(PEAK) P DC(AVG) ), then its average power would be delivered constantly to
the load, with its peak power equaling its average power. But since RF sig-
nals do vary (as any AC signal will), their peak power will be different from
their average power. This ratio between the peak and the average values of a
modulated signal is referred to as the peak-to-average ratio. The lower this
ratio, the closer to the P1dB level an amplifier can be driven without produc-
ing excessive intermodulation distortion (IMD) products, because the occa-
sional power peaks will be lower in amplitude with a modulation format that
has a lower peak-to-average ratio (see “P1dB compression point test” in Sec.
2.6.2). Thus the amplifier will not have to have as much of a power margin to
gracefully accept these higher-amplitude power spikes without creating inor-
dinate IMD. A signal with very high absolute peaks, but a low average pow-
er, will have a very poor peak-to-average ratio, forcing any amplifier that is
working with this type of signal to have a large amount of reserve power to
amplify these occasional peaks without excessive distortion. However, while
the peaks may be erratic over time—since the peaks are caused by the mod-
ulation shifting from certain constellation points to other constellation
points—the average power in a digital signal will be constant. This is due to
the digital signal’s encoding.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2004 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.