Interference, Multipath, and Scintillation - Understanding GPS: Principles and Applications - page 253

nated by the statistics of the CW signal. The probability density of a CW signal
amplitude is

1

()

Px = (6.3)

π 1 − x

2

This function is plotted in Figure 6.2. Observe in this plot that the CW signal
spends most of its time near the peak amplitudes rather than in the vicinity of the
zero crossing. The result is that the combination of the signals plus thermal noise
plus a strong CW signal spends very little time near the zero crossing.
Virtually all modern receivers are precorrelation ADC designs, so the ADC pro-
cess takes place prior to the digital correlation process. If the ADC is properly
designed, the digital correlation process that follows spreads the CW signal into a
wideband signal, while despreading the signal into a narrowband (data modula-
tion) signal with a bandwidth that is essentially the reciprocal of the predetection
integration time. This is a significant signal bandwidth reduction and CW interfer-
ence bandwidth expansion. Filtering the resulting narrowband signal provides sig-
nificant processing gain against the CW, but this is only true if the CW signal does
not first capture the ADC. The ADC is subjected to the full amplitude of CW inter-
ference in a precorrelation ADC receiver. In the case of a 1-bit ADC (limiter), there
is very little correlation possible in the presence of CW interference. The CW inter-
ference essentially captures the 1-bit ADC in a precorrelation ADC receiver design.
This increases the 1-bit ADC SNR degradation from 1.96 dB in the presence of
Gaussian noise to 6.0 dB in the presence of any constant envelope (including CW)
interference [3]. The nonuniform 2-bit ADC of Figure 6.1 reduces this degradation
to less than 0.6 dB in the presence of Gaussian noise, and in the presence of CW

7

6

5

4

3

2

1

0

− 1.5 − 1 − 0.5 0 0.5 1 1.5

x = CW jammer amplitude

Figure 6.2 Probability density of a CW (sinusoidal) signal.