Page 361 - Fundamentals of Radar Signal Processing

(5.57)

A similar analysis can be used to derive the improvement factor for a three-

pulse canceller; it is

(5.58)

To see how these formulas are used, consider the case where the clutter
spectrum is Gaussian with variance (in normalized radian frequency units) ,
that is, . Assuming σ π so that the continuous-time Fourier
ω
transform pair for Gaussian functions can be used to a good approximation, the

normalized autocorrelation function for c [m] at lag k is (Richards, 2006)

(5.59)

Using Eq. (5.59) in Eqs. (5.57) and (5.58) gives the improvement factor for a
Gaussian clutter spectrum with a two- or three-pulse canceller:

(5.60)

Table 5.1 shows the improvement factor predicted for two- and three-pulse
cancellers for the case of a Gaussian clutter power spectrum of various spectral
widths using Eq. (5.60). If the clutter spectrum is narrow compared to the PRF,

the improvement factor can be 13 dB or more even for the simple two-pulse
canceller. If the clutter spectrum is wide, much of the clutter power will be in
the passband of the MTI highpass filter and the improvement factor will be
small.