Page 215 - Fundamentals of Radar Signal Processing

major advantage of this technique is that it can be applied individually to each

pulse of data. Its major disadvantage is that it requires the transmitter/receiver
control and analog hardware be augmented to allow pilot signal insertion for
determining the correction coefficients. The pilot signal operation is performed
relatively infrequently on the assumption that ε and ϕ vary only slowly.
The phase rotation and integration technique of Eqs. (3.47) to (3.49), in
contrast, requires integration of at least three pulses with the transmitted phase

adjusted for each pulse. Thus, the technique requires both high-speed transmitter
phase control and more time to complete a measurement since multiple pulses
must be collected. The increase in required time implies also an assumption that
the scene being measured does not vary during the time required for the multiple
pulses; decorrelation of the scene degrades the effectiveness of the technique.
This method also places a heavier load on the signal processor, since the
integration requires N complex multiplies and N – 1 complex additions per time

sample, or a total of 4N real multiplies and 4N – 2 real additions, with N ≥ 3.
However, the integration method has one very important advantage: it does not
require knowledge of any of the errors ε, ϕ, γ, and κ. It also has the side benefit
that the integration of multiple pulses increases the signal-to-noise ratio of the
final result x(t). Given these considerations, it is often used in instrumentation
systems at fixed site installations, such as turntable RCS measurement facilities.

In these systems, N is often on the order of 16 to 64, and may even be as high as
65,536 (64K) in some cases.
Note also that Eqs. (3.47) to (3.49) implicitly assume that the phase
modulation θ(t) is the same for each pulse x (t). If θ(t) represents waveform
k
modulation (e.g., a linear FM chirp), this will be true; but if θ(t) contains a term
representing environmental phase modulation, for example due to Doppler shift,
then the technique assumes that the appropriate component of θ(t) is the same on

each of the pulses integrated. This is the case for stationary targets (assuming the
radar is also stationary). For constant Doppler targets, the frequency implied by
θ(t) will be the same from pulse to pulse, but the absolute phase will change in
general, so that the target response does not integrate properly. For accelerating
targets, the assumption will fail entirely. The phase rotation and integration

technique is therefore most appropriate for stationary or nearly stationary (over
N PRIs) targets. The algebraic technique does not have this limitation, since it
operates on individual pulses only.

3.4.3 Digital I/Q
Digital I/Q or digital IF is the name given to a collection of techniques that
form the I and Q signals digitally in order to overcome the channel matching
limitations of analog receivers. A number of variations have been described in
the literature. In general, they all share two characteristics. First, they use

analog mixing and filtering to shift the single real-valued input signal to a low
intermediate frequency prior to A/D conversion, greatly relaxing the A/D speed