Page 194 - Fundamentals of Radar Signal Processing

dimensional subvectors or two-dimensional submatrices of the datacube in

various dimensions. Figure 3.9 illustrates these relationships. The particular
operations depicted are discussed in upcoming chapters. For example, pulse
compression (see Chap. 4) is implemented as a one-dimensional convolution on
a single vector in the fast-time (range) dimension. Pulse compression can be
performed independently on each such range vector for every pulse and receiver
channel.

FIGURE 3.9 Correspondence between key radar signal processing functions
and operations on the radar datacube.

3.1.5 Dwells
A dwell, sometimes also called the time on target, is another term for a radar
data collection interval. Like the term CPI, a dwell can refer both to a time
interval and to the data collected within that time. It is defined in the IEEE
standard for radar definitions as “a data acquisition interval during which the

data is usually processed together for detection or measurement” (IEEE, 2008).
Consider a rotating radar like that shown in Fig. 2.15a and imagine that the
beamwidth is 3° and the antenna scans at a rate of 60° per second. A point target
would be in the beam for 50 ms during each scan. If the radar has a PRF of 2000
pulses per second, a target would be illuminated with 100 pulses during the
traversal of the mainbeam. Since it is known that the echo of a target, if present,
would be present in 100 successive pulses, it might be sensible to integrate 100

pulses at a time for SNR improvement before performing a detection test. In this
scenario, the dwell time would be the full 50 ms. If only 50 pulses at a time
were integrated, the dwell time would be considered to be 25 ms.
The idea of a dwell is not limited to mechanically scanning radars.
Consider a pulsed Doppler radar with an electronically scanned antenna pointed