UNIT IV RADAR SIGNAL PROCESSING

Radar Signal Processing Fundamentals, Detection strategies, optimal detection, Threshold

Detection, Constant False alarm rate detectors, Adaptive CFAR, pulse compression waveforms,
Compression gain, LFM waveforms matched filtering, radar ambiguity functions, radar
resolution, Detection of radar signals in Noise and clutter, detection of non fluctuating target in
noise, Doppler Spectrum of fluctuating targets, Range Doppler spectrum of stationary and
moving radar.

4.1 SIGNAL MANAGEMENT

Signal management includes everything associated with the waveforms and their
processing that is required for a radar to do its job of detecting and locating targets and
determining something about their nature.
Signal management starts with the design of a suitable waveform and its radiation into space,
the collection by the receiver of echo signals reflected from targets and other objects, the use
of signal processing to extract the desired signal and reject undesired echoes, the use of data
processing to extract information about the detected signals, the coordinated control of these
processes throughout the radar, and keeping within the resources and constraints that affect
signals and their management.

Figure.1 The various elements that enter into radar signal management.
Component Parts of Radar Signal Management

Signal Processing This is processing for the purpose of detecting desired echo signals and
rejecting noise, interference, and undesired echoes from clutter.

It includes the following:

Matched filter: to maximize the signal-to-noise ratio at the output of the radar receiver,
and thus maximize detectability of echo signals.

Detector/integrator: the means for processing in a convenient and efficient manner the
number of pulses received from a target so as to take full advantage of the total signal energy
received from a target.

Clutter reduction: to eliminate or reduce unwanted clutter by one or more methods, of

which filtering of moving targets based on the doppler frequency shift is the most important.

CFAR: used to maintain a constant false-alarm rate at the output of the threshold
detector when the radar cannot eliminate unwanted echoes.

Electromagnetic compatibility (EMC): the elimination of interference from other radars

and other electromagnetic radiations that can enter the radar receiver.

Electronic counter-countermeasures (ECCM): in a military radar, those methods

employed to reduce or eliminate the effectiveness of jamming, deception, and other hostile
electronic active and passive measures whose purpose is to degrade radar performance. ECCM
can be found throughout the radar, not just as part of the signal processing.

Threshold detection: the decision as to whether the output of the radar is a desired
signal.

Data Processing These are the processes that take place after the detection of the desired
signals for the purpose of acquiring further information about the target.

Target location: in range, angle, and sometimes radial velocity (from the doppler shift).
Location information is not generally thought of as either signal processing or data processing.
It is usually obtained as part of the detection process (since detection without location is of no
value).

Target trajectory: also called target track, which is the time history of the target's
location. Usually a prediction of the target's future position is included.
 Target recognition: the recognition of the type of target being viewed by the radar. It
might include the recognition of aircraft from birds, one type of aircraft or ship from another,
recognition of various types of weather, and information about the land and sea environment
(remote sensing).

Weapon control: in military systems, the use of the radar output for the control and
guidance of weapons.

Waveform Design The selection of the waveform depends on what is required of the radar for
detection in noise, clutter, interference, and electronic counter-countermeasures, as well as for
the extraction of information from the radar signal. Waveform design will affect the signal and
data processing.

Antenna This is not just for radiating and collecting radar signals, but is the means by which
angle information is obtained and by which the radar coverage is achieved. The antenna can act
as a spatial filter that can affect the spectral properties of wideband signals. It can also provide,
in some cases, the angle rate and extract a spatial doppler shift similar to the temporal doppler
shift. The target's tangential velocity obtained from the spatial doppler shift, along with the
radial velocity obtained from the more common temporal doppler shift, can provide the vector
velocity of the target.

Automatic Radar Control A radar often employs multiple waveforms and various signal
processing options to maximize performance under a variety of environmental conditions.
Radar control involves the automatic selection of the proper waveform and signal processing
according to the environment and interference (both natural and intentional) encountered by
the radar.

Sensor Integration The outputs from other radars covering the same region may be combined
to form tracks. Information from the civil aviation Air Traffic Control Radar Beacon Systems
(ATCRBS) or the military identification friend or foe (IFF), or other civil or military command and
control information can be used to assist in identifying the target. Noncooperative target
recognition (NCTR) based on special radar waveforms and processing, as well as signals and
information obtained by electronic warfare (EW) methods, such as electronic support measures
(ESM), may be used as part of an integrated military target-recognition system.

Resources for Signal Management - The resources for pursuing the management of signals
and extraction of information are

Energy Sufficiently large transmitted energy is important for detection of weak signals in
noise at long range and for obtaining accurate radar measurements.
 Bandwidth This is the classical measure of information and is especially important for
accurate range measurement and the temporal resolution of targets.

Time is necessary for accurate measurement of the doppler frequency. Time also is a
means for obtaining increased energy when peak power is a limitation. It is important for
achieving multiple functions from a single-beam radar within a required time, and for handling
the processing of many echo signals.

Space This applies to the physical aperture area required for an antenna. The larger the
antenna aperture the greater the echo energy at the receiver and the more accurate the spatial
measurements that can be obtained.

4.2 DETECTION CRITERIA

Detection of signals is equivalent to deciding whether the receiver output is due to noise
alone or to signal plus noise. This is the type of decision made by a human operator from the
information presented on a radar display. When the detection process is carried out
automatically by electronic means without the aid of an operator, the detection criterion must
be carefully specified and built into the decision-making device.

1. Neyman-Pearson Observer
The usual procedure for establishing the decision threshold at the output of the
radar receiver is based on the classical statistical theory of the Neyman-Pearson
observer. This is described in terms of the two types of errors that might be made in the
detection decision process.
One type of error is to mistake noise for signal when only noise is present. It occurs
whenever the noise out of the receiver is large enough to exceed the decision-threshold
level. In statistics this is called a Type I error. In radar it is a false alarm.

A Type II error occurs when a signal is present, but is erroneously considered to be

noise. The radar engineer would call such an error a missed detection.

It might be desired to minimize both errors, but they both cannot be minimized
independently. In the Neyman-Pearson observer, the probability of a Type I error is
fixed, and the probability of a Type II error is minimized.

The threshold level is set by the radar engineer so that a specified false-alarm
probability is not exceeded. This is equivalent to fixing the probability of a Type I error
and minimizing the Type II error (or maximizing the probability of detection), which is
the Neyman-Pearson test used in statistics for determining the validity of a specified
statistical hypothesis. The Neyman-Pearson criterion is employed in most radars for
making the detection decision, whether knowingly or not.
2. Likelihood-Ratio Receiver

The likelihood ratio is a statistical concept that has been used in radar detection
theory and information extraction theory to model optimum decision procedures. It
is defined as the ratio of two probability density functions, with and without signal
present,

where Psn is the probability-density function for signal plus noise and P n is the
probability density function for noise alone.

The likelihood ratio is a measure of how likely it is that the envelope v of the
receiver output is due to signal plus noise as compared with noise alone. If the
likelihood ratio is sufficiently large, it would be reasonable to conclude that a signal
is present.

3. Inverse Probability Receiver

This is another statistical concept and is based on the relationship known as
Bayes' rule for the probability of causes. As with the likelihood ratio, inverse
probability has been used as an analytical basis to model "optimum" receivers for
detection and information extraction. Inverse probability is different from the more
familiar direct probability that describes the chance of an event happening on a
given hypothesis. If an event actually happens (such as a voltage appearing at the
output of the radar receiver), the problem of forming the best estimate of its cause
is a problem in inverse probability.

4. Sequential Observer, Sequential Detection

It provides possibility of a quick decision and makes a detection decision with
fewer pulses. After a single sample of the receiver output, the sequential
observer makes one of three choices:
(1) The sample is due to the presence of signal and noise;
(2) It is due to noise alone; or
(3) It cannot be determined whether it is due to noise alone or signal-plus
noise.

If it can be decided that either No. 1 or No.2 applies, the test is completed and
the radar moves to the next resolution cell to repeat the operation. If the choice is No.
3, a decision cannot be made, and another observation is obtained and the choices
 examined again on the basis of the two observations. This procedure is repeated until a
decision can be made as to whether noise alone or signal-plus-noise is present.

The sequential observer fixes the probability of errors beforehand and allows the
number of observations (integration time) to vary. This procedure theoretically allows a
significant reduction in the average number of pulses (samples) needed for making a
decision.

It describes a two-stage detection process that can be employed with a phased-array

radar. The radar transmits a pulse or a series of pulses in a particular direction, but with
a lower threshold (and higher false-alarm probability) than normal. If no threshold
crossings are obtained, the antenna beam moves to the next position. If a threshold
crossing occurs, a second pulse or series of pulses is transmitted with higher energy, and
with a higher threshold. A detection is declared if the threshold is crossed on both
transmissions.

4.3 CONSTANT-FALSE-ALARM RATE (CFAR) RECEIVER

A target is detected when the output of the radar receiver crosses a predetermined
fixed threshold level set to achieve a specified probability of false alarm. When noise at the
receiver is due to internally generated noise of fixed level described by a gaussian probability
density function, the procedure for setting the threshold is well established. In many situations,
however, clutter echoes and/or hostile noise jamming can be much larger than receiver internal
noise. When this happens, the receiver threshold can be exceeded and many false alarms can
occur, which cause havoc with radar detection and tracking.
Constant False Alarm Rate (CFAR)receiver automatically raises the threshold level to keep
clutter echoes and external noise from overloading the automatic tracker with extraneous
information.

Constant False Alarm Rate (CFAR) detection also frequently referred to as “adaptive
threshold detection” or “automatic detection” is a set of techniques designed to provide
predictable detection and false alarm behavior in realistic interference scenarios.

CFAR is achieved, of course, at the expense of a lower probability of detection of desired

targets. In addition, CFAR can also produce false echoes when there is nonuniform clutter,
suppress nearby targets, and degrade the range resolution.

Cell Averaging CFAR The major form of CFAR has been the cell-averaging CFAR. It is illustrated
in Figure. It uses an adaptive threshold whose level is determined by the clutter and/or noise in
the vicinity of the radar echo. Two tapped delay-lines sample echo signals from the
environment in a number of reference cells located on both sides of the test cell (the range cell
of interest). The spacing between reference cells is equal to the radar range resolution (usually
the pulse width). The output of the test cell is the radar video output, which is compared to the
adaptive threshold derived from the sum of the outputs of the tapped delay lines defining the
reference cells. This sum, therefore, represents the radar environment to either side of the test
cell. It changes as the radar environment changes and as the pulse travels out in time. When
multiplied by a predetermined constant k, the sum provides an adaptive threshold to maintain
a constant false-alarm rate. Thus the threshold can adapt to the environment as the pulse
travels in time.

Figure Cell averaging CFAR

If the radar output is noise or clutter described by the Rayleigh probability density function, the
constant k which multiplies the sum of the tapped delay lines can be determined from classical
detection theory. When the statistics of the clutter are not known, which is often the case, the
value of k can only be estimated or some form of nonparametric detector used.

CFAR Loss The greater the number of reference cells (delay-line taps) in theCA CFAR the better
is the estimate of the background clutter or noise and the less will be the loss in detectability
(signal-to-noise ratio).

The CFAR loss is the signal-to-noise ratio required when CFAR is employed divided by the signal-
to-noise ratio required for fixed-threshold detection.

Nitzberg shows that the CFAR loss for single-pulse detection (N = 1) can be approximated by

wherePfa is the probability of false alarm.

4.4 ADAPTIVE CFAR
Adaptive CFAR is designed to improve the performance in non-homogenous clutter.
Adaptive CFAR is an extension of CFAR that dynamically adjusts the threshold based on the
local clutter environment. Instead of using a fixed threshold for the entire range, Adaptive CFAR
adapts to changes in clutter conditions by estimating the statistical properties of the clutter in
the vicinity of each cell under test. The goal is to maintain a constant probability of false alarms
(Pfa) while accounting for variations in clutter characteristics.

4.5 PULSE COMPRESSION IN RADARS

Pulse compression allows a radar to transmit a pulse of a relatively long duration and
low peak power to achieve large radiated power, but simultaneously to obtain the range
resolution of a short pulse. This is accomplished by utilizing frequency or phase modulation to
increase the signal bandwidth and then compressing the received echo waveform in a matched
filter.
The range resolution achievable with a radar system is

In a pulse compression system, transmitted waveform is modulated either in frequency or

phase, so that the B >> 1/τ. If τc represents the effective pulse width of the resultant
compressed pulse implying τc = 1/B, range resolution becomes

Thus the pulse compression radar can utilize a transmit pulse of duration τ and yet achieve a
range resolution corresponding to the compressed pulse of duration τc.

Dispersion factor or compression ratio (CR) is defined as

which indicates that the compression ratio also equals the time-bandwidth product of the
system.

There are many types of modulations used for pulse compression, but two that have wide
applications are the frequency modulation technique and the phase modulation technique.
Frequency Modulation in Pulse Compression

In the frequency modulation technique, the transmitter frequency is modulated to increase the
bandwidth of the transmitted waveform, while the received echo waveform is compressed.
There are two types of frequency modulation techniques using: linear frequency modulation
(LFM) and frequency stepping.

Linear Frequency Modulation (LFM) in Pulse Compression

The LFM or chirp waveform is the easiest to generate. It is more popular than any other coded
waveforms because the compressed-pulse shape and SNR are fairly insensitive to Doppler
shifts. Pulse compression in LFM achieves increased transmitter power and constant bandwidth
by a linear frequency modulation of the transmitted waveform and a receiver delay network
utilizing frequency-time characteristics of the transmitted waveform. But there are some major
disadvantages: it has excessive range-Doppler cross coupling errors, and it requires weighting to
reduce the sidelobes of the compressed pulse to an acceptable level.

The frequency-time characteristics of the transmitted signal are shown in Figure a. The
transmitted pulse is of duration τ = t2 − t1 and is linearly frequency modulated from f1 to f2 over
the pulse length by an amount ∆f as depicted in Figure b.

The effect of frequency modulation on the transmitted signal is shown in Figure c. The target
return signal will be similar to the transmitted signal.

The pulse compression filter in the radar receiver is matched to the transmitted waveform so
that the received signal experiences a frequency-dependent time delay as given in Figure d. It is
observed that the lowest frequency f1 is delayed the longest, while the highest received
frequency f2 is not delayed at all.

The final output is a short pulse of duration τ = 2/∆f and with a large amplitude of √CR, as given
in Figure e. The compressed pulse of Figure e displays the amplitude characteristics after
passage through the matched filter or dispersive delay line.

The sidelobes of the amplitude–time characteristics are often undesirable since they may result
in false detections. These sidelobes can be reduced by amplitude weighting of the received
signal, and will be discussed in a subsequent section. The effect of weighting the received signal
to lower the sidelobes, unfortunately, widens the main lobe and reduces the peak SNR
compared to the unweighted LFM pulse compression.
Figure. Linear FM pulse compression. (a) transmitted pulse, (b) transmitted frequency, (c)
transmitted waveform, (d) delay network, and (e) compressed pulse

Phase coded Modulation in Pulse Compression

In the phase-coded modulation technique, the long pulse is subdivided into an N number of
shortersubpulses of equal time duration T. The modulation, however, is not of the frequency of
each phase,but of the phase of each pulse selected in accordance with a phase code.

4.6 MATCHED FILTER

The matched filter is a filter that produces the maximum attainable SNR at the output
when both useful signal and additive white noise are present at the input. In radar applications
matched filters are used extensively to improve the SNR, and are of paramount importance.
The maximum SNR at the output is achieved by matching the radar receiver transfer function to
the received signal. Figure 8.1 shows the basic block diagram of a matched filter radar system.
Figure. Basic block diagram of a matched filter radar system.

The transmitted waveform is generated by a signal generator designated as s(t) . The signal
output s(t) is amplified, fed to the antenna, radiated, reflected from a target, and then received
by the receiver. The output of the receiver is fed into the matched filter after suitable
amplification. The matched filter impulse response h(t) is simply a scaled, time-reversed, and
delayed form of the input signal. The shape of the impulse response is related to the signal and,
therefore, matched to the input. The matched filter has the property of being able to detect the
signal even in the presence of noise. It yields a higher output peak signal to mean noise power
ratio for the input than for any other signal shape with the same energy content.

Consider that the transmit signal s(t) with white additive white Gaussian noise with a two-sided
uniform spectral density N0/2 that is passed through a matched filter with a frequency transfer
function H(ω). We want to find the H(ω) that will maximize the SNR at a given observation time
t0. More precisely,

Where s0 (t0) is the filter output corresponding to the input signal s(t) at the observation time t 0
n02( t) is the average power of the white noise at the output of the filter.
Let the Fourier transform of s(t) be S(ω). Then the output signal of the filter output prior to
envelope detection is

The power spectral density (in W/Hz) of the noise n(t) at the filter output, denoted by N(ω) , is

The average noise output power is then

When simplified, this yields

Substituting the above values in SNR equation, we get

We will now make use of the Schwartz inequality, which states that for any two complex signals
A(ω) and B(ω), the following inequality is true

This equality holds if and only if

Where ∗ denotes the complex conjugate, A0 is an arbitrary constant and can be assumed to be
unity.
Applying the Schwartz inequality to SNR equation yields

If we assume that equality occurs at t = t0 and A0 = 1, then

where we have used the expression of the energy of the signal given by

Thus the maximum output SNR depends only on the signal energy and input noise power. It is
also observed that the maximum SNR is not dependent on the modulation type or form of
transmitted signal. Consequently the signal bandwidth, type, and total energy can all be
independently selected when a matched filter is used. In fact, most of the modern radar uses a
matched filter in its receiver.
The equality in SNR equation holds when

Thus the matched filter derives its name from the fact that its transfer function is proportional
to the complex conjugate of the transmitted signal’s Fourier transform. The filter must change
with the change of the input signal at the filter input. Also the negative of time t indicates that
the impulse response is proportional to the input signal “running backward.” In other words,
the impulse response is the delayed mirror image of the conjugate of the signal.

4.7 THE RADAR AMBIGUITY FUNCTION

The ambiguity function is defined as the absolute value of the complex envelope of the
output signal obtained when the matched filter is fed by a Doppler-shifted version of the
original signal, to which the filter is matched. The mathematical functions called time-frequency
autocorrelation functions or ambiguity functions were first introduced by Woodward to allow
representation of the time response of a signal processor when a target has significant radial
velocity. The radar ambiguity function is used by radar designers to obtain a better insight
about how different radar waveforms may be suitable for the various radar applications. It is
also used to determine the range and Doppler resolution for a specific radar waveform.

Let g(t) be the envelope of the signal, then we define a special function of the complex
envelope, denoted by χ(τ,ωd), at time τ in response to a transmit waveform that has been
Doppler shifted by ωd Hz is given by

The amplitude |χ(τ,ωd)|is sometimes called the uncertainty function, and the square of the
amplitude |χ(τ,ωd)|2 is called the ambiguity function. The filter is matched to the signal
expected at a nominal center frequency and a nominal delay. The two parameters of the
ambiguity function are the additional delay τ and additional Doppler shift ωd . Therefore,
|χ(0,0)| becomes the output when the input signal is returned from a point target at the
nominal delay and Doppler shift for a matched filter. At any other values of τ and ω d other than
zero indicate a return from a target at some other range and velocity.

Now denote E as the energy of the complex signal envelope g(t) given by
We will assume that the complex signal envelope g(t) has a unit energy. For such normalized
signals the following properties of the ambiguity function can be listed:

Property (1) indicates that the maximum value of the ambiguity function occurs at (τ,ω d) = (0,0)
and for the normalized signal; the maximum value of the ambiguity function is 1.

Property(2) indicates that the total volume under the ambiguity function is a constant.

Property(3) implies that the ambiguity function is symmetric.

Property(4) indicates that multiplying the envelope of any signal by a quadratic phase will shear
the shape of the ambiguity function. This is applied to a pulse compression technique called
linear FM.

The ideal radar ambiguity function is represented by a spike of infinitesimal width that peaks at
the origin and is zero everywhere else, as illustrated in Figure. An ideal ambiguity function
provides perfect resolution between neighboring targets regardless of how close they may be
with respect to each other.

Figure. The ideal ambiguity function

4.8 RADAR RESOLUTION

The resolution of radar is its ability to distinguish between targets that are very close in either
range or bearing. Radar resolution is usually divided into two categories; • range resolution and
• angular (bearing) resolution.

Range resolution The range resolution of the radar defines is the ability to distinguish between
different targets placed in the same angular direction (bearing) but at different distances from
the radar. The signals reflected by two targets are shifted in the temporal domain.

After performing an FFT (Fast Fourier Transformation) two separate baseband frequencies
(frequency shift) corresponding to each target are obtained. The closer the two targets, the
smaller the difference between the baseband frequencies ∆f1 and ∆f2.

The range resolution dres can be written as a function of the signal bandwidth

Angular resolution is the ability of radar to distinguish between two separate targets placed at
different angles (bearings), but at the same range. Here, the frequency shift due to the signal
delay cannot be used to identify the signals from each of the targets. Again, special diversity is
needed, with multiple antennas at different locations.

The angular resolution θres indicates then the minimum angular separation that can be resolved
by the radar. For a single transmit antenna it can be calculated as:

N being the number of receive antennas and θ the angle of arrival.

4.9 DETECTION OF TARGETS IN CLUTTER

The various techniques that might allow the detection of desired targets in the presence of
undesired clutter echoes.
 1. Doppler frequency Shift The doppler frequency shift of the echo signal is widely used
for separating a moving target's echo signal from large, unwanted stationary clutter
echoes. Doppler filtering allows echoes of moving targets to be separated from those of
stationary clutter even though the clutter might be greater by many orders of
magnitude In spite of limitations, doppler filtering works well and can provide far
greater suppression of the clutter echoes than any other technique.

2. High Resolution The smaller the radar resolution cell, the less will be the competing
clutter echo. Short pulses (or the equivalent pulse compression) and narrow-beamwidth
antennas, therefore, can reduce the amount of clutter with which the target echo signal
must compete. High resolution usually is the preferred method for detection of ships.

High resolution is usually a good method for enhancing the signal-to-clutter ratio, even
when the statistics of the clutter are not Rayleigh. It is one of the few methods available
for increasing the target-to-clutter ratio when the target is stationary (no doppler shift).

3. Frequency Radar echoes from rain, sea (with horizontal polarization), and (to some
extent) land decrease with decreasing frequency. Although this favors the lower
frequencies, it is more difficult at the lower frequencies to achieve narrow beam widths
and high range resolution. Thus some of the benefits of reduced clutter at the lower
frequencies might be partially offset by the poorer resolution that results.

4. Polarization Circular polarization is often employed to reduce weather clutter in

microwave air-surveillance radars. Discrimination of a target located in the midst of rain
results because the echo from an aircraft (an asymmetrical scatterer) differs from that
of a raindrop (a circularly symmetric scatterer) when using circular polarization. This
difference in scattering can be used to enhance the echo from the target and suppress
echoes from rain.

5. Time Decorrelation Unlike receiver noise, clutter echoes are generally correlated from
pulse to pulse, and sometimes even from scan to scan. The techniques of rapid antenna
scan for detection of small targets in the sea, and time compression for detection of
moving targets in patchy land or sea clutter are examples of detection techniques.

4.10 DETECTION OF SIGNALS IN NOISE

The ability of a radar receiver to detect a weak echo signal is limited by the ever-present noise
that occupies the same part of the frequency spectrum as the signal. The weakest signal that
can just be detected by a receiver is the minimum detectable signal.
Threshold Detection

Detection of a radar signal is based on establishing a threshold at the output of the receiver. If
the receiver output is large enough to exceed the threshold, a target is said to be present. If the
receiver output is not of sufficient amplitude to cross the threshold, only noise is said to be
present. This is called threshold detection.

Figure 2.1 represents the output of a radar receiver as a function of time. The fluctuating
appearance of the output is due to the random nature of receiver noise.

Figure. Envelope of the radar receiver output as a function of time. A, B, and C represent signal
plus noise. A and B would be valid detections, but C is a missed detection.

When a large echo signal from a target is present, as at A in Figure, it can be recognized on the
basis of its amplitude relative to the rms noise level. If the threshold level is set properly, the
receiver output should not normally exceed the threshold if noise alone were present, but the
output would exceed the threshold if a strong target echo signal were present along with the
noise.

If the threshold level were set too low, noise might exceed it and be mistaken for a target. This
is called a false alarm.

If the threshold were set too high, noise might not be large enough to cause false alarms, but
weak target echoes might not exceed the threshold and would not be detected. When this
occurs, it is called a missed detection.

In early radars, the threshold level was set based on the judgment of the radar operator
viewing the radar output on a cathode-ray tube display. In radars with automatic detection
(electronic decision making), the threshold is set according to classical detection theory.

A threshold level is shown in Figure by the long dash line. If the signal is large enough, as at A, it
is not difficult to decide that a target echo signal is present. But consider the two weaker signals
at B and C, representing two target echoes of equal amplitude. The noise accompanying the
signal at B is assumed to be of positive amplitude and adds to the target signal so that the
combination of signal plus noise crosses the threshold and is declared a target.

At C the noise is assumed to subtract from the target signal, so that the resultant of signal and
noise does not cross the threshold and is a missed detection. The ever-present noise, therefore,
will sometimes enhance the detection of marginal signals, but it may also cause loss of
detection. The signal at C would have been detected if the threshold were lower. But too low a
threshold increases the likelihood that noise alone will exceed the threshold and be improperly
called a detection.

The selection of the proper threshold is therefore a compromise that depends upon how
important it is to avoid the mistake of (1) failing to recognize a target signal that is present
(missed detection) or (2) falsely indicating the presence of a target signal when none exists
(false alarm).

4.11 Fluctuating targets

A "fluctuating target" refers to a target whose radar cross-section (RCS) or radar
echo varies over time. Radar cross-section is a measure of how detectable an object is
by radar, and it depends on various factors such as the object's size, shape, material
composition, and orientation relative to the radar beam.

Fluctuating targets can be categorized into four different Swerling types : Swerling I , Swerling
II, Swerling III, Swerling IV.
Doppler Spectrum of fluctuating targets
For a fluctuating target, variations in the Doppler spectrum can provide
information about the target's motion, rotation, or other dynamic characteristics.

Figure. Spectrum of fluctuating and non fluctuating data

4.12 Range Doppler spectrum of stationary and moving radar

The Doppler spectrum is a graphical representation of the power distribution of the

radar signals across the Doppler frequency range. It shows how many targets and how much
clutter are present in each velocity bin. The shape and width of the Doppler spectrum depend
on the characteristics of the targets, such as their size, shape, orientation, and number. The
Doppler spectrum can be used to analyze the target motion and identify the dominant velocity
components.

The spectrum of the slow time signal from a single range bin consists of noise, clutter and on or
more target signals. The Figure shows the notional generic Doppler spectrum observed from
stationary radar. This spectrum is periodic with a period equal to the pulse repetition
frequency (PRF), so only the “principal period” from –PRF/2 to PFR/2 is shown.

Receiver noise is spread uniformly throughout the spectrum. Clutter occupies a portion of the
spectrum. The portion of the spectrum where clutter is the dominant interference is often
termed the clutter region. The portion where noise is the dominant interference is called the
clear region. The clear region is clear of clutter but not of all interference. A skirt region is
defined at the transition between the clutter and clear region; in the skirt region, both noise
and clutter are significant interference sources. Sattionary targets will appear at zero Doppler
shift, while moving targets can occur anywhere in the spectrum as appropriate to their radial
velocity relative to the radar.

Figure. The principal period of a notional generic Doppler spectrum for a stationary radar,
containing noise, clutter and target components.

For moving radar, the entire Doppler spectrum is shifted by the nominal radar to ground
Doppler shift of FD = 2v cos ψ / λ hertz while the main lobe is widened by BD = 2vθ sin ψ / λ
hertz, where ψ is the angle between the radar line-of-sight vector and the platform velocity
vector, and θ is the antenna beamwidth in radians.

In addition to this broadened main lobe clutter (MLC), side lobe clutter (SLC) and an altitude line
(AL) are now evident. Side lobe clutter is clutter from echos resulting from energy radiated and
received through the radar side lobes. It is thus weaker than the main lobe clutter.

Figure. Notional Doppler Spectrum for Moving radar platform.

Doppler processing
Doppler processing is the process of using a Doppler filter and Doppler spectrum to
distinguish moving targets from stationary ones. It involves transmitting a pulse or continuous
wave of a known frequency and recording the received signal. The Doppler frequency of the
received signal is then calculated by comparing it to the transmitted frequency. This signal is
then passed through a Doppler filter bank and its output is plotted as a function of the Doppler
frequency to obtain the Doppler spectrum. Peaks and valleys in this spectrum are associated
with target velocities and clutter levels, allowing for identification.