Pri Analysis and Deinterleaving by DR Wiley PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 76
At a glance
Powered by AI
The key takeaways are that Pulse Repetition Intervals (PRIs) are important for identifying radar signals, and deinterleaving signals from multiple radars. Radar constraints like range, resolution and bandwidth must be understood to properly analyze ELINT signals.

Pulse Repetition Intervals (PRIs) refer to the time between pulses in a radar transmission. Analyzing PRIs is important for identifying different radar systems and deinterleaving signals. Shorter PRIs allow for faster scanning but reduce maximum detection range.

Radar designers face constraints related to range, resolution and bandwidth. Long detection ranges require high transmitter power but reduce resolution. Wide bandwidths improve resolution but are difficult to implement. These constraints impact what ELINT signals can be expected.

PRI Analysis and Deinterleaving

Richard G. Wiley, Ph.D.


Research Associates of Syracuse, Inc
111 Dart Circle
Rome, NY 13441
315-685-3135; [email protected]
1

Pulse Repetition Intervals (PRIs) are often the key to


identifying the signals of many radar systems. The first step is
to deinterleave signals from multiple radar systems. This
briefing is a a brief introduction to PRI analysis and
deinterleaving from the ELINT/EW point of view

2
PULSE REPETITION INTEVAL (PRI)

ELINT Implications of Range Equations and Radar Constraints

The effects of the one-way range equation of ELINT and the two-
way range equation of radar on signal strength must be understood
and explored in order to appreciate the typical situations
encountered in ELINT and EW. Similarly, the constraints placed on
radar waveforms must be understood in order to correctly interpret
the functions and applications of the signals transmitted by radar
and also to be aware of the signal characteristics expected to be
encountered by ELINT. In many ways, understanding these aspects
of ELINT is what separates one who only observes signals from one
who both observes and analyzes signals.

Reference: ELINT, Chapter 2

4
Radar and ELINT Range Equations

PT GT G R O V 2
SR
(4S ) R R LT LR
3 4

PT GTE G E O2
SE
(4S ) R E LT LE
2 2

Ratio of ELINT Range to Radar Range


A significant aspect of these range equations is that the power
level transmitted by pulsed radar transmitters in order to detect
targets at long range is very high. This allows ELINT receivers to
detect radar signal at very long ranges even when observing the
sidelobes of the radar’s transmit antenna.
To simplify the discussion, suppose that the ELINT receiver
requires a signal level that is a factor G times the signal level
needed by the radar receiver, that is:
SE G (S R )
1/ 2
RE ª 4S 1 GTE G E LE º
RR « »
RR ¬ G V G T G R L R ¼
6
3
1 10
1/ 2 1/ 2
RE ª 4S º RE RR ª 4S º
RR « »
RR ¬ G RG ¼ B RR G R «¬ G »¼
0d
E
=3
GT
m:
ea
ELINT Range/Radar Range

100
in b
Ma
RangeRatioSLi
dB
RE/RR

RangeRatioMBi E
=0
: GT A
e
e lob
10 Sid
V  1 sq. m
G  30 dB
R
G 100
G  1
E

1 3
1 10 100 1 10
Ri
Range (km)
Figure 2-1 ELINT to Radar Range Ratio

2.2 Radar Constraints

ELINT signals of interest include radar signals of all types.


Sometimes, people concerned about ELINT attribute properties
to radar signals that are contrary to the constraints under which
radar systems must function. Avoiding this pitfall is an
important aspect of ELINT work. Understanding the
fundamental limitations faced by radar designers and the
associated ELINT implications is important. Consider this
statement: “Radars of the future could transmit noise
waveforms over GHz bandwidths and be undetectable by
ELINT receivers.” Should ELINT equipment be developed to
intercept and process this kind of signal? Probably not--
because signals like this would not be useful for tracking or
search radars in military applications.
8
Range Resolution related to Bandwidth

Range resolution in radar is inversely proportional to the


bandwidth of the signal (assuming that it is processed
coherently). The fundamental relationship is: c
'R
2B
Here c is the speed of light and B is the bandwidth of the signal
during the coherent processing interval; also called its
instantaneous bandwidth.
For example, to distinguish between two fighters in tight
formation 30m apart in range, BW must be about 5MHz. If one
postulates a value of B=1 GHz, the radar has a range resolution
of 15 cm. This means that the target echoes are resolvable in
15 cm range increments called range cells. The echoes from a
75m target are spread across 500 range cells.
9

3
1 10
Resolution (meters)
(meters)

100
RangeResolution

RngRes bi
Range

10

1
6 7 8
1 10 1 10 1 10
bi
Bandwidth B(MHz)
Bandwidth (MHz)

Figure 2.2. Range resolution Related to Radar Coherent Bandwidth


10
This spreading of the echoes across a multiplicity of range cells reduces the apparent radar
cross-section (and thus reduces the SNR available) in a single range cell. For this reason,
radar designs generally have range resolution appropriate for their function. This leads to
choosing coherent bandwidths of 10 MHz or less. (10 MHz corresponds to range
resolution of 15 m.) In this sense, there is no such thing as a “spread spectrum” radar—
what is transmitted is also received and the resulting range resolution is determined by the
bandwidth. What this means for ELINT is that the coherent bandwidth of radar signals is
likely to remain the same as it is now provided the radar performs the same task.

Range Resolution Required Resolution (m) Bandwidth (MHz)


30 5
1.Count A/C in attack 60 2.5
formation
2. Detect missile 15 10
separation at launch
3. Imaging of Ships, .5-1 150-300
Vehicles and Aircraft
4. High Resolution 0.15 1000
Mapping
11

Moving Targets and Integration Time Constraints


If a radar is to detect targets moving in a radial direction (toward or
away from the radar), the amount of time the target will be present in a
given range cell is determined by the target velocity and the range
resolution. This limits the coherent integration time of present day
radars to GR 'R
TCV 
v v
Here TCV is the maximum coherent integration time for a constant
velocity target with radial velocity v and įR is the change in range
during that time. If the target is accelerating in the radial direction,
the maximum integration time is now a quadratic function of both
velocity and acceleration

T ACC

v  v  2a(GR)
2
0.5



v  v  2a('R)
2
0.5

a 12
a
Constraints on Time-Bandwidth Product or Pulse Compression
Ratio
Because range resolution is determined by bandwidth and integration
time is determined by velocity, there is a natural limit on the product
of the instantaneous bandwidth and the duration of the coherent
processing interval or pulse width. This is called the "time-
bandwidth product." The radar's pulse compression ratio is limited to
no more than its time bandwidth product. By combining Equations
for range resolution and integration time it is easy to see that the time
bandwidth product is limited to:

Bv § ac · c
BT  ¨¨ 1  2  1¸¸ 
a o0
o
a © Bv ¹ 2v

13

6
1 10
BT Limit product BT

BT i  1 a=0 g

BT i  2 a=1 g
Maximum time-bandwidth

g
BT i  5 a=2
g
BT i  10 a=5
g
BT1 i 10
a=

Acceleration 0, 1,2, 5, 10 g's


Velocity=300m/s

5
1 10
4 5 6
1 10 1 10 1 10
bi
Signal Bandwidth B (Hz)
Bandwidth
Figure 2-414
Limit on Time x Bandwidth
Constraints on Doppler Resolution
If the radar coherently integrates the echoes in one range cell for the
entire integration time, the minimum doppler filter bandwidth, Bf, is
approximately the reciprocal of the integration time,.T, which is
either TCV for constant velocity targets or TACC for accelerating
targets:.
1
Bf |
T
However if the target is accelerating, the doppler shift changes.
Clearly there is a relationship between acceleration and the time the
doppler shift of the moving target remains within the doppler filter
bandwidth.
2aTf o 2aT
'f acc  Bf
c O
15

Because the coherent integration time is approximately equal


to 1/Bf, substituting Bf=1/T into 2-12 gives the maximum
allowable coherent integration time and the minimum doppler
filter bandwidth as

O 2a
T , Bf !
2a O

16
4
1 10
3
6.502 u10

3
a=10g
1 10
Doppler Spread( kHz)

'f i  1
100
'f i  2
a=1g
'f i  5
10
'f i  10

0.65
0.1 3
1 10 0.01 0.1 1

3 Ti 1
1u10

Coherent Integration time T (s)

Figure 2.5 Doppler Spread and Maximum Signal Bandwidth

17

4 3
1 10 1 10
3 1000
6.502u10

a=10g
le
t sca
3
1 10
a=5g
ad - ri gh
r Sp r e a=2g g
ple a=1
Doppler Spread( kHz)

100
Dop
'f i  1

Bandwidth (MHz)
100
'f i  2
Bi
'f i  5 Ma
10
xi m
'f i  10 um
Sig 10
n al B
and
w idt
1
h -le
f t sc
ale
0.65 1
0.1 3 1
1 10 0.01 0.1 1

.001 Ti 1

Coherent Integration time T (s)

Figure 2.5 Doppler Spread and Maximum Signal Bandwidth

18
The doppler filter bandwidth must be no wider than the spread of
doppler frequencies expected. Figure 2-5 also shows the
maximum radar signal bandwidth. For the case where acceleration
has a minimal effect on the integration time, the maximum
acceleration of the target can be expressed in terms of the radar
signal's bandwidth as
2
v
a max 2B 2
c( RF )

19

Long integration times require small target acceleration. The


radar designer must choose a bandwidth that suits the range
resolution required and integration to suit the target motion
expected. Long integration time implies either slow targets with
little acceleration or else poor range resolution. High
acceleration targets require wider signal bandwidths. An
aircraft target approaching at 300m/s and maneuvering at 3 g’s
needs a radar signal bandwidth of at least 2.5 MHz at 10 GHz.
Radar signals exhibit relatively constant characteristics during
coherent integration--important to know for ELINT analysis.
Tracking radars extend the coherent integration time when target
velocity and acceleration are known. Examining all possible
target velocities and accelerations requires huge processor
throughput and is generally not practical today.

20
Frequency Agility
From one coherent processing interval to the next, the radar can
change its carrier frequency without changing its range resolution
properties. The agility band is limited by the radar designer’s
ability to obtain sufficient power and to maintain beam width and
pointing angle--typically about 10% of the center frequency. (For
example, a 1 GHz agility band centered at 10 GHz.) What this
means for ELINT is that narrowband receivers have a low
probability of intercepting the complete radar transmission. If it is
sufficient to intercept only portions of the radar transmission,
narrowband receivers can be slowly tuned across the radar band
and the entire agility band can still be determined if the signals is
present for enough time. The coherent processing interval
determines the Doppler resolution. When FA is used with doppler
processing, the frequency is changed on a pulse-burst to pulse-
burst basis, not a pulse-to-pulse basis.
21

PRI Agility
Modern multifunction radar systems make use of multiple pulse
repetition intervals (PRI) values during one look at the target. It is
a requirement of today’s pulse doppler radars that the PRI remain
constant during each coherent processing interval. For moving
target indicating (MTI) radar designs, there is usually a sequence
of PRI values that must be completed during one processing
interval. This repeated sequence is known as "stagger" and ELINT
analysts call the period of the stagger the stable sum. This is
because when consecutive PRIs are added, the sum is constant
when one adds together the PRIs which make up the stagger
period--regardless of which PRI is selected as the starting point for
the sum.
22
MTI radars operate by subtracting (in amplitude and phase) the
echoes from one PRI from those in the next PRI. Stationary targets
have the same phase and amplitude and thus “cancel.” Echoes
from moving targets generally do not have then same amplitude
and phase and so do not cancel. However if the target moves an
integer multiple of half wavelengths in one PRI, the phase of the
second echo is shifted by a multiple of 360 degrees from the first
and the echoes cancel. Such speeds are “blind speeds.” Changing
the PRI changes the blind speed. A PRI sequence is selected to
detect targets regardless speed Moving target detection (MTD)
radar systems use a doppler filter bank to divide the frequency
region between the PRF lines into several filter bands (for example:
8 bands). This requires repeated constant PRIs (say 10 pulses at
one PRI and then 10 pulses at another, etc.) Multiple PRIs are
required due to range and velocity ambiguities and make visible
target ranges and velocities “eclipsed” by transmitted pulses (in
time) or spectral lines (in frequency).
23

For constant PRI and RF, the maximum unambiguous range (Ru)
and the maximum unambiguous velocity (Vu) are given by:
c(PRI ) c
Ru Vu
2 2( RF )( PRI )
Examples at 10 GHz:
PRI 1000 us, Vu=15 m/s and Ru=150 km
PRI 100 us, Vu=150 m/s and Ru=15 km
PRI 10 us, Vu=1500 m/s and Ru=1.5 km

As can be seen, the product of unambiguous range and velocity is a


constant. This means that the total ambiguity is fixed but changes
in PRI can increase the unambiguous range but decrease the
unambiguous velocity and vice versa. c2
RuVu
24
4( RF )
6
1 10
6
10
Inverse relationship of unambiguous
Rui  1 range and unambiguous velocity at
common radar frequencies
Rui  2

Unambiguous Range (m) Rui  31 105


22
5M
Rui  4 Hz
42
5M
Rui  5 1 .3 Hz
GH
3G z
Rui  6 5 .5
Hz
10
4 GH GH
1 10 15 z z
Rui  7 GH
35
GH z
z
Rui  8

1000
3
1 10
3 4
10 100 1 10 1 10

10 Vui  1  Vui  2  Vui  3  Vui  4  Vui  5  Vui  6  Vui  7  Vui  8 10


4

Unambiguous
25 Velocity (m/s)
Fi 2 7R /V l it R l t d

Frequency
Agility Band

(Depends
on
Frequency

Component
* Design,
ECM
Factors,
Designer
Ingenuity)

Coherent Processing Interval


(depends on radar mission)

Time

*BandwidthDepends
Determines Range Resolution Which
on Radar Mission

Figure 2-8. Modern frequency Agile Radar with 100% Duty Factor
26
USES OF PRI

27

UNAMBIGUOUS RANGE AND


VELOCITY DEPENDENCE
c

Analysis p. 147
28
RANGE-VELOCITY AMBIGUITY

Analysis p. 148
29

OPTIMUM PRI FOR MEDIUM PRF RADAR

30 Text p. 149
OPTIMUM PRI FOR MEDIUM PRF RADAR
Band “Be” Obscured at each PRF line

31

NOMINALLY CONSTANT PRI

32
PRI DRIFT

33 Analysis p. 153

CRYSTAL OSCILLATORS AND


COUNTDOWN CIRCUITS

34 Analysis pp. 191, 192


SCR-584

35

SEARCH RADAR PRI SELECTION

36
PRI STAGGER

Definition: Two or more discrete PRI intervals (elements) are alternating


in a periodic fashion.

• Desired Parameters
- Number of intervals
T T T T - Number of positions
Unmodulated Pulse Train - Interval values
- Sequence
H H
- Stable sum

T+H T- H T+ H T- H • Stagger Ratio


Typical Staggered Pulse Train
Two Interpulse Intervals Shown
• Stagger Versus Jitter

37

RADARS WITH STAGGER


Radar Pulse Width Average PRI Actual PRI’s Stagger Mode Stagger Ratio Stagger Purpose Radar Function
(μs) (μs) (μs)

1. 6,18 100 2500 5:7 To eliminate blind speeds Surveillance


3500
2. 4 3049 3032 89:90 To eliminate blind speeds Height Finder
3066
3. 6 3000 2954.55 0:97 To eliminate blind speeds Surveillance
3045.45 (almost 100:103)
4. 6 3000 2897 14:15 To eliminate blind speeds Experimental surveillance
3103
1000 613 5:7
1167
5. 24 3000 2750 11:13 To eliminate blind speeds Surveillance
3250
6. 3 1375 1250 5:6 To eliminate blind speeds Acquisition
1500
7. 20 5247 5000 0:91 To eliminate blind speeds Surveillance
5494 (almost 10:11)
8. 2 2777.9 2572.0 25:27:29 To eliminate blind speeds Air route surveillance
2777.8
2983.5
9. 1.4, 4.2 1250 1240 0.984 To identify second-time-around pulses Gap filter, surveillance and
1260 (almost 125:127) interrogator
10. 2 2632-3226 Unknown Unknown To eliminate blind speeds Air route surveillance
8-pulse stagger with three
programs
11. 42 1551.6 1408 (3) Almost 1033:1225:1073 3 pulses at each interval for double Detection; threat evaluation and
1667 (3) cancellation MTI to eliminate blind speeds target designation (long range
1460 (3) mode given here)

12. 6.7 4000 3571.4 (3) Exact order of 1 pulse 3 pulse canceller for MTI. Stagger to Surveillance
4405.1 (3) intervals is not known eliminate blind speeds
3745.3 (3)
4255.3 (3)
4081.6 (3)
13. 1-100 400 For first sequence only: Various Sequences To eliminate blind speeds. Has various Surveillance, tracking, kill
62.1 623.3 16:21:19:17:20:18 digital MTI processing including double assessment, missile guidance
2500 818.0 16:17:16:17 double-cancellation
740.1 16:19:16:19 38
662.0 16:21:16:21
701.1 16:17
DESCRIPTION OF PRI VARIATIONS

Nature of Pulse-to-Pulse PRI Variations

Periodic Random (non-periodic)

Discrete Continuous Discrete Continuous

Large Small Large Small Large Small Large Small

Type 1 2 3 4 5 6 7 8

(Large implies intentional, small implies incidental)

39

JITTERED PRI

Definition: Pulse repetition intervals are intentionally varied on


interval-to-interval basis in a random or pseudorandom
fashion. The variations are usually more than one percent.

• Intentional Jitter
- Discrete or continuous

• Desired Measurements
- Mean PRI
- Peak PRI deviation limits
- PRI distribution (histogram)
- Number of discrete PRIs

40
RADARS WITH JITTER
Pulse PRI Peak-to-Peak Jitter Peak-to-Peak Jitter Jitter Type Jitter Purpose Radar Function
Width (μs) (μs) (%)
(μs)

6, 18 3000 505 1.7 Random Anti-ECM and Sruveillance


1000 5 interference

26 4629 92.6 20 Random Anti-ECM and Target tracking


interference

200 4000 50 3.75 Random Anti-ECM and Long-range


interference surveillance

20 10204 999.9 9.8 Random Anti-ECM. Results from Decoy discriminator


5 10204 918.4 9.0 Or PRF being submultiple target tracking
400 6666 653.3 9.8 Programmed of RF which is jumping acquisition

0.9 416-1515 83-303 20 Unknown Unknown High resolution


(Variable) synthetic aperture
mapping

4-50 500-2777.7 60 2.2-12 Random To reduced inward Multifunction


4-2.67 3.3-4.0 None range gate stealers, anti-
interference, reduce
second-time around
41 echoes

PRI DWELL/SWITCH – PULSE DOPPLER

Definition: Rapid (automatic) switching between discrete PRIs with a dwell at each PRI

PRI = T1 PRI = T2

Dwell Time 1 Dwell Time 2

• Desired measurements

- Number of PRIs
- Value of PRIs
- Dwell times
- Total dwell time for sequence
- Dwell sequence
- Time to switch
42
SLIDING PRI
Definition: The pulse train has a PRI (PGRI) that is continuously changing in either
a monotonically increasing or decreasing manner between maximum
and minimum PRI limits.

• Desired Parameters

- PRI limits (min and max)


- Sweep waveform
- Sweep time (limits)

43

OTHER PRI TYPES 1

• Periodic Modulation

Definition: Pulse train consists of discrete or continuous intervals that


periodically increase and decrease, e.g., with sinusoidal,
sawtooth or triangular waveform

- Modulating waveform and rate


- Mean PRI and peak deviation limits

• Pulse Interval Displacement

Definition: Insertion of a different pulse interval into an otherwise


periodic pulse train

- Displacement value

44
OTHER PRI TYPES 2

• Interrupted Pulse Train

Definition: Intentional interruption of the pulse train with no apparent periodicity

- Range of on-period
- range of off-period

• Burst Pulse Train

Definition: Pulse train that is transmitted for some purpose for a relatively short
time and then is off for a relatively long time

- Burst definition
- Number of bursts per second
- Relationships of burst to scan

45

SCHEDULED PRIs

• Scheduled PRIs

Definition: PRIs are computer controlled, vary with the target environment and
function being performed by radar, and cannot be described by other
definitions

- Number of intervals
- Interval values
- Typical sequences
- Reason for sequence

46
MUTLIPLE PULSE GROUPS
• Constant and Cyclic Patterns

Definition: Pulse group characteristics remain constant or vary cylically in


predictable manner

- Number of pulses in group


- Pulse intervals
- Group position data

• Frames/formatted pulse trains (data encoded format)

Definition: Pulse train includes marker and data pulses

47

SUMMARY OF PRI TYPES

Analysis p. 151
48
DOPPLER EFFECT

v = radial velocity

c = 3(108) m/sec

fo = transmitted RF

v Doppler Shift (Hz)


km/hr @ 3 GHz @ 10 GHz
FIGURE 3-1. DOPPLER EFFECT 100 555.5 1851.8

cv ª 2v º 1000 5,555 18518.5


f fo x # f o x «1  »
1 c-v ¬ c¼
2000 11,111 37,037.0

2v
Doppler Shift f f  fo r x fo
d 1 c

49

50
FOURIER TRANSFORMS

51

IDEAL VS. ACTUAL SPECTRA


FOR CW SIGNAL

52
FM THEORY
V(t) A sin(2ʌ f c t  ș(t))
Phase Disturbance
Total Phase 2ʌ f c t  ș(t)
1 d 1 dș
Instantaneous Freq. (total phase) fc 
2 dt 2 dt

ASSUME ș(t) ǻijsin2ʌf m t

1 dș
f m ǻijcos2ʌf m t
2ʌ dt

Let ǻij ǻF/f m , then

1 dș
ǻFcos2ʌf m t
2ʌ dt

THEN :
ǻf
V(t) Asin(2ʌ f c t  sin2ʌ f m t)
fm

INDEX OF
MODULATION
53 " M"

BESSEL EXPANSION

V(t) Jo (m) sin Ȧc t  J (m)>sin(Ȧc  Ȧm )t  sin(Ȧc  Ȧm )t @


A 1

 J (m)>sin(Ȧc  2Ȧm )t  sin(Ȧc  2Ȧm t)@


2

 J (m) . . . . .
3

J0(m)

J1(m) J1(m)

J2(m) J2(m)

fc-2fm fc fc+2fm

fc-fm fc+fm
54
BESSEL FUNCTIONS

55

MOD. INDEX LESS THAN 1

FOR COHERENT SIGNALS:

ǻf
m  1 i.e. ǻf is small
fm

m
J o (m) | 1 J (m) |
1 2

J (m) | 0 J (m) | 0 ....... etc.


2 3

THEREFORE
m m
V(t) # A[sinȦc t  sin(Ȧc  Ȧm )t  sin(Ȧc  Ȧm )t]
2 2
A
V m ǻf
SB
Vc 2 2f m
mA/2 mA/2
in dB

V ǻf
fc-fm fc fc+fm 20 log SB 20 log
Vc 2f m
56
EXAMPLES

20 log ǻf  40 dB
2f m

ǻf 1 kHz
2f m

e.g. f m 1 kHz

ǻf 20 Hz

IF f c 10 GHz, STABILITY IS

20 Hz 2 parts in 109
10 x 109 Hz

(AT 1 kHz RATE)

57

RANGE AMBIGUITY RESOLUTION VIA MULTIPLE PRIs

12 μs = X
T1 = 40 μs

2 μs = Y
T2 = 30 μs

Actual Round Trip Echo Time is T = 92 μs

N1 T1 + X = T and N2 T2 + Y = T

Trial and Error N1 N2 N1 T1 + X N2 T2 + Y


Solution
1 1 52 32
1 2 52 62
2 2 92 62
2 3 92 92

Unambiguous Range c x ªLeast Common Multiple of T , T º


2 «¬ 1 2 »¼
c x >120 ȝs@
2 Analysis p. 196
58
ERICSSON PS-05/A MULTIMISSION RADAR

59

ERICSSON PS-05/A MULTI-MODE OPERATION (1)

60
ERICSSON PS-05/A MULTI-MODE OPERATION (2)

61

MTI VIDEO

62
MTI PHASE SHIFTS

63

MTI BLOCK DIAGRAM

64
BIPOLAR VIDEO

65

DOPPLER RETURNS

TRAIN

CAR
Typical images displayed on TPS-25 ground
Surveillance radar. Shown are target images
of: 1) a train, 2) an automobile, 3) a walking man,
and 4) a walking girl. (US Army photograph.)

MAN WALKING

WOMAN
WALKING
66
PULSED-OSCILLATOR MTI

= 2E sin(SfdT) cos [2Sfd(t + T/2) + Io]

Zeros at 0, ʌ, nʌ when

n
f
d T
cxn
so blind speeds are V
b 2RF x T

x PRF
2
Barton, p. 192
67

Page M50.ppt

68
BLIND SPEED ELIMINATION

No Stagger vb = n c/2(PRI)(RF)
1
T

Tİ 5 Vbn = Vb (7 + 5)/2


Tİ 7
6
T
Deep lobe
at 32/T

Tİ 63
Tİ 65
Null at
64/T

Ref:69Barton, page 222

IMPROVEMENT FACTOR OF CANCELLER

(S/C)out signal  to  clutter ratio at output of canceller


I
(S/C) signal  to  clutter ratio at input of canceller
in

Overall improvement factor I is found from:

1 1  1  1  ....
I I I I
1 2 3

I1, I2, I3 are the individual improvement factors calculated on basis of PRI, pulse
amplitude, pulsewidth, transmitter frequency, ……….. stabilities

70
INSTABILITY LIMITATIONS

71

CLUTTER STRENGTH

72
MTI + PULSE DOPPLER = MTD

8-Pulse Weighting
3-Pulse
Doppler And
Canceller
Filter Bank Magnitude

I,Q Data Target


Threshold Detection
From A/D
Converters

Zero Clutter Map


Magnitude
Doppler (Recursive
(I2 + Q2)1/2
Filter Filter)

Clutter 15 – 20 radar scans are


Memory needed to establish
Typical Applications the clutter map
New FAA ASR radars (10 pulse dwell)
AN/SPS-49 USN-adjunct to AEGIS (6-pulse dwell)
RAMP (Canada)

73

MTD PERFORMANCE
• Theoretical

RMS Clutter Width

Processor 0.01 PRF 0.1 PRF


MTI Improvement 1 canceller 25 dB 8 dB
Factor 2 cancellers 50 dB 12 dB
3 cancellers 72 dB 16 dB
FFT Improvement 8 pulses 35 dB 22 dB
Factor
MTI + FFT 1 canceller + 60 dB 28 dB
Improvement Factor 8 pulse FFT
2 cancellers + 80 dB 34 dB
8 pulse FFT
3 cancellers + 100 dB 36 dB

(Reference: NRL Report 7533, G.A. Andrews, Jr.)

• Practical
Performance of FAA ASR radar: 3 pulse MTI alone | 25 dB
3 pulse MTI + 8 pulse FFT | 45 dB

(Reference: Skolnik, Introduction to Radar Systems, 1980, p. 127-128)


74
ELINT IMPLICATIONS OF MTD

• Coherent carrier RF stability is necessary

• Constant PRIs Several PRIs of the same interval must be


Constant RF transmitted at the same RF (typically 4,
(for a certain 8, or 16 pulses for the FFT plus pulses
number of pulses) to fill the canceller. For example, a
three-pulse canceller plus an eight-pulse
FFT requires 10 pulses).

• “Stagger” to eliminate For these radars, the pulse interval


blind speeds stagger occurs not from pulse-to-pulse but
from pulse group-to-pulse group

• Long PRI MDT is generally used for long-range radars


where the low PRF creates very ambiguous
Doppler shifts.

75

PRI EXERCISES

1. The analyst found a signal at 6 GHz which had two-interval, two-position stagger. The
intervals were 500 and 550 microseconds. What is the average PRI? What is the
stagger ratio? What is H? What are the new blind speeds?

2. What is the improvement factor for MTI of a radar which has RMS jitter of 10 nanosec
and a pulse duration of 1.41 microsec?

3. A discrete random jitter PRI train was analyzed and the PRIs were found to be one of
the following 5 nominal values:

Nom
PRI (μsec)

2440.8
2428.7
2465.3
2453.1
2562.9

Is there a clock? If so, what countdowns are used and what is the clock frequency or
period? What common range mark is that closest to?

(This problem is discussed on p. 194-195 of analysis book.)

76
PRI EXERCISES #2 - ANSWERS
1. (500 + 550)/2 = 525 microsec = average PRI
R = 550/500 = 1.1 (11:10)
H = 550 – 525 = 25 microsec
Blind speed before stagger = nc/(2 • PRIave • RF)
(3 x 108 ) m / sec
VB  171.4 km / hr (106.5 mph )
2(525)(10 6 ) sec x 6(109 ) x 1 / sec

V/VB = (11 + 10/2 = 10.5)


V = (10.5) (171.4) = 1800 km/hr (1118.4 mph)

2. Improvement factor due to PRI instability is:

IdB = 20 log [ W / 2 't BW )], 'W = jitter, W = pulse duration,


B = bandwidth 2

IdB = 20 log [1.41 (10-6) sec/ 2 • 10(10-9)sec)]


= 20 log [102] = 40 dB

3. Periods Nearest Calculated


Period In Order Difference Countdown Clock Period
2440.8 2482.7 199 12.20452
2428.7 2440.8 12.1 200 12.20400
2465.3 2453.1 12.3 201 12.20447
2453.1 2465.3 12.2 202 12.20445
2562.9 2562.9 97.6 210 12.20428
12.204392 average
The differences 12.1, 12.3, 12.2 average 12.2
97.6 divided by 12.1 = 8
So use 12.2 to start for countdowns.
The average clock period is 12.204392 μsec so reciprocal is 81.93777 kHz (2000 yards, see p. 192.)

77

NOISE EFFECT ON PRI

T
Triggering Error 'T RISE x 'A
0.8 A

'A A

'T

TRISE/0.8 'A Slope A


Noise 'T (TRISE / 0.8)

78
PRI VARIATION DUE TO NOISE

V2
amplitude Noise Power 1
(Amplitude)2 Signal Power SNR
T
V Rise 1
Time 0.8 SNR
V2 V2  V2 2 V2
PRI Time1 Time 2 Time
T
V 2 Rise
PRI 0.8 SNR

79

BANDWIDTH EFFECT ON SNR

2
§
t · .35
SNR 3.125 ¨¨ V r ¸¸ tr |
¨
© PRI ¹
¸ Bandwidth

SNR Required for


Bandwidth (MHz) Rise Time 1 ns Jitter 10 ns Jitter 100 ns
Limit (ns) Jitter

0.1 3.5 Ps 81 dB 61 dB 41 dB
1.0 0.35 Ps 61 dB 41 dB 21 dB

10.0 35 ns 41 dB 21 dB X
100.0 3.5 ns 21 dB X X

80
AMPLITUDE INDUCED ERROR

81

AMPLITUDE COMPENSATED TRIGGER

82
PERFORMANCE OF TRIGGER CIRCUIT

83

DOPPLER SHIFT OF PRI

• In 1 PRI, the platform moves


VR • PRI

• Transmit time from transmitter to receiver changes by VR • PRI/c

• Example: VR = 600 M/S PRI = 3000 μs


3
Observed 'PRI = 600 x 3 x 10 r 6 ns
3 x 108

84
DELAY AND PULSE JITTER
Delay D2
Delay D1

Peak-to-Peak Jitter Peak-to-Peak Jitter


At Delay D1 At Delay D2

85

DELAYED SWEEP JITTER PHOTOS

~ 1 μs Jitter ~ 2 μs Jitter
Delay = 1 PRI Delay = 5 PRI

86
SYNTHESIS OF AVERAGE PRI

87

PRI DRIFT MEASUREMENT

88
REAL TIME RASTER DISPLAYS

89 Analysis p.74

DUAL AMPLITUDE AND TIME


DELAYS

90
DTE MODE-CIRCULAR SCAN
RADAR

91

RTR SIMULATION ON A
PERSONAL COMPUTER

92
MEAN PRI ESTIMATES

93

MINIMIZING THE SQUARED ERROR

94
RMS ERRORS COMPARED

95

96
97

NONCUMULATIVE AND
CUMULATIVE JITTER

98
CRAMER-RAO BOUNDS COMPARED

99

PRI ESTIMATION PERFORMANCE

100
USING THE WRONG JITTER
MODEL

101

PRI HISTOGRAMS

102
ACTIVITY IN 0.1S INTERVALS

103

INTERVALS FORMED BY
PULSE PAIRS

104
DELTA-T HISTORGRAM
(10% JITTER)

105

DELTA-T HISTOGRAM--
STAGGER

••• 4 5 7 4 5 7 4 5 •••

t0 = 0 t1 = 4 t2 = 9 t3 = 16 t4 = 20 t5 = 25 t6 = 27 t7 = 31 t7 = 37

A. (tn – tn-1) = 4, 5 or 7

B. (tn – tn-2) = 9, 11 or 12 (4 + 5, 4 + 7, 5 + 7)

C. (tn – tn-3) = 16 (4 + 5 + 7)

D. (tn – tn-4) = 20, 21 or 23

E. (tn – tn-5) = 25, 27, or 2

F. (tn – tn-6) = 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

A B C D E F
106
THREE POSITION STAGGER

107

DELTA-T HISTOGRAM:
TOA AUTOCORRELATION

n
f (t ) ¦ G (t  t n ) .....
n 1

t1 t2 . . . . . t3 t4 . . . . .
f
h (W) ³ f (t ) f t  W dt
f

f ª º
h (W) ³ «¦ G (t  t n ) ¦ G (t  t k  W)»» dt
f «¬ n k ¼

value only if t  t n 0 and


tn  tk  W 0 OR t n tk  W
EXAMPLE
t1 t = t2 t3 t4

t=0 W

108
t = t1 + W
DELTA-T HISTOGRAM:
TOA AUTOCORRELATION

h(W) ¦ ¦ G(t n  t  W)
n k k
IJ IJ
2 2
³ h(W) ³ ¦ ¦ G(t n  t k  W)
IJ IJ n k
1 1

A count of the number of


pulse pairs such that
W d tn  t ¢ W
1 k 2

THEREFORE:
A count of the number of pairs of pulses whose arrival
times differ by a value between W1 and W2 is equal to
the integral of the autocorrelation of the TOA’s

109

JITTER ANALYSIS MODEL

Center Frequency
(average PRF)

Jitter
Waveform FM Trigger Time of
Oscillator Generator Arrival
Peak Sequence
Amplitude

Periodicities
• Periods
• Amplitudes

¦ Drifts/Trends
• Slopes

Random Components
• Bandwidths
• Variances
110
• Probability Densities
INSTANTANEOUS
FREQUENCY ESTIMATION

500 700 600 500 400 500 PRIs (μs)

2500 Linear
2000 2000 2000 Interpolation
Freq
1428.5 1666.7

Midpoints
of Intervals
111

DEINTERLEAVING DEVICE

112
DEINTERLEAVING VIA DELTA-W HISTOGRAM

113

“PURE” VS. “IMPURE” INTERVALS

114
NUMBER OF EMITTERS DEINTERLEAVED

115

COMPLEX DELTA-W HISTORGAM - I

116
COMPLEX DELTA-W HISTOGRAM - II

117

COMPARISON OF DELTA-W HISTOGRAMS

118
EFFECT OF A NEAR MULTIPLE PRI

119

EFFECTS OF JITTER ON DELTA-W


HISTOGRAMS

120
Delta-T Histogram for Ten Interleaved Pulse Trains

Delta-T Histogram
100
Histogram Count

dhist b
50
.75˜ max( dhist )

0
5 4 4 4 4
8 10 1 10 1.2 10 1.4 10 1.6 10

6
int vb  PRI k˜ 10

PRI, Seconds

N 820 10 Interleaved Pulse Trains

121

Comparison of the Delta-T and Complex Delta-T Histograms

Comparing Delta-T Histograms


150
Complex Histogram Absolute Value

100

Delta-T Hisotgram bin Count


100
abchist b dhist b
0
1.05˜ max( abchist ) 1.05˜ max( dhist )

50

100

0
4 4 4 4
1 10 2 10 3 10 4 10


int vb  PRI k˜ 10
6
 int v  PRI ˜ 10
b k
6

N 820 PRI, Seconds 10 Interleaved Pulse Trains

Top Trace is the regular Delta-T Histogram;


Bottom Trace is the Complex Delta-T Histogram--Note how multiples of the PRIs are suppressed
The dots above the peaks indicate the true PRI values
122
Effect of Jitter on Delta-T Histograms
(Jitter=1 microsecond)

Comparing Delta-T Histograms


100 100
Complex Histogram Absolute Value

Delta-T Hisotgram bin Count


50

abchist b dhist b
50 0
1.05˜ max( abchist) 1.05˜ max( dhist )

50

0 100
5 4 4 4 4 4 4
5 10 1 10 1.5 10 2 10 2.5 10 3 10 3.5 10


intv b  PRI k˜ 10
6
 intv  PRI ˜ 10
b k
6

PRI, Seconds

7
Jitnc 0.5 Jitcum 0.5 N 820 width 5 u 10 10 Interleaved Pulse Trains
0 0

123

Effect of Jitter on Delta-T Histograms


(Jitter=2 microseconds)

Comparing Delta-T Histograms


100 100
Complex Histogram Absolute Value

Delta-T Hisotgram bin Count

50

abchist b dhist b
50 0
1.05˜ max( abchist) 1.05˜ max( dhist )

50

0 100
5 4 4 4 4 4 4
5 10 1 10 1.5 10 2 10 2.5 10 3 10 3.5 10


intv b  PRI k˜ 10  intv  PRI ˜ 10
6
b k
6

PRI, Seconds

7
Jitnc 1 Jitcum 1 N 820 width 5 u 10 10 Interleaved Pulse Trains
0 0
124
Effect of Jitter on Delta-T Histograms (Jitter=5 microseconds)

Comparing Delta-T Histograms


100 100
Complex Histogram Absolute Value

Delta-T Hisotgram bin Count


50

abchist b dhist b
50 0
1.05˜ max( abchist) 1.05˜ max( dhist )

50

0 100
5 4 4 4 4 4 4
5 10 1 10 1.5 10 2 10 2.5 10 3 10 3.5 10


intv b  PRI k˜ 10
6
 intv  PRI ˜ 10
b k
6

PRI, Seconds

7
Jitnc 2.5 Jitcum 2.5 N 820 width 5 u 10 10 Interleaved Pulse Trains
0 0

125

Complex Delta-T histogram: Original and Improved


Original Complex Delta-T Histogram Improved Complex Delta-T Histogram

Uniform Jitter=0.002

Uniform Jitter=0.02

Shift time origin


To avoid excessive
Phase variation
K Nishiguchi and M. Korbyashi,
"Improved Algorithm for
Uniform Jitter=0.2
estimating Pulse Repetition
Intervals,” IEEE Transactions on
Aerospace and Electronic Systems,
Vol. 36, No. 2, April 2000.
126
Example of Automated Peak Processing Results
Delta-T Hist. Complex Delta-T Input PRI Values

0 0
0
0 1· 10-4 0 1· 10-4 0 1· 10-4
1 1. 048· 10-4 1 1. 05·10-4 1 1. 05·10-4
2 1. 11·10-4 2 1. 11·10-4
2 1. 11·10-4
3 1. 15·10-4 3 1. 15·10-4 3 1. 15·10-4
4 1. 162· 10-4 4 1. 164· 10-4 6
PRI˜ 10 4 1. 163· 10-4
5 1. 176· 10-4 5 1. 178· 10-4 5 1. 177· 10-4
6 1. 19·10-4 6 1. 192· 10-4 6 1. 191· 10-4
pk 7 1. 21·10-4 pkc 7 1. 21·10-4 7 1. 21·10-4
8 1. 23·10-4 8 1. 23·10-4 8 1. 23·10-4
9 1. 26·10-4 9 1. 26·10-4 9 1. 26·10-4
10 0 10 0
11 0 11 0
12 0 12 0 This example based on the method of B.
Frankpitt, J. Baras, A. Tse, "A New Approach
13 0 13 0 to Deinterleaving for Radar Intercept
14 0 14 0 Receivers," Proceedings of the SPIE, Vol
5077, 2003, pages 175-186
15 0 15 0

Jitter =10 ns cumulative and 10 ns non-cumulative


Histogram Bin size 200 ns.127

Pulse Train Spectrum of Ten Interleaved Pulse Trains


k

PRF Spectrum

0.01
Amplitude

Xj
0.00011˜ max( X)

0.005

0
4 4 4 4 4 4
6000 8000 1 10 1.2 10 1.4 10 1.6 10 1.8 10 2 10

f j  PRF k This plot is the FFT of


PRF (Hz) TOA
N 8.705 u 10
3 10 Interleaved pulse Trains phase 2S ( )
T
PRF Resolution 10 Hz R. Orsi, J. Moore and R. Mahony, "Interleaved
Pulse Train Spectrum Estimation," International
Symposium on Signal Processing and its
applications, ISSPA, Gold Coast, Australia,
128
August 25-30, 1996
k

PRF Spectrum

0.03

Xj
Amplitude

.025 0.02

.015

0.01

0
4 4 4 4 4 4
4000 6000 8000 1 10 1.2 10 1.4 10 1.6 10 1.8 10 2 10

f j 1  PRF k  2˜ PRF k

PRF (Hz)
3
10 Interleaved pulse Trains N 1.741 u 10

Fewer Pulses--Degraded PRF Resolution (50 Hz)

129

Figure 13.10 Pulse Train Spectrum for a Shorter Record


k

PRF Spectrum

0.03

Xj
Amplitude

.03 0.02

.02

0.01

0
4 4 4 4 4 4
4000 6000 8000 1 10 1.2 10 1.4 10 1.6 10 1.8 10 2 10

f j 1  PRF k  2˜ PRF k

PRF (Hz)
10 Interleaved pulse Trains N 871

Fewer Pulses--Degraded PRF Resolution (100 Hz)


130
PULSE SORTING ALGORITHM
C
C C
C B B B
B
B B
B B B
A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A

3 Adjacent Matching Intervals

Step 1. Find 3 adjacent matching intervals


Step 2. Extend in both directions to discover other numbers of the pulse train
Step 3. Remove this pulse train and go back to Step 1.
If no more pulses can be removed, go to Step 4.
Step 4. Consider all pairs of pulses to search for intervals which match; go to Step 2.

131

SORTER SOFTWARE PERFORMANCE


(Pulses Pr ocessed)  10 (Pulses Wrong)
Score
Total Pulses  Noise Pulses
Amp On: 0.2 amp Tolerance
from pulse-to-pulse
Amp On 0% Jitter

100

90 Amp Off n
pO
Am
80
1% Jitter
ff
70 pO
A m
60
Score n
50 pO
Simulated Data Am
Average Density 200 pps
40
Amp Off
30

20
8% Jitter
10

1 10μs 100μs 1000μs


Time Tolerance
132
SIMULATION SCENARIOS

File Name I.P. #† PRI Variation

C2-3009-V05 2 30.0 0.5


1 70.0 0.5

C2-3009-V20 2 30.0 2.0


1 70.0 2.0

C3-3009-V05 1 3.0 0.5


3 120.0 0.5

C3-3009-V20 1 3.0 2.0


3 120.0 2.0

C4-3009-V05 1 50.0 0.5

C5-3009-V05 1 100.0 0.5


2 100.0 0.5
3 100.0 0.5

C5-3009-V20 1 100.0 2.0


2 100.0 2.0
3 100.0 2.0

† Denotes initial pulse number.


Table 1. Simulation scenarios.
Ref: Kofler and Leondes
133

FIXED GATE DEINTERLEAVING


RESULTS
File Name I.P.D. PRI % Misses
20 90.0 90.9
28 105.0 37.5
C2-3009-V05 30 60.0 50.0
35 69.9 81.8
63 180.1 0.0
69 90.0 0.0
20 90.3 81.8
C2-3009-V20 28 104.9 37.5
30 59.8 42.9
35 69.7 63.6
16 9.1 62.9
C3-3009-V05 . . .
. . .
. . .
473 36.0 100.0
7 3.7 3.2
C3-3009-V20 117 50.0 77.3
395 32.8 60.0
C4-3009-V05 No emitters detected
19 100.0 16.7
30 100.2 92.9
C5-3009-V05 47 199.9 25.0
48 99.9 75.0
56 200.0 0.0
19 99.9 16.7
30 100.7 92.9
C5-3009-V20 47 199.7 25.0
48 99.7 75.0
56 200.0 0.0
134
Ref: Kofler and Leondes
ADAPTIVE GATE DEINTERLEAVING
RESULTS

File Name I.P.D. PRI % Misses


C2-3009-V05 7 30.0 0.0
18 70.0 0.0
C2-3009-V20 7 30.0 0.0
18 70.0 0.0
C3-3009-V05 5 3.0 0.0
168 120.0 0.0
C3-3009-V20 5 3.0 0.0
168 119.9 12.5
C4-3009-V05 5 50.0 0.0
15 100.0 0.0
C5-3009-V05 20 100.0 0.0
25 100.0 0.0
15 100.0 0.0
C5-3009-V20 20 100.0 0.0
25 100.0 0.0
Kofler and Leondes

135

^ĐĞŶĂƌŝŽ
ŽŵƉůĞdžĞůƚĂͲd,ŝƐƚŽŐƌĂŵWĞĂŬƐ /ŶƉƵƚWZ/Ɛ

136
WƵůƐĞƚĂŐƐĨŽƵŶĚŝŶƉĞĂŬ
ηϮ

137

WƵůƐĞdĂŐƐŝŶWĞĂŬηϮǁŚŝĐŚŚĂǀĞƚŚĞ
ŵŽƐƚƉƌĞǀĂůĞŶƚWZ/WŚĂƐĞ

138
PRI ANALYSIS EXERCISE

Two signals are observed with the same angle of arrival but on different frequencies. The PRI of one is nearly
stable at 3000 μs. The PRI of the second jitters randomly with a mean value of 1500 μs and a peak-to-peak jitter
of about 20 μs. The analyst notices that the PRI’s of the second signal can be paired such that their sum is nearly
stable at 3000 μs; i.e., PRI #1 + PRI #2 = PRI #3 + PRI #4 = PRI #5 + PRI #6, etc. However, PRI #2 + PRI #3 z
PRI #4 + PRI #5. He also notices that the mean value of the second signal’s PRI is exactly one-half that of the
first signal’s PRI every time the two signals are reported. The first signal has a slow circular scan, the second a
faster sector scan. What conclusions might be drawn about these two radars?

What additional data would you request from the ELINT station?

139

PRI EXERCISE ANSWER

There is a good possibility that the second radar operates in PRI synchronism with
the first;
but at one-half the PRI. Alternate pulses are triggered by the master clock, the
intermediate
pulses are generated by “one shot” type delay circuit which is not stable.

The second radar may be a height finder using elevation sector scan and associated
with a long
range search radar.

Confirmation of this would be aided by using two receivers and making a recording
of both
Signals simultaneously to investigate whether the second signal is synchronized to the
first.

140
PRECISION PDWs

• Pulse Descriptor Words are computed from pre-detection


burst recordings

• Digitizer has “detected” presence of high SNR pulses,


and captured them

• Different capture and processing techniques apply to low


SNR pulses

• Standard PDWs computed are:


- Amplitude - Frequency
- Time of Arrival - Bandwidth
- Pulse Width

• Algorithms and accuracies are described

141 Condor Systems, Inc.

USEFULNESS OF PRECISION PDWs

• Reveals fine details of pulse train jitter patterns

• Permits very high accuracy computation of crystal


controlled PRIs with few pulses

• Can use very accurate pulse width to sort pulses

• Fine variations of frequency pulse to pulse reveal unique


emitter characteristics (e.g., frequency pulling effects due
to VSWR changes in antenna rotary joint, etc.)

• Amplitude droop in transponder pulse groups

• Precise antenna pattern scan envelope measurement

142 Condor Systems, Inc.


EXAMPLE OF PRE-DETECTION RADAR
PULSE RECORDING

143 Condor Systems, Inc.

CALCULATION OF AMPLITUDE, TOA,


PW

144 Condor Systems, Inc.


TOA MEASUREMENT ACCURACIES

• Digitizer time base determines ultimate accuracy

• Individual pulse time of arrival error determined by:


tr
't
2SNR
where 't RMS Error in TDOA
t r Pulse Rise Time
SNR Signal to Noise Ratio in Captured
Pulse Bandwidth

• Example: 30 ns rise time, 37 dB SNR yields RMS error of


300 picoseconds per pulse

145 Condor Systems, Inc.

PULSE WIDTH MEASUREMENT


ACCURACY

'pw 't 2 2
r  'tf
where 'pw RMS error in pulse width
't r RMS error of pulse risin g edge time
't RMS error of pulse falling edge time
f

Example: RMS errors of captured pulse edge times of


300 picoseconds yield 1.414 x 300 = 423
picoseconds RMS pulse width error per pulse.

Condor Systems, Inc.


146
EXAMPLE OF PULSE WIDTH
ACCURACY

147 Condor Systems, Inc.

PULSE FREQUENCY COMPUTATION

148 Condor Systems, Inc.


PULSE FREQUENCY ACCURACY
'f 1
T SNR TW
in
• Technique applies to high SNR cases (>+15 dB), sine
wave pulse
where 'f RMS frequency accuracy

T Integration time (~ pulse width)

SNR Input Signal  to  Noise Ratio in BW , W


in

W Input Pr e  det ection bandwidth

Example: 1 microsec pulse, 30 dB SNR, 20 MHz Bandwidth


yields RMS accuracy of 7 kHz.
Condor Systems, Inc.
149

EXAMPLE OF PULSE FREQUENCY


COMPUTATION

150 Condor Systems, Inc.


Pulse Bandwidth

151 Condor Systems, Inc.

EXAMPLE OF PULSE FREQUENCY


COMPUTATION

152 Condor Systems, Inc.

You might also like