High Angular Resolution

Techniques (Ch 12)

Phased Arrays
Doppler Beam Sharpening
Synthetic Aperture

PHASED ARRAYS

http://www.fas.org/programs/ssp/man/us
G. Brooker, Introduction to Sensors for Ranging and Imaging: SciTech, 2008. wpns/air/special/e3.html

1
 Resolution of a Single Aperture
 For imaging systems (not null steering trackers), the angular
resolution is limited by the beam divergence.
 Beam divergence 3dB (beamwidth) is a function of the wavelength
and the aperture size
k
 3dB 
d
where 3dB – 3dB Beamwidth
k – Constant (70 for  and 1.22 for rad)
 - Wavelength (m)
d – Aperture diameter (m)
 The for a weighted aperture, the cross range resolution, xr, is the
product of the beamwidth (rad) and range (m)
1.22 R
xr  R 3dB 
d

Requirement for Phased Arrays

 The wavelength is fixed by atmospheric window,
propagation effects or physical size constraints.
 It is difficult to make the antenna diameter, d, arbitrarily
large to obtain the required angular resolution because
of manufacturing limitations
 Use two or more transceivers in an array to synthesise
an effective linear aperture equal to the array baseline.
 Uses
 Phased array radars
 Phased array sonar (linear or 2D) underwater or in the air
 Long baseline radio telescopes

Acoustic arrays used for RASS research

2
 Transmitter Beam Synthesis
 Individual elements radiate
precisely in phase to produce
wave crests that move
forward in phase
 Interfere constructively to
produce a strong narrow
beam directed straight ahead
 If a linear phase shift is
applied across the array,
then the beam is
reconstructed at an angle

Receiver Beam Forming

 The power received by each element is the sum of the received power
scattered by target P from all the transmit elements
 The elements outputs are summed via lines of equal length to give Ea
Incoming Wave

 

N N-1 3 2 1
d d d
EN E2 E1

Ea

N
Ea   sin(t  k )
k 1

3
 Phase Shift Between Elements
 For an incoming signal at an angle  to the array
 The phase shift between adjacent elements is  (rad) where d (m) is
the element spacing

2d
 Sin rad


Moving Phase Reference to Geometric

Centre of Array
 The equation that describes the output voltage from the array for
phase shifts starting with element 1 at B is
N
E a   sint  k  1 
k 1

 It is usual to measure the phase shift relative to the geometric centre

of the array, not element 1.
 This displaces the phase front represented by line AB, down to CD

4
 Revised Equation Counting from Centre

 The equation that describes the output voltage for phase shifts
displaced to the geometric centre is
N
 N 1 
Ea   sin t  k  1  
k 1  2 
N
 N 1 
  sin t  k  
k 1  2 

Two Point Array

Ea = sin(t+/2) + sin(t-/2) The field intensity pattern, |Ea()|

Ea = sin(t).2cos(/2) for the antenna is the magnitude
sinA+sinB=2sin[(A+B)/2]cos[(A-B)/2] of the amplitude factor

sin( 2 / 2)
E a  sin(t )
sin(  / 2)
sin2A = 2sinAcosA

5
 Four Point Array

E a  sin   t  3 / 2   sin   t   / 2   sin   t   / 2   sin   t  3 / 2 

(1) (2) (3) (4)

See your book for the trig identities used to manipulate this equation

sin  4 / 2 
Ea  sin t 
sin  / 2 

The General Case

N Point Array
N
 N 1  sin  N / 2 
Ea   sin  t  k     sin t 
k 1  2  sin  / 2 

These equations give identical results if you plug in the numbers, but
I have never managed prove the relationship rigorously.

Bonus Quiz – provide a derivation showing the mathematical

equivalence of these two equations

6
 The Field Intensity Pattern
 Substituting for =2dsin/

 Nd 
Sin Sin 

Ea  Sin(.t ).   Amplitude factor

 d 
Sin Sin 
 
 The field intensity pattern Ea() is equal to the magnitude
of the Amplitude factor
 N d 
Sin Sin 
E a ( )    
 d 
Sin Sin 
 

The Field Intensity Pattern Continued

 N d 
Sin Sin 
E a ( )    
 d 
Sin Sin 
  

Nd
 Nulls where the numerator is zero Sin = 0, +/-,

+/-2 etc

d
 The denominator is zero at Sin = 0, +/-, +/-2
etc 

 Applying L’Hopitals rule where Ea = 0/0 we find maxima

each with value N, where sin = +/-nλ/d

7
 The Field Intensity Pattern cont…
• The maximum where sin = 0 is
called the Main lobe, all the other
lobes are called Grating lobes.
• If d/ = 0.5, the grating lobe does
not appear for n = +/-1 in real space
because sin>1 which is not
possible.
• If d/ = 1 the grating lobes appear at
+/-90, however as most real
radiating elements do not radiate
much at  = 90, the grating lobes
are suppressed.
• For a non scanning array, the best
element spacing is d = 
• For a scanned array the best
spacing is d < /2

Grating Lobes for N=10 and d/ = 1

8
 No Grating Lobes for N=20 and d/ = 0.5

Radiation Pattern: Linear Array

 The radiation pattern is defined as the normalised square of the
amplitude factor
 N d 
2 Sin 2  Sin 
G a ( ) 
Ea
   
N2  d 
N 2 Sin 2  Sin 
 
 For Nd = L (the length of the array) and for sin =  (small angle)

L 
Sin 2  Sin 
Ga ( )   
 L
2

 Sin 
 

9
 Linear Array cont…
 For d = /2, the half power beamwidth 3dB (rad) is as
follows
The beamwidth is smaller than the
1.73 normal 1.22/d which gives 2.44/N.
 3dB  rad This is because the sidelobes are so
N high in this case. See eqn 5.14

 If N is sufficiently large, the antenna will be equivalent to

a uniformly illuminated aperture, and the first sidelobe
will be 13.2dB down.
 For directive elements, the antenna pattern is the
product of the element factor Ge() and the array factor
Ga().

G() = Ge().Ga()

Arrays of Directive Elements

Directive Element
Pattern

Directive Element
Array pattern

Isotropic Array
Pattern

10
 2D Rectangular Array

Radiation Pattern: 2D Rectangular Array

 The radiation pattern may be approximated as the
product of the patterns of the two planes that contain the
principle axes of the antenna

G(,) = G()G()

 N d   M d 
Sin 2  Sin  Sin 2  Sin  
G ( ,  )       
 d   d 
Sin 2  Sin  Sin 2  Sin  
   

11
4x4 Rectangular Array of Isotropic Elements

Antenna Gain
 For large arrays, the non-scanned antenna gain can be
approximated by the gain of a uniformly illuminated aperture
4A
Go 
2
 For a scanned array, the gain is reduced by the scan angle o
because the projected aperture is reduced in size.

4A cos  o
G ( o ) 
2

12
 Beam Steering
•If the same phase is applied to
all the elements of the array, then
the main beam will be broadside
to the array and  = 0 Generators Phase Radiating

•The direction of the main beam

Shifters Elements

will be o if the relative phase

Antenna
difference between elements is

2d
 sin  o

Antenna
 (a)

(b)

Steered Radiation Pattern

 The radiation pattern is then

 N d 
2 Sin 2  ( Sin  Sin o ) 
G a ( ) 
Ea
   
N2  d 
N 2 Sin 2  ( Sin  Sin o ) 
  

 Grating lobes will occur at

d
( Sin g  Sin o )   n

 For a scan over +/-90, the element spacing should be d=/2
 For a practical array that can scan over +/-60, the spacing d>0.54

13
Corrections to Improve Range Resolution
Delay adjust
Phase shift only Delay adjust only And phase shift
Phase Time Time Phase
Shift Delay Delay Shift
1 1 1

2 2 2

3 3 1
1
4 5n 4 2
s
5 5 5n 1
s 5n
2 s
6 6 2

10 6.
ns 5n
s
Su
m
Si
gn Su
al m
Si
Su gn
m al
Si
gn
al

(a) (b) (c)

Half Power Beamwidth

 The half power beamwidth of a scanned array can be approximated
by the following formula (not valid near end-fire)

0.886
 3dB  rad
NdCos  o

 Taper is generally used to reduce the sidelobe levels.

Using cosine on a pedestal An = a0 + 2a1Cos(2n/N)
where 0 < 2a1 < a0

0.886   2a1  
2

 3dB  1  0.636    rad

NdCos  o   0  
a

14
 Active and Passive Arrays
 With the advent of low-cost MMICs, it is now practical to
manufacture individual transceiver modules to build active arrays
 Some of the best in the world are produced by CEA in Canberra,
Australia

CEAFAR Active phased Array

Scalable in size and power to meet a broad range

of applications, suitable from 'Corvettes to Cruisers';
Full 3D multifunction capabilities;
Advanced classification capabilities;
Optimised for littoral and open ocean;
Evolves to meet changing requirements;
Very high reliability, no in-mission maintenance.

15
 Matching and Mutual Coupling
 The impedance of the array elements varies with the scan angle
 Spurious lobes may appear due to the miss-match
 This is a difficult problem to solve analytically, and is often
determined experimentally by exciting a single element and
terminating all of the surrounding ones
 Coupling is proportional to 1/d for d=/2, so the pattern and
impedance are drastically altered by surrounding elements.
Generally the surrounding 5x5 or even 9x9 elements must be
considered.

http://www.zonamilitar.com.ar/foros/threads/rinc%C
3%B3n-de-aviones-especializados.23364/page-7

http://www.fas.org/programs/ssp/man/uswpns/air/special/e3.html

16
 Erieye video

http://www.youtube.com/watch?v=BOaxi2G0aNI

The Saab Erieye system uses an active phased array radar mounted in a
two sided array geometry contained in a large beam shaped structure
carried above the fuselage. The limitation of the two sided array is that it
can only cover two 120 degree sectors abeam of the aircraft, leaving 60
degree blind sectors over the nose and tail of the aircraft

Thinned Arrays
 Reducing the number of
elements leaves the main
lobe unaltered but degrades
the sidelobe level
 Thinning to 10% reduces the
main lobe level down to
10%, but leaves the sidelobe
level at 90%
 If the removed elements are
replaced by matched dummy
elements, then the pattern
remains unchanged, only the
gain is decreased

M. Skolnik, Radar Handbook: McGraw Hill, 1970.

17
 Advantages of Phased Arrays
 Inertialess rapid beam steering
 Multiple, independent beams
 Potential for large peak and
average powers
 Control of radiation pattern
 Graceful degradation
 Convenient aperture shape
 Electronic beam stabilisation

Courtesy Eli Brookner

MMIC Based Phased Arrays

Courtesy Eli Brookner

18
 Courtesy Eli Brookner

Pave Paws AN/FPS-115 Early Warning

Radar

 Frequency 433MHz
 Tx Power 284-440W
 1792 radiating elements
 Each array scans +/-60
 30m diameter
 6000km range for 10m2
target

 Note the 1980s graphical

interface

E. Brookner. (1985) Phased-Array Radars. Scientific American. 76-84.

19
 Pave Paws Radar Coverage

http://www.globalsecurity.org/space/systems/pavepaws.htm

Sea Based X-Band Radar

 Tracking and discrimination for ground-based
midcourse defence (Starwars mk II)
 Mass 50,000 tons, tall as a 24 story building
 Mechanically scanned X-Band phased array, 65%
populated
 45,000 Tx/Rx elements

http://www.mda.mil/news/11news0001.html

20
 Acoustic Phased Array: Paul Thompson

 16 radiating elements
 16 receiver elements
 3D imaging from a
single “ping”

P. Thompson, "Design and Construction of an

Ultrasound Imaging System Using Phased
Arrays," BE Honours Thesis, ACFR/AMME,
University of Sydney, Sydney, 2003.
Target Acoustic Image

Sidescan Sonar & Towfish

Operational Principle

Sidescan Array
G. Brooker, Introduction to Sensors for Ranging and Imaging: SciTech, 2008.

21
 Operational Principle
 A side-scan sonar antenna a short (50) linear transducer array
made of a piezo-electric material that is towed behind a ship

 The transducer is excited by a short (  3s) high voltage sinusoidal

stimulus at a frequency close to the resonant frequency of the array
which the array converts to vibrations and radiates into the water.

 The operational frequency is generally between 50kHz and 500kHz

with some short-range units operating up to 1MHz.

 The same array is used to receive any echoes. These are then
amplified and recorded to form an image. In modern systems, the
signal is digitised in the tow-fish and transmitted to the surface for
processing and display.

 The operational range for low frequency units (f  100kHz) is about

500m, this decreases to 50m at a frequency of 1MHz due to the
increasing attenuation of water with frequency.

Sound Attenuation in Water

Attenuation increases
by a factor of 10 from
100kHz to 1MHz

22
 Beam Pattern
 Because of its shape, the
array produces a fan beam
pattern
 narrow azimuth beamwidth
(typ. 0.75 to 1.5) determined
by the length of the array
 wide elevation beamwidth
(typ. 35 to 65) determined by
the vertical aperture of each
element.
 Arrays are placed on either
side of the tow-fish and
angled slightly downward to
produce the patterns shown

Sidescan Image: The Port Hunter

http://uboat.net/allies/mer
chants/1925.html

Sidescan
Bang Pulse Ocean Floor Playback
Water Krill & Fish etc http://www.youtube.com/watch?v
=AJoqh2SdQZk

23
 Effect of Shadowing

Because of the shallow

grazing angles, shadows
can add significantly to
the information available
from a sidescan image

J. Fish and A. Carr. (2001, March). Acoustics and Sonar (AUSS Ltd).
Available: http://www.marine-group.com/acoustic.htm

Signal Processing
 The standard matched-filtering principles developed for
radar are applied to sonar systems to ensure that the
maximum SNR is achieved.
 Most side-scan systems are real beam in that their cross
range resolution is a function of range xr = R.az
 Digital techniques can be applied to correct for phase
front curvature. This is known as focussing, and it can be
used to achieve a fairly constant linear beamwidth with
range (this is similar to SAR processing).
 Beam scanning techniques using phase shifters in the
arrays can be used to spotlight particular areas
 Simultaneous multi-frequency operation eg. 100kHz and
600kHz is possible for high resolution short range
operation and lower resolution long range operation.

24
 Pseudo 3D Images

If views are made from more than one perspective, they can be
combined into a pseudo 3D image as shown here for the Fritzen
S. Hultqvist. (2007). Swedish East Coast Wrecks. Available: http://www.abc.se/~m10354/uwa/wreck-se.htm

3D Sonar Imaging
 Sophisticated 2D sonar arrays
such as the one developed by
Thomson Marconi Sonar
(TMS) in Sydney can produce
short range 3D images with
voxel resolutions down to
1x1x1mm
 A group of 3 uniformly spaced
transmitters illuminates the
target with high frequency
(>1MHz) sound pulses.
 A sparse phased array made
up of 84 tiles each made up of
a random pattern of 32
hydrophone receivers receives
the echo.

25
Sonar Images

Synthetic Aperture
Radar

26
 Definition
 Synthetic Aperture Radar
(SAR) and Doppler beam-
sharpening (DBS) are
techniques that use the
forward motion of an aircraft
carrying a radar to improve the
cross-range resolution
 Both these techniques can be
used for sonar applications as
well

Space Based SAR

Doppler Beam Sharpening

 Doppler beam-sharpening uses the
decreasing radial velocity (hence
Doppler shift) across the beam
SIDE VIEW
footprint to synthesise improved
 slant range
cross range resolution elevation
h beamwidth

 For an aircraft flying at 250m/s, the

isovel (and isodop) lines at 1.25m/s ground
range gates
spacing are shown
PLAN VIEW
Decrease in Velocity Azimuth Angle azimuth
beamwidth
beam footprint

(m/s) (deg)
Dopple nstant

1.25 5.73
(isodop ift
r sh
s)
co
lines of

2.5 8.11 real aperture

resolution
3.75 9.93 Doppler
beam-sharpened
5 11.48 resolution

6.25 12.84
7.5 14.07

27
 Doppler Beam Sharpening
 Limitations include a trade off  The beam is scanned
between the “sharpening” and the physically to the one side of
observation time the direction of travel
 At 10GHz, 1.25m/s isodop lines are
 A reflectivity image is built up
83Hz apart requiring an observation
time of 12ms to resolve them, 3m
using the higher cross-range
travel time at 250m/s resolution
 At 94GHz, the lines are 800Hz apart
requiring an observation time of
1.2ms (0.3m)

Real Image Sharpened Image

Developing a Synthetic Aperture

Range

Azimuth
Doppler

Pulse
Width

28
 Generation of the Synthetic Aperture
 The term “synthetic aperture”
refers to the distance that the
sensor travels during the time
that the reflectivity data are
collected from a single point
 Energy from each point is
made to arrive in phase at the
output of the processor for all
of the samples to realise the

Range
narrow beamwidth.
 Good range resolution is
Azimuth
obtained using one of the
pulse compression techniques
discussed in the previous Point target signature from a
lecture moving target, before and after
pulse compression

Sampling the Aperture

29
 Synthesizing a Beam
 The process to determine the radiation pattern is similar to that used
for the fixed array
 The primary difference is that the signal received by each element is
due only to the received power scattered by target P from one
transmitter element
 This results in a slightly different radiation pattern for SAR. The
beamwidth is narrower, but the sidelobes are higher than that for the
equivalent phased array. For Le the synthetic array length
 L 
Sin 2  e Sin 
  
G SAR ( ) 

2
 e L 
 Sin 
  
 The half power beamwidth can be found by solving for GSAR()=0.5
and solving graphically (or using Newton)
Le 0.886
Sin   1.39
 2

Unfocussed SAR
 Aircraft motion that deviates from a
straight line and “range walk” is
compensated for.
Range

 One limiting condition for the largest

aperture Lmax is the point where the Azimuth

round trip phase error reaches /4 as Point target signature

determined in the diagram corrected for range walk
 Lmax
 Sin( / 2)
8 2
Lmax
Sin( / 2) 
2R
 for
 Lmax
2

8 4R
Lmax  R  / 2

30
Cross Range Resolution: Unfocussed SAR
 A second limiting condition is that the beamwidth is sufficiently wide
to illuminate the target at point P. Lmax  R 3dB
 The beamwidth is obtained by equating GSAR()=0.5 as before
Le 0.886
Sin   1.39
 2
0.886
Sin 
2Le
 The cross-range resolution cr = R = R.sin for small angles

R0.886
 cr 
2 Le
 Substituting for Le and simplifying Le  Lmax  R  / 2
 cr  0.3 R

Focussed SAR

Removal of the range curvature

from the returns from a point
target into a single range gate to
allow correlation in azimuth that
results in the improved cross-
Range

range resolution

Azimuth

31
 A Doppler Perspective
 A point scatterer enters the forward
edge of the beam. It will have Doppler
frequency:
2v r 2v
fd   cos( 3dB / 2) Target
 
 For small beamwidths, the Doppler
frequency decreases linearly to 0 and
then increases again.
Target just Target just
 The angle to the target as a function enters beam leaves beam
of time is
vt

R
 The Doppler frequency as a function of Synthetic
Aperture
time will then be
2v r 2v  vt 
f d (t )   cos  
  R

Doppler Perspective cont…

 Taking the derivative to obtain the rate of change of Doppler
frequency, or the Doppler slope
df d 2v v  vt 
 . sin  
dt  R R
 At t=0
df d 2v 2

dt R
 The total Doppler shift over time Td = time within the beam for
 = -3dB/2 to +3dB/2, assuming a linear change in frequency
2v 2
f d  .Td
R
 By analogy to the linear FM range resolution, the signal can be
passed through a matched filter to give a spectral resolution
f = 1/Td

32
 Doppler Perspective cont…
 The cross range resolution is then the optimised cross range resolution
of the real beam b = Le scaled by the ratio of the spectral resolution to
the whole Doppler shift

f f
 cr   b .  Le
f d f d
Substituting
R 1 R
 cr  Le 2
 Le 2 2
2 v Td Td 2 v Td
But Le = v.Td = Rb = R/D where D = antenna aperture

R R D D
 cr   . 
2 Le 2 R 2
The cross range resolution for focussed SAR is independent of the range R

Resolution Comparison
Frequency f = 94GHz
Aperture D = 120mm

33
 Distortion in SAR Images
Depression angle

Terrain with slopes steeper than these

tangentials will be imaged with layover

(a) (b) (c) (d)

Layover Weak Shadow Shadow Shadow

Return
Layover Foreshortening
Ground range projection

 layover, when the range to the top of an object is less than the
distance to its base
 foreshortening, when the near side of elevated objects appears
steeper than it actually is
 shadowing, when a tall opaque object blocks the signal path behind
it, and no returns are received

34
Distortion in SAR Images: Stretching

Depression angle

Sla
nt r
ang
e im
age
pla
ne

Ground range image plane

Shadowing in SAR Imagee

Sandia. (2004). X-band Synthetic Aperture Imagery, Sandia Labs, . Available:

http://www.sandia.gov/RADAR/imageryx.html

35
 Measurement
Coherence:
Speckle

Sandia Lab’s Miniature SAR

 Specifications
 Totalmass 10kg
 Range 15km
 Resolution 10cm

36
 China Lake
Airfield 3m

Sandia. (2004). X-band Synthetic Aperture Imagery, Sandia Labs, . Available:

http://www.sandia.gov/RADAR/imageryx.html

Another Airport that

isn’t Hong Kong

Sandia. (2004). X-band Synthetic Aperture Imagery, Sandia Labs, . Available:

http://www.sandia.gov/RADAR/imageryx.html

37
Sandia. (2004). X-band Synthetic Aperture Imagery, Sandia Labs, . Available:
http://www.sandia.gov/RADAR/imageryx.html Piers and a River 1m

Sandia. (2004). X-band Synthetic Aperture Imagery, Sandia Labs, . Available:

http://www.sandia.gov/RADAR/imageryx.html Pipeline Crossing a River 1m

38
 T-72 Tanks in Formation 10cm

Space Based SAR

 To achieve good angular resolutions
from real aperture space-borne radars is
impossible at lower frequencies because
the size of the antenna becomes
prohibitively large.
 With a SAR, the large synthetic aperture
results in a cross range resolution
independent of range cr = D/2 where D
is the antenna aperture
 The good range resolution r = c/2f is
achieved by transmitting a wide
bandwidth chirp.
 Because the trajectory of the satellite or
shuttle is precisely known and stable,
motion compensation is not required and
exceptionally high quality images can be
produced.

39
 Interferometry

 Because SAR is concerned with the phase

relationships between scatterers on the ground
 Two similar images are produced using offset
antennas, or on subsequent passes over the same
area,
 Interference patterns can be used to determine the
true height of the objects on the ground.
 In addition to being useful for mapping ground
features, this technology has a number of other
uses:
 Local deformation of the earth’s crust as an early
warning of earthquakes or volcanoes.
 Ground subsidence due to mining activities or
excessive use of ground-water

Interferometric SAR Image of the San Francisco Area

40
 Mississippi Delta

Oil rigs

Ship

41
 Magellan SAR
map of Venus

(26/11/04). Magellan Mission to Venus. Available:

http://www2.jpl.nasa.gov/magellan/

42
Sif Mons 2km High and 300km in diameter
3D Image produced by combining SAR and altimeter data
(26/11/04). Magellan Mission to Venus. Available: http://www2.jpl.nasa.gov/magellan/

Phased Array
Application

Performance of Sidescan
Sonar System

43
 Sidescan System Evaluation
ITC 5202 Transducer
Size of Unit 68.5x3.8cm
Array Dimensions 1.27x53cm shaded active array
Resonance Frequency 117kHz
Useable frequency range 111-126kHz
Beam pattern: 53cm line 1.5 at 117kHz
1.27cm line 60 at 117kHz
Efficiency >40%
Input power <5% duty cycle 1500W
Operating Depth Unlimited
Weight 4.3kg
Housing aluminium

Transducer
Array
Receive
-180dB rel 1V/Pa

Transmit
170dB rel 1Pa/V at 1m

44
 Worked Example

 What is the smallest target that can be detected by the

ITC-5202 transducer at a range of 500m?

 Operational frequency f = 117kHz

 Velocity of sound (assumed constant) c = 1522m/s
 Wavelength  = c/f = 13mm

Pulse Width and Range Resolution

 The quality factor is determined from the operational band
fr 117
Q   7 .8
f u  f l 126  111
 The rise time of any pulse generated by the transducer is related to
the resonant frequency and the quality factor
Q 7 .8
 rise    66.7 s
f r 117  103
 The minimum pulse-width must be at least twice the rise time if the
pulse it to reach its peak value. For a “rectangular” pulse, the total
pulse-width should be 5rise = 333s. This longer pulse gives
improved long range performance at the expense of the resolution
c 1522  333  10 6
R    0.25m
2 2

45
 Cross Range Resolution

 The cross range resolution at 500m (for the given azimuth

beamwidth of 1.5) is the product of the beamwidth in radians and
the range (no focussing)

1 .5
xr  R  500   13.1m
57.3

Pulse Repetition Frequency (PRF) and

Duty Cycle (DC)

 To operate out to a maximum unambiguous range of 500m, the

maximum pulse repetition frequency PRF is

c 1522
PRFmax    1.52 Hz
2 Rmax 2  500

 The transmitter power is limited to a maximum of 1500W for a Duty

Cycle of less than 5%. The duty cycle in this case is

DutyCycle  100 PRF   100  1 .52  333  10 6  0 .05 %

 which is much smaller than the limit, so the maximum power can be
applied to the transmitter.

46
 Pulse Compression

 If pulse compression is applied to obtain the best possible resolution

for the sonar but using the full duty cycle =5%, then
c 1522
R pc    0.05m
2 f 2  15  103
 And the duration of the pulse increases to

 5
 pc    32.9  10  3 s
100 f pr 100  1.52

Transmitter Power Density

 If the transducer was omnidirectional, then the power density at a
range of 1m would be the product of the electrical power Pelec , and
the conversion efficiency  divided by the surface area of a sphere
with a radius of 1m
Pelec 1500  0.4
I iso    47.7W / m 2
4 4
 The antenna gain, known as the Directivity Index (DI), which is
defined in terms of the power with respect to an isotropic radiator
can be calculated from the elevation and azimuth beam widths (in
radians)
4 4  57.32
G   458.4
 1.5  60
 The actual power density in the direction of the peak gain is the
product of the gain and the isotropic value

I  I isoG  47.7  458.4  21864W / m 2

47
 Sound Pressure Level
 The sound pressure level (SPL) or S is generally given in dB relative
to 1Pa at a range of 1m. This can be calculated from the power
density and the acoustic impedance of the water.
 The acoustic impedance, Z of water is the product of the density and
the velocity
Z   o c  1026.4  1522  1.56  10 6 kg / m 2 sec

 The relationship between the acoustic pressure, P in Pascals, the

power density, I in W/m2 , and the impedance, Z, is

P 2  IZ
 This can be re written for the acoustic pressure in Pa for the sound
pressure level, S, as follows

S  (10 6 P ) 2  1012 IZ

Sound Pressure Level in dB

 This is generally written in dB form

10 log10 S  20 log10 (10 6 P )  10 log10 1012  10 log10 ( IZ )

S dB  120  10 log10 ( 21864  1.56  10 6 )  120  105.3  225.3dB

48
 Sound Pressure Level from Graphs
 It can be seen that at 117kHz, the
transmitter voltage response is
170dB rel to 1Pa/Volt at 1meter.
 The electrical power input Pelec is
related to the RMS voltage V and the
transducer conductance G

Pelec  V 2G
 For a conductance G = 5.5k mho
from the transducer specification
table and a power Pelec = 1500W

Pelec 1500
V    522Vrms
G 5.5  10  3
 The sound pressure level SdB for
522V applied to the transducer is
S dB  170  20 log10 (V )  170  20 log10 (522)  224 .3dB

Transmission Loss
 As the signal propagates through the water, the sound pressure
level reduces because the wave is expanding on a spherical wave-
front and due to attenuation. The transmission loss in dB is H and is
determined as follows:
r2
H  20 log10   dB ( r2  r1 )
r1
 Because the sound pressure is determined relative to the level
existing at one meter from the effective centre of the sound source,
the equation can be rewritten for this reference distance as follows
H  20 log10 r   dB r
 The attenuation in dB/m is given by the following formula where f is
the frequency of the sound in kHz.

0.036 f 2 0.036  117 2

 dB   3 . 2  10 7 2
f   3.2  10  7  117 2  0.0329dB / m
f 2  3600 117 2  3600

49
 Target Strength T
 T in dB is defined by ratio of the reflected sound pressure scattered
by the target at a distance of one meter from the effective centre of
the scattered sound to the incident sound pressure on the target
Pr
T  20 log10
Pi
 This target strength is determined by its size, shape and the fraction
of sound that is re-radiated.
 If the scattering cross section is  square meters, then T in dB is
given by the following formula

T  10 log10
4
 As with the radar case, a sphere with a radius, a, much larger than
the wavelength will have a cross section equal to the projected area
a 2
2
a a
T  10 log10  10 log10    20 log10
4 2 2

Applying the Sonar Range Equation for a

Spherical Target

 For a spherical target, the echo sound pressure level E relative to

1Pa at a range of 1m from the receiver is easily calculated as
follows

E dB  S dB  2 H  T

a
EdB  S dB  40 log10 r  2 dB r  20 log10
2

E dB  225 .3  40 log10 r  0 .0656 r  12

50
 Range Eqn Applied to the Sea Floor
 Target strength will be the product of the range resolution and the
cross range resolution modified by a scaling factor to take into
account the reflectivity o of the surface.
 oA  o .R.XR  oR.r. az
T  10 log10  10 log10  10 log10
4 4 4
 and the echo SPL 1m from the receiver will be

 oR. az
EdB  S dB  40 log10 r  2 dB r  10 log10 r
4

E dB  225 .3  40 log10 r  0 .0656 r  42 .8  10 log10 r

SdB – 225.3dB
dB – 0.0328 dB/m
o – 0.1 m2/m2
az – 1.5 (0.0262 rad)
R – 0.25m

Noise Floor
 The noise level at sea is mostly generated by wind and wave action
on the surface. It is proportional to sea-state and inversely
proportional to frequency.
 From the table reproduced in the notes, we will assume sea state 3
generated by a wind speed of 15 knots
 Isotropic Noise pressure N1 (dB relative to 1Pa) into a 1Hz
bandwidth at a frequency of 1kHz is 65dB
 The frequency relationship to map the noise pressure level at 1kHz
to the transducer frequency is (see Fig 9.4 in your book)
N f  N1  17 log10 f kHz
 For a sea state 3 and the transducer frequency of 117kHz

N f  65  17 log10 117  30 dB

51
Noise Floor

52
 Noise Pressure Level: Pulsed System
 Because the noise floor is defined for an isotropic receiver into a
1Hz bandwidth
 The total noise pressure level in dB relative to 1Pa must take into
account the bandwidth  of the receiver (in Hz) and its directivity or
gain G.

LN  N f  10 log10   10 log10 G

LN  30  10 log10 (3  10 3 )  10 log10 ( 458.4)  38dB

 The minimum detectable signal level will be 13dB higher than this

Smin = 38+13 = 50dB

Noise Pressure Level: Pulse Compression

 The receiver bandwidth for a pulse compression system can be
approximated by the reciprocal of the uncompressed pulse width
(signal observation time
pc = 1/pc = 30Hz

 The minimum detectable signal level will be 13dB higher than this

Smin = 38+13 - 20 = 30dB

53
 Signal and Noise Levels

Targets Detectable at 500m

 The sea floor with a SNR = 25dB

 A sphere with a diameter of 1m with a SNR = 30dB
 A sphere with a diameter of 0.1m might be detectable but
with an SNR = 10dB it cannot produce a Pd = 0.9 and a
Pfa = 10-6

54
 Signal Level out of Transducer
 The actual voltage output by the
transducer is determined from the
transducer specifications. The
open circuit receiving response at
117kHz is –180dB rel 1V/Pa. For
a signal pressure of 50dB (the
sea-floor return at 500m), the
output is
20 log10 (V )  50  180   130 dB
V  10 130 / 20  320 nV
 This is very small, and so receiver
noise would be a consideration
when the actual detection
characteristics of the system were
being considered.

55