.. . .3. .

.‘o.

 

.m

.11 w.

um...

 

I»

9.71:4? .

 

mgr..-

3!"!

 

. .. 3..
. . .. n VmeriW.
W1. ..... ...v...«WMWWf
3. ,. . , . _,
fl .3... .. . H. _ . .
«WW. 3.:
n... Swan.
. 2M3 . .Mm
Sanka.” ”I 7
3...... 55
J“ z. .5...»

3..»...

 

 

.43....

.. x... bw .6.”
.} .. ..

u...

29,.

.. as 9 a...

‘
3v

7)
g:

 

f6
f

g

17.1 I,
a .nhmrn.
{Hy-.14

v.1... I:

 

Ari! .1
I: It I It":
I». 5 ‘11.. ..
cu. ..s....u...fl... 12, ..

r. ..
. 3. 33...}...
'rv'uc AI: 1“ e
i .z. .s. .
4.5m.

2.1...”
5.:

..

i .01“ 1.

vi... :Iarxrc. ..

.. ..L.Nu...ur1:fi!3
1..

  

 

I

 

 

r. n l LIBRARY
56. a.” Michigan State
University

This is to certify that the
thesis entitled

A UNIQUE APPROACH TO FREQUENCY-MODULATED
CONTINUOUS-WAVE RADAR DESIGN

presented by

Gregory Louis Charvat

has been accepted towards fulfillment
of the requirements for the

Master of degree in Electrical Engineering
Science

 

 

g/l/K/L/

Major Professor’s Signature

ie‘Cqu-u’ 0'3), 200»?

 

Date

MSU is an Affinnative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

nrr‘

I '5 ““41

 

'0'925 I'll

 

792511

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDateDue.p65-p.15

 

 

A UNIQUE APPROACH TO FREQUENCY-MODULATED CONTINUOUS-WAVE
RADAR DESIGN

By

Gregory Louis Charvat

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Electrical and Computer Engineering

2003

ABSTRACT

A UNIQUE APPROACH TO F REQUENCY-MODULATED CONTINUOUS-WAVE
RADAR DESIGN

By

Gregory L. Charvat

Frequency—Modulated Continuous-Wave (FMCW) Radar has traditionally been used in
short range applications. Conventional FMCW radar requires the use of expensive
microwave mixers and low noise amplifiers. A uniquely inexpensive solution was
created, using inexpensive Gunn oscillator based microwave transceiver modules that
consist of 3 diodes inside of a resonant cavity. However these transceiver modules have
stability problems which cause them to be unsuitable for use in precise FMCW radar
applications, when just one module is used. In order to overcome this problem, a unique
radar solution was developed which uses a combination of 2 transceiver modules to
create a precise FMCW radar system. This unique solution to F MCW radar is proven to

be capable of determining range to target, and creating Synthetic Aperture Radar images.

Cepyright by
GREGORY L. CHARVAT
2003

This Masters Thesis is dedicated to those individuals who continue to be a positive
influence in my life, and have always encouraged my work:

Dave and Rita Charvat
Louis and Jane Charvat

Gerard and Therese Nefcy

iv

I. 'E

“Look, Gail.” Roark got up, reached out, tore a thick branch off a tree, held it in both
hands, one fist closed at each end; then, his wrists and knuckles tensed against the
resistance, he bent the branch slowly into an are. “Now I can make whatever I want of it:
a bow, a spear, a cane, a railing. That’s the meaning of life.”

Ayn Rand, “The Fountainhead”

TABLE OF CONTENTS

LIST OF TABLES ............................................................................... vii
LIST OF FIGURES ............................................................................. viii
CHAPTER 1
OVERVIEW ....................................................................................... 1
CHAPTER 2
GENERAL THEORY OF OPERATION ....................................................... 5
2.1 Unmodulated CW Doppler Radar .............................................. 6
2.2 Frequency Modulated Radar .................................................... 9
2.3 Synthetic Aperture Radar (SAR) .............................................. 15
CHAPTER 3
PRACTICAL FMCW RADAR SYSTEMS, PAST AND PRESENT ..................... 22
CHAPTER 4
THEORY OF OPERATION .................................................................... 31
4.1 The MA87127-1 Gunn Diode Based Transceiver Module. . ..........32
4.2 A Unique FMCW Radar Solution ............................................ 36
4.3 System Design and Implementation .......................................... 42
4.3.1 The Front End Assembly ............................................... 44
4.3.2 The Power Supply ....................................................... 51
4.3.3 IF Chassis ................................................................. 55
4.3.4 Data Acquisition ........................................................ 65
4.3.5 Software ................................................................. 67
CHAPTER 5
EXPERIMENTAL RESULTS .................................................................. 82
5.1 Tuning Linearity Experimental Results ....................................... 83
5.2 Local Oscillator Amplitude ..................................................... 85
5.3 Standard Radar Targets ......................................................... 88
5.4 Single Beam Range to Target Experimental Results ....................... 90
5.5 Synthetic Aperture Radar Experimental Results ............................ 98
CHAPTER 6
CONCLUSION .................................................................................. 1 10
REFERENCES ................................................................................... 113

vi

LIST OF TABLES

Table 5.1: Single beam range to target experimental results summary ................... 92

Table 5.2: A list of the SAR images taken and their corresponding image plots. . ......100

vii

LIST OF FIGURES

Figure 1.1: Simplified coherent radar block diagram .......................................... 2
Figure 1.2: Simplified CW radar block diagram ............................................... 3
Figure 2.1: Block diagram of a CW Doppler radar system ................................... 6
Figure 2.2: Block diagram of a basic FMCW radar system .................................. 9
Figure 2.3: Linear ramp used to PM modulate the FMCW radar system .................. 10
Figure 2.4: Range profiles acquired in a straight line, where the antenna beam of the
radar system is perpendicular to the line ........................................... 16
Figure 2.5: Phase error progression of an n-element synthesized aperture ................ 17
Figure 3.1: Radio altimeter installed on an aircraft ........................................... 23
Figure 3.2: Rotating air variable capacitor chirp mechanism ................................ 24

Figure 3.3: Typical method for sealing altitude return signal level, processing return
distance, and displaying altitude used in radio altimeters ....................... 26

Figure 3.4: The video output of an FMCW radar system can be digitized, and with some
simple signal processing, the range to target data displayed in the frequency
domain ................................................................................. 27

Figure 3.5: Antenna coupling limits the sensitivity and dynamic range of FMCW radar

systems ................................................................................. 28
Figure 4.1: The MA87127-1 microwave transceiver module ............................... 32
Figure 4.2: MA87127-1 block diagram ......................................................... 33
Figure 4.3: MA87127-1 Physical Layout ...................................................... 34
Figure 4.4: Simplified block diagram of the unique FMCW radar solution ............... 36
Figure 4.5: System block diagram ............................................................... 42
Figure 4.6: FMCW radar system ................................................................ 43

viii

Figure 4.7: Front end assembly block diagram ................................................ 44

Figure 4.8: The front end assembly .............................................................. 45
Figure 4.9: Schematic of XCVRl and XCVR2 interconnections. . . .47
Figure 4.10: Transmit and Receive Antennas .................................................. 50
Figure 4.11: IF Chassis ............................................................................ 51
Figure 4.12: Power supply schematic ........................................................... 52
Figure 14.13: Physical layout of the power supply ............................................ 53
Figure 4.14: Block diagram of IF chassis ....................................................... 56
Figure 4.15: IF chassis physical layout ......................................................... 57
Figure 4.16: AMP2 schematic .................................................................. 59
Figure 4.17: AMP3 schematic .................................................................. 60
Figure 4.18: Video Amplifier Schematic ...................................................... 62
Figure 4.19: Modulation amplifier schematic ................................................. 64
Figure 4.20: Block diagram of the Data Acquisition signal chain .......................... 65
Figure 4.21: Schematic of NI6014 input and output connections ........................... 66
Figure 4.22: Continuous Operation and Display VI front panel ............................ 68
Figure 4.23: Continuous Operation and Display VI block diagram ........................ 70
Figure 4.24: Single Range Profile Capture VI front panel ................................... 71
Figure 4.25: Single Range Profile Capture VI block diagram .............................. 72
Figure 4.26: The SAR VI depends on the user to manually move the radar down a metal
track .................................................................................. 73
Figure 4.27: The Synthetic Aperture Radar VI front panel ................................... 75

Figure 4.28a: Synthetic Aperture Radar VI block diagram, showing the data acquisition

loop ................................................................................. 78

ix

Figure 4 .28b: S ynthetic A perture R adar V I block diagram, s howing the S AR p rocess

loop and the file save sequence ................................................. 80

Figure 5.1: XCVRl voltage versus frequency plot (relative to XCVR2). . .83
Figure 5.2: Power output of AMP2 versus frequency (relative to XCVR2), when ATTNl
is set to 20 dB ......................................................................... 86

Figure 5.3: Power output of AMP2 versus frequency (relative to XCVR2), when ATTNI
is set to 0 dB ........................................................................... 87

Figure 5.4: Trihedral comer reflector standard radar target ................................. 88

Figure 5.5: 20 stm and 30 stm trihedral comer reflector standard radar targets. . ...89

Figure 5.6: Single beam range to target experimental setup ................................. 90
Figure 5.7: Picture of the experimental setup, where the 30 stm target is in the
background, and the FMCW radar system is in the foreground ............... 91
Figure 5.8: Range profile of 30 stm target at 20 ft ......................................... 92
Figure 5.9: Range profile of 30 stm target at 25 fl ......................................... 93
Figure 5.10: Range profile of 30 stm target at 30 ft ....................................... 94
Figure 5.11: Range profile of 30 stm target at 35 fi ....................................... 95
Figure 5.12: Range profile of 30 stm target at 40 fi ....................................... 96
Figure 5.13: Synthetic Aperture Radar experimental setup ................................. 98

Figure 5.14: The Front End Assembly is manually pushed along the aluminum track and
lined up with a measuring tape in order to create a SAR image ............. 99

Figure 5.15: Both the 30 stm and 20 stm standard radar targets are placed down

range from the SAR system ...................................................... 100
Figure 5.16: Base line measurement before SAR processing .............................. 101
Figure 5.17: Base line measurement after SAR processing ................................ 102

Figure 5.18: 30 stm target located at a distance of 25 fi before SAR processing. . ...103

Figure 5.19: 30 stm target located at a distance of 25 fi afier SAR processing ....... 104

Figure 5.20: 20 stm target located at a distance of 25 it before SAR processing. . ....105
Figure 5.21: 20 stm target located at a distance of 25 ft after SAR processing ....... 106

Figure 5.22: 20 stm target located at a distance of 25 fl, and a 30 stm target located
at a distance of 40 ft before SAR processing .................................. 107

Figure 5.23: 20 stm target located at a distance of 25 ft, and a 30 stm target located
at a distance of 40 ft afier SAR processing .................................... 108

xi

Chapter 1

Overview

Continuous-wave (CW) radar is an old concept. CW radar uses a simple coherent design
topology. When frequency modulated (FM), CW radar can be used as a precise short
range radar. CW radar has been in widespread use since the mid 1930’s. CW radar is
advantageous for use in precision Doppler measurements. A great variety of microwave
parts are available for use in CW radar systems. A unique solution to CW radar was
created using a pair of inexpensive Gunn diode based microwave transceivers. Test
results on this unique solution were positive, proving that it was capable of short range to
target measurements, and a viable solution for use in a Synthetic Aperture Radar (SAR)

system.

CW radar is based on a coherent radar design topology. The fundamental difference
between a simple coherent radar (see figure 1.1) and a CW radar (see figure 1.2) is the
absence of a transmit mixer. In a CW radar, the transmit oscillator is connected directly
to the transmit antenna. Some of the transmit oscillator’s power is coupled off and used
to feed a receive mixer. A second antenna is connected to the input of that mixer. The
mixer beats in the difference between the transmit oscillator and any RF received in the
RF input port. The IF output of the mixer is the Doppler shift reflected from a moving

target. When this system is FM modulated, the radar can precisely indicate range to

target as a function offrequency. T his type 0 fradar is known as FMCW. T he time
difference between the FM modulated carrier being transmitted and the carrier that is
reflected off of a target can be measured on the IF output port of the receive mixer. This
time difference is in the form of frequency, and can be counted or analyzed as a function

of distance to target.

oooooooooooooooooooooo

 

 

 

 

 

 

oooooooooooooooooooo

ooooooooooooooooooooooo

ooooooooooooooooooooooo

 

 

 

Figure 1. 1: Simplified coherent radar block diagram.

 

 

 

Figure I . 2: Simplified CW radar block diagram.

Wide spread use of FMCW radar first occurred during the mid 1930’s [3]. Radio
altimeters mounted on aircraft made use of FMCW radar to measure the distance from
the aircraft to the ground directly below. Early radio altimeters used mechanical methods
to create the FM chirp needed to acquire range information. A typical design would
include an electric motor attached to an air variable capacitor used to modulate the
frequency of the transmit oscillator. The PM modulated transmission would allow range

data to be extracted from the return echo off of the ground below.

Many inexpensive microwave components and assemblies are currently available to
designers. One of the most inexpensive microwave components available today is a
transceiver module based around a Gunn oscillator with a varactor tuning diode, and a
Schottky mixer diode. This inexpensive microwave transceiver is widely known as the
Gunnplexer. A u nique approach to F MCW radar w as d esigned u sing the inexpensive

transceiver modules. This design incorporates two transceiver modules in a bistatic

configuration, where RF power from one is leaked into the other, creating a self tracking

coherent FMCW radar system.

Test results were positive. The unique solution to FMCW radar was proven effective in
measuring range to target. The unique radar solution was then successfully used to create
a SAR imaging system. This research has concluded that using two inexpensive
microwave transceiver modules in unison can yield extremely good performance for use
as an FMCW radar system at a low cost. This research has also concluded that the
unique solution to FMCW radar is capable of performing in the most advanced

applications, such as Synthetic Aperture Radar.

Chapter 2

General Theory of Operation

Two types of CW radar will be examined here. An introduction to Synthetic Aperture
Radar theory will also be presented. The first type of CW radar is the unmodulated CW
Doppler radar system. The second type of CW radar is the frequency modulated FMCW
radar system. CW radar systems have been proven useful in measuring precision
Doppler shifts on moving targets. FMCW radar systems have been proven useful in
measuring both Doppler shifts of moving targets and range to moving or fixed targets.
CW radar is often a simple cost effective solution to Doppler and short range

measurements. One useful application of CW radar is Synthetic Aperture Radar.

2.1 Unmodulated CW Doppler Radar System
Unmodulated CW radar is frequently used to precisely measure Doppler shifts of moving

targets. The block diagram of a simple CW radar system is shown in figure 2.1.

 

 

 

 

 

 

Figure 2.1: Block diagram of a CW Doppler radar system.

Looking at figure 2.1, the fixed oscillator OSCl drives the power amplifier PAl. The
output carrier wave from PAl is fed to the transmit antenna, and radiated out to the
moving target. The carrier reflected off the moving target is Doppler shifted by a

frequency of f 4 Hz given by the equation [1]:

_ 2M,
C

f. (2.1)

Where: f, = transmitted frequency (Hz)
c = velocity of propagation (m/sec)

v, = relative velocity of the moving target to the radar (In/sec)

The reflected signal fiom the target is amplified by LNAl. The output of LNAl is fed
into the RF port of MXRl. Some power fi'om the output of PAl is coupled into the local
oscillator (LO) port of mixer MXRl. The product of the reflected signal from the target

andthe carrier outputofPAl are acquired fromthe IFportofMXRl. T hus IFport

output of MXRl is the Doppler shift f d due to the moving target. This is shown by

equations 2.2 through 2.4.

Frequency reflected from the moving target:

f reflected = foscr + f d (2-2)

Frequency output of the IF port of MXRl (neglecting the high frequency mixer product

because of the high frequency cut-off of most practical microwave mixers):

le = freflected - fOSCl (2'3)

And it is proven, the frequency at the IF port of MXRl:

le = fd (2-4)

The amplitude of signal reflected from the moving target is a function of the radar range

equation:
P443) = Pt(dB) + Gt(dB) + Gr(dB) + 0-013) ‘ 20103(f05c1 ) " (2 5)
— 40log(R) — 3o log(47r) + 2010g(c) '
Where: PM?) = relative logarithmic power collected by the receive antenna in dB

PM” relative logarithmic power transmitted by the transmitter in dB

GM” = logarithmic gain of the transmit antenna in dB
GM,” = logarithmic gain of the receive antenna in dB

0.13 = radar cross section (RCS) of the moving target in stm

R = range to target in meters

The basic principles of CW Doppler radar are not difficult to understand. The CW
Doppler radar system described here can be constructed easily with coaxial parts from
manufacturers such as Mini-Circuits and Miteq. A radar system such as this is very

effective in performing Doppler measurements.

2.2 Frequency Modulated Radar

When modulated, a CW radar system is capable of providing range to target information.
When a CW radar system is FM modulated, the range to target information provided is in
the form of heat frequencies. This is known as FMCW radar. The beat frequencies on
the video output of an FMCW radar system correspond to multiple targets and their
ranges. FMCW radar systems are also capable of measuring the Doppler shift of a

moving target. The block diagram of a basic FMCW radar system is shown in figure 2.2.

 

 

 

 

 

 

 

 

Figure 2. 2: Block diagram of a basic FM C W radar system.

Looking at figure 2.2, OSCl is FM modulated with a triangular ramp input with a period
of 2 ms. The triangular ramp is an alternating linear ramp with both positive and

negative slopes as shown in figure 2.3.

FM Modulation Linear Ramp
30 I r I 7 I I I I

 

Positive Slope Linear Ramp

 

 

 

E \ Negative Slope Linear Ramp
$15 _ / -
>
10 ~ d
5 — .
o I I l I I I I I I
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (seconds) x 10-3

Figure 2.3: Linear ramp used to FM modulate the FMCW radar system.

The PM output of OSCl is fed into PAl. PAl amplifies OSCl to an appropriate transmit
level. The output of OSCl is radiated out of the transmit antenna. The FM modulated
carrier is reflected off of the target at a range of R meters. The reflected signal is delayed
in time on its way to and from the target. The reflected signal is received by the receive
antenna, and amplified by LNAl. The output of LNAl is fed into the RF port of MXRI.
Some power from PAl is coupled into the LO port of MXRl. When the LO and the RF
are multiplied together in MXRl, the IF output of MXRl is the range to target in
frequency. This range to target in terms of frequency is known as the beat frequency.

The greater the beat frequency on the IF output port of MXRl, the greater the range to

10

target. If the target is moving, then the Doppler shift of the moving target is added onto
the beat frequency present on the IF port of MXRl. The relationship between frequency,
FM chirp bandwidth, range to target, and Doppler frequency shift can be found using the
equations for both a positive and a negative linear ramp modulation waveform [2]. When
a positively sloped linear ramp FM modulates OSCl, the beat frequency at the IF port of

MXRl is represented by:

BMMR

C

fb zf; =

 

- fd (2.6)

Where: f b = the beat frequency at the IF port of MXRl

f; = the beat frequency at the IF port of MXRl when OSCl is modulated with
a positive linear ramp.

Af = chirp frequency deviation

f," = FM modulation rate

R = range to target

When a negatively sloped linear ramp FM modulates OSCl, the beat frequency at the IF

port of MXRl is represented by:
_ 8A mR
fb =11, =—’§——+fd (2.7)

Where: f b’ = the beat frequency at the IF port of MXRl when OSCl is modulated with

a negative linear ramp

From the equations above, the range to target is found using:

11

C

= SAflm

 

R (f,) (2.8)

Where: <fb> = the average frequency difference

If the target is moving, the velocity of the target can be found using:

xi

View) (2.9)

The amplitude of the return signals can be approximated using the logarithmic radar

range equation 2.5.

The basic principle of FMCW radar is somewhat difficult to conceptualize. The most
important concept explained here is that a shift in time corresponds to a shift in
frequency. This is because the radar is frequency modulated in time. The current value
of the transmitted frequency is different than what was transmitted 2 ns ago. These small
and subtle frequency differences make up the beat frequencies on the IF output of MXRl ,

and hence the range to target information in the form of low frequency beats.

FMCW radar systems are capable of measuring distances of very close targets with high
accuracy. T he r ange r esolution o btainable b y an F MCW r adar s ystem depends on its
transmitted chirp bandwidth. The more frequency that is covered during a chirp, the
higher the range resolution. It is also important to transmit an extremely linear frequency
chirp. Any nonlinearities in the frequency chirp reduces range resolution and maximum

unambiguous range.

12

The maximum range resolution obtainable for an FMCW radar system depends on the
chirp bandwidth:
B = 2Af (2.10)

Where: B = chirp bandwidth

The maximum range resolution is determined using the equation:

5R=— 2.11
23 ( )

Where: (57? = maximum range resolution

The In aximum range resolution 0 nly o ccurs w hen round trip time for the signal, T , is
substantially less than the time it takes for one full chirp:

T=2'—I—z—<<-L (2.12)
c

m

Where: T = round trip time for the radar signal to reach its target and be reflected back

to the radar

The maximum range at which the radar can detect and accurately measure range is
known as the maximum unambiguous range. The maximum unambiguous range can be
determined using the following equation:

C

m :3": (2.13)

Where: Rm = the maximum unambiguous range

13

FM tuning linearity is extremely important to characterizing both range resolution and
maximum unambiguous range. Both are adversely affected by poor FM chirp linearity.
The PM chirp linearity accuracy required when given a desired range resolution and

maximum unambiguous range can be determined using the equation:

fi— = g! = chirp nonlinearity (2-14)

Where: 6f = deviation of chirp from linear

An example of determining the required chirp nonlinearity for a desired range resolution
and maximum unambiguous range:
An FMCW radar system is required to have a range resolution of 2 feet, and a
maximum unambiguous range of 150 feet. Thus, the chirp must not be any more

than 1.33% nonlinear.

14

2.3 Synthetic Aperture Radar (SAR)

An FMCW radar system can be used as the radar unit in a Synthetic Aperture Radar
system. A SAR system is capable of creating detailed radar images. These images are
created by combining range profiles from the FMCW radar unit at many different
locations. A range profile is defined as the range to target information acquired from a
radar system. When a number of individual range profiles are combined together in a

certain way through a SAR process, it is possible to create a detailed radar image.

For this application, a focused array SAR system will be explored [6]. In a focused array
SAR system, a large number of range profiles are collected and spaced at a defined
interval. The large number of range profiles are collected by moving the radar in a
straight line, and projecting the antenna beams outwards in a perpendicular direction to

that line, as shown in figure 2.4.

15

reader“
' System;

.,__d

I

1‘1 1

:. TravelinginaMx.’
; Straight Line I ‘

--Locafionsf'i

Radaf‘.i...‘j

//. Antenna;
:i‘::Beam"

\

/

 

 

 

 

aWhere: Previous _ ’
~Range PrOfileS '
EWereAcquired _

_

 

Figure 2. 4: Range profiles acquired in a straight line, where the antenna beam of the

radar system is perpendicular to the line.

16

‘ ‘ ; .1 ' ; .DeltaiRitranstatesi
.3. ‘j'toiphaso: ' ..
' .3 Zprwréesiéfi Phil f

     

 

 

 

Figure 2. 5: Phase error progression of an n-element synthesized aperture.

Of all the range profiles acquired, every N number of range profiles collected are used to
synthesize an N element focused array. The array of N elements is focused onto each

point Pl. at a range R, downrange from the radar and centered in the middle of the array
(see figure 2.5). When the aperture is looking at the point B. , a progressive phase error

occurs between the center of the array and each element on either side of the center of the

17

array, as shown in figure 2.5. This progressive phase error is approximated using the

equation:

 

Ill

7R.

2
¢n 2 [d") radians (2.15)
Where: ¢n = phase error

d" = distance from center of N element array

n = n’th element from the center of the N element array

R‘. = range to point B.

The N element array is then focused on to point P. at range R, using the phase error

1

equation 2.15. This focusing procedure is achieved by using the following equations:

x; = x" cos ¢n + y" sin (in (2.16)
Where: x; = the phase compensated real component of the n’th element in the N
element array
x" = the uncompensated real component of the n’th element in the N element
array
y" = the uncompensated imaginary component of the n’th element in the N

element array

18

y. = y. COS¢. - X. sin ¢.. (217)
Where: y; = the phase compensated imaginary component of the n’th element in the N

element array

The sum of the real and imaginary components of the phase compensated N element

array are then summed:

Xzfix' (2.18)

n
n=l

Where: X = the sum of all the phase compensated real components of the N element
array

N = N element array length

hi)»; (2.19)

n=1
Where: Y = the sum of all the phase compensated imaginary components of N element

array

The magnitude of the sum of real and imaginary components of the N element array is

taken using the equation:

.9:sz +Y2 (2.20)

Where: S = the magnitude of the focused N element array on point B. at range R,

19

This magnitude is then saved as a new range bin which represents the distance R). The
entire process is repeated again for range R)+1 , where the resulting magnitude is stored as
a new range bin which represents the distance Rm . The process is repeated until the N

element array has produced a focused magnitude for every single range bin.

After the N element array is finished focusing on every range bin, it moves on. The radar
is moved forward by one increment and a range profile is acquired. The new range
profile, plus the previous N-l range profiles are used to synthesize and focus another N

element array, where the procedure described above is performed again.

This entire process is performed repeatedly, producing lines of focused range bins, until
the user of the radar system decides to stop. The resulting focused range bins and their
accompanying focused magnitudes are then used to construct a detailed Synthetic

Aperture Radar image.

20

Summary:

FMCW radar is a proven technology capable of precisely ranging targets at close ranges.
It is also capable of measuring the Doppler shift of a moving target. The basic FMCW
radar system description and accompanying equations described above are extremely
important to understanding the concept and some design aspects of this type of radar
system. An FMCW radar system also has the potential to be used as a Synthetic Aperture
Radar. The principles of SAR explained in this section are essential to understanding the

application of FMCW in a SAR system.

21

Chapter 3
Practical FMCW Radar Systems, Past

and Present

FMCW radar has been in use for many years. Its widespread use first started in the mid
1930’s, where it was used in aircraft radio altimeters. There were many design
challenges in implementing early FMCW radar systems. Those challenges included
generating a linear chirp, acquiring frequency domain range data, and antenna coupling
issues. Many of the design issues of early FMCW systems are still a challenge today.
These include antenna coupling, and creating a precision chirp using an inexpensive
oscillator. Regardless of the design challenges of FMCW radar, it has proven an

effective way to measure range and Doppler shift in many applications.

FMCW radar w as first used primarily in aircraft radio altimeters. The radio altimeter
used FMCW radar to measure the distance from the aircraft to the ground below (see
figure 3.1). Unlike a conventional altimeter, which indicates altitude above sea level, the
radio altimeter indicates altitude above anything on ground, including mountains, trees,

and other terrain features. One example of an early radio altimeter was the Western

22

Electric Type IA radio altimeter [3]. The Type 1A was introduced in 1936. It had a

maximum height range of roughly 5000 feet.

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.1 : Radio altimeter installed on an aircraft.

Early radio altimeters used mechanical methods to generate a linear frequency chirp.
Typical mechanical frequency modulation methods included agitating or moving crucial
parts 0 f a V HF o r microwave o scillator. T hese p arts that w ere m echanically agitated
were usually made of a non-conducting dielectric inside of a microwave cavity oscillator,
or an air variable cap. An example of early radio altimeter modulation schemes is the
Raytheon model AN/APN-22 [4]. Developed for the US Navy in 1954, this radio
altimeter used a CW cavity magnetron transmitting tube which was feeding a horn
antenna, connected together using waveguide. The magnetron would chirp from 4.2 GHz

to 4.4 GHz using a mechanical vibrating reed system. The vibrating reed varied the

23

frr

TE:

Ill:

be

capacitance across the straps of the anode cavity on the magnetron assembly. This
vibrating reed w as c ontrolled b y a m agnetic c oil, 8 irnilar to that o f a loudspeaker. A
second coil was also connected to the reed, allowing for feedback control of the vibrating

reed.

One of the more interesting methods of producing a chirp is in using a rotating air
variable capacitor [5]. A small electric motor would be connected to an air variable
capacitor, which would rotate, producing FM modulation in the transmitter oscillator
circuit (see figure 3.2).

...................................................................

...................................................

..........................

 

.................

 

 

aaaaaaaaaaaaaaaaa

 

 

 

 

..........................

 

.................................................
...................................................................

...................................................................

Figure 3.2: Rotating air variable capacitor chirp mechanism.

Modern FMCW radar systems typically use a solid state varactor diode to modulate the
frequency of the transmit oscillator. The varactor diode is typically placed in the
resonant tank circuit of a free-running oscillator. With the varactor diode placed in the
tank circuit, the free-running oscillator becomes a voltage controlled oscillator (V CO).
Varactor diodes are not the most linear way of shifting the frequency of a VCO, however
they are the least expensive solid state solution to creating an F M chirp. Great care must

be taken in determining the most linear range of operation of a varactor tuned VCO when

24

used as the transmit oscillator of an FMCW radar system. The Bendix ALA-51A is an
example of radio altimeter that uses a varactor diode to modulate its transmit oscillator
[1]. The yttrium-iron-garnet (Y IG) oscillator is a good alternative to the varactor VCO.
A YIG oscillator is a ferromagnetic tunable resonator, in which the frequency is varied by
changing a DC current controlling a magnetic field, for this reason YIG oscillators have
excellent tuning linearity. YIG oscillators can be used as VCO’s, and will produce a low
distortion linear chirp, ideal for FMCW radar. Because of their excellent tuning linearity
linearity, YIG oscillators are often used as the sweep L0 in most modern spectrum
analyzers and network analyzers. However you will not find them in most FMCW radio
altimeters because YIG oscillators are very expensive, and are not suitable for a cost

effective F MCW radar solution.

Acquiring and processing range data from an FMCW radar system has always been a
challenge. Early, and many current, FMCW radar systems use a simple frequency
counter. There is a difference, however, between early and modern frequency counters.
The early counters would use a frequency to voltage converter of some type [4] [5], and
display the range data on an analog meter display. Modern radio altimeters, such as the
Bendix ALA-51A, use a conventional digital frequency counter. Both early and modern
radio altimeters use an interesting trick in the video signal chain to scale the return
amplitudes of high and low altitudes. Low altitudes beat in as relatively low audio range
frequencies. High altitudes beat in as relatively high audio range and low RF range
frequencies. The amplitude return from the ground of a low altitude beat is much higher

than that of a high altitude beat. Thus, the designers of radio altimeters use a high pass

25

filter in the video chain to attenuate the low altitude beat signals, and amplify the high
altitude beat signals [1]. This high pass filter typically has a cut off response of 6 dB per
octave. It is placed in series with the video signal chain, before the frequency counter as
seen in figure 3.3. The output of the high pass filter is fed into a limiting amplifier. That
amplifier is fed into a frequency counter which counts 0 V crossings. The output of the

frequency counter is fed into an altitude display indicator.

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.3 : Typical method for scaling altitude return signal level, processing return

distance, and displaying altitude used in radio altimeters.

FMCW radar range data can also be processed by digitizing the video output. Multiple
target returns can be acquired by processing the digitized video output of an FMCW radar
system. This is achieved by performing a Fast Fourier Transform on the digitized
waveform (see figure 3.4). The resulting frequency plot indicates the target ranges in
terms of frequency, and target return amplitudes. With modern digitizer technology, it is
possible to expand the applications for FMCW radar systems, beyond simple range to

large target applications.

26

 

 

 

 

 

 

 

 

 

Figure 3. 4: The video output of an FMCW radar system can be digitized, and with

some simple signal processing, the range to target data displayed in the frequency

domain.

Antenna c oupling is the o ne p roblem that h as always plagued F MCW r adar d esigners
since the 1930’s to present day. Antenna coupling limits the sensitivity of the receiver,
and the dynamic range of the entire radar system [1]. Coupling can be measured from
the transmit antenna to the receive antenna by measuring the power at the receive antenna

that is picked up by the transmitter as shown in figure 3.5.

27

'iiiiii33?;3i13i??iii331iiiTrians'mit?iiii
.IIOSC1WIIhIIZZIiIIZi'Z‘I’Ant'e'n'n‘a’Illil

   
   
  

'::aKnown::‘-""'

.::output:::

:::power:::. ::::::::::..

‘1:::1:::::‘ .iummedx.

'fArEItennaxr.
Coupling

" :::::::::::::::.' fiReceivéjtifi
':::0utputto:t"':" iiA'nfe’r-ma'iiiii
.iiRFiPoweri. ......:::::t:i:iiii:
"'iMeteriiiiiiiiii"'

Figure 3. 5: Antenna coupling limits the sensitivity and dynamic range of FMC W

radar systems.

The problem of antenna coupling can be uniquely solved in FMCW radar systems. The
coupling produces a beat frequency in the video chain of the radar system. This beat
frequency corresponds to the delay between the transmitter to transmit antenna, through
the coupling path, and through the receive antenna back to the receiver. This delay and
its corresponding beat frequency can be nullified by inserting a delay line between the
transmitter and transmit antenna, or the receiver and receive antenna. An example of this
is shown in figure 3.1. This method nullifies the beat fi'equency seen in the video signal
chain c aused b y the antenna c oupling. H owever, this In ethod c an n ot s olve p roblems

such as saturation in the receiver front end due to too much power coupled from the

28

transmitter at the receive antenna. Problems such as these are best solved with better

antenna arrangements, and decoupling techniques.

Delay lines have the potential to cause problems. Modern radio altimeters and other
FMCW radar systems operate at high microwave frequencies, such as the Bendix ALA-
51A. Delay lines are typically waveguides or microwave coax cables. When a delay line
is inserted before the LNA in a receiver, it introduces losses which add directly to the
noise floor of the receiver. This increased noise floor causes a further reduction in
sensitivity and dynamic range. Careful consideration must be used in the placement of

delay lines.

29

Summary:

FMCW radar systems have been in use since the mid 1930’s. Modern FMCW radar
systems are much different, and also very similar in many ways to the early models.
FMCW radar continues to be primarily used in the application of radio altimeters. But
with modern data acquisition and signal processing methods, F MCW radar systems show

potential for other applications.

30

Chapter 4

Theory of Operation

A unique FMCW radar system was designed by the author utilizing inexpensive, readily
available microwave transceiver modules. This system is based around a pair of WA-
Com MA87127-1 X-band microwave transceiver modules. These modules are
configured in an unusual way such that a deliberate coupling is created between the
transmit and receive horn antennas. This deliberate coupling facilitates a unique

approach to FMCW radar.

31

4.1 The MA87127-1 Gunn Diode Based Transceiver Module
The MA87127-1 transceiver module is an inexpensive X-band microwave transceiver

solution. It is small and compact, as shown in figure 4.1.

saris/sans

 

Figure 4.1: The MA8 712 7-1 microwave transceiver module.

The MA87127-1 is composed of three major components, VCO, mixer, and circulator as
shown in figure 4.2. The VCO is fed into port 1 of the circulator. Port 2 of the circulator
is connected to the WR-90 waveguide flange input/output port of the transceiver. Port 3
of the circulator is connected to the RF input of the mixer. Some power is coupled off
the VCO and fed into the LO port of the mixer. The IF output of the mixer is connected

to a small solder terminal on the outer case of the transceiver.

32

 

 

 

. . ...IFW .............................

Figure 4. 2: MA8712 7-] block diagram.

VCOl is a varactor controlled Gunn diode oscillator. A varactor diode is placed inside of
a cavity Gunn oscillator as shown in figure 4.3. A bias voltage on the varactor diode
between, roughly, 0 and 20 V controls the frequency of the Gunn oscillator. A second
bias voltage of approximately 10 V is needed to cause the Gunn diode to oscillate at the

frequency of the cavity that it is placed in.

33

.........................................................

2:33:33223323:2::sizes...""....zzzzzzz:3::.¢....:.3c..¢u1a..332333
::::::::::::::sycamzxzxzz::::::M‘agnet::::::::::z::z
::Gunn:Diode;:::::

 

 

 

 

Figure 4.3: MA8712 7-] Physical Layout.

CPLRl is a symbolic representation of the coupling action that occurs between the Gunn
oscillator diode and the Schottky mixer diode placed within close proximity (see figure
4.3). The close proximity of the Schottky mixer diode to the Gunn diode causes a

coupling to occur.

MXRl is created by the coupled power from CPLRl. The coupled power from the

Gunn diode oscillator causes the Schottky mixer diode to switch on and off. This

switching action causes the Schottky mixer diode to operate as a single balanced mixer.

34

CIRCl is a ferrite circulator place inside of the resonant waveguide cavity that contains
VCOl and MXRl. CIRCl is basically a large magnet precisely placed inside of the
resonant cavity. CIRCl causes RF power from VCOl to exit the input/output port, and

causes RF power coming into the input/output port to be transferred into MXRl.

When looking at figure 4 .2, it appears as though one transceiver module alone can be
utilized as an F MCW radar system. However, it was found in lab tests that the pass band
of the IF port on MXRl starts to roll off around 1 MHz, causing little to no response at
audio frequency, which is were most beats from a short range FMCW radar system will
be located. The transceiver module’s receiver worked most efficiently above 30 MHz,
where the loss due to the mixer was found to be the least. The lack of an acceptable low
frequency to DC response from MXRl renders one individual transceiver module useless
for short range FMCW radar use. This problem is common for most microwave

transceiver modules of this type.

Another problem often encountered with this type of transceiver module is frequency
drift. According to the specifications, the MA87127-l will drift at a rate of 350 KHz per
1 d egree C elsius. It w as also found in lab tests that the transceiver m odule drifts o ff

frequency when it encounters sever VSWR reflections on its input/output port.

The tuning linearity of the varactor diode controlled Gunn oscillator is not very good.

For practical uses, a linear tuning range must be found by experimentation after the

transceiver module is installed into whatever system it might be used in.

35

4.2 A Unique FMCW Radar Solution
The MA87127-1 X-band microwave transceiver module is not capable of accurate
FMCW radar operation by itself. However, when two MA87127-1 (or similar)

transceiver modules are used, the unique FMCW radar design solution can be obtained.

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4. 4: Simplified block diagram of the unique FM C W radar solution.

36

A simplified block diagram of the unique FMCW radar solution is shown in figure 4.4.

XCVRl is centered at frequency f, and FM modulated with a linear chirp m(t), which

causes a frequency deviation of ¢(t). The frequency deviation ¢(t) can be found using

the following equations [15]:

d¢
E’ = kfm(t) (4.1)

Where: k f = the frequency deviation constant, expressed in radians per second per unit
of m(t)
m(t) = the m odulation w aveforrn which is frequency modulating X CVRl (in

this case, it is a linear ramp)

The frequency deviation constant, k f , can be found using the equation:

k f = 271 d (4.2)
Where: f d = the frequency deviation constant of the modulator, and is expressed in

hertz per unit of m(t)

The frequency deviation ¢(t) can be expressed using the equation:

(15(1) = k f [m(a)da + ¢o (4.3)

Where: I: time
t0 = initial time

¢O 2 phase deviation at time t = to

37

Finally, the output of XCVRl is represented with the equation:

TX, (t) = A, cos[279",t + ¢(t)] (4.4)

The output of XCVRl is fed into the transmit antenna. The transmitted signal is reflected
off of the target. The target is situated at a range R and moving at a velocity v (if it is
moving). The range R and velocity v correspond to a time difference and Doppler shift
between the original transmit signal and that which was picked up by the receive antenna
and fed into XCVR2. This time difference corresponds to a beat frequency difference f,
as was proven in section 2.2. Thus, the reflected signal from the target is represented by
the equation:

TX”, (I) = A, cos[2zgf,t + ¢(:)+ 2473:] (4.5)

XCVR2 is set to a fixed frequency of f2. XCVR2 is radiating a fixed fi'equency carrier
at that frequency which can be represented by the equation:

TX, (z) = A, cos[279’2t] (4.6)

As explained in section 4.1, the IF output of each transceiver module is a product of its
VCO frequency and any RF power that is coming into the input/output port of the

module. Because of this, the IF output of XCVR2 can be calculated:
”720): TX1b(t)TX2(t)

= A3 cos[27gf,t + ¢(t)+ 271fbt]cos[27f2t]

38

A22 cos[27§flt + ¢(t)+ 279‘}; + Zryfztl-r

”:3 cos[27;f,t + ¢(t)+ 271,: — 271,1]

 

(4.7)
+

 

The higher frequency term can be dropped. This is a practical consideration since the IF
output port of the transceiver modules is not capable of producing X-band microwave

signals. Thus, the IF output of XCVR2 can be simplified as:

A: cos[27;f,t + ¢(t)+ 279‘}; — 2799] (4.8)

11:20): 2

 

Simultaneously, some power from XCVR2 is coupled into XCVRl. Power from XCVR2
is coupled out using CPLR2 and output through ATTl, CIRCl, and into CPLRl. The
coupled power injected into CPLRl is fed into XCVRl. The resulting frequency

response at the IF port of XCVRl is calculated using the equation:
11710): TX2(’VX1(’)
= A: cos[2zy’2t]cos[279fit + ¢(t)]

A22 cos[271f2t+ 27rflt + ¢(t)]+

 

(4.9)

A; coslzzrzr — hit-9150)]

+

 

Like XCVR2, the higher frequency term can be dropped. Thus, the IF output of XCVRl

can be simplified as:

”:3 cos[279"2t — 27m: — ¢(t)] (4.10)

 

1F.(t)=

39

IF, (I) is fed into the input port of a limiting amplifier, LIMl. The output of MM] is
used as the LO drive of MXRl. IF2(t) is fed into the input port of an LNA, which is
represented by LNA]. The output of LNAl is fed into the RF input port of MXRl.
IF, (t) and IF2 (t) are multiplied together in MXRl. The IF output of MXRl is amplified
by a video amplifier. The resulting product from MXRl can be represented by the
equation:

Video Output = IF, (t)le (t)

 

if c.4242. _ 24,: —¢(t)1008124f1t + .(.)+ 274% - 249’]

[if cos[2rg’.z —2o’,t -¢(t)+279’.t +¢(‘)+2’?fb’ ‘ ”734+

 

+

 

”if cos[27;f,t + ¢(t)+ 271,: — 2793: - ¢(t)+ 271,; — 2293:]

4
[if cos[2nf,,t]+

 

 

”:3 cos[— 471,: — ¢(:)+ 2473:] (4.11)

The IF port of MXRl is not capable of reproducing the high fi'equency term from

equation 4.8. Therefore the video output of the radar system can be expressed as:

A“ cos[2zgj,t] (4.12)

Video Output = 4

 

4O

It is clear from the equation above, that the video output is the beat frequency difference

f , due to distance from target R and velocity of target v. Thus, we have an FMCW radar

system using two inexpensive microwave transceiver modules.

It should be noted that when the FM modulation k = 0 , the system will function as a CW

Doppler radar. Thus causing the video output equation to become:

A4
4

 

Video Output = cos[27rfdt] (4.13)

Where f , becomes f d when there is no chirp, from equations 2.6 and 2.7.

A unique and low cost FMCW radar system has been proven. This system is
theoretically capable of both FMCW and CW Doppler operation. Unwanted antenna
coupling described in chapter 3 is used to the advantage of this radar system. The system
deliberately forces antenna coupling for use as a reference LO for MXRl. One further
advantage to this system is that the delay lines required to cancel any further unwanted
coupling can be installed in series with IF] and IF2, allowing for the use of low

frequency and low cost delay lines.

41

4.3 System Design and Implementation

A unique FMCW radar solution was presented in section 4.2. The actual implementation
of this system will be explored in this section. The system is composed of five major
components. These components include the front end assembly, power supply, IF
chassis, data acquisition and trigger, and the software. The entire system block diagram

is shown in figure 4.5. A picture of the system can be seen in figure 4.6.

 

 

   
 
   
   

  

 

     
  
  

 

 

 

    

Ramp

Input I
|
|
I

Misc. ______ _L__.__._1 _______ L ______
Power I Mode 1 i. Transmit ‘1'
Lines l Doppler Selector I I .X£\«31_ Antenna
1 Freq. , 1
POWER | Tune , i
SUPPLY I I
Modulation
I Amp 1
I
l
|

l
I
I
I
Antenna i
I
I
I
I

IIF CHASSIS
Recewe
I
1 Video
I Amp 1F
: Video a _______ F 301399359123.

OUtpUt MXRI AMP3 LNAI JI

____—__——___..

7 I
| | IFFI'
FFT v1
III-6014 | 1 Window H H :
|

l
_________________ I L_____.___§J_' "LA REA

 

 

 

 

 

 

 

 

Figure 4. 5: System block diagram.

42

 

Figure 4. 6: FMCW radar system.

43

4.3.1 The Front End Assembly

The unique solution to FMCW radar has a very unusual front end. The front end
assembly creates the chirp, transmits it, and receives the reflected signal. The front end
assembly is also where the forced coupling between XCVRl and XCVR2 occurs. A
block diagram of the front end assembly is shown in figure 4.7, and apicture of it is

shown in figure 4.8.

 

 

 

 

 

 

 

 

Figure 4. 7: Front end assembly block diagram.

 

 

TRANS2

XCVR2
Al 1N1

 

Receive - =—.-
Antenna ' ' '

 

CLPRl

 

Transmit
Antenna

 

 

 

XCVRl

TRANS 1

 

Figure 4. 8: The front end assembly.

The Front end assembly creates a transmit chirp using an X-band microwave transceiver
module, XCVRl. The transmit chirp is created when the varactor diode controlled Gunn
oscillator of XCVRl is modulated with a linear ramp which is fed in from the IF chassis
to VCO CTRLl. The input/output port of XCVRl is connected to the directional coupler
CLPRl. CPLRl is then fed into the transmit antenna. The chirp signal is transmitted out
of the transmit antenna, bounced off of the target, and is reflected back into the receive
antenna. The receive antenna is fed into the directional coupler CPLR2. CPLR2 is then
fed into the input/output port of an X—band microwave transceiver module, XCVR2.
XCVR2 is set to a fixed frequency through biasing VCO CTRL2, and converts the entire
chirp bandwidth down to an intermediate frequency. That intermediate frequency is then

outputted to the IF chassis through IF2.

45

The forc
frequean
attenuate
waveguic
circulato:
transitior
couples 5

an intern

The Gun
from Gu

oscillator

XCVR]
microwa
Were des

The forced coupling occurs between CPLRl and CPLR2. XCVR2 is set to a fixed
frequency. Some power from XCVR2 is coupled out to the adjustable microwave
attenuator ATTl through CPLR2. That power is then fed out of ATTl and into the
waveguide to coax transition TRANSZ. The coaxial port of TRANSZ is fed into a
circulator, CIRC 1. The output of CIRCl is then fed into a second waveguide to coax
transition, TRANSl. The waveguide end of TRANS] is fed into CPLRl. CPLRl then
couples some of the power directly into XCVRl. That power is then converted down to

an intermediate frequency IF 1 , and outputted to the IF chassis through IFl Out.

The Gunn oscillators of both XCVRl and XCVR2 are powered with a DC bias voltage
from Gunn Oscillator Bias Input. Both transceivers are biased such that their Gunn

oscillators are in oscillation.

XCVRl and XCVR2: Both XCVRl and XCVR2 are M/A-Com MA87127-1 X-band
microwave transceiver modules. The operational details of these transceiver modules
were described in chapter 4.1. The connections made to XCVRl and XCVR2 are shown

in figure 4.9.

46

  
   
    
  
   
   
  

 

 

 

 

 

 

E ....................................
1...;i:;:.:...‘A +10V:Bias:::; ““““““
Biasinputand; .....
VCQCTRLanut -------
99999810.".195 7"

.............. g [RF inputioutpm .1
‘ 'i 3 (WOWWW‘IQ
""" . [F'?F‘99::;x;s:
AIF.‘I§3'.§
'OutputA .............
iiiiiffiiBNdFé'ro'é-Ie...
..:.;. Bendix 10-6Male. i3 MIDICII’CUIIS
,1, _ ,,; """" : .31T24 """""""""
Biaslnput and : . . . miM‘ ........
.:vco CTRL2 input : .........
CO“"°¢T°'2.,5:D' i ......... _ '
,Iamnpuooutpmi
.5. 313' ‘ ,‘WRQOWaveguide
 . is: ..... ,f'eoseézsea
.IF2;;;:g ..........
...... Output“;
‘BNcFé‘rfia‘Ie » "If?
.................... MW'CIWWS ' i; Q j;
I T2512iifT"z..,'ji.iij‘iji

Figure 4. 9: Schematic of XC VRI and XCVR2 interconnections.

Looking at figure 4.9, XCVRl is biased with +10 V from the power supply which is fed
into the XCVR module through pin A of Bias and VCO CTRLl Connector 1. This
connector is a Bendix model 10-6 connector. The bias voltage is filtered with a 47uF

cap, and then tied to the Gunn diode bias terminal on XCVRl. The bias ground is fed

47

from the power supply to XCVRl through pin B of the 10-6 connector. The ground wire
is then tied to the chassis of XCVRl. The VCO control voltage, VCO CTLRl, originates
from the IF chassis. VCO CTRLl is fed in through pins D and C of the 10-6 connector.
Pin D connects to the center conductor of a RG174 coax cable. Pin C connects to the
shield of that cable. At the end of the length of RGl74, the center conductor is tied to the
varactor diode terminal on XCVRl. The shield of the coax is tied to the chassis of
XCVRl. The IF output of XCVR], IFl, is fed out of the module through a solder pin.
That pin is tied to a 1 K Ohm resistor, which is tied to ground. The 1 K Ohm resistor
protects the IF output of XCVRl. The IF output pin is then connected to the primary
winding of a 2:1 impedance matching transformer. The transformer is a Mini-Circuits
model T2-l. The secondary winding is then connected to a BNC female connector. That

BNC connector is the IF] output port.

Much the same as XCVRl, XCVR2 is biased with +10 V from the power supply which is
fed into the XCVR module through pin A of Bias and VCO CTRL2 Connector 2. This
connector is a Bendix model 10-6 connector. The bias voltage is filtered with a 47uF
cap, and then tied to the Gunn diode bias terminal on XCVR2. The bias ground is fed
from the power supply to XCVR2 through pin B of the 10-6 connector. The ground wire
is then tied to the chassis of XCVR2. The VCO control voltage, VCO CTLR2, originates
from the IF chassis. VCO CTRL2 is fed in through pins D and C of the 10-6 connector.
Pin D connects to the center conductor of a RG174 coax cable. Pin C connects to the
shield of that cable. At the end of the length of RG1 74, the center conductor is tied to the

varactor diode terminal on XCVR2. The shield of the coax is tied to the chassis of

48

XCVR2. The IF output of XCVR2, IF2, is fed out of the module through a solder pin.
That pin is tied to a 1 K Ohm resistor, which is tied to ground. The 1 K Ohm resistor
protects the IF output of XCVR2. The IF output pin is then connected to the primary
winding of a 2:1 impedance matching transformer. The transformer is a Mini-Circuits
model T2-1. The secondary winding is then connected to a BNC female connector. That

BNC connector is the IF 2 output port.

CLPRl and CLPR2: CLPRl and CLPR2 are both X—band directional couplers of
unknown part numbers and an unknown origin (see figure 4.8). They were purchased as
government surplus. When tested, both couplers had a coupling factor of -10 dB. Both

couplers use WR90 waveguide flange connections for all ports.

TRANS] and TRANSZ: TRANSl and TRANSZ are both WR90 to type N female
connector transitions (see figure 4.8). These transitions were tested, and found to be
functional over the entire operational bandwidth of the radar system. The transitions are

government surplus, and the part numbers and origin of these transitions is unknown.

‘ Transmit Antenna and Receive Antenna: The transmit and receive antennas are both
M/A-Com model number MA86551 (see figure 4.10). Both are horn antennas with a
WR90 waveguide feed. The operational frequency range of these antennas is 8 GHz to
12.4 GHz. The H-plane and E-plane 3 dB beamwidths are specified to be 25 degrees.

The nominal gain of these antennas is 17 dB.

49

Receive Antenna

Transmit Antenna .

 

Figure 4.10: Transmit and Receive Antennas.

ATTNl: ATTNl is a waveguide attenuator. Both the input and output ports of this
attenuator are WR90 waveguide flange. The attenuator is adjusted with a manual knob,
which can be seen in figure 4.8. The part number and origin of this attenuator is

unknown since it is a government surplus component.

CIRCl: CIRCl is a ferrite circulator set up as an isolator. CIRCl is a coaxial part,
where both the input and output ports are SMA female connectors (see figure 4.8). The
isolation of CIRCl was measured to be 10 dB. The part number and origin of this

circulator is unknown.

50

4.3.2 The Power Supply

The power supply is a crucial part of the system. The power supply provides a number of
voltages to power the various components of the system. These voltages include +10 V
for the transceiver modules, +12 V for the RF amplifiers, +24 V for the modulation
amplifier, and +/- 5V for the video amplifier. The power supply is located in the IF

chassis (see figure 4.11). Figure 4.12 shows the power supply schematic.

Power
A supply

Components

    

 

 

IF Signal
Chain

 

 

 

02/1 'i/ZGI

Figure 4.11: IF Chassis.

51

 

 

   
   
  
   
  

j ,Linelnp-Powor-Onelnternationai . ' j ,‘I,";j . +5 ‘H+5'V

: PW" . * ‘ _ModeI#HB_24-1.2-Aj - - . _ . M0333. . .
GNDF Output +24V12A - .. AA
' ‘ 3 - - 024L-osoo com

 

 

 

- LineIn : I

 

 

 

 

- -sense Com TZAIVI-sensq Converter
GND ‘ ' I ‘
4,2. . . 1 . . .

 

 

Bendix 10-6
Female Connector

ooooooooooooooo

PowerCord—I—C) WOW“

 

x I‘Arm osa ' i
‘ ' ' F‘ .78L05‘Qm‘ o : +1‘2v

 

 

 

 

 

 

i
. 5 land VCO » 01
1 SO, .'§CTRL1‘ '...,L.GN F

 

................

 

 

c>———~——~<>

 

 

 

+10,V In ' : """"""" ‘ T
. l
. gem Quip“. 3""
to §VCO CTRL2 01m
_ ~ goonnactorz . Z .
Bo'ndix 1036 F0018“

Conn-star. ‘ ' ' '

Figure 4.12: Power supply schematic.

120 VAC line voltage is fed in through a standard 3 prong power cord. The ground
terminal of the cord is tied directly to chassis ground. The hot wire of the power cord is
tired directly to a fuse as soon as it enters the IF chassis. The fuse is then tied to a main
power switch. The other terminal of the power switch is tied to one of the line in
terminals of the 24 VDC linear power supply. The other line in terminal on the power
supply is tied directly to the neutral wire of the power cord. The linear power supply
output is 24 VDC at 1.2 A. The output common terminal of the power supply is tied
directly to ground. The +24 VDC output of the power supply is tied directly to a 1 A
firse. The other end of the fuse is fed to a power on indicator lamp, and then distributed
throughout the system to anything requiring 24 VDC. A 12 V linear regulator is tied to

the 24 VDC rail of the power supply. The regulator provides 12 V to anything that

52

requires it. A single transistor 10 V regulator is connected to the 12 V regulator output.
The 10 V regulator provides bias voltage for XCVR] and XCVR2. A DC to DC
converter is also connected to the 12 V rail. The DC to DC converter creates +5 V and -5
V for the video amplifier. Careful attention is paid to laying out the line voltage wires in

close proximity to the DC lines. The physical layout of the power supply is shown in

 

 

 

 

 

 

 

 

 

 

figure 4.13.
10 V Regulator | l 12 V Regulator
\
‘ ' 24 V
D 9. .
BC to rt C 4 .a, g Lrnear
onve er - , Power
I 111;. 4,, ' , Supply
,3, ‘1
,
Power .1
Switch ., *

 

 

 

Figure 14.13: Physical layout of the power supply.

Linear Power Supply: The linear power supply is a Power-One model number HB24-
l.2-A international series. This power supply is capable of supplying 24 VDC at 1.2 A of
current. It also has remote sensing capabilities, which are simply bypassed in this

application by tying —sense to the com, and +sense directly to the +24 V output terminal.

53

12 V Regulator: The 12 V regulator is a general purpose linear chip, the 78L12. This

particular 78L12 uses a large TO-204 case for heat dissipation.

DC to DC Converter: The DC to DC converter is an Astec model number AAlOU-024-
050. This DC to DC converter is capable of dropping up to 36 VDC down to +5 VDC
and -5 VDC. It is used exclusively to power the video amplifier. In this particular
application, the video amplifier does not draw enough current to turn on the DC to DC
converter, so load resistors were tied directly between the outputs and ground to provide

enough current drain to turn on the converter in a stable operational mode.

10 V Transistor Regulator: The 10 VDC transistor regulator is a simple zener diode
and NPN transistor regulator. This particular regulator uses a large 2N3055 NPN
transistor. This particular transistor also uses a TO-204 case, similar to the 12 VDC

regulator, for heat dissipation.

54

4.3.3 IF Chassis

The IF chassis is a critical part of the system. The IF chassis serves two purposes. It
multiplies the IF from XCVRl and XCVR2, producing a video output that consists only
of target beat fi'equencies and Doppler shifts. It also amplifies and offsets the ramp
modulation input to appropriate levels for the varactor diode VCO in XCVRl. An inside
view of the IF chassis is shown in figure 4.11. The block diagram of the IF chassis is

shown in figure 4.14.

55

- Spa... 5
: Supply: . .
' 'InpUts . 1 i '

    

: ‘ "\ch cm.
. _ [Voltage Output

.....

 

IARar'n'p 4'
' , Input '/

 

' ' I ' Modulation

 

  

 

‘ :IF3IE Input

 

AmP2 ' ' i i AMP1

 

l
'11-;VCOCTRL2'?
1 /Voltage Output

 

 

—_-*———e_———_'_——'———'_-fi

Video. L i T

. Output : TIF’21nput .

_ —.* —. — _ — — — ~ ‘ — — ..u *

Figure 4.14: Block diagram of IF chassis.

The IF chassis physical layout is shown in figure 4.15.

56

   

 

Video Amp ' 1. 1 ,
Video Output A . V

Figure 4.15: IF chassis physical layout.

  

Inputs

 

 

 

 

When multiplying the IF from XCVRl and XCVR2, the IF chassis takes in IFl from
XCVRl and amplifies it to a mixer LO drive level of +7 dBm, and fed into the double
balanced mixer, MXRl. This is achieved using two general purpose RF amplifiers,
AMPl and AMP2. IF2 takes a slightly different rout when it is fed into the IF chassis.
IF2 is fed directly into a low noise amplifier, LNA]. The output of LNAl is then fed into
a general purpose RF amplifier, AMP3. The output of AMP3 is fed into the RF port of
MXRl. When the L0 and RF are mixed in MXRl, the resulting IF is then amplified
with a video op amp, the Video Amp. The output of the Video Amp is fed out of the IF
chassis through the Video Out connection. The Video Out is the range to target and

Doppler shift information in the form of beat frequencies.

57

Two operational modes can be chosen using the Mode Selector. The Mode Selector is a
two way switch. It can connect the VCO CTRL] output port on the IF chassis to either
the Modulation Amp or a tuning potentiometer. When the Mode Selector is in the CW
Doppler position, the VCO CTRL] port is connected to a tuning potentiometer. The
tuning potentiometer is capable of tuning XCVR] across its entire tuning range. When
the Mode Selector is in the FMCW position, the VCO CTRLl port is connected to the
output of the Modulation Amp. The Modulation Amp is a simple op-arnp which
amplifies the modulation from an external ramp generator, and DC offsets that
modulation waveform. The DC offset feature is adjustable, and allows the FM
modulation to be centered around any desired frequency within the tuning range of
XCVR]. The output port VCO CTRL2 on the IF chassis is tied to ground, since XCVR2

is left un-modulated.

AMP]: AMP] is a general purpose surplus RF amplifier. It has a gain of approximately
20 dB between the frequency ranges of 10 MHz and 200 MHz. It requires a +12 VDC
connection to the p ower s upply. T he R F input and o utput c onnections are b oth B NC
female connectors. Other specifications on this amplifier are unknown since it is a

surplus item of an unknown origin.

AMP2: AMP2 is a pair of Mini-Circuits MAV-2 MMIC amplifiers in series on a strip of
50 Ohm microstrip. The gain of these two amplifiers in series is approximately 25 dB
from 100 KHz to above 1 GHz. The maximum 1P] output of this amplifier is +13 dBm

at -12 dBm input. However, the IP] is adjustable. The IP] is adjusted using the

58

potentiometer connected to the LM317 voltage regulator. The LM317 allows for
adjustment of the bias on the second MAV-2 amplifier. By lowering the bias on the
MAV-2, the IP] can be lowered. A low IP1 causes AMP3 to work as a limiting
amplifier. This limiting amplifier drives the LO of MXR]. The RF input and output

connections are both BNC female connectors. The schematic of this amplifier is shown

 

 

 

 

 

 

  

 

 

 

 

 

   

 

 

 

in figure 4.16.
LM317
. C) 3 ' T - I" ; ; our ' . 2
+12 V Input, 1K 4. 7 HF . l , . . . 11 UP
. ' . T . . , . '. .240.
FTP—“'1: 2N2222A ' . ‘
T 10 V , . . . . ’ 5'.K
0.01 uF . , .0 01
. I . I I I I 0—1 I uF
Feedthru Cap 9 g 'Feedthru cap? C. ‘ '
_ ' , _ . . . 1 . I.
BNC FemIale » 3 - - . BNC Female
E ............. . _ 0 01 ...............
g Cantorf. I3}, Lfig E, Cgantar
i...§r.‘.°.'.'.‘i..i ,UsmP MAv-2 ' “3‘”? ' WW 2 . ”5"”? 5“”

Figure 4.16: AMP2 schematic.

MXR]: MXR] is a double balanced coaxial mixer. This mixer has an HP part number
of HP]0514A. This mixer requires an LO drive level of +7 dBm. It has an insertion loss
of approximately 6 dB, and an input IP1 of +1 dBm. All of the ports, RF, LO, and IF are

BNC female connectors.

59

LNA]: LNA l is a surplus RF low noise amplifier of an unknown origin. It has a gain
of 25 dB across a frequency range of roughly 10 MHz to 250 MHz. Its IP1 output power
is +10 dBm with -15 dBm of input power. Both the input and output ports are SMA

female connectors.

AMP3: AMP3 is nearly identical to AMP2. AMP3 is a pair of Mini-Circuits MAV-2
MMIC amplifiers in series on a strip of 50 Ohm microstrip. The gain of these two
amplifiers in series is approximately 25 dB from 100 KHz to above 1 GHz. The IP1
output of this amplifier is +13 dBm at -12 dBm input. The RF input connection is an
SMA female c onnector. T he RF 0 utput c onnection is a BNC female c onnector. T he

schematic of this amplifier is shown in figure 4.17.

 

+12 V Input G

      
   

1K 4.7uF - . 1K ~ 4.7uF

 

." 1.0V ,

  

 

 
 
       

0.01.uF 0.01 uF’
. 31‘41—1 . . 1‘1
Feedthru Ca? g Feedthru Ca} (1
SMA Female ' ' . ‘ ' ' ' ‘ ' ' BNC Female
. 0.01 . . . ‘ _ """"""""
. ._ uF 5° . 5,0»th
». 0.01 , ’ 1 [>3
3 :' g F 1 50.0hm F 50; 50 Ohm
599999....3 ”Strip ' MAv-z. 'u "3MP Msz u USMP £999.99“ .....

Figure 4.1 7: AMP3 schematic.

60

Video Amp: T he Video Amp is the obsolete M axim semiconductor M AX404. T his
video 0p-amp has a maximum bandwidth of 80 MHz, and is built in an 8 pin DIP
package. In the Video Amp circuit, the MAX404 is set up for an AC gain of 8. The
circuit is setup to take a source impedance of 50 Ohms, which is terminated by a 50 Ohm
resistor on the MAX404 input terminals. The output impedance of this circuit is 50
Ohms set by a resistor in series with the output terminal of the MAX404. A schematic of

the Video Amp is shown in figure 4.18.

61

t:t+;5voc:mpm.o """ ;: :p'
dd; :‘FeedthrucapP (i
p::::i;é'i;:;:i'_::__L . """""

-------

: seminar~ ‘ i ' i : :1: i : ; ' ; : :BNC‘Fema'é

 

 

+/

Sheild L :

OCA--.-.- q.-

‘-

 

--.cnpoounogoupoyn

..... ”.‘...~--

 

 

 

" 'Feédthmca'pp g f""'.'.i‘

T ‘I'

 

 

? #5 vocmpu‘tO ‘ ‘ 9 ‘

Figure 4.18: Video Amplifier Schematic.

Modulation Amp: The Modulation Amp shares a circuit board with the 10 V regulator
and the DC to D C converter c omponents o f the p ower s upply (see figure 4 .13). T he
Modulation Amp is based around an LM324 single supply quad op-amp. Only one op-

amp out of the 4 available on the chip is used. The LM324 was chosen because of its

62

 

reliable single ended performance, and ability to swing over 20 V from its output. The
schematic of the Offset Amplifier and the Modulation amplifier is shown in figure 4.19.
The LM324 is powered with a single supply of 24 VDC. The modulation input is
brought in through the Ramp Input port on the IF chassis. The Ramp Input port is
terminated with a 2.2 K Ohm resistor. The Ramp Input is then AC coupled through a 1
uF c apacitor into the LM324 op-arnp. T he relatively high 1 uF c apacitor w as chosen
because the design requires the use of a low frequency modulation input. At the input of
the op-amp, a DC offset is introduced. This DC offset is adjustable with a 10 K Ohm
trimmer pot. The op-amp is setup for a voltage gain of approximately 11. The output of
the op-amp is fed into the FMCW mode side of the Mode Selector switch. The Mode
Selector switch is a 2 way switch, capable of transferring the VCO CTRLl Output port
between the op-amp and a DC tuning potentiometer. The DC tuning potentiometer is
setup as a voltage divider, where the input is regulated to a voltage output of 20 VDC. It
is capable of outputting anywhere between 0 VDC and 20 VDC. The output of this

circuit is fed into the CW Doppler mode position of the Mode Selector switch.

63

+24 VDC input Q-<*—i ‘

 

 

 

I +5 VDCInput DC “ A 2‘2'UF“
Wm - ‘r—ll—j
' .' , , fl1'OK , 0.1'uF "'
streamers $45K 9 ' i _ - ., . - .1
:"“f"f'f""- 1 uF :>' ’ . , . Mode VCO CTRL1 Output
' C)—

 

 

§snoud §

    
 

 

 

 

..‘MuF .Behdthyp‘et
. ~ ~ FemaleConnector» ,
. L .VC'O CTRL2 Output Connector
. ' ; '. o
' E .C "i
Bendix Type 10-6

Female Connector

Figure 4.19: Modulation amplifier schematic.

64

4.3.4 Data Acquisition

The data acquisition system acquires the Video Output from the IF Chassis, and triggers
the Ramp Generator (see figure 4.20). Data acquisition is achieved by use of a National
Instruments NT60l4 PCI card installed in a PC. Analog input channel 0 from the NI6014
is connected directly to the Video Output fi'om the IF Chassis (see figure 4.21). Digital
10 channel 0 from the NI6014 is connected directly to the trigger input of an HP3312A
function generator set to produce a 100 Hz linear ramp. The analog input being used on

the NI6014 is capable of 200 K samples per second, at a resolution of 16 bits.

 

 

 

 

 

 

 

 

 

 

 

Rarnp Video
Output Input
\/
I— — — — — -— — —- ——————————————— -—-.
I I
------------- "1 I

: Ramp : . . ' ' I I
I Generator . .9 Digitizer >__I . I I .
I ””312" I ' Nl°6014 ' I Output to
I . ' I ~ ’ w ' I PC and
I I . ' I | software
I I Trigger I I 5
i DATA ACQUISITION ————————— _ * _ _ '7' I

Figure 4.20: Block diagram of the Data Acquisition signal chain

65

 

I Center T1 1 z .

Vidw'm’m O E g \ I I” ‘ Pin 68, ACHO Input
from IF ' : K" J J ) .
Chassis . , I __-..--._~- - __-

C>—'——J Pin 62. AISENSE
SheIIdE . ‘

Pin Conne‘ctiOns to‘ the
BNC Female NI6014 IIO Connector

Center
Tngger ' '

Outgut to 0
HP 312A ‘ ' '

' Sheila

..c-- ......................

BNC Female

Pin 52. DIOO

 

 

Pin 50. DGND

Figure 4. 2]: Schematic of N160] 4 input and output connections.

66

4.3.5 Software

The software written for this radar system was coded using National Instruments
LabVIEW Student Edition. LabVIEW is a symbolic programming language created by
National Instruments. LabVIEW code looks much like a schematic, where various pieces
of code are wired together to form a program. LabVIEW interfaces directly with
National Instruments data acquisition products so that complex data acquisition and
signal processing programs can be created using the symbolic programming environment
of LabVIEW. A LabVIEW program that interfaces with instruments outside of the
computer is known as a Virtual Instrument (V I). In the case of this radar system, all VI’s

were written to interface with the National Instruments NI6014 data acquisition PCI card.

Three VI’s were created to run the radar system. A VI was created to continuously
operate and display range information. A second VI was created to perform a single
range profile capture. A third VI was created to operate the radar system as a Synthetic

Aperture Radar.

Continuous Operation and Display: The Continuous Operation and Display VI was
created in order to continuously operate the radar system and display range data in near
real time. This program is used to monitor and adjust the radar system characteristics. It
is also used for aligning standard radar targets. The V1 causes the NI6014 to trigger the
ramp generator and sample the Video Output from the IF Chassis. This sequence of

operations is done repeatedly until the program is stopped.

67

The V1 front panel (see figure 4.22) consists of a sample rate digital control, number of
samples digital control, a scaled sample window control, two waveform graphs, and a
stop button. The sample rate digital control sets the sample rate of the digitizer (up to
200 KSPS) on the NI6014. The number of samples digital control sets the number of
samples collected by the NI6014. The scaled sample window control selects between 7
different scaled sample window types, including Hanning, Hamming, Blackman-Harris,
Exact Blackman, Blackman, Flat Top, Four Term Blackman-Harris, and Seven Term
Blackman Harris. Waveform graph 1 displays the time domain waveform that is fed into
a Fast Fourier Transform (FFT) sub VI. Waveform graph 2 displays the magnitude of the

FFT of the Video Output data. The stop button stops the program.

Wmetorm erh 2
ma;

  

- mu.

 

 

' Tine
Number 0! Senate: Sample Rate FFT Sraroandow Type
71m 9 20mm 3" 0 STOP,

 

 

Figure 4.22: Continuous Operation and Display V7 front panel.

68

The V1 block diagram is shown in figure 4.23. When the V1 is run, the program first
turns off the ramp generator trigger line. It then waits 250 ms and turns on the ramp
generator trigger line. On the rising edge of the ramp trigger line, the ramp generator
starts a linear ramp which modulates the radar system. When the ramp generator is
triggered, the data acquisition case statement is set to true and instructs the digitizer on
the NI6014 to collect a specified number of samples at a specified rate which is selected
by the user. After data is acquired, the data is then put through a selectable scale
window. After the scale window, the data is padded with 9000 0’s. This padded data is
plotted on waveform graph 1 then fed into an FFT sub VI. The absolute value of the real
output of the FFT sub V1 is then plotted on waveform graph 2. After plotting the FFT
data, the program starts over again in a while loop and executes repeatedly until the stop

button is pressed on the front panel of the VI.

69

 

 

 

 

 

 

While
Loop \

 

 

 

 

 

Sequence ‘W

 

 

 

 

 

 

 

 

 

Figure 4. 23: Continuous Operation and Display VI block diagram.

Single Range Profile Capture: The Single Range Profile Capture VI was created in
order to capture and save a single magnitude range profile from the radar system. A
range profile is defined as the raw range to target(s) information from a radar system.
The V1 triggers one ramp from the ramp generator to modulate the radar system and at
the same time it digitizes the Video Output from the IF chassis. This data is then
displayed and stored in a spreadsheet file. The VI only runs one time and then must be

reset.

The V1 front panel (see figure 4.24) consists of a sample rate digital control, number of
samples digital control, a scaled sample window control, and two waveform graphs. The
sample rate digital control sets the sample rate of the digitizer (up to 200 KSPS) on the

NI6014. The number of samples digital control sets the number of samples collected by

70

the NI6014. The scaled sample window control selects between 7 different scaled sample
window types, including Hanning, Hamming, Blackman-Harris, Exact Blackman,
Blackman, Flat Top, Four Term Blackman-Harris, and Seven Term Blackman Harris.
Waveform g raph 1 displays the time d omain w aveform that is fed into a F ast F ourier
Transform (FFT) sub VI. Waveform graph 2 displays the absolute value of the real

output of the FFT of the Video Output data.

 

 

 

qubetdsm 5.“ Hi. FFT smwm Type
groom 9200000 :20

 

 

Figure 4.24: Single Range Profile Capture ”front panel.

The V1 block diagram is shown in figure 4.25. This V1 is very similar to the Continuous
Operation and Display VI. When the V1 is run, the program first turns off the ramp

generator trigger line. It then waits 250 ms and turns on the ramp generator trigger line.

71

On the rising edge of the ramp trigger line, the ramp generator starts a linear ramp which
modulates the radar system. When the ramp generator is triggered, the data acquisition
case statement is set to true and instructs the digitizer on the NI6014 to collect a specified
number of samples at a specified rate which is selected by the user. After data is
acquired, the data is then put through a selectable scale window. After the scale window,
the data is padded with 9000 0’s. This padded data is plotted on waveform graph 1 then
fed into an FFT sub VI. The magnitude of the complex output from the FFT sub V1 is
then plotted on waveform graph 2. After plotting the FFT data, the VI then saves the data
as a spreadsheet file and prompts the user for a file name and location to store it. The V1

then stops execution.

l">'l|l"‘.']|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Case
Sequence
..I
m _- , , _ _ , . - _-j____“_.__‘__fi

 

 

Figure 4. 25: Single Range Profile Capture VI block diagram.

72

Synthetic Aperture Radar: The Synthetic Aperture Radar VI was written in order to
create detailed SAR radar images using the unique FMCW radar solution. Creating a
detailed SAR image would prove the viability of the unique radar solution for use in

many advanced applications.

The SAR VI works with the hardware and the user to create a SAR image. The SAR VI
controls the radar system by triggering the ramp generator, and by digitizing the Video
Output from the IF chassis. The V1 prompts the user to manually move the radar system

down a metal track (see figure 4.26).

 

Figure 4.26: The SAR VI depends on the user to manually move the radar down a

metal track.

The program acquires data at one location on the track at a time. The user specifies how

many “beams” are to be collected. Beams refers to range profiles, the term “beams” was

73

used during the programming of the VI. The user then specifies the spacing between
each beam, and the number of beams per synthetic aperture. The VI prompts the user to
move the radar down the track one increment, and then collects data when the user hits
the ‘Enter’ key. The radar is then chirped, and data is collected. The user is prompted to
do the same again. This process is repeated until the specified number of beams (range

profiles) are collected.

The V1 front panel consists of many controls and output indicators (see figure 4.27).
There are a total of 3 waveform graphs, 6 digital controls, and 3 digital indicators. The
waveform graph at the top of the VI front panel plots all of the real components from
each beam of data provided by the data acquisition loop. The waveform graph just below
that one plots all of the imaginary components from each beam of data provided by the
data acquisition loop. The waveform graph at the very bottom of the VI front panel is
used to plot the magnitude of the output of every single focused range profile calculated
by the SAR process. These waveform graphs are used simply to indicate weather or not
the data acquisition loop has successfully acquired all of the needed range profiles used to
make the SAR process function properly, and to check weather or not the SAR process
has s uccessfully occurred. T he s ample rate digital c ontrol s ets the s ample rate 0 f the
digitizer (up to 200 KSPS) on the NI6014. The number of samples digital control sets the
number of samples collected by the NI6014. The scaled sample window control selects
between 7 different scaled sample window types, including Hanning, Hamming,
Blackman-Hanis, Exact Blackman, Blackman, Flat Top, Four Term Blackman-Harris,

and Seven Term Blackman Harris. The total number of beams digital control sets the

74

total number of range profiles that will be collected by the VI. The spacing digital
control sets the spacing between each range profile. The number of beams per aperture
sets the number of elements in the N element synthetic arrays that will be created by the
SAR process. The current beam count digital indicator shows the user what range profile
number has just been acquired. The length of array digital indicator shows the user the
length dimension of the SAR process output array. The width of array digital indicator

shows the user the width dimension of the SAR process output array.

   
   
 

Heel Cavemen at FFT
15000

 

Number olSurvbs Smut: Rue FFT Sume/nbw TWO lotuNurbelolQeam Spectrum] NuthelolBeum Dam
gram (Imam 1p 11113 52008333 1):»

Imagery [moaned at FF!
umo- '

 

SAR Procen Odom FELI.

 

Figure 4.27: The Synthetic Aperture Radar VI front paneL

75

The SAR VI runs only once. It first acquires the specified number of range profiles. It
then performs a SAR process on the acquired data. It then saves the data before and afier

SAR process.

The data acquisition loop acquires the specified number of ‘beams’ inputted by the user
(see figure 4.28a). The term ‘beams’ refers to range profiles, but happened to be used in
the programming of the SAR VI. When the program begins execution, it sets the ramp
generator trigger to 0, waits 250 ms, then sets the ramp trigger to +5 V. The ramp
generator modulates the radar. At this point in time, the program acquires the specified
number of samples at the specified sample rate. The data is windowed by the specified
FFT window type, padded with 9000 0’s, then run through an FFT sub VI. The output of
the FFT is split into real and imaginary arrays. The user is then prompted to move the
radar and hit the ‘enter’ key to continue. When the enter key is hit, the process starts over
again. The end result is a number of real and imaginary 1D arrays equal to that of the
total number of beams inputted by the user. These 1D arrays are outputted from the data
acquisition FOR loop to create two 2D arrays (real and imaginary) of range profile versus
radar p osition. O ne additional p rocess takes p lace inside 0 f the d ata acquisition F OR
loop. The absolute value of the real data is taken, so that the voltages of the frequency
components can be determined. This result is outputted from the FOR loop to the file
save sequence. This result is also scaled to the logarithmic radar range equation 2.5, and

outputted from the FOR loop to the file save sequence. This data is saved in order to

76

show the data before the SAR process, so that it can be compared to data after the SAR

process.

77

.EeeN aehwmiuea Saw efi “£333 .Ewhwawfi
«83 E :3: 2:55.. seesaw. anew.» «Emma

 

-I 1 III. 0

...
1..
m
w
Q
r
_
I

.2.
1.

 

 

            

$5
E9012~§§
unaiuzeiszzo.

...,..m_

fig

 

 

main;.M.a.u.u.H.n.u.u.M.

 

 

 

 

 

 

 

 

 

78

The Synthetic Aperture Radar process loop uses the data provided by the data acquisition
loop to create a SAR image. The range profile data from the data acquisition loop is
provided as two 2D arrays, real and imaginary data versus radar position. The SAR
process loop (see figure 2.28b) executes the exact procedure described in section 2.3,
using equations 2.15 through 2.20. The result from the SAR process loop is a 2D array of
synthetic focused aperture magnitudes versus radar position. This array is outputted from
the SAR process loop to the file save sequence. The SAR process loop also scales the
resulting SAR magnitudes to the logarithmic radar rage equation 2.5. This result is

outputted from the SAR loop and sent to the file save sequence.

79

.3553.» 9:3. 03 as $5. fiee~ memeehfi 20m. 05 Mam: e5.
$2.5: «83 E 0.3% 2:55.. Basin £va 2:.5

 

 

 

 

 

Eng—:— 92 03mg

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8O

The file save sequence saves the data in a spreadsheet file from before and after the SAR
process loop (see figure 4.28b). The file save sequence saves all of the data sequentially.
First the absolute value of the real data fi'om the data acquisition loop is saved. Next the
data scaled to the radar range equation from the data acquisition loop is saved. Then the
magnitude data from the SAR process loop is saved. Finally, the magnitude data scaled
to the radar range equation from the SAR process loop is saved. The various data saved
from the file 8 ave s equence is imported into M icrosott E xcel, where it is plotted in 3
dimensions. The 3D plots show the radar image, produced before and alter the SAR

process. After all of the files are saved, the VI stops execution.

81

Chapter 5

Experimental Results

Four experiments were conducted using the tmique FMCW radar solution. The tuning
linearity was tested. The L0 injection amplitude feeding MXRl was tested. Two
standard radar targets were designed and fabricated in order to test the radar system. The
radar targets were then used to test the range linearity. And finally, the standard radar

targets were used to test the system as a Synthetic Aperture Radar.

82

5.1 Tuning Linearity Experimental Results

The fiequency of XCVR1 is voltage controlled. Ideally, the relation between voltage and
frequency should be perfectly linear. However, the relationship between voltage and
frequency is not. A plot of XCVR1 tuning voltage versus frequency relative to XCVR2

was plotted in figure 5.1.

 

 

Tuning Linearity of XCVR1

 

100 + Tuning Linearity of I
XQLJ

Frequency (MHz)

 

 

I 0 5 10 15 20
I VCO CTRL1 Voltage (Volts)

 

 

Figure 5.1: X C VR1 voltage versus frequency plot (relative to X C VR2).

In an FMCW radar system, tuning linearity is important. This was shown in equations
2.10 through 2.14. In the case of XCVR1, the most linear tuning range was found to be
between 90 MHz and 160 MHz. This provides 70 MHz of transmit chirp bandwidth for

the FMCW radar system. The corresponding tuning voltage for this frequency range is 5

83

V to l 9 V. T he amplitude o f the ramp generator, and the o ffset p otentiometer in the
Modulation Amplifier were carefully set to accommodate the tuning voltage range of 5 V
to 19 V. From these results, it was determined that the radar will perform the best if

XCVR1 is modulated between the voltage range of 5 V to 19 V.

84

5.2 Local Oscillator Amplitude

The output of AMP2 is the LO drive signal for MXRl. AMP2 is setup as an adjustable
limiter, where the output power limit is adjusted through a bias control potentiometer on
the second stage of AMP2. The amplitude must be powerful enough and flat across the
entire range of 90 MHz to 160 MHz (relative to XCVR2) in order for MXRl to operate
properly. The amplitude flatness depends on the coupling between XCVR1 and XCVR2,
controlled by ATTNl. The amplitude flatness also depends on the bias control

potentiometer inside of AMP2.

An experiment was conducted with ATTNl set to 20 dB of coupling attenuation. The
bias control of AMP2 was set to full bias. A plot of the amplitude on the output of AMP2

was made (see figure 5.2).

85

 

 

A-.- __ _. _.__._L*___”,I

Output of AMP2 With ATI'N1 set to 20 dB I

 

 

 

 

-2 L ATI'N1 setto 20 dB I

 

 

 

Amplitude (dBm)

O

I

I

/

\
O
a
5
9.,
>
Z
'0
N
g
5‘.

 

 

 

 

'19 09 e° <19 .90 .39 \09 .359
Frequency (MHz) Relative to XCVR2

I_w‘_-

 

 

Figure 5. 2: Power output of AMP2 versus frequency (relative to X C ”(2), when

A TTNI is set to 20 dB.

The L0 drive amplitude from the output of AMP2 was not flat over the tuning range of

XCVR1. In order to overcome this problem, ATTNl was set to 0 dB, maximum

coupling. A second plot was made of the output power from AMP2, shown in figure 5.3.

86

H—i —__..*

 

Output Power of AMP3 With 0 dB of Front End Coupling Attenuation

I
I
I
I
I
I
I
CF Output Power of AMP3 I I
I
I
I
I
I

 

With 0 dB of Front End
Coupling Attenuation

 

 

Amplitude (dBm)
_l-.l AA ..L—L—l AA
0103 VG) (DO—‘NOO #010) VCDCD

 

20 70 120 170 }
Frequency (MHz) l

._ .m.._w _* __ *. ___________.M__.J

Figure 5.3: Power output of AMP2 versus frequency (relative to XCVR2), when

ATTNI is set to 0 dB.

When ATTNI is set to 0 dB, the amplitude output of AMP2 is much more flat over the
entire frequency range of XCVR1 (relative to XCVR2). It was decided that for optimum

performance, ATTNl will be set to 0 dB.

A further adjustment was made to the bias control on the second stage of AMP2. The
bias control potentiometer inside of AMP2 was adjusted so that the maximum output
power of AMP2 over the entire frequency range of XCVR1 will be limited to 8 dBm. It
was found that these adjustments caused the LO drive level for MXRl to be the most flat

over the entire frequency range of XCVR1.

87

5.3 Standard Radar Targets
Two standard radar targets were created. Both of these targets are trihedral corner
reflectors, with radar cross sections of 20 stm and 30 stm. The trihedral comer

reflectors consist of three equally sized metal square plates, with dimension L (see figure

5.4).

 

 

 

 

 

Figure 5. 4: T rihedral corner reflector standard radar target.

The maximum radar cross section of a trihedral corner reflector target can be calculated

using the equation [1]:

 

1271‘] (5.1)

ad3(m) = 1010g[ 2.2

Where: 0am“) = maximum radar cross section in stm

L = length of the squares that make up the trihedral

88

From equation 5.1, the 20 stm target was found to have a length of 69.9 mm. The 30
stm target was found to have a length of 393.07 mm. Both targets were fabricated

using 2 mm thick sheet aluminum, and are shown in figure 5.5.

     

/ .
30 stm Target 20 stm Target

Figure 5. 5: 20 stm and 30 stm trihedral corner reflector standard radar targets.

89

5.4 Single Beam Range to Target Experimental Results

A single beam range to target experiment was conducted to test the range linearity of the
unique FMCW radar solution. In this experiment, the radar antennas were pointed down
the length oflarge room. A measuring tape was laid down on the floor. A standard
target m ounted o n a tripod w as p laced at v arious ranges from the radar antennas (see
figures 5.6 and 5.7). The single range profile capture VI was then used to indicate and

save the range to target measured by the radar system.

 

/
~ /
. /
, / .
/
/ /
/ ,
/ itfitenna
/ / Beam
/
/ /
/ / . . . ’ . ,
' / ' so deem Standard
:(K A RadarTarget _
Radar \ . Range T0 (Corner Reflector) _

\.

Figure 5. 6: Single beam range to target experimental setup.

90

 

 

Figure 5. 7: Picture of the experimental setup, where the 30 stm target is in the

background, and the FM C W radar system is in the foreground.

The single range profile capture VI was set to sample 1000 samples at a rate of 200
KSPS. The ramp generator was set to produce a triangular waveform at a frequency of
100 Hz. From section 5.1, the FMCW radar system had a sweep bandwidth of 70 MHz.
From this information, the resulting frequency range data is in the form of 14 Hz/foot for
a round trip. This translates to 0.714 it per x axis unit in the resulting range data
displayed by the continuous operation and display VI. Range to target information
displayed by the VI was noted and plotted. The difference between actual distance to
target and the measured distance to target was also noted. The results are summarized in

table 5.1.

91

Actual Distance

500

450 I

400

Amplitude (V‘10,000)
é 2' § 9 § ‘9

0!
O

O

(m
20
25
30
35
40

Measured Distance

(:9
30.702
38.199
44.132
50.337
55.692

Ran e Profile

Range Difference (ft) 3193
10.72 figure 5.8
13.199 figure 5.9
14.982 figure 5.10
15.337 figure 5.11
15.962 figure 5.12

Table 5.1: Single beam range to target aperimental results summary.

Range to Target

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

92

-~ \ [3393?
up- .2
I .
I/ - .
\ l I" \/ \
D \
I
20 40 60 80 100 120
Beat Freqeuncy (Hzl20)

Figure 5. 8: Range profile of 30 stm target at 20ft.

WWTM

AmplltudeN‘10.000)
§ 53 § § § §

s§§

O

 

Beat Frequency (H1120)

Figure 5. 9: Range profile of 30 stm target at 25 ft.

93

Range to Target

§_§§§§§§

Amplitude (v'1o.ooo)
273

§

8

 

Beet Frequency (Hzml)

Figure 5.10: Range profile of 30 stm target at 30 ft.

94

Wuhan!

§H§§§§§§

Amplitude (V'10.000)

s§§

O

 

Beet Frequency (Hz/20)

Figure 5.11: Range profile of 30 stm target at 35 fi‘.

95

Range to Target

 

450

 

B
l
I

 

l
9
I

 

 

‘é

 

 

i

 

 

I’— "#"X'OTT
I . 599951 l

 

§

r..__._... _ ”—7....

 

 

Amplitude (V'10,000

5'3
I

 

 

d
O
O

 

 

I

 

 

 

 

 

 

 

 

O

 

 

Beat Frequency (Hz/20)

Figure 5.12: Range profile of 3 0 stm target at 40ft.

From the results in table 5.1, and the magnitude range profiles in figures 5.8 through
5.12, differences in actual ranges to target and ranges measured by the radar are evident.
These differences in range are due to the delay caused by the cables running from IFl and
IF2 to the IF Chassis. These range differences are also caused by the delays in the
various amplifier components inside of the IF Chassis. Due to the range differences

found, a mean range difference was calculated and found to be:

2 range _ differences

Mean range difference = 5

 

= 13.984 it (5.2)

96

From this experiment, it w as found that the mean range difference is 1 3.984 ft. T his
mean range difference was then factored into the Synthetic Aperture Radar VI, so that the

aperture could be more accurately focused.

97

5.5 Synthetic Aperture Radar Experimental Results

From the previous sections, 5.1 through 5.4, the unique FMCW radar solution has been
tested and adjusted for optimum performance, and standard targets were created to
characterize performance. The final experiment in this study of the unique FMCW radar

solution is the implementation of a Synthetic Aperture Radar imaging system.

For this experiment, the radar was setup in a large loading clock. The Front End was
mounted on a 12 foot long aluminum track, supported by steel saw horses (see figure
5.13). The radar was moved down the track manually, and lined up with a measuring
tape (see figure 5.14). Standard radar targets were placed down range from the system

(see figure 5.15).

”$5.33 i/QGQ 3

 

Figure 5.13: Synthetic Aperture Radar experimental setup.

98

15/11/2003

 

Figure 5.14: The Front End Assembly is manually pushed along the aluminum track

and lined up with a measuring tape in order to create a SAR image.

I 30 stm
T

arget

 

I

 

20 stm
Target

 

 

 

 

Figure 5.15: Both the 30 stm and 20 stm standard radar targets are placed down

range from the SAR system.

99

Images of the standard targets were taken with the SAR VI. In all of the experiments, the

SAR VI was set to the following parameters:

Number of Samples = 10,000
Sample Rate = 200 KSPS

FFT Sample Window Type = 3
Total Number of Beams = 138
Spacing (fl) = 0.08333

Number of Beams per Aperture = 32

Four different SAR images were taken using the standard targets. A list of the SAR

images taken is shown in table 5.2.

Data Before SAR
Target lmaged Process SAR Image
Nothing, base line measurement figure 5.9 figure 5.16
30 stm target at 25 ft figure 5.11 figure 5.17
20 stm target at 25 ft figure 5.13 figure 5.18
20 stm target at 25 ft, and 30 stm
target at 40 ft figure 5.15 figure 5.19

Table 5. 2: A list of the SAR images taken and their corresponding image plots.

The X axis units for the SAR images are in 0.71411 increments with a small range offset

of 3.866 ft. The Y axis units are in 1 inch increments. The Z axis magnitude units are in

100

Volts*lOK at 50 ohms. The radar system is located at the far right hand side of the

image, facing the left hand side of the image.

Cl 200-300

I 100—200

I0—100

 

 

Figure 5.16: Base line measurement before SAR processing.

101

  
  

.S19 9? 77*
fszs {03500-4000I

I .3000-3500‘
I2500-3000I
I2000—2500I
i 01500-2000 ‘
$55 £23332”
561

S31
$37

 

7849

Figure 5.1 7: Base line measurement afler SAR processing.

A b ase line m easurement w as taken with n othing p resent in front 0 f t he radar s ystem
except for the miscellaneous clutter in the location where the SAR data was captured.
The base line measurement was taken as a reference, and for use in locating the ideal
location to place standard targets. The before SAR processing data can be seen in figure
5.16, where the absolute value of the real range profile data is plotted. The data after
SAR processing can be seen in figure 5.17. From the results shown in figure 5.17, it is
apparent that the radar is picking up the various clutter located at the far end of the range

(this clutter can be seen visually in figure 5.15).

102

‘ I.
I .'I it .
”III"

 

 

 

D 300-400
D 200-300
I 100-200
I 0-100

 

Figure 5.18: 30 stm target located at a distance of 25 fl before SAR processing.

103

 

I fichoodl
137000-3000 ,
“000-7000 I
I5000-6000
.4000-5000
03000-4000 ;
02000-3000 '
I 1000-2000
Imo

 

 

 

Figure 5.19: 30 stm target located at a distance of 25 fl after SAR processing.

A scan was taken of a 30 stm standard target at a range of 25 feet from the radar
system. The data before SAR processing can be seen in figure 5.18, where the absolute
value of the real range profile data is plotted. The data after SAR processing can be seen
in figure 5.19. From these plots it is clear that the SAR process has successfully created a

very strong image of the standard target at the correct location.

 

Figure 5.20: 20 stm target located at a distance of 25 ft before SAR processing.

105

      
 

325 i i-i.
1 I2500-3000
II2000-2500I
1:11500-2000I
5‘3 131000-1500}
- I, . 3‘9 II500-1000
77:1. $55 3.9500 _.I

‘0-

7

, ’7
rim/f Uri/1’.
. min" , 11"

Figure 5.21 : 20 stm target located at a distance of 25 fl after SAR processing.

A scan was taken of a 20 stm standard target at a range of 25 feet from the radar
system. The data before SAR processing can be seen in figure 5.20, where the absolute
value of the real range profile data is plotted. The data after SAR processing can be seen
in figure 5.21. From these plots it is clear that the SAR process has successfiilly created a

very strong image of the standard target at the correct location.

106

   

92839 “2:333999999

Figure 5. 22: 20 stm target located at a distance of 25 ft, and a 30 stm target

located at a distance of 40 ft before SAR processing.

107

IE 3500-40007

., $25 I

1 “0003500
.I2500-3000
Ilzooo-zsoo
. . 131500-2000;
5 - 1.549 101000-1500
'- .. S55 I500-1000
861 Iran; .

$37
; s43

    
 

 

 

Figure 5. 23: 20 stm target located at a distance of 25 ft, and a 30 stm target

located at a distance of 40 fl after SAR processing.

And finally, a scan was taken of both targets, the 20 stm standard target at a range of 25
feet, and the 30 stm standard target at a range of 40 it from the radar system. The data
before SAR processing can be seen in figure 5.22, where the absolute value of the real
range profile data is plotted. The data after SAR processing can be seen in figure 5.23.
From these plots it is obvious that the SAR process has successfully created a very strong
image of the standard targets, and is capable of differentiating between them downrange

and cross range.

108

Summary:

After measuring and adjusting the tuning linearity and LO drive power level to MXRl,
accurate measurements from the unique solution FMCW radar were made. The system
was found to be capable of measuring range to target accurately, with some offset range.
That offset range was then factored into the Synthetic Aperture Radar VI, allowing the
system to acquire focused SAR images. In this section, the capabilities of the unique

solution to FMCW radar have been shown.

109

Chapter 6

Conclusion

Continuous-wave radar is a proven design concept. The coherent design topology of
traditional CW radar systems allows for simple implementation. CW radar is capable of
providing the Doppler shift data of a moving target. When CW radar is frequency
modulated, it is capable of providing range to target information. FMCW radar systems
use expensive microwave coaxial components. However, it has been shown in the
previous chapters that a lower cost alternative can be created with the use of two

inexpensive microwave transceiver modules.

CW radar has been in use for many years. Wide spread use of FMCW radar occurred due
to its practical application in radio altimeters. These FMCW radar altimeters have been
in use since the mid 1930’s. Similar radio altimeters are in use to date. Due to their
application, very simple means of data acquisition and processing have been used in radio

altimeters.
FMCW radar systems tend to use expensive coaxial microwave parts. Expensive parts

are n ot a c ost effective solution for high v olume low p ower r adar s olutions. F or this

reason, a low cost FMCW radar solution was created. This low cost solution took

110

advantage of Gunn diode based microwave transceiver modules commonly found in

motion sensors and radar speed guns.

This unique FMCW radar solution has been proven capable of determining range to
target. More importantly, the unique solution has proven itself very effective in the
advanced radar application of Synthetic Aperture Radar. Detailed images were created of

standard targets at various ranges, using data from the unique solution to FMCW radar.

Future work on this unique solution to FMCW radar should include creating a method for
increasing the transmit chirp bandwidth, increasing transmit power, automating the
sliding track used to create SAR images, and testing the ability to use multiple sensors in
close proximity. One way of increasing the chirp bandwidth would be to phase lock
XCVR1 to a highly accurate low frequency chirp reference. An increased chirp
bandwidth would cause an increase in range resolution. Increasing the transmit power
would in effect cause an increase in system sensitivity. Since the noise figure of the
transceiver modules is unchangeable, an increase in transmit power is required in order
for the system to detect smaller radar targets. Automating the metal track used to create
SAR images would minimized phase errors caused by moving the radar system manually.
Testing multiple radar units in close proximity must be done to determine the feasibility
of deploying multiple inexpensive radar sensors. This test would be necessary for
determining to what extent the radar sensors interfere with each other, and if it is possible

to prevent such interferences.

111

Test results show that this unique solution to FMCW radar has great potential for more
advanced applications. The use of inexpensive radar solutions will become increasingly
important in the near future for use in high volume applications where a reliable range to
target sensor is needed. I t is the conclusion of the author that this unique solution to
frequency-modulated continuous-wave radar is a practical low cost alternative to

previous FMCW radar designs.

112

References

[1] M. I. Skolnic, “Radar Handbook.” New York: McGraw-Hill, 1970.

[2] J. L. Eaves, and E. K. Ready, “Principles of Modern Radar.” New York: Van
Nostrand Reinhold Company, 1987.

[3] M. P. G. Capelli, “Radio Altimeter,” [RE Transactions on Aeronautical and
Navigational Electronics, vol. 1, pp. 3-7; June 1954.

[4] F. T. Wimberly and J. F. Lane, “The AN/APN-22 Radio Altimeter,” IRE
Transactions on Aeronautical and Navigational Engineering, vol. 1, pp. 8-14; June
1954.

[5] A. Black, K. E. Buecks, and A. H. Heaton, “Improved Radio Altimeter,” Wireless
World, vol. 60, pp. 138-140; March 1954.

[6] G. W. Stimson, “Introduction to Airborne Radar.” El Segundo, California: Hughes
Aircraft Company, 1983.

[7] “The ARRL Handbook, 71St Edition.” Newington, Connecticut: The American
Radio Relay League, Inc, 1994.

[8] C. A. Balanis, “Antenna Theory, Analysis and Design, 2"d Edition.” New York:
John Wiley & Sons Inc, 1997.

[9] J. J. Carr, “Secrets of RF Circuit Design, 2"d Edition.” New York: McGraw-Hill,
Inc, 1997.

[10] S. E. Craig, W. Fishbein, and O. E. Rittenbach, “Continuous-Wave Radar With
High Range Resolution and Unambiguous Velocity Determination,” IRE
Transactions on Military Electronics, vol. MIL-6, no. 2, April 1962, pp. 153—161.

[11] J. D. Kraus, “Antennas, 2nd Edition.” New York: McGraw-Hill, Inc, 1988.

[12] U. L. Rohde, J. Whitaker, and T. T. Bucher, “Communications Receivers, 2nd
Edition.” New York: McGraw Hill, Inc, 1997.

113

[13] W. K. Saunders, “Post-War Developments in Continuous-Wave and Frequency-
Modulated Radar,” IRE Transactions in Aerospace and Navigational Electronics,
vol. ANE-8, no. 1, March 1961, pp. 7-19.

[14] P. Vizrnuller, “RF Design Guide, Systems, Circuits, and Equations.” Boston:
Artech House, 1995.

[15] R. E. Ziemer, and W. H. Tranter, “ Principles of Communications, Systems,
Modulation, and Noise, 4th Edition.” New York: John Wiley & Sons, Inc., 1995.

114

   

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII