FPGA BASED SIGNAL PROCESSOR FOR

MULTIFUNCTION RADAR.

BY
DHIVYA.R
REG.NO. : 15703008

INTERNAL GUIDE: EXTERNAL GUIDE:

Mrs.MALATHI Mr.J.MADHAVAN
Senior Lecturer Scientist – ‘B’
SRM Engg. College LRDE , Bangalore.
 ABSTRACT

The existing methods used to generate Inphase (I) and Quadrature phase (Q)
components that are used to extract the target information from the reflected signals, uses
oscillator that has fixed frequency and mixer. The disadvantage of these methods is
whenever the input sampling frequency of the component is changed; the oscillator
(hardware) has to be changed. While the Digital Amplitude Phase Demodulation (DAPD)
that contains Direct Digital Synthesis (DDS) and a mixer can ameliorate such limitations.
DDS, otherwise known as Numerically Controlled Oscillator, uses phase accumulator,
which can generate various frequencies based on the value of the phase accumulator.

The I & Q signals which are then passed through the signal processing chain to
extract range, velocity and elevation information about the target. All the existing radars
are processor based those are limited for low and moderate frequency applications. While
the Field Programmable Gate Array (FPGA) based Signal processor can work under wide
range of frequencies including passband level. Thus FPGA based signal processor can be
useful in Space Time Adaptive Processing (STAP) applications.

Pulse compression is a method to discriminate two targets that are closer in range.
Doppler filtering is a method to extract the target information. The objective of the
present investigation is two fold: a) to generate I and Q components using DAPD, b) to
implement FPGA based signal processor by implementing Pulse compression and
Doppler filtering in it and to test the performance.

All these algorithms are implemented using XILINX software and tested in the
Virtex II FPGA kit.
 ACKNOWLEDGEMENT

I consider my privilege to express my gratitude to Shri KU Limaye,

scientist ‘H’, Director LRDE, for giving me permission to carryout the project

work at LRDE. I am thankful to Mr.Ravind.S,Scientist ‘G’, divisional officer and

Project Director, Radar-E for providing me an opportunity to work in the

division.I express my greatest thanks to my external guide Mr.Madhavan.J,

scientists ‘B’ for providing guidance, suggestions and corrections, in every stage,

which made my project to come out with flying colors.I am also thankful to

Mr.Rajani N, LRDE for his guidance, support and encouragement.

I sincerely regard my thanks to Prof.Venkataramani.R, Principal of SRM Institute

of Science and Technology, Kattankulathur , Prof.Jayashri.S, Head of

Department, Department of Electronics and Communication for providing a good

environment to work.

I sincerely regard my thanks to Mrs.Malathi.M, Senior Lecturer,

Department of Electronics & Communication, SRMIST for her valuable

guidance, support and encouragement. It is pleasure to express my whole hearted

thanks to our project coordinator Mr.Ganesan.S. My sincere thanks to all faculty

members and the non-teaching staff for extending their support and co-operation

for the completion of my project.Last but not the least, I wish to thank all my

family members and friends for their kind support.

Abstract
Acknowledgement
Table of contents
Chapter 1
Radar Fundamentals 1
1.1. Radar 1
1.2. Basic Radar 1
1.3. Electronically Steered Phased Array Radar 2
1.4. Simple Form of Radar Range Equation 3
1.5. Range To a Target 5
1.6. Pulse Repetition Frequency 6
1.7. Doppler Effect 7
1.8. Doppler Filter Banks 8
1.9. Pulse Doppler Radar 9
1.10. Basic Radar Measurements 9
1.10.1. Measurement of A Point Target 10
1.10.2. Measurement of Distributed Target 11
1.11. Radar Clutter 13
1.12. Application of A Radar 13
Chapter 2
Digital Amplitude Phase Demodulation 15
2.1. Introduction 15
2.2. Digital Amplitude Phase Demodulation or Digital I/O Demodulation 16
2.3. Blocks in DAPD 17
2.3.1. Chirp Signal Processing 17
2.3.2. Direct Digital Synthesis 17
2.3.3. LowPass FIR Filter 18
2.3.4. Decimation. 20

Chapter 3
Direct Digital Synthesis 22
3.1. Overview 22
3.2. DDS Advantages 22
3.3. Theory of Operation 23
3.4. Trends in Functional Integration 27
3.5. Calculating the Frequency Tuning Word 29
3.6. Understanding the Sample Output of a DDS Device 29
3.7. Jitter and Phase Noise Considerations in a DDS System 30
3.8. Direct Clocking of a DDS 31
3.9. DDS as a Clock Generator 31
3.10. DDS Vs Standard PLL 32
Chapter 4
Pulse Compression 34
4.1. Introduction 34
4.2. Concept of Operation 36
4.3. Types of Pulse Compression 37
4.3.1. Analog Pulse Compression 37
4.3.1.1. Correlation Processor 37
4.3.1.2. Stretch Processor 38
4.3.2. Digital Pulse Compression 38
4.3.2.1. Frequency Coding 39
4.3.2.2. Binary Phase Codes 40
4.3.2.3. Poly Phase Codes 42
4.3.3. Linear Frequency – Modulated Waveforms 44
4.3.3.1. Merits of Chirp 44
4.3.4. Pulse Compression Matched Filter 45
4.3.4.1. Pulse Compression Algorithm 45
4.3.4.2. Pulse Compression Implementation 46

Chapter 5
Doppler Filtering 47
5.1. Introduction 47
5.2. Doppler Effect 47
5.2.1. Doppler Filter Bank 48
5.3. Bandwidth of the Filters 49
5.4. Passband of the Filter Bank 50
5.5. Filtering Requirement 52
Chapter 6
System Requirements 53
6.1. The Hardware and Software Requirements 53
6.1.1. Hardware 53
6.1.2. Software 53
6.2. MATLAB 53
6.3. XILINX 54
Chapter 7
System Testing 56
Chapter 8
Results 57
Chapter 9
Conclusions 68
Chapter 10
Future Enhancements 69
References 70
List of Figures
List of Abbreviations
 LIST OF ABBREVIATIONS

Term Full Form

DDS Direct Digital Synthesis

PRF Pulse Repetition Frequency

DAPD Digital Amplitude Phase Demodulation

LFM Linear Frequency Modulated

VLIF Very Low Intermediate Frequency

FIR Finite Impulse Response

RF Radio Frequency

A/D Analog to Digital Converter

D/A Digital to Analog Converter

VHDL Very High Speed Integrated Circuit Hardware Description Language

FPGA Field Programmable Gate Array

PROM Programmable Read Only Memory

RAM Random Access Memory

LPF Low Pass Filter

NBF Narrow Band Filter

SFW Stepped Frequency Waveform

I&Q InPhase & Quadrature Phase

FFT Fast Fourier Transform

 LIST OF FIGURES

Figure 1.1: Time Domain Measurement of Range to a Target 5

Figure 1.2: Signal Returns in Range Measurement 6

Figure 1.3: Doppler Frequency Measurement using Doppler 8

Filter Bank

Figure 2.4: General block diagram for DAPD 15

19
Figure 2.5: An N-tap FIR Filter
Figure 2.6: A Basic Decimator 20
21
Figure 2.7: Frequency domain view of Decimation
23
Figure 3.8: Simple Direct Digital Synthesizer
Figure 3.9: Frequency-tunable DDS System 24
Figure 3.10: Signal flow through the DDS architecture 25
Figure 4.11: Range resolution 35
36
Figure 4.12: Pulse compression
Figure 4.13: LFM pulse as segments of progressively higher 37
frequency.
41
Figure 4.14: Binary Phase Code of Length 7
Figure 4.15: a) A frequency – modulated pulse over time 44
b) The frequency of the same pulse over time
Figure 5.16: Received signals are applied in parallel to a bank of 48
filters.
Figure 5.17: Overlapping of passbands 49
Figure 5.18: Passband of the filter bank 51
 CHAPTER 1

RADAR FUNDAMENTALS

1.1. Radar

The word “RADAR” is an acronym coined from the expression “RAdio Detection
And Ranging”. RADAR can detect the presence of target and measure its position, and it
can measure the velocity of the target.

Equipment that uses the principles of radar is called a radar system. Such a radar
system may be small enough to be installed in an automobile spotlight, such as a speed –
detection radar, or large enough to require one or more buildings to enclose a single
complete radar system.

1.2. Basic Radar

Radar is an electromagnetic system for the detection and location of the reflecting
objects such as aircraft, ships, spacecrafts, people, and the natural environment. It
operates by radiating energy into space and detecting the echo signal reflected from the
object, or the target. The reflected energy that is returned to the radar not only indicates
the presence of a target, but by comparing the received echo signal with that was
transmitted, its location can be determined along with other target-related information. It
can operate in darkness, haze, fog, rain, and snow. Its ability to measure distance with
accuracy and in all weather is one of its most important attributes.

1.3. Electronically Steered Phased Array Radar

 A phased array is a directive antenna made up of number of individual antenna or
radiating elements. Its radiation pattern is determined by the relative amplitude and phase
distribution of the current at each of its elements. The phased array antenna has the
advantage of being able to have its beam electronically steered in azimuth and elevation
by changing the phase of the current at each element. The beam of a large fixed phased
array antenna therefore can be rapidly steered from one direction to another without the
need for mechanically moving a large and heavy antenna. Typical phased array radar for
microwave radar might have several thousand individual radiating elements using phase
shifters that allow the beam to be switched from one direction to another in several
microseconds, or less.

Electronically steerable phased array is of interest because they can provide agile,
rapid beam steering Potential for large peak and large average power. Each element can
have its own transmitter and receiver. The power aperture product can be large,
especially at lower frequencies.

• Multiple target tracking can be accomplished either by generating multiple,

simultaneous, independent beams or by rapidly switching a single beam to view more
than one target in sequence.

The disadvantage of phased array radar is that it is complex and can be of high cost.

A linear array consists of antenna elements arranged in a straight line in one

dimension. A planar array is a two-dimensional configuration of antenna elements
arranged to lie in a plane. In both the linear and planar array, the element spacing usually
is uniform. The planar array may be thought of as a linear array of linear arrays. Most
phased arrays of interest for radar are planar.

1.4. Simple Form of Radar Range Equation

 The radar range equation relates the range of radar to the characteristics of the
transmitter, receiver, antenna, target, and the environment. It is useful not only for
determining the maximum range at which particular radar can detect a target, but it can
serve as means for understanding the factors affecting radar performance.

If the transmitter power is Pt and radiated by an isotropic antenna, the power density
at a distance R from the radar is equal to the radiated power divided by the surface area
4πR 2 of an imaginary sphere of radius R, or

Pt
Power density at range R from an isotropic antenna = (1.1)
4πR 2

Power density is measured in units of watts per square meter. Radars, however,
employ directive antennas to concentrate the radiated power Pt in a particular direction.
The gain of the antenna is measure of the increased power density in some direction as
compared to the power density that would appear in that direction from an isotropic
antenna. The maximum gain G of the an antenna may be defined as

Maximum power density radiated by a directive antenna

G=
Power density radiated by loss less isotropic antenna with same power input

Pt Gt
Power density at a range R from a directive antenna = (1.2)
4πR 2

The target intercepts a portion of the incident energy and scatters it in various
directions. It is only the power density scattered in the direction of the radar that is of
interest. The radar cross section of the target determines the power density returned to the
radar for a particular power density incident on the target. It is denoted by σ` and is often
called in short as target cross-section, radar cross section of target.

Pt Gt σ
Reradiated power density back at the radar = ° (1.3)
4πR 4πR 2
2
 The radar cross section has units of area, but it can be misleading to associate the
radar cross section directly with the target physical size. Radar cross-section is more
dependent on target shape than on its physical size.

The radar antenna captures a portion of the echo energy incident on it. The power
received by the radar is given as the product of the incident power density times the
effective area Ae of the receiving antenna. The effective area is related to the physical
area A by the relationship Ae = ρ a A , where ρ a = antenna aperture efficiency. The

received signal power Pr (watts) is then

Pt Gt σ
Pr = ° ° Ae (1.4)
4πR 4πR 2
2

The maximum range of the radar Rmax is the distance beyond which the target cannot be
detected. It occurs when the received signal power equals the minimum detectable signal
S min . Substituting S min = Pr and rearranging

1
 P GtA σ  4
Rmax =  t 2 e  (1.5)
 (4π ) S min 

This is the fundamental radar range equation. (It is also called, for simplicity, the
radar equation or range equation). The important parameters are the transmitting gain and
the receiving effective area. The transmitter power Pt has not been specified as either the

average or the peak power. It depends on how S min is defined. Here Pt denotes peak
power.

If the same antenna is used for both transmitting and receiving, as it usually is in
radar, antenna theory gives the relationship between the transmit gain G and the receive
effective area Ae

4πAe 4πρ a A
G= = (1.6)
λ2 λ2
 c
Where λ= wavelength. (Wavelength λ = , where c=velocity of propagation and
f
f =frequency)

1.5. Range to a Target

Figure 1.1: Time Domain Measurement of Range to a Target

The most common radar signal, or waveform, is a series of short duration, somewhat
regular shaped pulses modulating a sine wave carrier. (This is called sometimes pulse
train). The range to a target is determined by the time TR it takes for the radar signal to
travel to the target and return back. Electromagnetic energy in free space travels with the
speed of light, which is c = 3 × 10 8 m / s . Thus the time for the signal to travel to a target
located at a range R and return back to the radar is 2R/c.the range to a target is then

cT R
R= (1.7)
2

With the range in kilometers or in nautical miles, and T in microsecond.

R(km) = 0.15TR (µs ) or R(nmi ) = .081TR (µs )

 Each microsecond of round trip travel time corresponds to a distance of 150 meters,
164 yards, 492 feet, 081 nautical miles, or .093 statute mile. It takes 12.35µs for the radar
signal to travel a nautical mile and back.

1.6. Pulse Repetition Frequency

Figure 1.2: Signal Returns in Range Measurement

The maximum unambiguous range beyond which targets are not expected often
determines the pulse repetition frequency. When echoes appear from beyond the
maximum unambiguous range, especially for some unusually large target or clutter
source or when anomalous propagation conditions occur to extend the normal range of
the radar beyond the horizon. Echo signal that arrive at a time later than the pulse
repetition period are called second time around echoes. They are also called multiple time
around echoes, particularly when they arrive from ranges greater than 2 Run . The apparent
range of these ambiguous echoes can result in error and confusion.
 Changing the PRF of the radar can recognize ambiguous range echoes. When the
PRF is changed, the unambiguous echo remains at its true range. Ambiguous range
echoes, however, appear at different apparent ranges for PRF.

1.7. Doppler Effect

When a reflected radar signal is received from a radially moving target an apparent
shift in the RF transmitted frequency is measured. This frequency shift is then a measure
of the relative radial velocity of the target. For every half wavelength per second that a
target’s range decreases, the RF frequency phase of the received signal advances by the
equivalent of one cycle per second (one hertz). The formula for the Doppler shift is as
follows:

fd = (-2 Vr )/C (1.8)

Where fd = Doppler frequency

Vr = relative radial velocity

C = Speed of light

When the moving target’s direction is at an angle to a stationary radar’s beam, the
measured Doppler frequency must be corrected by the cosine of the angle that is formed
by the radar’s and target’s lines of direction. If the radar is airborne, the measured
Doppler frequency must be corrected by the cosine of the off-bore sight angle as well as
by the cosine of the angle that is due to the difference in altitude between the radar and
the target.

1.8. Doppler Filter Banks

 Figure 1.3: Doppler Frequency Measurement using Doppler Filter Bank

A Doppler filter bank is set of contiguous filters for detecting targets. A filter bank
has several advantages over the single filters.

Multiple moving targets can be separated from one another in a filter bank. This can
be particularly important when one of the echo signals is from undesired moving clutter,
such as rainstorm or birds with a non-zero Doppler ship. When the clutter and target echo
signal appear in different Doppler filters, the clutter echo need not interfere with the
detection of the desired moving target.

A measure of the targets radial velocity can be obtained from the filter number in
which detection occurs. It might be ambiguous, but a change in the prf can resolve the
ambiguity in the radial velocity, Just as changing the prf can resolve range ambiguities.

The narrow band Doppler filters exclude more noise than do the MTI delay line
cancellers by providing coherent integration.

1.9. Pulse Doppler Radar

 The radar that can detect small aircraft at long ranges, even when their echoes are
buried in strong ground clutter. It can track them either singly or several at a time while
continuing to search for more. The important thing is it can detect and track moving
targets. The difference is in the radar features such as Coherence, Digital processing, and
Digital control.

A radar that increase its PRF high enough to avoid the problems of blind speed is
called pulse Doppler radar more precisely, a high PRF pulse Doppler radar is one with no
blind speeds within the Doppler space. In some situations, however, it may be beneficial
to operate at slightly lower PRF and accept both range and Doppler ambiguities. Such
radar is called a medium PRF pulse Doppler radar. Thus there are three different types of
pulse radars that use Doppler’s. They differ in their PRFs and the type of ambiguities
they are willing to tolerate. These are

1. The MTI with no range ambiguities and many Doppler ambiguities.

2. The high PRF pulse Doppler with just the opposite many range
ambiguities and no Doppler ambiguities

3. The medium PRF pulse Doppler radar has both the Range and Doppler
ambiguity.

1.10. Basic Radar Measurements

Radar can obtain a target location in range and azimuth, and sometimes elevation.
After several observations of the moving target over a period of time, the target
trajectory, or track, can be obtained.
The resolution capabilities of radar usually determine whether a target is considered
as a point scatter or a distributed target

1.10.1. Measurement of a Point Target

 The basic measurements that can be made for a point target when only a single
observation is made are range, radial velocity, direction (angle), and, in some special
cases, tangential velocity.

• Range

The measurement of distance, or range, can be obtained from the round-trip time TR
required for a radar signal to travel to the target and back. The range r is given by
cT R
R= . In many radar applications the target range is the most significant
2
measurement that is made. No other sensor has been able to compete with radar for
determining range to distant target, especially in accuracy, ability to make a measurement
over very long or very short distances and under adverse weather conditions. Long-range
air surveillance radar might measure range to an accuracy of many tens of meters, but
accuracies of few centimeters are possible with precision systems. In most precise
systems, the accuracy of a range measurement is limited only by their accuracy with
which the velocity of propagation is known. The spectral bandwidth occupied by the
radar signal is fundamental resource required for accurate measurement. The greater the
bandwidth, the more accurate can be the range measurement.

• Angle Measurement

Almost all radars utilize the directive antennas with relatively narrow beam widths.
A directive antenna not only provide the large transmitting gain and large receiving
aperture need for detecting weak echo signals, but its narrow beam width allows the
targets direction to be determined accurately. It can do this by noting the direction the
antenna points when its received echo signal is maximum. Typical microwave radar
might have a beam width of one or a few degrees. The narrower the beam width, the
greater the mechanical and electrical tolerance that are required of the antenna.Angular
accuracy can be much better than the antenna beam width. Angle accuracy depends on
the electrical size of the antenna (size measured in wave lengths) with SNR typical of
those required for reliable detection, the angular location of a target can be determined to
about1/10th of beam width. The best precision monopulse tracking radars used for range
instrumentation can determine angle to 0.1mrad rms (0.006 degree) if the SNR is large
enough and if the proper efforts are taken to minimize errors.

• Radial Velocity

Measurement of the radial component of velocity in many types of radar is obtained

from the rate of change range. This is known as range rate. The classical method for

finding the radial velocity is based on v r =

(R2 − R1 ) . It is found from the range
(T2 − T1 )
R1 measured at time T1 and the range R2 at time T2 . However, this method of finding
range rate is not considered here as a basic radar measurement even though it may be
widely used. Instead the Doppler frequency shift is the basic method for obtaining radial
velocity. It can be made on the single observation. Using the classical expression for the
Doppler frequency shift, f d , the radial velocity v r is given as v r = λf d 2

Where λ = wavelength. It can be shown from theoretical expressions that the radial
velocity accuracy derived from the Doppler frequency shift can be much better than that
found from the range rate, assuming the time between the two range measurements in the
range two method is same as the time duration of the Doppler frequency measurement.

1.10.2. Measurements on a Distributed Target

With sufficient resolution in appropriate dimension, the size and shape of a

distributed target can be ascertained. It should be recalled that resolution and accuracy are
not the same. Range resolution requires that the entire bandwidth be occupied
continuously without gaps in the signal frequency spectrum. Range accuracy, however,
only requires, as a minimum, that there be adequate spectral energy at the two ends of the
spectral bandwidth.
Radial profile the targets profile (and size) in the range dimension can be obtained when
the radar range resolution cell is smaller in size than the targets dimensions. To obtain the
cτ
radial a profile of a target it is required that << D , where D = the targets radial
2
component and τ =pulse width. Good resolution in the range dimension requires large
spectral bandwidth.

• Tangential (Cross—Range) Profile

With sufficient resolution in the angle dimension, the tangential profile of a
distributed target can be determined. This can provide the angular size of the target and
location in angle of the scattering centers. If the range is known, the location of scatters in
the tangential dimension can be determined since the cross range (tangential) dimension
is equal to the product of range and angle (the latter in radians).

• Size And Shape

When the tangential profile is obtained at each range resolution cell, the target
image (size and shape) is formed.

• Symmetry
The response of the target to change in polarization of the radar signal can provide
a measure of the symmetry of the target. (The polarization of a radar signal is determined
by the orientation of the electric field). If a sphere were directly viewed by radar with a
rotating linearly polarized signal, there would be no change of the echo signal when the
polarization is changed.

1.11. Radar Clutter

Radar is the term used by radar engineers to denote unwanted echoes from the
natural environment. It implies that these unwanted echoes “clutter” the radar and make
difficult the detection of unwanted targets. Clutter includes echoes from land, sea,
weather (particularly rain), birds and insects. At the lower radar frequencies, echoes from
ionized meteor trails and aurora also can produce clutter large clutter echoes can mask
echoes from desired targets and limit radar capability. When clutter is much larger than
receiver noise, the optimum radar waveform and signal processing can be quite from that
employed when only receiver noise is the dominant limitations of sensitivity. Radar
echoes from the environment are not always undesired, reflection from storm clouds, for
example, can be a nuisance to a radar that must detect aircraft, storm clouds containing
rain are what radar meteorologist wants to detect in order to measure rain fall rate over a
large area.
Echoes from land and sea are examples are of surface clutter. Echoes from rain and
chaff are examples of volume clutter.

1.12. Application of Radar

Radar has been employed to detect targets on the ground, on the sea, in air, in space,
and even below ground. The major areas of radar application are briefly described below.

• Military

Radar is an important part of Air-Defense systems as well as the operation of

offensive missiles and other weapons. In Air Defense it performs the functions of
surveillance and weapon control. Surveillance includes target detection, target
recognition, target tracking, and designation to a weapon system. Weapon control radars
track targets, direct the weapon to an intercept, and asses the effectiveness of the
engagement. The military has been the major user of the radar.

• Remote Sensing

All radars are remote sensors: however, this term is used to imply the sensing of
the environment. Four important examples of remote sensing are 1.weather observation,
which is a regular part of TV weather reporting as well as a major input to national
weather prediction 2.planetary observation, such as the mapping of Venus beneath its
visually opaque clouds. 3. Short range below ground probing and 4. Mapping of sea ice
to route shipping in an efficient manner.

• Air Traffic Control (ATC)

Radar have been employed around the world to safely control air traffic in the
vicinity of airports (air surveillance radar, or ASR), and en route from one airport to
another (air route surveillance radar, or ARSR) as well as ground traffic and taxing
aircraft on the ground (airport surface detection equipment, or ASDE).. The air traffic
control radar beacon system widely used for the control of air traffic, although not a
radar, originated from military and uses radar like technology.

• Aircraft Safety And Navigation

The airborne weather avoidance radar outlines regions of precipitation and

dangerous wind shear to allow the pilot to avoid hazardous conditions. Low flying
military aircraft relies on terrain avoidance and terrain following radars to avoid
colliding with obstructions or high terrain. Military aircraft employ ground mapping
radars to image a scene.

CHAPTER 2

DIGITAL AMPLITUDE PHASE DEMODULATION (DAPD)

2.1. Introduction:

In the radar signal generator (RSG), the digital generation unit generates the
frequency modulated (chirp) very low IF for pulse compression. The modulation may be
Linear Frequency Modulation (LFM) or Non-linear Frequency Modulation (NLFM).
The very low IF (VLIF) at the RSG is 10 MHz and the bandwidth (BW) is 2.5 MHz.
Sampling is done by 20 MHz clock. DAPD removes the VLIF carrier and generates the I
and Q components digitally. The complete DAPD is done in RSG/RSR.

2.2. Digital Amplitude Phase Demodulation (DAPD) or Digital I/ Q Demodulation:

The general block diagram for DAPD is given below

Figure 2.4: General block diagram for DAPD

The 5 MHz VLIF signal is digitized at 20 Mega samples per second. The sampled
signal is filtered by an image rejection filter (low pass FIR filter) and decimated by 8 to
get the I and Q signal. If I(t) is the input signal to be sampled, then the sampled signal is
given by

X(n) = I(nTs) (2.9)

If I(f) is the spectrum of I(t), then the spectrum of the sampled signal is given by
X(f) = K [I(f – mfs)] (2.10)
where fs = 20 MHz
The image filtered complex signal Y(n) has a sampling rate of 20MHz.The input real
signal I(t) is converted into complex signal Y(n) by the use of DDS. Y(n) is decimated by
8 to get z(n) = y(8n).The spectrum of z(n) is given by

Z(Ω) = K1 Y(Ω / 8) (2.11)

Hence the demodulation to get I and Q signal is done by digital filtering and simple
decimation process. The resultant I and Q signals are not exactly orthogonal due to
residual image. The image rejection depends on the filter characteristics H(Ω).The I and
Q signals are further time multiplexed and sent to signal processor (SP) along with other
synchros for further processing.

2.3. Blocks in DAPD

The DAPD or Digital I/Q demodulation has four functional blocks.

1. Chirp signal generation
2. Direct Digital Synthesis (DDS)
3. Low pass FIR filter
4. Decimator.
2.3.1. Chirp signal generation

A Chirp or Linear FM (LFM) waveform is pulsed frequency modulation scheme

in which a carrier is swept over a wide frequency band during a given pulse interval.
The advantages of Chirp waveform are
• Range resolution,
• Lower power to detect weak targets.

The chirp signal is generated in MATLAB using the following equation.

S(t) = A Cos[ 2Πfct + (Bt2 /2(pw))] (2.12)

Where,
Amplitude (A) = 1,
Center frequency (fc) = 5 MHz,
Bandwidth (B) = 2.5 MHz,
Sampling frequency (fs) = 20 MHz,
Pulse width (pw) = 50 µsec,
Time (t) = 0: 1/fs: pw.
The chirp signal is shown in Graph (1). Each input pulse is sampled at the rate of 20 MHz
and hence 1001 samples are obtained for a single pulse. The sampled coefficients are
then passed to the multiplier block.

2.3.2. Direct Digital Synthesis (DDS)

Direct digital synthesis (DDS) is a technique for using digital data processing blocks as a
means to generate a frequency- and phase-tunable output signal referenced to a fixed-
frequency precision clock source. Based on the reference clock source (100 MHz), DDS
generates a 5 MHz sine waveform and cosine waveform.The DDS is simulated in
XILINX and the output is shown in waveform (1).
The sampled coefficients of each pulse (1001 coefficients) from the signal
generation block is multiplied with the sine waveform and cosine waveform output of the
DDS block to get the in phase (I) and quadrature phase (Q) components i.e. real and
imaginary signal.The output is shown in waveform (2).

2.3.3. Low pass FIR filter

The real and imaginary signal are passed through the low pass FIR filter in which
the FIR filter is loaded with the LPF coefficients. This co-efficients has the low pass filter
characteristics. The cut off frequency of the low pass filter is fc/2 MHz.The low pass FIR
filter output is shown in waveform (4).
The LPF coefficients (32 coefficients) are generated in MATLAB using the
“Remez algorithm”. B = REMEZ (N, F, A) returns a length N+1 linear phase (real,
symmetric coefficients) FIR filter which has the best approximation to the desired
frequency response described by F and A in the minimax sense. F is a vector of
frequency band edges in pairs, in ascending order between 0 and 1. 1 corresponds to the
Nyquist frequency or half the sampling frequency. A is a real vector the same size as F
which specifies the desired amplitude of the frequency response of the resultant filter B.
The desired response is the line connecting the points (F(k), A(k)) and (F(k+1), A(k+1))
for odd k; REMEZ treats the bands between F(k+1) and F(k+2) for odd k as "transition
bands" or "don't care" regions. Thus the desired amplitude is piecewise linear with
transition bands. The maximum error is minimized.The convolved output of the signal
and the LPF co-efficients is shown in Graph (2).
The implementation of low pass FIR filter is done in VHDL using Xilinx software.
Fundamentally, an FIR is a type of digital filter with a very simple structure. It is nothing
more than a chain of delay, multiply, and add stages. Each stage consists of an input and
output data path and a fixed coefficient (a number which serves as one of the
multiplicands in the multiplier section).
The FIR filter equation for the output y(n), in terms of the input x(n) for an FIR of
arbitrary length N is,

y(n) = a0 x(n) + a1 x(n-1) + a2 x(n-2) + … + aN-1 x(n-N-1) (2.13)

Application of the z-transform leads to the transfer function, H(z), of an N-tap FIR filter.

H(z) = a0 + a1 z-1 + a2 z-2 + … + aN-1 z-(N-1) (2.14)

Increasing the number of taps in an FIR increases the sharpness. The response of the FIR
is completely determined by the coefficient values. By choosing the appropriate
coefficients it is possible to design filters with just about any response: lowpass,
highpass, bandpass, bandreject, allpass, and more.
 Figure 2.5: An N-tap FIR Filter

2.3.4.Decimator

The image filtered complex signal Y(n) is then passed to the next stage for
decimation. Here the complex signal Y(n) is decimated by 8 to get Z(n).The function of a
decimator is to take data that was sampled at one rate and change it to new data sampled
at a lower rate. The data must be modified in such a way that when it is sampled at the
lower rate the original signal is preserved. A pictorial representation of the decimation
process is shown in Figure below.
 Figure 2.6: A Basic Decimator

Notice that the decimator has two parts. An input section that samples at a rate of Fs and
an output section that samples at a rate of (1/m)Fs, where m is a positive integer greater
than 1. The structure of the basic decimator indicates that for every m input samples there
will be 1 output sample. Here the value of m is 8.The decimated output is shown in
waveform (5).

Figure 2.7: Frequency domain view of Decimation

The decimation process is expected to translate the spectral information in Figure

(a) such that it can be properly contained in Figure (c). The problem is that the Nyquist
region of Figure (a) is three times wider than the Nyquist region of Figure (c). This is a
direct result of the difference in sample rates. There is no way that the complete spectral
content of the Nyquist region of Figure (a) can be placed in the Nyquist region of Figure
(c). This brings up the cardinal rule of decimation: The bandwidth of the data prior to
decimation must be confined to the Nyquist bandwidth of the lower sample rate. So, for
decimation by a factor of m, the original data must reside in a bandwidth given by
Fs/(2m), where Fs is the rate at which the original data was sampled. Thus, if the original
data contains valid information in the portion of the spectrum beyond Fs/(2m),
decimation is not possible.

CHAPTER 3

DIRECT DIGITAL SYNTHESIS (DDS)

3.1. Overview

Direct digital synthesis (DDS) is a technique for using digital data processing blocks
as a means to generate a frequency- and phase-tunable output signal referenced to a
fixed-frequency precision clock source. In essence, the reference clock frequency is
“divided down” in a DDS architecture by the scaling factor set forth in a programmable
binary tuning word. The tuning word is typically 24-48 bits long, which enables a DDS
implementation to provide superior output frequency tuning resolution.
Today’s cost-competitive, high-performance, functionally integrated, and small
package-sized DDS products are fast becoming an alternative to traditional frequency-
agile analog synthesizer solutions. The integration of a high-speed, high-performance,
D/A converter and DDS architecture onto a single chip (forming what is commonly
known as a Complete-DDS solution) enabled this technology to target a wider range of
applications and provide, in many cases, an attractive alternative to analog-based PLL
synthesizers. For many applications, the DDS solution holds some distinct advantages
over the equivalent agile analog frequency synthesizer employing PLL circuitry.

3.2. DDS advantages

• Micro-Hertz tuning resolution of the output frequency and sub-degree phase tuning
capability, all under complete digital control.
• Extremely fast “hopping speed” in tuning output frequency (or phase); phase-
continuous frequency hops with no over/undershoot or analog-related loop settling
time anomalies.
• The DDS digital architecture eliminates the need for the manual system tuning and
tweaking associated with component aging and temperature drift in analog
synthesizer solutions.
• The digital control interface of the DDS architecture facilitates an environment where
systems can be remotely controlled, and minutely optimized, under processor control.
• When utilized as a quadrature synthesizer, DDS afford unparalleled matching and
control of I and Q synthesized outputs.

3.3. Theory of Operation

In its simplest form, a direct digital synthesizer can be implemented from a

precision reference clock, an address counter, a programmable read only memory
(PROM), and a D/A converter.

Figure 3.8: Simple Direct Digital Synthesizer

In this case, the digital amplitude information that corresponds to a complete

cycle of a sine wave is stored in the PROM. The PROM is therefore functioning as a sine
lookup table. The address counter steps through and accesses each of the PROM’s
memory locations and the contents (the equivalent sine amplitude words) are presented to
a high-speed D/A converter. The D/A converter generates an analog sine wave in
response to the digital input words from the PROM.
The output frequency of this DDS implementation is dependent on
1. The frequency of the reference clock, and
2. The sine wave step size that is programmed into the PROM.

While the analog output fidelity, jitter and AC performance of this simplistic architecture
can be quite good, but it lacks tuning flexibility. The output frequency can only be
changed by changing the frequency of the reference clock or by reprogramming the
PROM. Neither of these options support high-speed output frequency hopping.
With the introduction of a phase accumulator function into the digital signal chain, this
architecture becomes a numerically controlled oscillator, which is the core of a highly
flexible DDS device. An N-bit variable-modulus counter and phase register is
implemented in the circuit before the sine lookup table, as a replacement for the address
counter. The carry function allows this function as a “phase wheel” in the DDS
architecture.

Figure 3.9: Frequency-tunable DDS System

To understand this basic function, visualize the sine wave oscillation as a vector
rotating around a phase circle each designated point on the phase wheel corresponds to
the equivalent point on a cycle of a sine waveform. As the vector rotates around the
wheel, visualize that a corresponding output sine wave is being generated. One revolution
of the vector around the phase wheel, at a constant speed, results in one complete cycle of
the output sine wave. The phase accumulator is utilized to provide the equivalent of the
vector’s linear rotation around the phase wheel. The contents of the phase accumulator
correspond to the points on the cycle of the output sine wave. The number of discrete
phase points contained in the “wheel” is determined by the resolution, N, of the phase
accumulator. The output of the phase accumulator is linear and cannot directly be used to
generate a sine wave or any other waveform except a ramp.
Therefore, a phase-to amplitude lookup table is used to convert a truncated
version of the phase accumulator’s instantaneous output value into the sine wave
amplitude information that is presented to the D/A converter. Most DDS architectures
exploit the symmetrical nature of a sine wave and utilize mapping logic to synthesize a
complete sine wave cycle from ¼ cycle of data from the phase accumulator. The phase-
to-amplitude lookup table generates all the necessary data by reading forward then back
through the lookup table.
 The phase accumulator is actually a modulus M counter that increments its stored
number each time it receives a clock pulse. The magnitude of the increment is determined
by a digital word M contained in a “delta phase register” that is summed with the
overflow of the counter. The word in the delta phase register forms the phase step size
between reference clock updates; it effectively sets how many points to skip around the
phase wheel. The larger the jump size, the faster the phase accumulator overflows and
completes its equivalent of a sine wave cycle. For an N=32-bit phase accumulator, an M
value of 0000…0001(one) would result in the phase accumulator overflowing after 232
reference clock cycles (increments). If the M value is changed to 0111…1111, the phase
accumulator will overflow after only 21 clock cycles, or two reference clock cycles. This
control of the jump size constitutes the frequency tuning resolution of the DDS
architecture.
 Figure 3.10: Signal flow through the DDS architecture

The relationship of the phase accumulator and delta phase accumulator form the
basic tuning equation for DDS architecture:
FOUT = (M (REFCLK)) /2N (3.15)
Where: FOUT = the output frequency of the DDS
M = the binary tuning word
REFCLK = the internal reference clock frequency (system clock)
N = the length in bits of the phase accumulator

Changes to the value of M in the DDS architecture result in immediate and phase
continuous changes in the output frequency. In practical application, the M value, or
frequency tuning word, is loaded into an internal serial or byte-loaded register, which
precedes the parallel-output delta phase register. This is generally done to minimize the
package pin count of the DDS device. Once the buffer register is loaded, the parallel-
output delta phase register is clocked and the DDS output frequency changes. Generally,
the only speed limitation to changing the output frequency of a DDS is the maximum rate
at which the buffer register can be loaded and executed. Obviously, a parallel byte load
control interface enhances frequency-hopping capability.

3.4. Trends in Functional Integration

One of the advantages to the digital nature of DDS architecture is that digital
functional blocks can readily be added to the core blocks to enhance the capability and
feature set of a given device. For general-purpose use, a DDS device will include an
integrated D/A converter function to provide an analog output signal. This “complete-
DDS” approach greatly enhances the overall usefulness and “user-friendliness”
associated with the basic DDS devices. The present state of the art for a complete-DDS
solution is at 300 MHz clock speeds with an integrated 12-bit D/A converter. Along with
the integrated D/A converter, DDS solutions normally contain additional digital blocks
that perform various operations on the signal path. These blocks provide a higher level of
functionality in the DDS solution and provide an expanded set of user-controlled
features. The individual functional blocks are described below:
• A programmable REFCLK Multiplier function include at the clock input, multiplies
the frequency of the external reference clock, thereby reducing the speed requirement
on the precision reference clock. The REFCLK Multiplier function also enhances the
ability of the DDS device to utilize available system clock sources.
• The addition of an adder after the phase accumulator enables the output sine wave to
be phase-delayed in correspondence with a phase tuning word. The length of the
adder circuit determines the number of bits in the phase tuning word, and therefore,
the resolution of the delay. In this architecture, the phase tuning word is 14-bits.
• An Inverse SINC block inserted before the D/A converter compensates for the
SIN(X)/X response of the quantized D/A converter output, and thereby provides a
constant amplitude output over the Nyquist range of the DDS device
• A digital multiplier inserted between the Sine look-up table and the D/A converter
enables amplitude modulation of the output sine wave. The width of the digital
multiplier word determines the resolution of the output amplitude step size.
• An additional high-speed D/A converter can be included to provide the cosine output
from the DDS. This allows the DDS device to provide I and Q outputs, which are
precisely matched in frequency, quadrature phase, and amplitude. The additional D/A
converter may also be driven from the control interface and used as a control DAC
for various applications.
• A high-speed comparator function can be integrated which facilitates use of the DDS
device as a clock generator. The comparator is configured to convert the sine wave
output from the DDS D/A converter into a square wave.
• Frequency/phase registers can be added which allow frequency and phase words to be
pre-programmed and their contents executed via a single control pin. This
configuration also supports frequency-shift keying (FSK) modulation with the single-
pin input programmed for the desired “mark” and “space” frequencies.
The growing popularity in DDS solutions is due to the fact that all of this performance
and functionality is available at a reasonable price and in a comparatively small package.
The following is the guideline for the level of performance available from the dual 12-
bit/100 MHz complete-DDS solution
-100 MHz external reference clock
- Frequency tuning word length = 12 bits
- Phase tuning word length = 12 bits which provides .0847 degrees of phase delay control
resolution.
- Output frequency bandwidth (assuming one-third of REFCLK rate) = 33.33 MHz
-Output amplitude control = zero output to full scale in 204 steps (12-bit control word)

3.5. Calculating the Frequency Tuning Word

The output frequency of a DDS device is determined by the formula:

FOUT = (M (REFCLK)) /2N

Where: FOUT = the output frequency of the DDS

M = the binary tuning word
REFCLK = the internal reference clock frequency
N = the length in bits of the phase accumulator
The length of the phase accumulator (N) is the length of the tuning word, which
determines the degree of frequency tuning resolution of the DDS implementation. Let’s
find the frequency tuning word for an output frequency of 5 MHz where REFCLK is
100 MHz and the tuning word length is 12 bits (binary). The resulting equation would be:

5 MHz = (M (100 MHz)) /212

Solving for M…

M = (5 MHz (212))/100 MHz

M= 11001100
Loading this value of M into the frequency control register would result in a frequency
output of 5 MHz, given a reference clock frequency of 100 MHz.
3.6. Understanding the Sampled Output of a DDS Device

An understanding of sampling theory is necessary when analyzing the sampled

output of a DDS based signal synthesis solution. The Nyquist Theorem dictates that there
is a minimum of two samples per cycle required to reconstruct the desired output
waveform. Images responses are created in the sampled output spectrum at fclock ± fout.
The 1st image response occurs at fclock - fout . The 2nd , 3rd , 4th , and 5th images appear
at fclock + fout, 2fclock – fout, 2fclock + fout, 3fclock – fout respectively. In the case of
the fOUT frequency exceeding the fCLOCK frequency, the 1st image response will
appear within the Nyquist bandwidth (DC - ½ fclock) as an aliased image. The aliased
image cannot be filtered from the output with the traditional Nyquist anti-aliasing filter.
In typical DDS applications, a lowpass filter is utilized to suppress the effects of the
image responses in the output spectrum. In order to keep the cutoff requirements on the
lowpass filter reasonable, it is an accepted rule to limit the fout bandwidth to
approximately 40% of the fclock frequency.

3.7. Jitter and Phase Noise Considerations in a DDS System

The maximum achievable spectral purity of a synthesized sine wave is ultimately

related to the purity of the system clock used to drive the DDS. This is due to the fact that
in a sampled system the time interval between samples is expected to be constant.
Practical limitations, however, make perfectly uniform sampling intervals as
impossibility. There is always some variability in the time between samples leading to
deviations from the desired sampling interval. These deviations are referred to as timing
jitter. There are two primary mechanisms that cause jitter the system clock. The first is
thermal noise and the second is coupling noise. Thermal noise is produced from the
random motion of electrons in electric circuits.
Any device possessing electrical resistance serves as a source for thermal noise.
Since thermal noise is random, its frequency spectrum in infinite. In fact, in any given
bandwidth, the amount of thermal noise power produced by a given resistance is
constant. This fact leads to an expression for the noise voltage, Vnoise, produced by a
resistance, R, in a bandwidth, B. It is given by the equation:

Vnoise = √(4kTRB) (3.16)

Where Vnoise is the RMS (root-mean-square) voltage,

k is Boltzmann’s constant (1.38x10-23 Joules/K),
T is absolute temperature in degrees Kelvin (K),
R is the resistance in ohms, and B is the bandwidth in hertz.

The implication here is that whatever circuit is used to generate the system clock
it will always exhibit some finite amount of timing jitter due to thermal noise. Thus,
thermal noise is the limiting factor when it comes to minimizing timing jitter.
The second source of timing jitter is coupled noise. Coupled noise can be in the form of
locally coupled noise caused by crosstalk and/or ground loops within or adjacent to the
immediate area of the circuit. It can also be introduced from sources far removed from
the circuit. Interference that is coupled into the circuit from the surrounding environment
is known as EMI (electromagnetic interference). Sources of EMI may include nearby
power lines, radio and TV transmitters, and electric motors, just to name a few.

3.8. Direct Clocking of a DDS

The output signal quality of a direct digital synthesizer is dependent upon the
signal quality of the reference clock that is driving the DDS. Important quality aspects of
the clock source, such as frequency stability (in PPM), edge jitter (in ps or ns), and phase
noise (in dBc/Hz) will be reflected in the DDS output. One quality, phase noise, is
actually reduced according to 20 LOG (Fout/Fclk).
Reference clock edge jitter has nothing to do with the accuracy of the phase increment
steps taken by the phase accumulator. These step sizes are fixed by the frequency
“tuning” word and are mathematically manipulated with excellent precision regardless of
the quality of the clock. In order for the digital phase step to be properly positioned in the
analog domain, two criteria must be met:
• Appropriate amplitude (this is the DAC’s job)
• Appropriate time (the clock’s job)

3.9. DDS as a Clock Generator

A clock generator should produce a precisely timed logic pulse train with very low
edge jitter and a fixed duty cycle. The logic output levels should be compatible with the
devices that it will be “clocking”. Precise timing implies a very high-Q oscillator; low
edge jitter implies high noise immunity. While these attributes are relatively easy to
accommodate for a single frequency, for example, a crystal clock oscillator. But it is hard
for a designer to accommodate the need for multiple clock frequencies that need to be
changed frequently or rapidly and that have no integer relationship with each other. This
is where the DDS shines! With a single precise pulse train that times the assembly of new
sine wave samples, the DDS can output 2N-1 discrete frequencies (where N is the DDS
resolution in bits). These frequencies range from dc to one-half the input clock frequency
at intervals of 1/2N.The DDS output is a sampled sine wave containing many extraneous
frequency components .The amount of jitter resulting from an unfiltered sampled sine
wave is equal to 1 input clock cycle.

3.10. DDS vs Standard PLL

PLL (phase-locked loop) frequency synthesizers are long-time favorites with designers
who need stable, programmable high & low frequencies, and high quality signal sources
or clocks. They are well understood, widely available, and inexpensive. But DDS has
some additional features compared to a PLL.

Extremely fast frequency changes make a DDS thousands of times more agile than a
PLL. This makes DDS a natural choice for frequency- hopping and spread-spectrum.
• Frequency resolution is extraordinary! Up to one-millionth of a Hertz
• Fundamental output frequency span > 40 octaves (.000001 Hz to 150 MHz)
• Effortless ultra high-speed digital phase modulation (PSK) and FSK
• Perfect, exactly repeatable synchronization of multiple DDS’s (allowing quadrature
and other phase offset relationships to be easily accomplished)

Applications requiring any of the above traits should evaluate DDS as a possible
solution. (As an example, consider dielectrophoresis: This phenomenon is utilized in
micro-biology studies to separate, move and rotate individual cells or bacteria in a
polarized medium using non-uniform traveling fields, typically under a microscope. The
traveling waves are emitted from microelectrodes that are excited by synchronized
signals from two DDS's. Rotation and movement of particles is accomplished by
connecting the synchronized signals of relative differing phases to successive electrodes
(0o, 90 o , 180o, 270 o, etc.), which in turn generate the traveling field in which the
particles move. Differing particles are affected differently by various wavelength signals,
and as such, it is desirable to generate signals over a wide frequency range).

DDS, by virtue of its extremely wide output frequency span, phase offset
capability and precise synchronization, is an ideal vehicle to generate the synchronized
signals from 1 kHz to 50 MHz typically used in this technique. One major difference
between a PLL and a DDS is the PLL’s ability to lock its output to the input phase of a
reference clock. A standard PLL can easily lock its VCO to a 10 MHz input signal and
provided a phase locked 20 MHz output signal. The DDS can get extremely close to the
20 MHz output frequencies but requires an internal clock speed that is at least twice that
of the output frequency.
 CHAPTER 4

PULSE COMPRESSION

4.1. Introduction

Pulse compression is a signal processing technique designed to maximize the

sensitivity and range resolution of radar systems. The sensitivity of radar depends on the
energy transmitted in the radar pulses. This can be expressed in terms of the average
transmitted power- that is, the peak power multiplied by the transmitter duty cycle.
Although the peak transmitter power may be as high as several hundred kilowatts, since
most radars transmit very short pulses (typically a couple of microseconds long), the
average transmitted power may be much less than 1% of this value. Clearly this is not an
efficient use of the available transmitter power.

Transmitting longer pulses improves the radar's sensitivity by increasing the

average transmitted power. However, simply lengthening the radar pulse has the effect of
degrading the range resolution of the radar, since the radio pulse is just spread over a
larger distance. A technique is needed for increasing the average power without
compromising resolution.

Range resolution is the ability to detect two targets in space that are close to each
other. If the transmitted pulse is wide enough to simultaneously dwell on both targets, the
returned signal will appear to the radar as single target. The key to solving this problem is
the realization that the range resolution of a radar does not necessarily depend on the
duration of the transmitted pulse; in fact, it depends on the bandwidth of the pulse. For a
simple rectangular pulse, the bandwidth is just 1/T, where T is the pulse duration.
However, by manipulating the amplitude and/or phase within the pulse, its bandwidth can
be altered without changing its duration. In other words, the radar resolution can be
changed independently of the average transmitted power. This manipulation of the
transmitted pulse is known as pulse coding, or in some cases, pulse modulation.
 Long pulse Figure 4.11: Range resolution Short pulse

Pulse compression refers to a family of techniques used to increase the bandwidth of

radar pulses. In the radar receiver, these echo pulses are `compressed' in the time domain,
resulting in a range resolution which is finer than that associated with an uncoded pulse.
Many methods exist to achieve this, including binary phase coding, polyphase coding,
frequency modulation, and frequency stepping. Methods involving changing the
frequency of the transmitted pulse are often referred to as `chirp' pulses, in analogy to the
sound of a frequency-modulated audio signal.

The echo received from two targets which are close to each other is shown in
Graph(3).The pulse compressed output is shown in Graph (4) ,in which two peaks are
recognized. Thus we infer that there are two targets in space close to each other.

The main disadvantage of pulse compression is the appearance of range side lobes
around the main signal peak, which have the effect of spreading out echoes from
surrounding range gates, and introducing range ambiguities. This effect can be minimized

by the use of complementary codes, which are carefully chosen pairs of codes whose
range side lobes cancel out under ideal conditions.
4.2 Concept of operation

The pulse compression is a practical implementation of a matched-filter system.

The coded signal may be represented either as a frequency response H(ω) or as an
impulse time response h(t) of a coding filter. The coded signal is obtained by exciting the
coding filter H(ω) with a unit impulse.
The received signal is fed to the matched filter, whose frequency response is the complex
conjugate H*(ω) of the coding filter. The output of the matched-filter section is the
compressed pulse, which is given by the inverse Fourier transform of the product of the
signal spectrum H(ω) and the matched-filter response H*(ω).

Figure 4.12: Pulse compression

With chirp, the radio frequency of each transmitted pulse is increased at a

constant rate throughout its length. So every echo naturally has the same linear increase
in frequency. The received echoes are passed through the filter. The echo is visualized as
consisting of a number of segments of equal length and progressively higher frequency.
The first segment having the lowest frequency takes longest to get through the filter.
The second segment takes less time than the first, the third less time than then second etc.
The increments of frequency are such that the difference in transit time for successive
segments just equals their width. If the segments are 0.1 microsecond wide, the first
segment takes 0.1 microsecond longer to go through than the second, it, in turn takes 0.1
microsecond longer to go through than the third, etc., as a result in passing through the
filter, the second segment catches up with the first, the third segment catches up with the
second, the fourth catches up with the third and so on. All segments thus combine and
emerge from the filter at one time. The output pulse is only a fraction of the width of the
received echo, yet has many times its peak power.

Figure 4.13: LFM pulse as segments of progressively higher frequency.

4.3 Types of pulse compression

4.3.1 Analog Pulse Compression

4.3.1.1 Correlation Processor

Pulse Compression is accomplished by adding frequency modulation to a long pulse

at transmission, and by using a matched filter receiver in order to compress the received
signal. As an example, using LFM within a rectangular pulse compresses the matched
filter output by a factor ξ = Bτ`, which is directly proportional to the pulse width and
bandwidth. Thus, by using long pulses and wideband LFM modulation we can achieve
large compression ratios. This form of pulse compression is known as “correlation
processing”.

4.3.1.2 Stretch Processor

 Stretch processing, also known as “active correlation”, is normally used to process
extremely high bandwidth LFM waveforms. This processing technique consists of the
following steps: First, the radar returns are mixed with a replica (reference signal) of the
transmitted waveform. Low Pass Filtering (LPF) and coherent detection follow this.
Next, Analog to Digital (A/D) conversion is performed; and finally, a bank of Narrow
Band Filters (NBFs) is used in order to extract the tones that are proportional to target
range, since stretch processing effectively converts time delay into frequency. All returns
from the same range bin produce the same constant frequency.

The reference signal is a LFM waveform that has the same LFM slope as the
transmitted LFM signal. It exists over the duration of the radar “receive-window”, which
is computed from the difference between the radar maximum and minimum range.

4.3.2 Digital Pulse Compression

There are three main digital pulse compression techniques namely frequency codes,
binary phase codes, and poly-phase codes. The pulse compression is based on the
autocorrelation function since in the absence of noise; the output of the matched filter is
proportional to the code autocorrelation. Given the autocorrelation function of a certain
code, the main lobe width (compressed pulse width) and the side lobe levels are the two
factors that need to be considered in order to evaluate the code’s pulse compression
characteristics.

4.3.2.1 Frequency coding (Costas Codes)

Construction of Costas Codes can be understood from the construction process of

Stepped Frequency Waveforms (SFW). In SFW, a relatively long pulse of length τ` is
divided into N sub pulses, each of width τ1 (τ` = Nτ1). Each group of N sub pulses is
called a burst. Within each burst the frequency is increased by ∆f from one sub pulse to
the next. The overall burst bandwidth is N∆f. More precisely,
τ1 = τ`/N (4.17)
And the frequency for the ith sub pulse is
fi = f0 + i∆f (4.18)
where i varies from 1 to N and f0 is a constant frequency and f0 >> ∆f. It follows that the
time-band width product of this waveform is
∆f τ` = N2 (4.19)
Costas codes are similar to SFW, except that the frequency for the sub pulses are selected
in a random fashion, according to some predetermined rule or logic. For this purpose,
consider the N*N matrix shown below.

0 1 2 3 4 5 6 7 8 9
•
•
•
•
•
•
•
•
•
•

In this case, the rows are indexed from I= 1,2…N and the columns sub pulses and the
columns are used to denote the frequency. A “dot” indicates the frequency value
assigned to the associated sub pulse.
 For a matrix of size N*N, there are a total of N! possible ways of assigning the
“dots”.The sequence of “dots” assignment for which the corresponding ambiguity
function approaches an ideal or a “thumback” response are called Costas codes. A near
thumback response was obtained by Costas using the following logic: only one frequency
per time slot (row) and per frequency slot (column). Therefore for an N*N matrix the
number of possible Costas codes is drastically less than N!.

Three-dimensional plots for the ambiguity function of Costas signals show the near
thumback response of the ambiguity function. All side lobes, except for few around the
origin, have amplitude 1/N. Few side lobes close to the origin have amplitude 2/N, which
is typical of Costas codes. The compression ratio of a Costas code is approximately N.

4.3.2.2 Binary phase codes (Barker Codes)

In this case, a relatively long pulse of width τ` is divided into N smaller pulses,
each is of width ∆τ = τ`/N. Then, the phase of each sub-pulse is randomly chosen as
either 0 or π radians relative to some CW reference signal. It is customary to characterize
a sub-pulse that has 0 phase (amplitude of +1volt) as either ”1” or “+”. Alternatively, a
sub-pulse with phase equal to π (amplitude of -1volt) is characterized by either “0” of “-”.
The compression ratio associated with binary phase code is equal to ξ = τ`/∆τ, and the
peak value is N times larger than that of the long pulse. The goodness of a compressed
binary phase code waveform depends heavily on the random sequence of the phase for
the individual sub-pulse.

One family of binary phase codes that produce compressed waveforms with constant
side lobe levels equal to unity is the Barker code. A barker code of length n is denoted as
Bn. There are only seven known Barker Codes that share this unique property. Since
there are only seven Barker Codes, they are not used when radar security is an issue.
 + + + - - + +

Figure 4.14: Binary Phase Code of Length 7

In general, the autocorrelation function (which is an approximation for the

matched filter output) for a BN Barker Code will be 2N∆τ wide. The main lobe is 2∆τ
wide; the peak value is equal to N. There are (N-1)/2 side lobes on either side of the
main lobe.

The most side lobe reduction offered by a Barker code is –22.3dB, which may not be
sufficient for the desired radar application. However, Barker codes can be combined to
generate much longer codes. In this case, a Bm code can be used within a Bn code (m
within n) to generate a code of length mn. The compression ratio for the combined Bmn
code is equal to mn.

The main drawback is that such codes are only known for limited values of M, and
finding additional codes with the minimum peak side lobe for any given length involves
an exhaustive search with a size growing exponentially with M.

4.3.2.3 Poly phase codes (Frank Codes)

Codes that use any harmonically related phases based on a certain fundamental phase
increment are called poly-phase codes. This coding technique is demonstrated using
Frank codes. In this case, a single pulse of width τ` is divided into N equal groups; each
group is subsequently divided into other N sub-pulses each of width ∆τ. Therefore, the
total number of sub-pulses within each pulse is N2, and the compression ratio is

ξ = N2 (4.20)

As before, the phase within each sub-pulse is held constant with respect to some CW
reference signal.

A frank code of N2 sub-pulses is referred to as an N-phase Frank code. The first step
in computing a frank code is to divide 360° by N, and define the result as the
fundamental phase increment ∆ϕ. More precisely,

∆ϕ = 360°/N (4.21)

Note that the size of the fundamental phase increment decreases as the number of
groups is increased, and because of phase stability, this may degrade the performance of
very long Frank codes. For example, a 4-phase frank code has N=4; and the fundamental
phase increment is ∆ϕ which is equal to 360°/4 = 90°.

The phase increments within each row represent a stepwise approximation of an up-
chirp LFM waveform. The phase increments for subsequent rows increase linearly
versus time. Thus, the corresponding LFM chirp slopes also increase linearly for
subsequent rows.

The original Frank code has two important properties,

1. The code is perfect (having an ideal periodic correlation function)
2. The aperiodic autocorrelation exhibits relatively low side lobes.

The main drawback is that the phase increments of the Frank code are higher in the
central part of the code and lower close to the code ends. Thus, band limiting results in
amplitude intensifying the ends of the code relative to the code center and thus yields a
decrease in main lobe width and higher side lobes.

4.3.3 Linear Frequency – Modulated Waveforms

The power of the pulse compression concept comes from the waveforms used.
We concentrate on a popular pulse compression waveform called the linear frequency
modulated (or chirp) pulse. The modulation within each pulse (in this case frequency
modulation) is the critical element of the pulse compression waveform. The modulation
provides the basis and power of the compression concept.

There are two ways to represent the pulses in a pulse train,

a) A frequency – modulated pulse over time
b) The frequency of the same pulse over time

Figure 4.15: a) A frequency – modulated pulse over time

b) The frequency of the same pulse over time
 The modulated sinusoidal signal in figure (a) is the pulse, which is characterized by
its pulse width and is called the uncompressed pulse width, T. The figure (b) shows the
frequency change within the pulse as a function of time. The characteristics of interest in
figure (b) are the bandwidth of the modulation within the pulse, B. The bandwidth is
simply the difference between the highest and lowest frequencies within the
uncompressed pulse.

4.3.3.1 Merits of chirp

Linear frequency modulation has the advantage of enabling very large

compression ratios to be achieved. In addition, it is comparatively simple. No matter
when a pulse is received or what its exact frequency is, it will pass through the filter
equally well, and with the same amount of compression.

4.3.4 Pulse compression matched filter

Code : C(k) , k = 0 to (N-1)

Matched filter impulse response: h (n) = C* (N-n)

Matched filter output: y (n) = x (n) * h (n)

↑___ Convolution

Y (ω) = X(ω) • H(ω)

↑___ Multiplication

y (n) = IFFT [X(ω) • H(ω)]

Fast convolution is achieved by computing using Fast Fourier Transform (FFT).

 4.3.4.1 Pulse compression algorithm

Length of pulse compression code = N

1. Choose suitable number B = 2n approximately equal to 2N or more.

2. Perform FFT for the first B range cells.
3. Apply spectral multiplication function on above.
4. Perform IFFT
5. The first (B– N+1) samples are valid and the last (N-1) samples discarded.
6. Now choose the next block of B range cells starting from (B-N+2)
7. Repeat step (2) to step (5) for this block.
8. Continue block by block till all range samples are processed.

4.3.4.2 Pulse compression Implementation

Rcells
0 100 200 300 400 500

256 point FFT

0 255
Spectral Multiplication
Block 1 0 255
256 point IFFT
0 255
Valid Output
0 156

256 point FFT

157 412
Spectral Multiplication
157 412
Block 2
256 point IFFT
157 412
Valid Output
157 313
Zero pad
Block 3
314 500
The 256 point IFFT output is shown in waveform (7) and the 156 valid outputs are shown
in waveform (8).

CHAPTER 5

DOPPLER FILTERING

5.1 Introduction

There are two basic reasons for sensing Doppler frequencies. One is to separate,
resolve returns received simultaneously from different objects. The other is to determine
range rates. Sensing Doppler frequencies and detecting differences between them is done
with a bank of Doppler filters. One of the keys to understanding their operation is a good
understanding of the Doppler effect.

5.2 Doppler effect

The Doppler effect is a shift in the frequency of a wave radiated, received or

reflected from an object. A wave radiated from an object is compressed in the direction of
motion and is spread out in the opposite direction. In both cases, the greater the object’s
speed, the greater the effect will be. Since frequency is inversely proportional to
wavelength, the more compressed the wave is, the higher its frequency is and vice versa.
Therefore, the frequency of the wave is shifted in direct proportion to the object’s
velocity. In the case of a radar, Doppler shifts are produced by the relative motion of the
radar and objects from which the radar’s radio waves are reflected. If the distance
between the radar and the reflecting object is decreasing, the waves are compressed.
Their wavelength is shortened and their frequency is increased. If the distance is
increasing, the effect is just the opposite.

The Doppler shift is given by,

fd = (-2 Vr )/C (5.22)

Where fd = Doppler frequency

Vr = relative radial velocity

C = Speed of light.

5.2.1. Doppler filter bank

A radar can detect the echoes from many different sources simultaneously, and in the
process sort them out on the basis of differences in Doppler frequency. This is done by
applying the received signal to a bank of filters commonly referred to as Doppler filters.

Figure 5.16: Received signals are applied in parallel to a bank of filters.

Each filter is designed to pass a narrow band of frequencies. Ideally it produces an

output only if the frequency of the received signal falls within this band. Actually
because of filter side lobes, it may produce some output for signals whose carrier
frequencies lie outside the band. If the return is to be sorted by both range as well as
Doppler frequency, a separate filter bank must be provided for each range increment.
Moving up the bank from the lower end, each filter is tuned to a progressively higher
frequency. If the output lies in the initial filter banks, then the signal is inferred as the
clutter or noise. The Doppler filter output of single target with clutter is shown in Graph
(5). If the output lies in the successive filter banks, then it is inferred as the target. Based
on where the output lies in the filter bank, the target is decided as the slow moving target
or fast moving target and its range is determined. The Doppler filter output of two targets
with different range and velocity is shown in Graph (6).

To minimize the signal to noise ratio occurring when adjacent filters straddle a
targets frequency, the center frequencies of the filters are spaced so the pass bands
overlap. Thus if a targets Doppler frequency gradually increases, an output is produced,
first, primarily from one filter, next more or less equally from that filter and the next filter
up the line, then primarily from that second filter and so on.

Figure 5.17: Overlapping of passbands

5.3 Bandwidth of the Filters

A narrow band filter achieves its selectivity by integrating the signals applied to it
over a period of time. The width of the band of frequencies passed by the filter depends
primarily upon the length of the integration time; tint .the relationship between the
bandwidth and the tint is learned from the spectrum of a sinusoidal signal of duration τ
(single pulse), which has a sin x/x shape. Each point on this plot corresponds to the
output the signal would produce from a narrow band filter that integrates the signal
throughout its entire duration. The number of pulses that must be integrated to achieve a
given bandwidth is equal to tint times the PRF. Hence the 3-db bandwidth of a filter
equals the PRF divided by the number of pulses integrated.

5.4 Passband of the Filter Bank

If the PRF is greater than the spread between the maximum positive and negative
Doppler frequencies for all significant targets, or if the radar is not pulsed, enough filters
must be included in the bank to bracket the anticipated Doppler frequencies. For example
if the PRF were 180 kilohertz, the maximum anticipated positive Doppler frequency 100
kilohertz, and the maximum anticipated negative Doppler frequency –30 kilohertz, then
the passband of the filter bank would have to be at least 100 + 30 = 130 kilohertz wide to
pass the return from all targets.

On the other hand, if the PRF is less than the anticipated spread in Doppler
frequencies, the passband of the bank should be made no greater than the PRF. The
reason is that the spectral lines of the pulsed signal occur at intervals equal to the PRF,
and it is desirable that any one target appears at only one point in filter banks passband.

Depending on the targets Doppler frequency, the spectral line falling within the
passband in this case may not be the target’s central one (carrier frequency). It may be
one of the lines (side band frequencies) above or below it. Since the lines are
harmonically related, which one it is doesn’t matter. The important point is that for each
target one and only one line falls within the passband. This requirement is satisfied when
the width of the passband equals fr.the Doppler filters making up the bank may be either
analog or digital.
 Figure 5.18: Passband of the filter bank.

For analog filtering, the radar return is translated to a relatively low intermediate
radio frequency. Each filter typically uses one or more quartz crystals. Its response to an
incoming signal is analogous to that of a pendulum’s response to a succession of
impulses. For digital filtering, the IF output of the receiver is translated to video
frequencies by applying it to a pair of synchronous detectors, along with a reference
signal whose frequency corresponds to that of the transmitter.

The outputs of the detectors represent the inphase (I) and quadrature (Q)
components of the return. Quadrature components are needed to preserve the sense of the
Doppler frequencies.

Digital filtering is a clever way of adding up (integrating) successive samples of a

continuous wave so that they produce an appreciable sum only if the waves frequency
lies within a given narrow band.
5.5 Filtering requirement

The FFT is an algorithm, which vastly reduces the amount of processing necessary to
form a bank of digital filters than DFT. Its efficiency is achieved primarily by choosing
the parameters of the bank so that they are harmonically related and consolidating the
formation of the filters in to a single multiple step process. By making the number of
filters, N, equal to a power of two and the number of samples summed equal to N, further
processing is accomplished in log 2 N steps. The basic processing instruction for
performing the individual partial summation consisting of a phase rotation, a complex
addition and a complex subtraction is called the FFT butterfly. The Doppler filtering is
implemented in Xilinx using 16 point FFT. The output is shown in waveform (9).

CHAPTER 6

SYSTEM REQUIREMENTS

6.1 The Hardware and Software requirements

6.1.1 Hardware

• Pentium or higher processor

• SVGA Monitor
• 128 MB RAM
• Mouse or a pointing device
• Virtex-II FPGA Kit

6.1.2 Software

• Operating system – Any OS

• Programming – MATLAB, VHDL.
• Synthesis Tool – ModelSim XE, XILINX

6.2 MATLAB

MATLAB is a high-performance language for technical computing. It integrates

computation, visualization, and programming in an easy-to-use environment where
problems and solutions are expressed in familiar mathematical notation.

MATLAB is an interactive system whose basic data element is an array that does
not require dimensioning. This allows you to solve many technical computing problems,
especially those with matrix and vector formulations, in a fraction of the time it would
take to write a program in a scalar non-interactive language such as C or Fortran.
MATLAB provides a full programming language that enables us to write a series of
MATLAB statements into a file and then execute them with a single command. In this
project, the chirp signal generation, generation of low pass filter coefficients, pulse
compression and Doppler filtering is done using MATLAB. The output graphs are shown
in the result.

6.3 XILINX

The EDA Tool used for implementing this project is XILINX. As the size and
complexity of digital systems increases, more and more computer aided design tools are
being introduced into the hardware design process. The early paper and pencil design
methods have given way to sophisticated design entry, verification and automatic
hardware generation tools such as XILINX, Altera. In the design process much of the
work of transforming a design from one form to another is tedious and repetitive. Design
automation tools can help the designer with design entry, hardware generation, and
verification and design management. And Modelsim XE was used for the simulation,
Modeling and Testing.
 Initially the basic components such as counters, D Flip-flops, ROM, multipliers,
shift registers are designed using VHDL. Then the Phase accumulator block of DDS is
built using the modulus-n counter, the sine/cosine lookup table is implemented using the
ROM. The implementation of DDS is done using VHDL in XILINX software.

The FIR filter is designed using VHDL in XILINX and is loaded with the
Lowpass filter coefficients generated using MATLAB, which obtains the characteristics
of a Lowpass filter. Thus the filter now acts as a Lowpass FIR filter. The multiplier
block, decimator block of DAPD is also implemented in XILINX and simulated and
tested using Modelsim XE.

The Pulse compression and Doppler filtering modules are also implemented in
XILINX and simulated and tested using Modelsim XE. The implementation of pulse
compression is done using 256-point FFT core and the test bench is written for the core to
work. The Doppler filtering is designed using the16-point FFT core and the test bench is
written for the same to work and simulated and tested using modelsim XE. The output
waveforms are shown in the result.

CHAPTER 7

SYSTEM TESTING

The DDS, DAPD, Pulse Compression and Doppler Filtering are tested in real time
using the Virtex-II FPGA kit. The testing is done by programming the flash PROM
available on the development kit with the design bit file generated by the EDA tool.

The Virtex-II Platform FPGA solution is the result of the largest silicon and software
R&D effort in the history of programmable logic, with the goal of revolutionizing the
 design of complex single-chip sub-systems in terms of engineering productivity, silicon
efficiency, and system flexibility

The Virtex-II Platform FPGA family is a complete programmable solution that allows
digital system designers to rapidly implement a single-chip solution with density up to 10
million system gates, in weeks rather than months or years. The inherent flexibility of
Xilinx FPGA devices allows unlimited design changes throughout the development and
production phases of the system, with important benefits in improved productivity,
reduced design risk, and higher system flexibility. This further accelerates the industry
from custom ASICs to FPGAs in fields such as optical networking systems; gigabit
routers, wireless cellular base stations, modem arrays, and professional video broadcast
systems.The Virtex-II solution combines the most flexible FPGA architecture, advanced
process technology, powerful software synthesis technology, and robust IP library, to
provide the most complete system integration solution today. It incorporates all required
power supplies, various configuration mode pin interfaces, user IOs on separate
connectors. It also includes detachable features like digital inputs, outputs on LEDs,
matrix keyboard, 7 segment displays, Alpha-numeric LCD Display, RS232 interface,
making Virtex-II the ideal solution for tomorrow’s high-performance system designs.

8. RESULTS
Graph 1 : LFM Chirp signal
Graph 2: Filter output
Graph 3: Signal return from two targets

Graph 4 : Pulse Compression output: Two targets

 Graph 5: Doppler Filter Output: Single target with clutter

Graph 6: Doppler Filter Output: Two target with different range and velocity
Waveform 1: Direct Digital Synthesis (DDS) waveform

Waveform 2: Multiplier (Signal and DDS output multiplied) output

Waveform 3: Scaled Multiplier output

Waveform 4: Lowpass FIR filter output

 Waveform 5: Decimated output

Waveform 6: DAPD final output

Waveform 7: FFT output (256 values)
Waveform 8: Pulse compression output (156 valid outputs)
Waveform 9: Doppler filtering output
 CHAPTER 9

CONCLUSION

“Just as all things come to an end, so does this project, but the journey towards the
destination has all through, been an invaluable learning experience”. The primary
objective was to develop a FPGA based signal processor for multifunction RADAR. The
Design and Analysis of this project was done using various steps. First the VHDL code
was synthesized using the XILINX software. Next it is simulated using the Modelsim
XE.The real time testing was done using the Virtex-II FPGA kit.

Initially the Chirp signal was generated using MATLAB and the inphase (I) and
quadrature phase (Q) components were generated using DAPD or Digital I/Q
Demodulation. The required frequency to generate these components was generated
using the Direct Digital Synthesis (DDS).The implementation of DAPD and DDS was
done in XILINX and tested in Virtex-II FPGA kit.

The Pulse Compression, which is done to discriminate two targets that are closer
in range and Doppler Filtering, which is used to extract the target information, such as
range, velocity was implemented in XILINX and simulated using Modelsim XE.The
performance was tested using the Virtex-II FPGA kit.

CHAPTER 10

FUTURE ENHANCEMENTS

“The scope of betterment knows no bounds, and is based on the developer’s ability to
think in an innovative way”. In this project two signal processor functions, pulse
compression and Doppler filtering is implemented. The present work can be implemented
in the following areas

• Magnitude and Logarithmic computation

 • Constant False Alarm Rate (CFAR)
• Side Lobe Blanking (SLB)

Magnitude and Logarithmic computation

This is to normalize the Doppler filter output, which will avoid the signal processor
saturation.

Constant False Alarm Rate (CFAR):

This algorithm is to maintain the false alarm rate. Cell Averaging CFAR or Greatest
of CFAR etc can be used to maintain the false alarm rate. These algorithms are also used
to avoid clutter (which is the unwanted targets).

Side Lobe Blanking (SLB):

This is to avoid the jamming. Jamming is misguiding the radar by sending more
power through the side lobe. Here if signal processor finds any detection in side lobe the
detection will be rejected. We can put a null in that direction through side lobe
cancellation technique.

All the FPGA based signal processing algorithms can be used in STAP. STAP is a
technique used to avoid jamming and by using this technique we can customize the beam.
The beam can be put at any point, as we require.
 REFERENCES

1. A Technical Tutorial on Digital Signal Synthesis, Analog Devices, Inc

2. Bhasker.J : VHDL PRIMER, Pearson Education Asia, 3rd Edition
3. George W. Stimson :Introduction to Airborne RADAR, SciTech publishing,Inc.,
Second Edition
4. Merrill l. Skolnik : Introduction to RADAR systems, TATA McGraw-HILL,3rd
Edition
5. Perry Douglas.L : VHDL,Pearson Education Asia, 3rd Edition.
6. www.dataweek.com
7. www.cdtechware.com