Chapter 1

Introduction
Wave types
A. Acoustic or sound waves:
These are mechanical vibrations of the particles that make up the medium. They are
longitudinal waves and require a material medium to propagate.
B. Electromagnetic waves:
The combination of an electric field and a magnetic field, which are perpendicular to
each other and vary over time, generates an electromagnetic wave. This wave travels in
a direction perpendicular to the electric field ( → ) and the magnetic field ( → ).
E H

They consist of photon flux.

They exhibit the same optical properties, such as reflection and refraction.

Propagation Direction

RADAR = Radio Detection and Ranging

SONAR = Sound Navigation and Ranging
Both systems operate based on the same principle of functioning:
An electromagnetic wave (RADAR) or an acoustic wave (SONAR) is emitted.
This wave is reflected by the target and then captured by the receiving system.

Key Properties:
1 / 27
RADAR
Wavelength ( ): λ 1 cm < λ < 1 m

Frequency ( ):
f 300 MHz < f < 30 GHz

Wave speed ( ): c 3 × 10
8
m/s

SONAR
Frequency ( ):
f 1 kHz < f < 1 MHz

Wavelength ( ): Depends on the medium.

λ

In water ( ):
v = 1500 m/s

1.5 mm ≤ λ ≤ 1.5 m

Applications
Surveillance
The purpose is to intercept targets, such as airplanes and missiles.
Fire Control
Primarily in the military domain: ground-to-air or air-to-air operations (e.g., combat
aircraft).

Radars for Aircraft and Satellites

Examples:
SAR (Synthetic Aperture Radar) for imaging.
Satellites for espionage.

Instrumentation Radars
Landing Radars: Onboard aircraft for assisting with landings.
Weather Radars: For meteorological observations.
Missile Guidance Radars: For controlling the trajectory of missiles (e.g., ARIANE or
other projectiles).

Civilian Applications
Road Traffic Radars: For monitoring and controlling vehicle speeds.
Air Traffic Control Radars: To ensure safe navigation of aircraft.

Astronomical Radars
Used in space exploration.
2 / 27
 Observing the movement of distant planets such as Venus and Mercury.

Principle
Definitions
Radar
Purpose: Detect the presence of targets and calculate specific parameters, such
as:
Distance
Position (3D coordinates)
Velocity

Key Components:
1. Emitter
2. Receiver
3. One or more antennas
4. Processing system (for data analysis and exploitation)

Noise B(t)

Information S(t)

Transmitter Receiver

S(t) + B(t)

Processing System

Main Types of Radars

Monostatic Radars

3 / 27
 Obstacle

Transmitter
Signal

Receiver

Obstacle

Transmitter
The signal reflects from the object
and some of it reaches the antenna

Receiver

The transmitter and receiver are co-located, often sharing the same antenna.
The radar sends out a signal, and the same system detects the echoes reflected
from the target.
Commonly used in conventional radar systems due to its simplicity and reliability.

Bistatic Radars

4 / 27
 Obstacle

Transmitter

Receiver

The transmitter and receiver are at different locations.

The radar relies on the transmitted signal reflecting off the target and being picked
up by the distant receiver.
Useful for stealth detection and applications requiring spatial separation.

Multistatic Radar

5 / 27
 Obstacle

Transmitter

Processing Center

Receiver 1 Receiver 2 Receiver n

Involves multiple receivers and/or transmitters.

Signals are transmitted from one or more locations and received at multiple sites,
improving detection reliability and target tracking accuracy.
Often used for advanced surveillance systems and in cluttered environments.

Different Perspectives of Radar Operators

Basic Concepts
S/B = Signal-to-Noise Ratio (SNR):
2
⟨S (t)⟩
τ
S/B| dB = 10 log 10 ( )
2
⟨B (t)⟩ τ

Or:
∫ γ S (f ) df
Δf
S/B| dB = 10 log 10 ( )
∫ γ B (f ) df
Δf

Where:
⟨X(t)⟩ τ is the average value of over time
X(t) τ

τ is the duration of the signal, given by:

1
τ =
Δf

Δf is the bandwidth of the signal

6 / 27
 γ S (f ) is the spectral density of S(t) (it is the distribution of power over frequency)

Characteristics of Radar Echoes

The bandwidth of : Determines the resolution of the radar.
S(t)

Very low energy signals (low SNR): Challenging to detect.

Complex signals: Contain information about the target (e.g., its shape or structure).

Pure Detection and/or Target Recognition

Two Key Aspects
1. Pure Detection:
Focuses solely on identifying the presence of a signal.
2. Target Recognition:
Involves identifying specific characteristics or details about the target.

Pure Detection

S(t) + B(t)
Receiver

Signal Model:
The received signal is modeled as , where:
S(t) + B(t)

: The actual signal to detect

S(t)

: Gaussian noise (
B(t) ) with the characteristics:
B(t) ∼ N (0, σ )
2

Mean = 0
Variance = σ
2

The probability density function of the noise is: B(t)

2
1 −
B

2
P B (B) = ⋅ e 2σ

√ 2πσ

Detection by Thresholding
A threshold is set to determine whether a signal is present ( ) or absent (only
S(t)

B(t)).
Performance depends on:
7 / 27
 The strength of relative to
S(t) B(t) .
The chosen threshold value.
S(t) + B(t)

1
Threshold h

Probability of False Alarm: It is the likelihood of mistakenly identifying noise as a valid

signal, this means that it is the probability of the noise to exceed the threshold.
Therefore the probability of false alarm is:
+∞

P f a = P τ (B(t) > h) = ∫ P B (v) dv

h

Probability of Detection: it is the likelihood of correctly detecting the presence of a

signal, this means that it is the probability of the noisy signal crossing the threshold.
Therefore, the probability is:
∞

P d = P τ (S(t) + B(t) > h) = P τ (B(t) > h − S(t)) = ∫ P B (v) dv

h−S(t)

h - S(t) h

If we increase the threshold h, will decrease since fewer noise fluctuations will
Pf a

exceed a higher threshold. Also will decrease because of that since even valid signals
Pd

may fail to reach the higher threshold, leading to missed detections. But if we decrease
the threshold h, the opposite will happen, both and will increase. (since the lower
Pf a Pd

8 / 27
threshold makes it easier for the noise fluctuations to reach it. Also the lower threshold
will detect smaller signals)

Noise B(t) S/B

Signal S(t)

Threshold h

The above diagram shows how Pd and Pf a depend on the threshold and the S/B ratio.

9 / 27
This plot shows the relationship between Pd (y-axis) and S/B (x-axis) for different
values of :Pfa

Each curve corresponds to a different false alarm probability. Lower generally

Pfa

shifts the curve downwards.

For a fixed , as S/B increases (stronger signal relative to noise), the probability of
Pfa

detection also increases.

Pd

For a given , increasing the threshold reduces , but higher S/B is required to
Pd Pfa

maintain . Pd

Target Recognition
And that is done by:
1. Building a Database: Involves studying a set of obstacles and storing relevant data
to form a reference database.
2. Signal Comparison: The received signal is compared against the reference
database to identify the target.

Extracting key parameters

To build the database, we need to extract key parameters such as surface area, volume,
permittivity, conductivity, etc...). These parameters are often referred to as "signatures"
( ) and are usually extracted by studying the interaction between the
P1 , P2 , … , Pn

electromagnetic wave and the obstacle. For example if n=3, then we have three
parameters:

10 / 27
Processing the retrieved data

Signal reflected
by the obstacle
Data Processing +
Measurement of signatures Decision
Classification

Summary:
1. We study of the wave/obstacle interaction to find a model of the diffracted field.
2. From the diffracted field to classify the object

Basic Notions about Radar

11 / 27
 Obstacle

d
Transmitter

Receiver

1. Detect the target: Using the parameters Pfa , Pd and S/B.

2. Calculate the distance d:
2d c ⋅ Δt
Δt = ⟹ d =
c 2

where:
Δt is the difference of time between sending the signal and receiving it reflected by
the obstacle.
cis the speed of light
d is the distance between the antenna and the obstacle
3. The Obstacle's Direction:

12 / 27
Parabolic antennas are generally used:

-3dB

This is the radiation pattern of a parabolic antenna. is the opening angle, is the
θo l

diameter of the antenna and G maxis the maximum gain. We can approximate the
opening angle:
λ
θ o ≈ 70 ⋅
ℓ

where:
λ is in meters[m]

l is in meters [m]

θo is in degrees ∘
[ ]

The other type of antennas is the cylindrical reflector antenna:

The horizontal and vertical angles can be approximated:

λ
θ v ≈ 70 ⋅
ℓ

λ
θ h ≈ 70 ⋅
h

13 / 27
 4. The speed of an obstacle:

λf o
vr =
2

where fo is the doppler shift and is the speed of the obstacle.

vr

Pulse Radar

Sinusoidal Generator

Pulse Generator
Receiver

Important Quantities
Peak Power: It is the power emitted during , denoted as .
τ Pc

The Energy of a pulse:

W = Pc ⋅ τ

Average Power Emitted:

W Pc ⋅ τ
P av = =
Tr Tr

Wavelength ( ): It depends on:

λ ↔ f

Propagation medium
Technical characteristics
Receiver Sensitivity: This is the minimum power detectable by the radar. Therefore
it is the minimum required for the radar to detect a signal. (denoted as
S/B )
(S/B) o

14 / 27
 G (Power Gain) S/B

Δf (pass-band)

It is the power of the noise at the output of the amplifier

P Bs :

It is the power of the noise before entering the amplifier

P Be :

The relation between the noise power at the input and output is:
P Bs = P Be ⋅ G ⋅ F = kT Δf ⋅ G ⋅ F

Where:
It is the thermal power of the noise (at the input)
P Be :

Boltzmann's Constant =
k : 1.38 × 10
−23
J/K

Temperature of the noise in Kelvin (

T : )
= 273 + T
∘
C

Noise factor
F :

It is the signal's power at the output of the amplifier

P Ss :

It is the signal's power before entering the amplifier

P Se :

The relation between them is:

P Ss = G ⋅ P Se

Then the signal to noise ratio is:

G ⋅ P Se
(S/B) =
kT Δf ⋅ G ⋅ F

P Se
=
kT Δf ⋅ F

Therefore it is necessary for the signal's power at the output of the amplifier to be:
P Se ≥ (S/B) ⋅ kT Δf ⋅ F
o

15 / 27
 Thus, the sensitivity is:
S = (S/B) ⋅ kT Δf ⋅ F
o

Blind Distances
At the beginning, there is a protection mechanism for the receiver during the
transmission of emitted pulses.
If Δt (round-trip time: radar ↔ obstacle) ≤ τ : Reception is impossible.
When the time has elapsed, the distance can be determined as follows:
τ

2d cτ
≤ τ ⟹ d ≤
c 2

This is the blind distance—the distance below which the radar cannot detect objects.
Practically, this distance is even larger due to the protection mechanisms for the
receiver:

Gas Ionization Control

Receiver

gas

16 / 27
 Conductivity of gas

10%
t
deoin

deoin

deionization time

Therefore, the blind distance is:

c
d ≤ (τ + Δt deion )
2

Ambiguous Distances

d max

Tr

17 / 27
 2d max
= Tr
c

Therefore, the ambiguous distance is:

c ⋅ Tr
⟹ d ≥
2

If we assume an isotropic antenna, we can draw its blind spots (in red) as following:

North

Ambiguous Distances

West East

South Blind Distances

Resolution Cell

Gmax

Gmax - 3dB

Echoes are merged

18 / 27
 Emitted Signal

Reflected Signal by obstacle 1

Reflected Signal by obstacle 2

Echoes are not merged if:

δt > τ

And are merged if:

δt ≤ τ

2
⟹ Δt 1 − Δt 2 = (d 2 − d 1 ) ≤ τ
c

cτ
⟹ d2 − d1 ≤
2

This is the resolution cell, which is the zone where there will be confusion of echoes.

Radar Equations
Gain G(θ, ϕ)

It represents an increase in radiated power in a specific direction ( ) compared to

θ, ϕ

isotropic radiation. A power source emitting equally in all directions is called an

Pt

omnidirectional antenna, radiating across the entire space:

19 / 27
 Pt

The radiation intensity in a given direction ( θ, ϕ ) is given by:

Pt
ψ(θ, ϕ) = [W/sr] (Watt/steradian)
4π

Where Pt is the transmitted power.

In real case scenario:

Therefore:
ψ(θ, ϕ)
G t (θ, ϕ) =
Pt

4π

E.I.R.P. (Equivalent Isotropic Radiated Power)

20 / 27
 Pt

The E.I.R.P. is defined as:

E.I.R.P. = P t ⋅ G t (θ, ϕ)

For any real-world problem, the E.I.R.P. simplifies analysis by considering an equivalent
isotropic source (easier to study) that would radiate the same power density as the real
source in a specific direction ( )
θ, ϕ

Power Density at Distance d

The emitted power density at a distance is given by:

d

P t ⋅ G t (θ, ϕ) 2
Pi = [W/m ] (1)
2
4πd

Pt

Radar Cross Section (RCS)

21 / 27
 Pi Obstacle

Pt d

Pr

The radar cross section (denoted as or effective surface area) represents the
σe

equivalent area of the target that reflects the radar signal.

The received power density by the radar from the target is modeled as if the target
reflects isotropically (the power reflected by the target is ): Pi ⋅ σe

Pi σe
P inc = (2)
2
4πd

Where Pi is the power emitted by the radar and is incident to the object

Effective Surface of the Antenna

Pr Pinc

Definition
The effective area of an antenna in the direction ( ), denoted as θ, ϕ , representsA e (θ, ϕ)

the equivalent area that captures power from an incident electromagnetic wave in that
direction. That area is measured in squared meters and is related to the incident
[m ]
2

power density on the antenna and received power by the antenna:

2
P r = P inc ⋅ A e (θ, ϕ) [m ] (3)

Where:
Pr : Received power by the antenna
P inc : Incident power density on the antenna [W/m ]
2

22 / 27
Antenna Reciprocity (Carson's Theorem)
The reciprocity theorem states that an antenna exhibits the same directional
properties for both transmission and reception.
The gain is identical for emission and reception.
G(θ, ϕ)

Relationship Between Gain and Effective Area

It has been demonstrated that the effective area A e (θ, ϕ) of an antenna and its gain
G(θ, ϕ) are related by the following formula:
G r (θ, ϕ) G r max 4π
= =
A e (θ, ϕ) A e max λ2

2
λ
⟹ A e (θ, ϕ) = G r (θ, ϕ) ⋅ (4)
4π

Received Power
If we substitute equations (2) and (4) in equation (3) :
2
Pi σe λ
⟹ Pr = ⋅ G r (θ, ϕ) ⋅ (5)
2
4πd 4π

Then we substitute equation (1) in equation (5) :

2
P t ⋅ G t (θ, ϕ)σ e ⋅ G r (θ, ϕ)λ
⟹ Pr =
(4π) 3 d 4

Where:
Pt ​: Transmitted power.
G t (θ, ϕ): Transmit gain of the antenna.
G r (θ, ϕ): Receive gain of the antenna.
σe : Radar cross section of the target.
d: Distance between the radar and the target.
λ: Wavelength of the signal

Calculate the SNR

23 / 27
 Transmitter

PBe PBs
S/B
PSe PSs

We proved before that:

P Se
S/B =
kT Δf ⋅ F

And since P Se actually is Pr we can say that:

2
P t ⋅ G t (θ, ϕ)σ e ⋅ G r (θ, ϕ)λ 1
S/B = ⋅
3 4
(4π) d kT Δf ⋅ F

2
P t ⋅ G t (θ, ϕ)σ e ⋅ G r (θ, ϕ)λ
⟹ S/B = (6)
3 4
(4π) d ⋅ kT Δf ⋅ F

S(t)

Applying Fourier Transform on the radar's signal:

∞
−j 2 πf t
S(f ) = ∫ S(t)e dt
−∞

τ /2
−j 2 πf t
= ∫ S(t)e dt
−τ /2

sin(πf τ )
= τ ⋅
πf τ

24 / 27
 S(f)

Therefore if increases, the noise power increases and if

Δf Δf decreases, the signal's
power decreases. Thus:
1
Δf ≈
τ

Then we'll replace Δf in equation (6) with 1

τ
⋅ L :
2
P t ⋅ G t (θ, ϕ)σ e ⋅ G r (θ, ϕ)λ
⟹ S/B =
1
(4π) 3 d 4 ⋅ kT F ⋅ ⋅ L
τ

Where is the loss factor. We can calculate this loss factor using this formula:
L

L = L ds + L prop + L misc
dB dB dB
dB

Where:
L ds : Loss due to the system.
dB

L prop : Propagation losses.

dB

L misc : Miscellaneous losses.

dB

Atmospheric Loss
Noise for round-trip path:
Contributing Factors:
Standard constitution (mixing of , , etc.). O2 H2

Excitation of molecules.
Local conditions (pressure, temperature).
Absorption.

Radar Range Equation

25 / 27
We proved that:
2
P t ⋅ G t (θ, ϕ)σ e ⋅ G r (θ, ϕ)λ τ
S/B =
(4π) 3 d 4 ⋅ kT F L

And of course the SNR should be equal or greater than the sensibility:
S/B ≥ (S/B)
o

From here we can calculate the range of the radar:

2
4
Pt ⋅ τ ⋅ Gt ⋅ Gr ⋅ λ ⋅ σe
d range = √
3
(4π) ⋅ kT F L ⋅ (S/B) o

Signal-to-Noise Ratio in Decibels

S/B = (P t ⋅ τ ) dB⋅J + (G e ) dB + (G r ) dB + 2(λ) dBm + (σ e ) 2 − 3(4π) dB − 4(d) dBm − (kT F L) dB
dB dBm

Summary Table (Bilan):

Parameter Increases S/B Decreases S/B

Pt ✓
τ ✓
Ge ✓
Gr ✓
λ ✓
L ✓
d ✓

Explanation of Ambiguities and Key Points:

1. Noise and Losses:
The losses include system inefficiencies ( ), propagation losses (
L ds ), and other L prop

factors ( ), such as atmospheric absorption. These are critical for determining

L misc

how much signal power is lost before reaching the target and returning to the
receiver.
2. Radar Cross-Section ( ): σe

This is a measure of how effectively a target reflects radar signals. It depends on

the size, shape, and material properties of the target.
3. Wavelength ( ): λ

A longer wavelength ( ) improves

λ since it is less affected by atmospheric
S/B

attenuation.
26 / 27
4. Decibel Conversion:
Parameters like , , and are often expressed in decibels for ease of calculation
Pt τ Ge

when combining gains and losses in a radar system.

5. Radar Range Equation:
The fourth root dependence on , , and means that even small improvements
Pt τ Ge

in these parameters can significantly increase the radar range.

27 / 27