1

University of Management and Technology Lahore

School of Engineering
Department of Electrical Engineering
Design and simulation of a Transmitter-Receiver loop for Frequency Modulated - Continuous
wave Radar for Advanced Driver Assistance System (ADAS)
Semester: Fall 2019
Course: Communication Systems-EE410
Complex Engineering Problem (CEP), Maximum Marks 15/100
Resource Persons: Jameel Ahmad & Faran Awais Butt
Emails: {jameel.ahmad,faran.butt}@umt.edu.pk
Submission Date: Monday Week 13
————————————————————————————————–

TABLE I
C OURSE L EARNING O UTCOMES (CLO S ).

CLO Description
Use theorems such as Parseval/Rayleigh Energy’s theorem and tools such as Fourier transform to represent and analyze signals
CLO 1
in time and frequency domain as well as understand the characteristics of distortion-less communication channel. (C4)
Identify and differentiate various blocks in Amplitude and Frequency/Phase Modulation/Demodulation.
CLO 2
Analyze the working of AM/FM MODEMS using time and frequency domain analysis. (C4)
CLO 3 Learn and evaluate performance parameters such as Bandwidth, Power and Signal-to-Noise Ratio of communication receivers. (C6)
CLO 4 Design AM and FM transmitter and receiver to achieve the desired performance specifications. (C6)

TABLE II
P ROGRAM L EARNING O UTCOMES (PLO S ) FOR CEP.

PLO Description
PLO 2 Problem Analysis
PLO 3 Design and Development of Solutions

TABLE III
A SSESSMENT RUBRIC AND D ELIVERABLES (M AX . M ARKS :15).

Assessment Criteria CEP Attribute[1] CLO/PLO Excellent Average Adequate

Depth of Analysis
Derivation of Transmitted and received signal in time domain 1/2 1 0.5 0.1
& Interdependence
Plots of Transmitted Chirp in Time and Frequency domain Depth of Analysis 1/2 1 0.5 0.1
Waveform of Input and output of each block of Radar Depth of Analysis 2/2 2 1 0.5
Beat Frequency Spectrum for single and multiple targets Depth of Analysis 3/2 1 0.5 0.3
Link Budget and SNR Calculation (Excel sheet) for 5.8GHz Radar Interdependence 3/2 2 1.5 0.3
Circuit Level Implementation of 2.4 or 5.8 GHz Radar Depth of Knowledge 4/3 3 1 0.5
2-D Range-velocity Graph Depth of Knowledge 4/3 2 1 0.2
Report and Viva Voce - 3 1.5 0.5
 2

Objectives of Complex Engineering Problem

In the modern era, most automotive manufacturers have adopted radar systems not only for premium cars but
for all their vehicles. New security solutions such as Adaptive Cruise Control (ACC) or Automatic Emergency Braking
(AEB) are now possible. To make self-driving feasible, several sensors such as the camera, LIDAR and RADAR are
used. Radar sensors are especially important since, unlike other sensors, they provide exact information about the
distance and speed of neighboring vehicles regardless of the weather conditions. The aforementioned systems are
normally referred to as ADAS (Advance driver-assistance systems) and include a vast range of applications that aim
to help and guide the driver to avoid accidents and dangerous situations: forward collision avoidance, parking sensors
or pedestrian protection systems. Illustration of ADAS system is given in Fig. 1 where vehicles seem connected to
each other through ADAS. In this Complex Engineering Problem (CEP) for EE410:Communications Systems course,
students will go through an in depth analysis of linear FMCW radars by modeling an end-to-end radar system for ADAS
related application. Students will also design an FMCW radar system with triangular chirp pattern. Student will also
propose a model that can be implemented in the local industry. Simulation setup can be done in MATLAB Simulink or
simRF Toolbox or RF Blockset or other software such as Keysight ADS or NI AWR VSS. A Technical Report should
be submitted having 10-20 pages, supported with literature review of 3 IEEE journal/Transactions per group.
Link budget calculations would also be performed. Bonus points will be given to those students who will do
a hardware implementation at circuit level and propose a non-linear chirp pattern design for the proposed
framework with its pros and cons.
Each group should review 3 research papers from IEEE Conferences or journals from IEEE or Elsevier. As a future
work, students can read paper on interference from neighboring vehicles as it is a serious concern [2], [3], [4] in ADAS.

Index Terms
FMCW Radar, ACC, ADAS, Beat Frequency, Cruise Control, Link budget

I. I NTRODUCTION

Fig. 1. Radar based ADAS systems

Frequency Modulated Continuous Wave radar (FMCW radar) is a special type of radar sensor which radiates
continuous transmission power like a simple continuous wave radar (CW Radar). In FMCW radars the frequency
modulation spreads the transmitted energy over a large modulation bandwidth B. Large bandwidth helps to detect
targets with high resolution. The power spectrum is nearly rectangular over the modulation bandwidth. This makes
interception difficult. It is resistant to jamming because it is a deterministic signal and its form is only known to
the user. FMCW (sometimes called LFMCW or Linear FMCW) is the most commonly used scheme in automotive
radar today. In Linear FMCW: TX signal has frequency changing linearly with time (i.e. chirp). In FMCW Radars, the
transmitted frequency is linearly changed during the run time to the target and back to the Radar, and the received
signal is shifted by a time delay to the initial signal. By mixing the current transmitted signal with the reflected signal,
the frequency difference caused by the run time can be defined. Due to the known modulation parameters of the
transmitter, the run time of the signal can be calculated, which is proportional to the distance of the object. For
a precise distance measurement, an extremely highly linear modulation of the transmitted frequency is necessary,
because each non-linearity of the modulation will decrease the accuracy of the FMCW Radar.
An FMCW system is shown in Fig. 2 which consists of a transmitter, a receiver, a mixer and an analog to digital
converter (A/D). An FM signal is transmitted and received through antennas, and the transmitted and received signals
are multiplied in the time domain(mixing) and processed. An FMCW radar measures the range by examining the
 3

beat frequency in the de-chirped signal. To extract this frequency, a de-chirp operation is performed by mixing the
received signal with the transmitted signal. After the mixing, the de-chirped signal contains only individual frequency
components that correspond to the target range. In addition, even though it is possible to extract the Doppler
information from a single sweep, the Doppler shift is often extracted among several sweeps because within one
pulse, the Doppler frequency is indistinguishable from the beat frequency.

Fig. 2. Structure of an FMCW radar system.

In Fig.3, the duration in time T it is half of the period of the low frequency modulation waveform fm . Generally
the duration in time T it is much greater than the return time of the echo signal td . The low frequency modulation
waveform signal can be Triangle, Sawtooth, Sinusoidal, or other periodic shape signal. In practical applications the
frequency of the modulation signal fm could be between 10 Hz and 1 kHz. The basic waveform that the FMCW
radar transmits and receives is the saw tooth waveform. The triangular waveform offers some additional
advantages that students are expected to know. The first sub-figure in Fig.3 shows a triangular chirp pattern.
The received signal is a time-delayed copy of the transmitted signal where the delay, ∆t , is related to the range.
Because the signal is always sweeping through a frequency band, at any moment during the sweep, the frequency
difference,fb , is a constant between the transmitted signal and the received signal. It is usually called the beat
frequencyfb or intermediate frequency IF . Because the sweep is linear, one can derive the time delay from the beat
frequency and then translate the delay to the range.

TABLE IV
FMCW RADAR SYSTEM PARAMETERS .

Parameter Definition Value

f0 Operating Frequency (GHz) 77
fV CO Operating Frequency Range (GHz) 75-80GHz
Chirp specification(B,T) Bandwidth, duration B = 1150M Hz, T = 7.33µsec
Ns Number of Sweeps of Chirp 64
rr range resolution(m) 1
vT Maximum Target Speed (km/h) 120 + sum of last two digits of your student ID

To measure the range and Doppler, an FMCW radar typically performs the following operations:
1) A chirp signal is launched into the free space using a transmit antenna. A chirp signal is an FM-modulated
signal of a known stable frequency whose instantaneous frequency varies linearly over a fixed period of time
(sweep time) by a modulating signal.
2) The Tx antenna with Gain Gt transmits chirp and radiates the signal into space with suitable Transmit power
Pt .
 4

Fig. 3. Chirp signal specification

3) The signal propagates to the target, gets reflected by the target, and travels back to the radar.
4) The receiving antenna with Gain Gr collects the signal with received Power Pr .
5) The received signal is de-chirped using a mixer resulting in Intermediate frequency IF and passed to ADC
and saved in a buffer.
6) Once a certain number of sweeps fill the buffer, the Fourier transform is performed in both range and Doppler
to extract the beat frequency as well as the Doppler shift. One can then estimate the range and speed of the
target using these results. Range and Doppler can also be shown as an image and gives an intuitive indication
of where the target is in the range and speed domain.

II. L INK B UDGET C ALCULATION FOR 5.8GH Z FMCW P RACTICAL R ADAR

ADIsimRF from Analog Devices is an easy-to-use RF signal chain calculator. It calculates signal levels, distortion
and noise for signal chains with up to fifty elements. Data are presented in tabular and graphical form and can
be plotted vs. input power, gain and frequency. Data can also be exported to Excel. ADIsimRF includes frequency-
dependent device models for most of ADI’s RF components along with many DACs and ADCs and some transformers
and filters from third-party vendors. To assist in device selection, component selection tables are included along with
a Device Selection Wizard.
The proposed RF subsystem is bi static radar type which incorporates transmit and receive paths. It is implemented
to increase the isolation between the two paths. The transmit path starts with the voltage controlled oscillator
(HMC431) which is able to output a 2 dBm signal to sweep in the range of 5.6 to 6 GHz according to modulation
voltage signal from the function generator. The swept signal is amplified by way of a low noise amplifier (HMC320)
which provides a power amplification of 13 dBm at the operating frequency. After that the swept signal is divided by
way of a splitter (SCN-2-6+) which introduces a 3.5 dBm insertion loss. A portion of the swept signal is coupled to
the mixer as reference signal. The final transmitted power of the system is 11.5 dBm which is transmitted through
a 14 dBi TX antenna. The received path starts with a series of low noise amplifiers (LNA). This amplified signal
from the RX antenna is inputted to the mixer where it is down converted by mixing with the reference signal. This
down converted signal is amplified at the gain stage and passed through low pass filter before it is sampled by the
PC/ADC/Base band signal processing. The complete system level diagram is shown in Fig 5.
Once the transmit power of the radar is realized, Frii’s equations is used to determine the received power. The
goal is to design the receiver to detect a 0.3 meter squared target ranging at distances between 5 to 50 meters.
With an operating frequency of 5.8 GHz the power received ranges from -52.4 dBm to -92.4 dBm for target ranges
from 5 to 50 meters. To reduce the amount of amplification required at the baseband gain stage, RF amplification is
implemented at the RF receiver path with a cascade of three low noise amplifiers (HMC320) with each providing a
power amplification of 13 dBm at a noise figure of 2.5 dBm. The baseband circuit is designed to amplify the incoming
signal and filter any noise received from the IF frequency of the RF circuit. The link budget table below summarizes
 5

the power received, mixer power output and mixer output voltage for targets at 5 and 50 meters. From this the gain
required at the gain stage can be determined.
For weak signals detected at target ranges as far as 50m, amplification is required for the built in ADC of the PC’s
microphone to detect noticeable measurements. For the ADC of the microphone a minimum 25mV peak to peak
signal was determined to be sufficient. To detect targets at the maximum range of 50 meters, amplification provided
by the gain stage should be x40. Any signal higher than 1.2 volts peak to peak would be limited by the limiter circuit.
To characterize the receiver path of the RF subsystem Analog Devices provides ADIsimRF to calculate total gain,
noise figure, signal to noise ratio (SNR) and peak to peak output voltage.
Students can download ADIsimRF from Analog Devices ( http://adisimrf.download.analog.com/ADISimRF/publish.htm)
and use this software for calculating the Link Budget of 5.8GHz FMCW Radar implemented using Analog Devices
components listed along with various power levels. The snapshot of ADIsimRF is shown in Fig. 4. The implemented
radar is shown in Fig.5. Students can also fill in the Link Budget Sheet as shown in Fig.6 and make it part of their
report.

Fig. 4. Link Budget of 5.8GHz FMCW Radar System

Fig. 5. Block Diagram of 5.8GHz FMCW System using Analog Devices Components

III. P ROBLEM S TATEMENT AND S TUDENT S UBMISSION

Although FMCW radar systems have been on the horizon for many years, its full potential has not yet been
leveraged for emerging applications. Our aim in this CEP is to make use of this radar system for automotive
 6

RF System (Link Budget) Calculations

System Variables Variable Units Equation Value
Frequency f0 MHz 315
Speed of Light c m/s 299792458
Wavelength λ m λ = c/f0 0.951722089

Block Variable Units Equation Value

PA Power PPA dBm 10
TX Match Loss LMatchT dB
TX source PTX dBm 10
TX connector loss LConT1 dB (from Connector Loss sheet) -0.57
TX cable loss LCabT dB (from Cable Loss sheet)

TX connector loss (remote antenna) LConT2 dB (from Connector Loss sheet)

TX power PT dBm PT = PTX(C&C Loss) 9.43

TX antenna gain GT dBi -15
Effective (Isotropic) Radiated Power EIRP dBm EIRP = PT GT -5.57
Distance d m 175
Channel Medium Loss Factor L0 dB (from Medium Loss sheet) 0
Free Space Loss LFS dB LFS = (λ/4πd)2 -67.27475527
Power at RX Antenna, Free Space Path PChanFS dB PChanFS = LFSL0 EIRP -72.84475527
Flat Earth Loss (Includes Ground Bounce) LFE dB (from Ground Multipath sheet) -92.73416576
Multipath Loss LMP dB
Obstruction Loss LObs-Total dB 0
Power at RX Antenna, Flat Earth Path PChanFE dB PChanFE = LFEL0LMPLObs EIRP -98.30416576
RX antenna gain GR dBi -15
RX connector loss LConR1 dB -0.57
RX cable loss LCabR dB
RX connector loss (remote antenna) LConR2 dB
RX power, Free Space Path PRFS dBm PRFS = PChanFS GR(C&C Loss) -88.41475527
RX power, Flat Earth Path PRFE dBm PRFE = PChanFE GR(C&C Loss) -113.8741658 -114
Sensitivity of Rx

Receiver Sensitivity Calculations Variable Units Equation Value

RX Noise Figure NF dB 7
Operating Temperature T0 K 290
Effective Noise Temperature Te K Te = T0(NF - 1) 1163.442978
Boltzmann's constant k J/K 1.38E-23
Receive Bandwidth BWRX MHz 0.012
Antenna Temperature TAnt K 300
Noise Power (at RX) Pn dBm Pn = k (TAnt + Te)BWRX -126.1556386
Signal to Noise Ratio SNRRX dB SNRRX = PRX/Pn 12.15563863

Fig. 6. RF System’s Link Budget Calculation Shee.

applications. Students are expected to submit following in their final report for Complex Engineering Problem. Each
item should be treated as a separate section in the report.
1) Write time domain expression for transmitted linear FMCW signal and received signal.
 7

2) Determine the expressions for a linear FMCW radar system for items given in table V.

TABLE V
FMCW PARAMETERS

Parameter Definition
Pr
Pt
Radar Equation
R Range(m)
∆R Range resolution(m)
vr Velocity of the moving Target(m/s)

3) Design and simulate a generic complete end to end linear FMCW radar system from transmitter to
target and then to receiver. Use the parameters given in table IV. You may assume other parameters
that are not explicitly mentioned but you should properly mention your assumptions if any. Develop
the system model of FMCW Radar in MATLAB. You may take the help from following link of Mathworks
https://uk.mathworks.com/help/phased/examples/automotive-adaptive-cruise-control-using-fmcw
-technology.html
4) Given the configuration parameters in table IV, simulate and generate Range-Doppler response pattern
for linear chirp pattern FMCW using MATLAB. A sample of 2D range-velocity diagram is shown in
Fig. 8 [5]. The two-dimensional FFT process gives a 2D range-velocity image (FFT heatmap). Typically,
detection of objects is done on this image. After detection, the range and relative speed of the objects
are easily calculated (ref. to Fig.7).
5) Repeat step 3 and 4 for triangular chirp pattern.
6) Calculate Link Budget for receiver path for targets at 5 meters and A + 30 meters using ADIsimRF from
Analog Devices for 5.8GHz FMCW radar shown in Fig.5. A in this case is the last two digits of your
student ID. For example if the last two digits of your ID are 50, it should be 30+ 50=80 meters.
7) Propose circuit level implementation of FMCW radar based ADAS systems for vehicles in Pakistan.You
can read paper titled: Implementation of FMCW Radar for Training Applications by Oguz Baarslan and
Ercan Yaldlz [6] or the paper titled: S-Band 2.4GHz FMCW Radar by Iulian Rosu and Filip Rosu by
visiting (https://qsl.net/va3iul/) for some details how circuit level implementation can be carried out
for 2.4GHz radar. This is an optional part but highly recommended if a students want to earn a better
grade in EE410.
8) Students are expected to write an abstract/short summary (1-2 page max) of the subject matter by
doing a literature review of ”FMCW Radars for ADAS systems”. This should include a ”summary” of
three papers, ”critical analysis” of the paper and your ”suggestions” based on your critical analysis.
The papers can be either conference papers or journal articles or patents but they should be published
within last 5 years. Each group is expected to email us 5 different research papers in this domain. Out
of those five papers, we will short list three research articles for each group.
Important Note: Students must submit CEP report by Monday of Week 13 (the week before lab vivas).

IV. S UBMISSION RULES AND R EGULATIONS

1) 3 students are allowed to make a group.
2) Although team work is encouraged and also carry weight in this assessment, each individual would also be
evaluated separately.
3) A complete report has to be submitted by each group.

R EFERENCES
[1] E. A. Board, “Manual of accreditation.” Pakistan Engineering Council Islamabad, 2014, pp. 1–63.
[2] M. Barjenbruch, D. Kellner, K. Dietmayer, J. Klappstein, and J. Dickmann, “A method for interference cancellation in automotive radar,” in
2015 IEEE MTT-S International Conference on Microwaves for Intelligent Mobility (ICMIM), April 2015, pp. 1–4.
[3] ——, “A method for interference cancellation in automotive radar,” in 2015 IEEE MTT-S International Conference on Microwaves for Intelligent
Mobility (ICMIM), April 2015, pp. 1–4.
[4] J. Mun, H. Kim, and J. Lee, “A deep learning approach for automotive radar interference mitigation,” in 2018 IEEE 88th Vehicular Technology
Conference (VTC-Fall), Aug 2018, pp. 1–5.
[5] M. Song, J. Lim, and D. Shin, “The velocity and range detection using the 2d-fft scheme for automotive radars,” in 2014 4th IEEE International
Conference on Network Infrastructure and Digital Content, Sep. 2014, pp. 507–510.
[6] O. Başarslan and E. Yaldız, “Implementation of fmcw radar for training applications,” in 2017 4th International Conference on Electrical and
Electronic Engineering (ICEEE), April 2017, pp. 304–308.
 8

FMCW 2D FFT processing in a nutshell

1 2 3 N 1
ADC data
corresponding to chirps
are stored as the rows
Tc
of a matrix

3
2 A doppler-FFT along
A range-FFT on the column resolves
each row resolves each column
Chirp index

objects in range (‘range-bin’) in

velocity
velocity

range
range
44

Fig. 7. FMCW 2D FFT processing in a nutshell

Fig. 8. Range-Velocity Diagram

 Appendix-A
Information about components used in 5.8GHz FMCW Radar

ADIsimRF receiver path calculation for target at 50 meters

Range Power Received Mixer Power Output Mixer Voltage Output

5m -52.4 dBm -20.4 dBm 60.3 mVpp
50m -92.4 dBm -60.4 dBm 603 uVpp

Component Selection

This section will cover the process involved in selecting components that meet system design
specifications detailed in the previous section. The selection process has been divided into RF
components and baseband components. The table below summarizes the main components
selected in this process.
Device Model Manufacturer Specification
Function Generator XR-2206CP Exar/Jameco Frequency Range:
0.01-1MHz
I = 14 mA
Voltage Controlled HMC431LP4ETR Analog Devices Power Output: 2dBm
Oscillator Vcc = 3V
I = 33mA
Power Splitter SCN-2-65+ Mini-Circuits Insertion loss: 3.5 dBm
Passive
Low Noise Amplifier HMC320MS8GE Analog Devices Noise Figure: 2.5 dBm
(13 dBm) Power Gain: 13 dBm
I = 40 mA
Mixer HMC218BMS8GE Analog Devices Conversion Loss: 7 dBm
LO Power: + 13 dBm
Passive
Antenna 5.8GHz 4-Patch Ripafire Gain = 14 dBi
Array Bandwidth: 350 MHz
Range: 5.6 – 5.95 GHz
Gain Stage Amplifier TL972IP Texas Instruments
Active Low Pass Filter MAX291CPA+ Maxim Integrated Corner Frequency Range:
0.1 Hz - 25 kHz

HMC431LP4ETR Voltage Controlled Oscillator:

The Voltage Controlled Oscillator (VCO) generators the RF signal transmitted by the radar. The
VCO requires many considerations when selecting. For range measurements applications an
optimal VCO would have a tuned voltage to oscillate linearly in conjunction to a linear output in
the frequency of operation. Constant power output over this range would be ideal. In addition the
VCO should have low phase noise. The VCO chosen for the design was the HMC431LP4ETR
manufactured by Analog Devices. The graphs below, taken into account for the analysis, are
obtained from the datasheet provided by Analog Devices. The expected operating frequency is
between 5.6 to 5.95 GHz. To operating in this range the tuning voltage would span between 2 to
6 volts. Based on the sensitivity vs. tuning voltage graph in this range there is a deviation of 50
MHz/Volt. Although this is not perfectly linear it is acceptable for radar applications. The output
power remains constant at just above 2 dBm for the entire frequency range.
XR-2206CP Function Generator:

Once a VCO has been selected a function generator must be taken into account. Initially the
Teensy 3.1 in conjunction with the MCP4921 DAC was taken into consideration but due to size
and power constraints another option was chosen. The XR-2206CP is a monolithic function
generator Exar that is able to produce the required triangle waveform to drive the VCO without a
microcontroller and digital to analog converter.

HMC320MS8GE Low Noise Amplifier:

The Low Noise Amplifier (LNA) is required before the splitter and the RF input of the mixer to
allow for enough amplification of the signal entering the baseband circuit. When selecting a
LNA, it is critical to choose one that provides high power amplification and relatively low noise
figure which would reduce the signal to noise ratio introduced during cascading of multiple
LNAs. Also an LNA that does not require a complicated bias network would simplify the
assembly process. The HMC320MS8GE by Analog Devices was chosen for the radar design
which provides the option for configuration with three bias conditions. The graphs below are
obtained from the datasheet. To maximize power amplification of each LNA, a high power bias
is chosen.

SCN-2-65+ Power Splitter

When selecting a power splitter it is important to pay close attention to insertion losses and a
number of other key factors. The selection of the SCN-2-65+ Power Splitter by Mini-Circuits is
justified in the following table. For a frequency range for 5500-6500MHz the insertion loss
remains low between 3.5 and 3.8 dB. The splitter can provide a large isolation of about 17 dB
which is good. Also the splitter had the added benefit of easy soldering and being low on cost.
HMC218BMS8GE Mixer

The ideal mixer would be a double balanced mixer which has the advantages of good port
isolation, noise protection and linearity. Fortunately Analog Devices provides the
HMC218BMS8GE that meets these specifications in a passive double balanced topology. The
graph below shows the conversion loss for a given LO drive of the mixer. With an expected LO
drive of about +11dBm for our system design, the conversion loss remains constant at 6.5 dBm
under the expected operating frequencies.

TL972IP Amplifier (Gain Stage) and MAX291CPA+ (Low Pass Filter)

The gain stage and active low pass filter in consideration is a modified version from the
baseband design implemented in lab 1 for the quarter 1 radar which had a cut off frequency of 15
kHz. Modifications include the cascading of 2 amplifiers (TL972IP) in the gain stage to achieve
the required amplification of x40 and an improved dedicated active lowpass filter IC
(MAX291CPA+) which can provide a maximally flat passband response as a 8th order
Butterworth filter.