A Study on mmWave Sensor Technology

and its Applications

Ansary Shafew Eui-jik Kim Dongwan Kim
2022 13th International Conference on Information and Communication Technology Convergence (ICTC) | 978-1-6654-9939-2/22/$31.00 ©2022 IEEE | DOI: 10.1109/ICTC55196.2022.9952690

Department of Electronics Engineering Division of Software Department of Electronics Engineering

Dong-A University Hallym University Dong-A University
Busan, South Korea Chuncheon, South Korea Busan, South Korea
[email protected] [email protected] [email protected]

Abstract— Sensing technology has undergone numerous sensor is significantly more widely used than other sensors.
evolutions along with the rising daily use of various types of However, sensing technology formerly disapproved of the
sensors in the modern world. The mmWave sensor is the usage of mmWave. The lack of appropriate methods for the
technology that has advanced the most in the last few years. generation, transmission, channelization, and reception of
LiDAR, ultrasonic, and image sensors are additionally useful electromagnetic waves in the millimeter range was one of the
sensing technologies, but they have significant limitations in primary causes [6].
terms of high accuracy, long distance measurement and
potential interference from external environmental factors. TABLE I. SUMMARY OF ATTRIBUTES OF DIFFERENT SENSORS
However, mmWave sensors have fewer drawbacks than current
sensors and are broadly used in a wide range of industries, Features LiDAR mmWave Imaging Ultrasonic
including RADAR technology, medical applications, meteor- Primary Technology Laser beam Radio Wave Light Sound Wave
ological applications etc. This is paving the way for overcoming Range ~200m ~250m ~200m ~5m
their restrictions. The overall operation of the mmWave sensor
Resolution Good Average Very Good Poor
was addressed in this paper, along with its use and research in
many different fields. Affected by weather
Yes Yes Yes Yes
conditions
Keywords— sensor, millimeter Wave, FMCW, amplitude Affected by lighting
No No Yes No
conditions
I. INTRODUCTION Detects speed Good Very Good Poor Poor
Sensors are used by the majority of physical systems to aid Detects distance Good Very Good Poor Good
with logical decision-making. In industrial, residential, and Interference
Good Poor Very Good Good
professional settings, sensors (i.e., Millimeter wave (mmWave), susceptibility
LiDAR, passive infrared, ultrasonic, vision and thermal Size Bulky Small Small Small
imagers) are utilized to determine position and occupancy.
They promise a secure, productive, and independent
environment and are well regarded in terms of development Additionally, insufficient research was done on the laws
and research. These sensors use various methods, including governing mmWave propagation in a nonhomogeneous
mmWave, lasers, infrared light, and sound propagation, to atmosphere. However, contemporarily, mmWave sensing
find objects and their locations. LiDAR is frequently technology has come a long way due to extensive research by
employed to quantify the separation between an object and a the academic community. Today, the mmWave sensor is
sensor [1]. Imaging sensors have a wide array of applications utilized vastly in radiometry, cloud sensing, automobile
due to their adjustable sensitivity to different wavelengths of RADAR, and RADAR-based communication systems for the
light. Exploiting those attributes, imaging sensors are used for Concealed Weapons Detection (CWD) and various industries
image processing, video data collection, thermal imagery etc. as theory and experience continue to uncover new benefits of
Ultrasonic sensors are excellent for a wide range of sensing mmWave [6]. The defining attributes of the mmWave sensor
tasks in both Autonomous Vehicles (AVs) and industrial are its availability, ease of use, application, smaller size,
applications. They can detect granular, liquid, solid, or sensing capabilities, etc. Some of the key traits of various
powdered particles. Ultrasonic sensors use sonic transducers sensors compared to the mmWave sensor are compiled in
that can emit sound waves with a frequency of 40 kHz to 70 Table 1 [2].
kHz. Human ears are unaffected in this frequency range This paper in its novel approach, aims to amalgamate and
because they are outside the realm of human audibility [2]. outline the introduction, definition, working principle, appli-
The sensors discussed here are all able to measure distance. cations, and recent research of the mmWave sensor into one
However, LiDAR, imaging, and ultrasonic sensors are less place. Thus, hopefully contributing in the future to other
resistant to environmental factors [3] thus, less ideal in terms research conducted on this subject.
of precision and constant measurement. These sensors do not
have object penetration capability; therefore, fog, snow, or The contents of this paper are organized as follows.
even lingering dust on top of the sensor plane might Section II features a brief introduction to mmWave sensors
significantly alter the findings. Due to their often very limited and discusses their working principle. Section III assesses the
functional range, these sensors cannot be employed for long- recent strides made in the application and research field of
range measurement [4]. mmWave sensors. Finally, Section IV concludes the article.

MmWave sensors are also employed as a RADAR device II. GENERAL WORKING PRINCIPLE
to quantify the distance, velocity, and angle between the Conventional RADAR detects an object by detecting the
sensor and the object [5]. They are easier to implement due to periodic pulse that bounces off the object and returns to the
the smaller size and availability and have fewer environmental receiver sensor [7]. As a result, traditional sensors operate
drawbacks than the other sensors. That is why the mmWave more slowly and offer fewer data [7]. The mmWave sensor

978-1-6654-9939-2/22/$31.00 ©2022 IEEE 1771 ICTC 2022

Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.
 works by emitting an electromagnetic wave and measuring the The general main components of an FMCW sensor are
round-trip time delay of that wave from an object [8]. The shown in the simplified block diagram in the Fig. 3.
wavelength of mmWave ranges from 1 millimeter to 10
millimeters and has a frequency range of 30 to 300 GHz [9].
This sensor can operate without interference and can measure
a greater distance since the atmosphere only faintly attenuates
millimeter waves [2].
The mmWave sensors generally use Frequency Modulated
Continuous Wave (FMCW) to extend their ability and range
and thus, it is the most common approach when it is used as a
RADAR device [8], [9]. FMCW is a continuous signal that
can detect several attributes such as velocity, range, and angle Fig. 3. General FMCW sensor components
of an object and provide real-time information [8].
In Fig. 3, it is shown that a synthesizer generates a chirp
and it is relayed through the TX antenna. After that, the chirp
signal is reflected from an object of d distance away and
received through the RX antenna. The equation of the
transmitted signal can be expressed as,
ௌ
‫ݕ‬ଵ = ‫் ܣ‬௑ ‫{ ݊݅ݏ‬2ߨ ቀ݂௖ ‫ ݐ‬+ ଶ ‫ ݐ‬ଶ ቁ} (1)
Here, ‫் ܣ‬௑ is the amplitude of the transmitted signal, ݂௖ is
the carrier frequency, and ܵ is the frequency increase rate.
From Fig. 4(a), the TX and RX chirp can be observed where,
the RX chirp is a time delayed version of the TX chirp.
Considering the time delay, the equation of the received signal
can be expressed as,
Fig. 1. Typical application of mmWave sensor on AVs ௌ
‫ݕ‬ଶ = ‫ܣ‬ோ௑ ‫[ ݊݅ݏ‬2ߨ ቄ݂௖ (‫ ݐ‬െ ߬) + ଶ (‫ ݐ‬െ ߬)ଶ ቅ ] (2)
Due to the linear increase in the frequency of FMCW
signals over time, this signal type can be utilized to calculate Here, ARX denotes the amplitude of the received signal and
the separation between a sensor and the object. It is also IJindicates the round-trip time delay between an object and the
denoted as a chirp signal. It is possible to see the chirp signal sensor. Now, these two signals from (1) and (2) are fed into
by looking at the frequency as a function of time. An easy the mixer and the mixer’s output will have the instantaneous
approach to portray a chirp is with a frequency vs time plot (f- frequency which is equal to the difference of the instantaneous
t plot), which in Fig. 2(a) shows how time and frequency work frequencies of the two input signals (y1 and y2). The phase as
together as well as how they fluctuate while the amplitude A well, will be equal to the difference between the two signals.
is constant. This signal is called the Intermediary Frequency (IF) signal.

Fig. 2. Chirp signal showing (a) frequency fluctuation over time and (b)
linear increase of frequency over time

From Fig. 2(b), it can be approximated that the signal

frequency f increases linearly with time. This plot can be Fig. 4. TX and RX signal showing (a) time delay between TX and RX
chirps and (b) IF signal
defined as an amplitude vs time plot (A-t plot). In this
example, a chirp is defined as having a carrier frequency fc and Since the mixer’s output is equal to the difference between
sweeps a bandwidth B, in duration Tc, which is equivalent to the instantaneous frequencies of the two input signals, as
the slope or frequency increase rate S. In this case, slope S illustrated in Fig. 4(b), the IF signal is a straight line. As a
specifies how quickly the chirp ramps up. result, when a singular object is in front of the sensor, an IF
signal with a steady tone of frequency is generated. From Fig.

1772
Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.
 4(a,b), this frequency can be assumed as SIJ. Here, SIJ = S2dc-1 the signal can be converted in to range component. This is
and IJ 2dc-1 where d corresponds to the distance between the denoted as range-FFT. The peaks in the range-FFT correlating
object and the sensor and c is the speed of light. As a result, to every chirp will be in the same spot but in a different phase.
the output signal can be calculated from the equation below, The calculated phase difference is Ȧ which refers to a
ௌ displacement in the object of vTc. Thus, Ȧ  ʌIIJ, and by
‫ݕ‬௢௨௧ = ‫ܣ‬ூி ‫{ ݊݅ݏ‬2ߨ ቀܵ߬‫ ݐ‬+ ݂௖ ߬ െ ଶ ߬ଶ ቁ} (3) deriving it further,
Here, ‫ܣ‬ூி is the amplitude of the IF signal. After calcu- ߱=
ସగௗ
=
ସగ௩்೎
ఒ ఒ
lating and acquiring the IF signal, Fast Fourier Transform (FFT)
is applied to extract the frequency of the IF signal. And, ‫ = ݒ‬ସగ்
ఒఠ
(5)
೎
In a situation where two objects are approximately located From (5), it is possible to calculate the object's velocity
together, it could be quite complicated to distinguish them using the phase difference between two successive chirps.
properly, let alone calculate the distance. In that case, Range
Resolution is applied. Range Resolution is the capacity to
resolve two items that are relatively tightly spaced apart. The
two objects can be distinguished by increasing the length of
the IF signal. According to the fourier transform theory,
lengthening the IF signal improves resolution. Furthermore,
the bandwidth needs to be raised correspondingly to match the
elongation of the IF signal. The IF spectrum of a longer IF
pulse has two distinct peaks. Again, according to the fourier
transform theory, a frequency component with a separation
Fig. 6. Velocity unambiguity measurement illustration
greater than 1/T Hz can be resolved using an observation
window of Tc. Accordingly, the value of ¨f will be greater
However, there will be ambiguity since the velocity
1/Tc. If the frequency difference between two IF signal tones
measurement is based on a phase difference. As illustrated in
fulfill this relationship, then the frequency difference can be
Fig. 6, the measurement is unambiguous only if _Ȧ_ʌ.
resolved. Since, ¨f = (S2¨d)-c and B=STc, the resolution of
two objects separated by a distance of ¨d will be, Evidently, the maximum relative velocity that can be
௖ ௖ determined by two chirps which are Tc space apart is,
ο݀ > ଶௌ் = ଶ஻
೎ ఒ
௖ ‫ݒ‬max =
4்೎
(6)
So, ݀௥௘௦ = ଶ஻ (4)
Here, Ȝ is the wavelength of the chirp. Higher vmax requires
The range resolution dres depends only on the bandwidth shorter transmission time between chirps.
swept by the chirp. As a result, an FMCW sensor with a few
GHz chirp bandwidth will have a range resolution of a few The mmWave sensor can calculate the angle of a reflected
millimeters. signal with the axial plane, as shown in Fig. 7. This angle is
denoted as the Angle of Arrival (AoA). The angle is estimated
As seen in Fig. 5, a mmWave sensor sends two chirps based on the phase shift in the range-FFT or Doppler-FFT
delayed by Tc to calculate velocity. In order to determine the peak, which occurs during a small change in distance between
object's range, each reflected chirp is additionally subjected to the sensor and object. As a result, angle estimation needs a
FFT processing. minimum of two RX antennas. The phase variation in the 2D-
FFT peak caused by the contrasting distances of ¨d between
the object and each of the antennas is used to compute the
AoA.

Fig. 7. Differential distance from an object to RX antennas

Hence, the phase change can be derived by,

ʌǻd
Ȧ= Ȝ
(7)
Assuming a planar wavefront, ¨d = d sin ș can be calculated.
Here, ݀ is the distance between the RX antennas as illustrated
Fig. 5. Difference in phase of two consecutive chirps with same peaks in Fig. 8. So, after measuring ߱ and the AoA, ș can be
computed,
Since the frequency of the chirps increase linearly with ȜȦ
respect to time, the distance between the sensor and the object ș = sin-1 ( ʌd ) (8)
can be precisely estimated. Thus, the frequency component of

1773
Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.
 It should be noted that the correlation between Ȧ and ș is A. AV Applications
non-linear, and ǻG depends on VLQ ș in this case. This can be The mmWave sensors have a wide array of applications
expressed as a non-linear dependency. Only when ș has a and AV is one of those fields. Conforming to the Society of
smaller value, then sin ș is approximated with a linear Automotive Engineers, the Advanced Driver Assistance
function (sin ș ~ߠ .). As a result, the estimation is more System (ADAS) is a six-tiered system that categorizes the
accurate when șhas a small value and the accuracy depends various levels of autonomy [10]. It includes fully autonomous
on AoA. Thus, when ș , Ȧis most sensitive to changes. or self-driving vehicles and also those that are only driven by
When the value of ș increases, the sensitivity of Ȧ to ș humans [11]. AVs always depend on an amalgamation of
reduces. Therefore, the estimation of ș is more prone to error. sensors and software to assess their surroundings and navigate
without the assistance of a human in order to reach increasing
autonomy levels. To achieve enhanced autonomous mobility
and transportation that is secure for riders and pedestrians
[10], scientists and engineers from various disciplines and
backgrounds are participating in improving mmWave sensor
technology and tackling all the related issues that come along
with it.
In intelligent automobile systems, the 4D mmWave sensor
is employed as a multi-lane, multi-object tracking device.
mmWave sensor is becoming increasingly common for use in
ADAS for parking assistance, collision detection, and driving
Fig. 8. Illustration of AoA measurement using 2 RX antennas
safety. The mmWave frequencies used by the sensor systems
for AVs are 24, 74, 77, and 79 GHz, and they can be used for
The sensor's maximum AoA can be used to define the Short-Range, Medium-Range, and Long-Range RADAR
sensor's max angular Field of View (FOV). While measuring (SRR, MRR, and LRR) respectively [12]. In contrast to MRR
ș, there could be some ambiguity as well. As illustrated in Fig. and SRR, which are utilized for side view detection or parking
9, the unambiguous angle measurement requires _Ȧ_ၨ assistance, LRR can be utilized to find distant objects or
Fulfilling this parameter, the maximum FOV measurable by targets ahead of an automobile [13]. MmWave sensors are
two RX antennas that are in d distance from each other can be among the many sensor technologies that are widely
measured from the equation below, accessible, and because of their simplicity, they are frequently
utilized in AVs. The majority of AV sensor systems currently
Ȝ
ߠ௠௔௫ = sin-1 ( 2d ) (9) in use, employ a collection of miniature antennas that can
produce one or more antenna lobes (e.g., a 24 GHz sensor with
A distance of d between two antennas RI Ȝ JLYHV WKH horn antennas and a 77 GHz sensor with Printed Circuit Board
maximum FOV (+/-ၨ . antennas). System-on-Chip designs have made this more
typical since they enable various approaches, such as digital
beam shaping, to reduce the receiver's spatial FOV and lessen
electromagnetic interference from other sensing devices
running at related center frequencies. This kind of technology
is now being used in the automotive industry in terms of Lane
Change Assistance, Adaptive Cruise Control, Collision
Avoidance Systems, and Blind Spot Detection to increase
Fig. 9. Angular FOV unambiguity measurement illustration
safety due to the development of smart.
The mmWave sensors are seeing many advancements in
Targets are found using the outlined equations for the AV research field as well. O. Toker and S. Alsweiss,
velocity, range and angle. The mmWave sensor is an optimal demonstrated [14] how moving pedestrians can be detected by
solution for systems that require sensing the presence and a mmWave sensor technique and distinguished from other on-
behavior of people or animals since it can detect location and road vehicles. For ADAS and AV applications and in
motion accurately. Consequently, it has some penetration particular for the safety of pedestrians, this is a significant
capabilities that enables it to be used in security systems to issue. Without requiring a lot of computer power, this method
detect humans or concealed weapons. makes use of the sensor's capacity to function in a variety of
lighting and environmental situations (such as day, night, fog,
III. APPLICATIONS AND RESEARCH OF MMWAVE SENSOR
etc.) [14]. Zhang et al. demonstrated [15] an all-weather,
Considering its numerous uses and advantages, the highly precise pedestrian trajectory perception system based
mmWave sensor is extensively used in the affinitive on the mmWave sensor called MRPT. More particularly, a
industries. This sensor is impervious to diverse atmospheric track-before-detection-based algorithm is suggested to treat
conditions like dust, smoke, and fog when compared to other pedestrian detection and tracking concurrently, as opposed to
sensors [2]. In addition, it can function in specific frequencies regarding them as two independent stages in conventional
to reduce interference while being unaffected by other sources methods. This is applied to increase the tracking performance
and intense lighting. With all these benefits, it can be utilized in the face of high backdrop clutters. In-depth simulation and
in a variety of situations where other sensors give off less- experimental results from that research showed that MRPT
than-ideal readings due to many external variables. These outperforms conventional approaches in terms of pedestrian
attributes also unfurl the potential applications of this sensor detection and tracking in low Signal-to-Noise Ratio situations.
in terms of research and development, which has led to Steffen Heuel et al. described [16] a pedestrian detection
vigorous research on this subject matter. system using the 24GHz mmWave sensor. They showed

1774
Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.
 tremendous progress in 24 GHz mmWave sensor develop- has different chemical compositions. It is anticipated that the
pment and accurate measurement with high resolution even continued research on the variation of reflected signals would
with multiple targets. J. Hasch et al. provided [17] a summary produce findings and enable the mmWave sensor to make
of the most recent developments in mmWave technology for more precise measurements.
use in automobile radar applications, as well as seamlessly
engaged two actual silicon-based radar chips. They also A. Ahmad et al. proposed [22] employing the FMCW
highlighted many cutting-edge packaging and antenna sensor to monitor numerous people's vital signs without
solutions. Their system performed well and had promise. making physical touch. To divide objects in the range-azimuth
Three distinct implementations of car radar sensors were plane, they used the FMCW waveforms' built-in range-gating
shown, illustrating the progression from current to next to capability and numerous receive channels. Douglas T. Petkie
future generation solutions. et al. presented [21] a 228 GHz heterodyne radar system that
can measure respiration rates at standoff range of up to 50
B. Meteorological Applications meters and sequentially measure respiration and cardiac rates
The mmWave sensors are also used in meteorological at a distance of 10 meters. They also reviewed the most recent
applications as well. Such as cloud presence, location, advancements in hardware and signal processing, as well as a
physical makeup, turbulence, and dynamics which can all be wide range of research aimed at improving the system's
determined qualitatively using a mmWave sensor. The performance in a variety of prospective field applications.
polarization capabilities of the mmWave sensor can be used More research is constantly happening to improve certain
to examine the deformation of clouds. Cloud microphysical aspects in different fields to make them better and more
data can be retrieved by combining the mmWave sensor and reliable, owing to the attributes of the mmWave sensor.
LiDAR [18]. D. Security Applications
B. Hattenhorst et al. proposed [19] a system to monitor Additionally, devices with mmWave sensors can provide
gases and aerosols and suggested a method of monitoring security in spaces like restrooms or lockers that are off-limits
gases and aerosols that makes use of an ultrawideband FMCW to video filming. By identifying people's presence and the
sensor that operates between 72 and 92 GHz at a sampling rate direction they are moving in, the mmWave sensor can enhance
of 500 Hz. Additionally, utilizing information from molecular automatic doors with many clever features. By doing so, the
spectroscopy, they outlined a model for simulating the door will only open when essential personnel are in proximity
refractive index of arbitrary gases at mmWave frequencies. and the air conditioners will only run when the room is
The gas under test can then be identified thanks to simulations. inhabited. At the same time, the mmWave sensor reduces
Through a number of studies, they evaluated how well the energy consumption [23], [24].
sensor performed in terms of refractive index. The sensor's
accuracy, precision, and robustness make it ideal for the In the research sector, Z. Li et al. discovered [25] and
industry's in-process real-time gas or aerosol categorization proved a novel and useful side-channel attack for screen
monitoring. According to R. A. Krofli and his team's descr- content utilizing a low-cost radio-frequency sensor to estimate
iption [20], when employed separately, the 3 mm and 8 mm the liquid crystal's nematic state. They created WaveSpy, an
wavelength sensors may analyze the physical structure, end-to-end portable through-wall screen attack system, by
dynamics, and small-scale turbulence of clouds. They utilizing the connection between the contents of the screen and
discussed two distinct techniques for determining ice mass the states of the liquid crystal arrays in displays. Even when
profile profiles using radar/lidar and radar/IR radiometer data. the victim's screen is put in an isolated area, WaveSpy's low-
They obtained novel results from an aerial system to show that cost, energy-efficient, and lightweight mmWave sensor can
ram-wave radars are highly suitable for aviation and satellite remotely collect the liquid crystal state response to a set of
devices due to their small size and weight. mmWave stimuli and assist screen content inference. Under
realistic conditions, WaveSpy's screen content type detection
C. Medical Applications accuracy was 99 percent accurate and its top-3 sensitive
The mmWave sensors have major applications in medical information retrieval success rate was 87.77 percent. The
systems in terms of contactless measurement. The majority of development and evaluation of the silent mmWave sensor
currently used measurement tools rely on physical contact. performing in the 88-100 GHz band were given by B.
They must be fastened to the patient's body to measure and Kapilevich et al. [26]. This sensor's scanning portion is
monitor. Patients who require constant observation for a focused on a heterodyne setup with NF = 9–10dB and a
prolonged period of time may not always find this convenient. bandwidth up to 12 GHz. It can pick up signals typical of
However, in the case of mmWave sensor use, patients have passive mmWave imaging, like noise. Depending on the
some degree of privacy because it does not use optical vision antenna system, the sensor can be set up for indoor and
to track human activity. Component mmWave sensor can pick outdoor imaging settings. It was created and proven to work
up on human movements, breathing, heartbeats, and minor with the MATLAB 7 data gathering system using the
body movements [21]. Applications including monitoring LabView interface. Examples of indoor and outdoor photos
blood pressure, emotions, and sleep make use of this that were rebuilt using this sensor were shown, demonstrating
mmWave sensor characteristic. High precision is a benefit of its suitability for needs related to homeland security.
short wavelengths. Movements as small as a millimeter and a IV. CONCLUSION
half can be detected by a mmWave sensor operating at 60GHz
or 77GHz (with a corresponding wavelength in the range of 4 Modern mmWave sensors offer various functions,
mm) [21]. The development of mmWave technology for vital applications, and research in fields like AV, meteorological,
sign monitoring is ongoing. The difference in reflected signals medicine, and security. Although different sensors can be
between individuals is one of the biggest problems. The employed for the same purpose (e.g., LiDAR, Ultrasonic,
reflection is influenced by the skin's composition, tissue, and imaging sensors, etc.), they are less resistant to environmental
type [21]. The body contains different amounts of water and factors and relatively harder to use in terms of size and

1775
Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.
 accessibility. Since mmWave sensors can operate without Tracking for Autonomous Urban Driving," in IEEE Transactions on
interference in various weather situations by exploiting their Instrumentation and Measurement, vol. 71, pp. 1-17, 2022, Art no.
8000117.
penetration ability, they are also more compatible with current
[16] S. Heuel and H. Rohling, "Pedestrian classification in automotive radar
technology and use cases. The sensor can also deliver more systems," 2012 13th International Radar Symposium, 2012, pp. 39-44.
accurate data with higher transmission rate, employ wider [17] J. Hasch, E. Topak, R. Schnabel, T. Zwick, R. Weigel and C.
spectrum, and withstand interference better. Extremely high Waldschmidt, "Millimeter-Wave Technology for Automotive Radar
frequencies allow for simultaneous, non-interfering use for Sensors in the 77 GHz Frequency Band," in IEEE Transactions on
multiple short-range applications. A significant drawback is Microwave Theory and Techniques, vol. 60, no. 3, pp. 845-860, Mar.
that the sensors still require the assistance of a perceptual 2012.
system to make more reliable decisions. Nevertheless, by [18] Cadence System Analysis "mmWave RADAR Applications and Advantages".-
Available:-https://resources.system-analysis.cadence.com/blog/msa2022-
performing additional research in this area, there may be a mmWave-RADAR-applications-and-advantages. [Accessed: 07-Jul-2022].
wide range of prospective applications and the creation of
[19] B. Hattenhorst, L. Piotrowsky, N. Pohl and T. Musch, "An mmWave
more affordable and dependable solutions to a wide range of Sensor for Real-Time Monitoring of Gases Based on Real Refractive
issues. Index," in IEEE Transactions on Microwave Theory and Techniques,
YRO൰൳QR൫൫SS൯൪൭൭-൯൪൮൮1RY൬൪൬൫
ACKNOWLEDGMENT [20] R. A. Krofli and R. D. Kelly, “Meteorological research applica tions of
This work was supported by the National Research MM-wave radar,” Meteorol. Atmos. Phys., vol. 59, nos. 1–2, pp. 105–
121, Mar. 1996.
Foundation of Korea (NRF) grant funded by the Korea
[21] Douglas T. Petkie, Erik Bryan, Carla Benton, and Brian D. Rigling
government. (NRF-2021R1F1A1062443) "Millimeter-wave RADAR systems for biometric applications", Proc.
SPIE 7485, Millimetre Wave and Terahertz Sensors and Technology
REFERENCES II, 748502, Sep. 17, 2009.
[22] A. Ahmad, J. C. Roh, D. Wang and A. Dubey, "Vital signs monitoring
[1] D. Wang, C. Watkins, and H. Xie, “MEMS Mirrors for LiDAR: A of multiple people using a FMCW millimeter-wave sensor," 2018
Review,” Micromachines, vol. 11, no. 5, p. 456, Apr. 2020. IEEE Radar Conference (RadarConf18), 2018, pp. 1450-1455.
[2] J. Vargas, S. Alsweiss, O. Toker, R. Razdan, and J. Santos, “An [23] A. Luukanen, R. Appleby, M. Kemp, and N. Salmon, “Millimeter-
Overview of Autonomous Vehicles Sensors and Their Vulnerability to wave and terahertz imaging in security applications,” in Terahertz
Weather Conditions,” Sensors, vol. 21, no. 16, p. 5397, Aug. 2021. Spectroscopy and Imaging. New York, NY, USA: Springer, 2013, pp.
491–520.
[3] S. Royo and M. Ballesta-Garcia, “An Overview of Lidar Imaging
Systems for Autonomous Vehicles,” Applied Sciences, vol. 9, no. 19, [24] S. Björklund, H. Petersson, A. Nezirovic, M. B. Guldogan and F.
p. 4093, Sep. 2019. Gustafsson, "Millimeter-wave RADAR micro-Doppler signatures of
human motion," 2011 12th International RADAR Symposium (IRS),
[4] M. Bijelic, T. Gruber and W. Ritter, "A Benchmark for Lidar Sensors
2011, pp. 167-174.
in Fog: Is Detection Breaking Down?," 2018 IEEE Intelligent Vehicles
Symposium (IV), 2018, pp. 760-767. [25] Z. Li et al., "WaveSpy: Remote and Through-wall Screen Attack via
mmWave Sensing," 2020 IEEE Symposium on Security and Privacy
[5] C. Waldschmidt, J. Hasch and W. Menzel, "Automotive Radar — From
(SP), 2020, pp. 217-232.
First Efforts to Future Systems," in IEEE Journal of Microwaves, vol.
1, no. 1, pp. 135-148, Jan. 2021. [26] B. Kapilevich, B. Litvak, M. Einat and O. Shotman, "Passive mm-wave
Sensor for In-Door and Out-Door Homeland Security Applications,"
[6] F. J. Yanovsky, “Millimeter Wave Radar: Principles and
2007 International Conference on Sensor Technologies and
Applications,” In Millimeter Wave Technology in Wireless PAN,
Applications (SENSORCOMM 2007), 2007, pp. 20-23.
LAN, and MAN, Editors: Shao-Qiu Xiao, Ming-Tuo Zhou, Yan Zhang,
Chapter 10, pp.305-376, ISBN: 9780849382277, Auerbach
Publications CRC Press, 2008, 464 pp.
[7] S. D. Blunt and E. L. Mokole, "Overview of RADAR waveform diversity,"
in IEEE Aerospace and Electronic Systems Magazine, vol. 31, no. 11, pp. 2-
42, Nov. 2016.
[8] G. M. Brooker, "Understanding Millimeter Wave FMCW Radars," in
First International Conference on Sensing Technology, Nov. 21–23,
2005, Palmerston North, New Zealand (G. Sen Gupta, S. C.
Mukhopadhyay, and C. H. Messom, eds.), pp. 152–157.
[9] B. van Berlo, A. Elkelany, T. Ozcelebi and N. Meratnia, "Millimeter
Wave Sensing: A Review of Application Pipelines and Building
Blocks," in IEEE Sensors Journal, vol. 21, no. 9, pp. 10332-10368, 1
May 1, 2021.
[10] K. Bimbraw, "Autonomous cars: Past, present and future a review of
the developments in the last century, the present scenario and the
expected future of autonomous vehicle technology," 2015 12th
International Conference on Informatics in Control, Automation and
Robotics (ICINCO), 2015, pp. 191-198.
[11] SAE International. [Online]. Available: https://www.sae.org/.
[Accessed: 07-Jul-2022].
[12] W. Buller, B. Wilson, J. Garbarino, J. KeOO\ % 7KHOHQ țĮȚ % 0
%HO]RZVNLµ5DGDU&RQJHVWLRQ6WXG\¶ȈİʌIJİȝȕȡȓȠȣ
[13] J. Steinbaeck, C. Steger, G. Holweg and N. Druml, "Next generation
radar sensors in automotive sensor fusion systems," 2017 Sensor Data
Fusion: Trends, Solutions, Applications (SDF), 2017, pp. 1-6.
[14] O. Toker and S. Alsweiss, "mmWave RADAR Based Approach for
Pedestrian Identification in Autonomous Vehicles," 2020 SoutheastCon,
2020, pp. 1-2.
[15] Z. Zhang, X. Wang, D. Huang, X. Fang, M. Zhou and Y. Zhang,
"MRPT: Millimeter-Wave RADAR-Based Pedestrian Trajectory

1776
Authorized licensed use limited to: Uppsala Universitetsbibliotek. Downloaded on September 10,2024 at 17:03:25 UTC from IEEE Xplore. Restrictions apply.