2020 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2020), 20 June 2020, Shah Alam,

Malaysia.

IR-UWB Radar Sensor Based Object Distance

Determination as a Preliminary Input for Human
Motion Detection
Terence Jerome Daim Razak Mohd Ali Lee
Faculty of Engineering Faculty of Engineering
Universiti Malaysia Sabah Universiti Malaysia Sabah
Kota Kinabalu, Malaysia Kota Kinabalu, Malaysia
[email protected] [email protected]

Abstract—Detection of human motion is a method of diffraction, shadowing, scattering and reflection caused by a
detection in which various methods and instruments are person or object on the effective area of the transmitted
combined to assess and identify the condition of a human electromagnetic signals. The received electromagnetic
motion. The function of human motion detection in its signals then travels and are received back by the IR-UWB
original form is not immediately provided by IR-UWB
radar sensors. The changes in signal amplitude as well as the
radar sensors readily available on the market for
purchase nowadays. The distance information between the Time of Arrival (ToA) of the reflected UWB impulse are used
radar and any detected object observed within range is one of to detect and identify the various states of human motion.
the data sources widely used to formulate human motion Human motion detection techniques were originally
detection. This paper introduces an investigation work on based on vision sensors such as video camera surveillance
distance information as generated by the IR-UWB radar and infrared motion detectors. However due to the limitation
sensor with the purpose of creating a new human motion of visible line-of-sight (LOS) required in these techniques,
detection technique based on the IR-UWB radar sensor. The the research focus on human motion detection is gradually
precision of calculating an object's distance precisely a diverted into using electromagnetic waves signals based,
human right hand is practical for the reason described.
where the received signals changes are assessable due to the
Keywords— distance information, IR-UWB radar, human presence of person or object.
motion detection Most common off-the-shelf IR-UWB radar sensors
readily available in the market for purchase nowadays do not
straight away provide the function of human motion detection
I. INTRODUCTION in its original form. Generally, the data output from these
sensors are required to be augmented or manipulated in one
Human motion detection is a type of detection where
way or another using various processing techniques in order
combination of various techniques and tools used in order to
to get the sensor to have the function of human motion
determine and recognize the state of a human motion. Such
detection.
detection has many merits towards a number of possible
One of the data output that has been commonly used in
applications that many of them have been studied and
the process of formulating human motion detection is the
developed. Among the many techniques developed to sense
distance information between the IR-UWB radar sensor and
human motion, detection by using IR-UWB radar sensor has
any detected object within range. In general and ideal
recently become an area of research focus [1]-[2]. The reason
condition, the distance information as produced by the IR-
is in its current form factor and hardware specifications, a
UWB radar sensor will corresponds accordingly to the
typical IR-UWB radar sensor has most of modern consumer
distance of a detected object. Ideally, the greater the distance
applications require features such as size, transportability,
of a detected object from the IR-UWB radar sensor, the
low power consumption and limited processing resources to
greater the distance information that the IR-UWB radar
detect complex, human motion in contact-less manner. In
sensor will register. However, this is not as simple as to use
addition, being a technology that offers unique capability of
the data output directly in real condition as the accuracy in
penetrability, high temporal resolution [3] and emits non-
providing distance information may not be straightforward
ionizing electromagnetic waves [4]-[7], IR-UWB radar
and useful. Therefore some form of processing algorithm
sensor is a promising merger of radar and UWB technology
needs to take place in order to fulfil the requirement.
that can also be employed in a number of useful application.
This paper presents a preliminary investigation work to
Some of the applications are such as through-wall
determine the relationship between the calculated distance
surveillance, post-disaster search & rescue operation and
information based on the data output by an IR-UWB radar
contactless measurement for medical application [8]-[13].
sensor and the true distance of a detected object for the
In order to approximate the state or form of a person or intention of developing a new IR-UWB radar sensor based
object, human motion detection using IR-UWB radar sensor human motion detection technique.
in general takes advantage of the natural occurrence of

978-1-7281-6133-4/20/$31.00 ©2020 IEEE 201

Authorized licensed use limited to: University of Melbourne. Downloaded on August 02,2020 at 18:20:50 UTC from IEEE Xplore. Restrictions apply.
2020 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2020), 20 June 2020, Shah Alam, Malaysia.

IR-UWB Radar Sensor

Transmitter UWB Impulse

Signal transmission

Transmitted UWB Impulse signal

Tx Antenna
Analog to
Receiver Digital Object
Converter
Reflected UWB Impulse signal

Bandpass
filter Rx Antenna

Digitized reflected
received UWB Impulses
signal

Fig.1 IR-UWB radar sensor basic operation

II. METHODOLOGY B. Experiment Setup

Fig. 2 shows the IR-UWB radar sensor used in this work.
A. System Modelling The manufacturer of the IR-UWB radar sensor is NOVELDA
Fig. 1 shows the basic operation of an IR-UWB radar and the specific model used is X4M300 [14]. Fig. 3 depicts
sensor commences with transmission of UWB impulse the experiment setup that is used to investigate the distance
signal ( ). If the transmitted signal hit an object within its between IR-UWB radar sensor and a detected object within
path and coverage, the reflected UWB impulse signal then range. In order to allow the IR-UWB radar sensor to sense
journeys back and are received by the IR-UWB radar sensor and measure a detected object distance, this work has selected
through its receiving antenna. The received signal will then three different objects to serves as individual target. The
be sampled by an analog to digital converter and then through objects selected are first a 6cm square cube that is made of
a bandpass filter for the final output processed signal. The plastic and has aluminum reflective surfaces on all sides. The
following is the expression of the final processed received second object is also a 6cm square cube made of plastic but
signal ( ) : with transparent surfaces on all sides. The third object is a
human right hand. The purpose of testing an objects with
reflective and transparent surfaces respectively is to test the
( ) = ( − )+ ( ) radar sensor capability in sensing objects with different types
of surfaces.
(1) For the purpose of this experiment, the measurement for
each of the individual target begins from 5cm distance from
Where is the delay and is the amplitude of -th the IR-UWB radar sensor and gradually increase with 10cm
received signal, is the digitized sample index while and interval until reaching maximum distance of 205cm
( ) represent the number of UWB impulse signals reflected respectively. The distance information is calculated based on
back from the surroundings and associated noise the data output extracted from the IR-UWB radar sensor. The
respectively. When an object is detected at -th received calculation sequence is described in the next section. Fig. 4,
signal, the distance from the IR-UWB radar sensor to the 5 and 6 show examples of the individual object positioned in
object, can be calculated by equation (2) below: this experiment respectively.

×
= (2)

Where c is the speed of light, 3 × 10 / , is the delay

due to the reflected UWB impulse signals on the object. Since
the signal has to journeys twice the distance to the object and
back, a divisor of two is required.

Fig.2 IR-UWB radar sensor

978-1-7281-6133-4/20/$31.00 ©2020 IEEE 202

Authorized licensed use limited to: University of Melbourne. Downloaded on August 02,2020 at 18:20:50 UTC from IEEE Xplore. Restrictions apply.
2020 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2020), 20 June 2020, Shah Alam, Malaysia.

IR-UWB Radar Sensor

Reflected UWB Impulses

Transmitted UWB Impulses
Object
Processing Terminal

Measured distance

Fig.3 Test setup for distance between an object and IR-UWB radar sensor.

Fig.4 Example placement of a square cube reflective object (a) top view; (b) side
i

Fig.5 Example placement of a square cube transparent object (a) top view; (b) side
i

Fig.6 Example placement of a human right hand (a) top view; (b) side view

978-1-7281-6133-4/20/$31.00 ©2020 IEEE 203

Authorized licensed use limited to: University of Melbourne. Downloaded on August 02,2020 at 18:20:50 UTC from IEEE Xplore. Restrictions apply.
2020 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2020), 20 June 2020, Shah Alam, Malaysia.

C. Distance Information Calculation Sequence III. RESULT

Fig. 7 shows an example of data output by the IR-UWB Fig. 8, 9 and 10 shows the results of the experiment
radar sensor without any form of configuration. If an object conducted for the distance measurement of a square cube
within the sensing path or area of the IR-UWB radar sensor, reflective object, a square cube transparent object and lastly
the pattern of the normalized amplitudes of this initial data a human right hand respectively. Table 1 shows the
output will change according to the position and movement corresponding distance with respect to the object position
of the object. The information provided by this initial data number.
output in this form does not fulfil the requirement in this work
where a distinctive distance information between the IR- TABLE I. CORRESPONDING POSITION NUMBER AND DISTANCE
UWB radar sensor and a detected object is required. Position Number # Distance (cm)
Therefore some form of processing algorithm needs to take 1 5
place in order to fulfil the requirement. 2 15
3 25
4 35
5 45
6 55
7 65
8 75
9 85
10 95
11 105
12 115
13 125
14 135
15 145
16 155
17 165
18 175
Fig.7 Example IR-UWB radar sensor data output without any form of 19 185
configuration 20 195
21 205

A program specifically developed for the purpose of this

work to transform the raw data output into distance
measurement is used. The following describe the sequence of
measurement program when performing a measurement:

i. Start and initiate the IR-UWB radar sensor.

ii. Start the measurement program and wait until the
normalized amplitudes of the raw data output is in a
stagnant condition.
iii. Record the data points initial normalized
amplitudes, where = 1 … is the data point
number. is the maximum data point number.
iv. Position an object on the measurement point (Figure
3, 4 and 5) for at least 5 seconds or until the
normalized amplitudes is in a stagnant condition. Fig.8 Calculated distance for a reflective square cube object
v. Record the data points normalized amplitudes again,

vi. The difference, is then calculated by using the

following equation:

= | − | … | − |

vii. The estimated distance of the detected object will be

determined by comparing data points and then
select the corresponding highest calculated
amplitudes difference.

Fig.9 Calculated distance for a transparent square cube object

978-1-7281-6133-4/20/$31.00 ©2020 IEEE 204

Authorized licensed use limited to: University of Melbourne. Downloaded on August 02,2020 at 18:20:50 UTC from IEEE Xplore. Restrictions apply.
2020 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2020), 20 June 2020, Shah Alam, Malaysia.

REFERENCES

[1] J. Park and S. H. Cho, "IR-UWB Radar Sensor for Human Gesture
Recognition by Using Machine Learning," 2016 IEEE 18th
International Conference on High Performance Computing and
Communications; IEEE 14th International Conference on Smart City;
IEEE 2nd International Conference on Data Science and Systems
(HPCC/SmartCity/DSS), Sydney, NSW, 2016, pp. 1246-1249.
[2] T. Ha and J. Kim, "Detection and Localization of Multiple Human
Targets Based on Respiration Measured by IR-UWB Radars," 2019
IEEE SENSORS, Montreal, QC, Canada, 2019, pp. 1-4.
[3] A. Yarovoy, L. Ligthart, J. Matuzas, and B. Levitas, “UWB radar for
human being detection,” IEEE Aerosp. Electron. Syst. Mag., vol. 21,
no. 3, pp. 10–14, Mar. 2006.
[4] B. Lee, S. Lee, Y. Yoon, K. Park and S. Kim, "Adaptive clutter
suppression algorithm for human detection using IR-UWB radar,"
2017 IEEE SENSORS, Glasgow, 2017, pp. 1-3.
[5] Admir Muric, Christos Anestis Georgiadis, Fisayo Caleb Sangogboye,
Fig.10 Calculated distance for a human right hand and Mikkel Baun Kjærgaard. 2019. Practical IR-UWB-based Occupant
Counting evaluated in Multiple Field Settings. In Proceedings of the
1st ACM International Workshop on Device-Free Human Sensing
For a square cube reflective object, the best achieved (DFHS’19). Association for Computing Machinery, New York, NY,
measurement points with the least errors are at position USA, 48–51.
number 1,2,3,9 and 15. These also corresponds to a Mean [6] B. Benamrouche and D. Dragomirescu. 2017. Energy efficient IR-
Absolute Percentage Error (MAPE) of 17%. As for UWB transceiver for autonomous wireless sensor network. In
Proceedings of the Second International Conference on Internet of
transparent square cube object and human right hand, the best things, Data and Cloud Computing (ICC ’17). Association for
achieved measurement points with the least errors are at Computing Machinery, New York, NY, USA, Article 129, 1–4.
position 10 and 17 and 3, 7, 9 and 10 respectively. The [7] Xiaohui Chen, Sheng Xu, Huarui Yin, Weidong Wang, and Guo Wei.
calculated MAPE for transparent square cube object is 27.6% 2011. Potentials of IR-UWB technology for ubiquitous computing.
while human right hand is 11.7%. From these results, it can Personal Ubiquitous Comput. 15, 1 (January 2011), 75–84.
be determined that human right hand achieved the best result [8] A. Pantelopoulos and N. G. Bourbakis, “A Survey on Wearable Sensor-
Based Systems for Health Monitoring and Prognosis,” IEEE Trans.
among the three objects when calculating the distance Syst. Man, Cybern. Part C (Applications Rev., vol. 40, no. 1, pp. 1–12,
between the object and the IR-UWB radar which resulted 2010.
with the least MAPE of 11.7%. [9] L. Lampe and K. Witrisal, “Challenges and recent advances in IR-
UWB system design,” ISCAS 2010 - 2010 IEEE Int. Symp. Circuits
IV. CONCLUSION Syst. Nano-Bio Circuit Fabr. Syst., pp. 3288–3291, 2010.
[10] J. R. Fernandes and D. Wentzloff, “Recent advances in IR-UWB
In this paper, a preliminary investigation work to transceivers: An overview,” ISCAS 2010 - 2010 IEEE Int. Symp.
determine the relationship between the calculated distance Circuits Syst. Nano-Bio Circuit Fabr. Syst., pp. 3284–3287, 2010.
information based on the data output by an IR-UWB radar [11] A. Alarifi et al., “Ultra Wideband Indoor Positioning Technologies:
sensor and the true distance of a detected object has been Analysis and Recent Advances,” Sensors, vol. 16, no. 5, p. 707, 2016.
carried out. The purpose of which the result of this work will [12] Z. Irahhauten, H. Nikookar, and G. J. M. Janssen, “An overview of
be used as a preliminary requirement for the development of ultra wide band indoor channel measurements and modeling,” IEEE
Microw. Wirel. Components Lett., vol. 14, no. 8, pp. 386–388, 2004.
a new IR-UWB radar sensor based human motion detection
[13] G. Llano, J. Reig, L. Rubio, A. P. García, and U. P. De Valencia,
technique. As can be found from the result section, the “Ultra-wideband Frequency Analysis : State-of-the-art , Measurements
accuracy in determining the distance of an object specifically and Modeling,” pp. 147–154, 2009.
a human right hand is usable for the mentioned purpose. [14] Novelda, "X4M300 Datasheet", Datasheet, www.xethru.com
However, the present of large errors at certain position points
has to be addressed in future works.

978-1-7281-6133-4/20/$31.00 ©2020 IEEE 205

Authorized licensed use limited to: University of Melbourne. Downloaded on August 02,2020 at 18:20:50 UTC from IEEE Xplore. Restrictions apply.