Content uploaded by Goutam Kumar Sahoo
Author content
All content in this area was uploaded by Goutam Kumar Sahoo on Mar 12, 2024
Content may be subject to copyright.
IoT-Based Road Accident Rescue System
Implementation for Smart City Applications
Harshit Srivastava
Dept. of ECE, NIT Rourkela
Odisha, India
harshit_srivastava@nitrkl.ac.in
Goutam Kumar Sahoo
Dept. of ECE, NIT Rourkela
Odisha, India
goutamkr_sahoo@nitrkl.ac.in
D. D. Senanayake
Dept. of ECE, NIT Rourkela
Odisha, India
119ec0210@nitrkl.ac.in
B. A. U. Maduwantha
Dept. of ME, NIT Rourkela
Odisha, India
119me0350@nitrkl.ac.in
M. G. C. Samudrini
Dept. of ECE, NIT Rourkela
Odisha, India
119ec0192@nitrkl.ac.in
A. I. Nanayakkara
Dept. of ECE, NIT Rourkela
Odisha, India
119ec0194@nitrkl.ac.in
Santos Kumar Das
Dept. of ECE, NIT Rourkela
Odisha, India
dassk@nitrkl.ac.in
Poonam Singh
Dept. of ECE, NIT Rourkela
Odisha, India
psingh@nitrkl.ac.in
Abstract—Road accidents are one of the biggest problems in
the world, in which many precious lives have been lost. This
work proposes a road vehicle accident detection system with
location alerts to rescue accident victims. The main hardware
modules include the MPU9250, a 9 degree-of-freedom (9-DoF)
micro electro mechanical system (MEMS) based inertial measure-
ment unit (IMU), Arduino nano with ESP8266 microcontroller
and GPS. The IMU’s three axis accelerometer, gyroscope and
magnetometer data are programmed to determine the orientation
and position of the vehicle. A Wi-Fi-based communication is
established using ESP8266 to send data received from sensor units
to Google Firebase cloud servers in real-time. The performance of
the developed device has been evaluated using a laboratory setup
and also in real-time driving scenarios. The developed sensor
module performs well on accident detection and emergency alert
generation, which can be used in vehicles to save many lives in
the event of an accident through its automatic alert service.
Keywords—Road Accident, Inertial Measurement Unit (IMU),
Smart city, Internet of Things (IoT), Road Safety.
I. INTRODUCTION
Road traffic accidents result in significant economic dam-
ages for both individuals and their families as well as for
entire countries. Governments may incur costs as a result
of car accidents, including those for necessary medical care,
rehabilitation support, and property damage. According to
the WHO, road traffic accidents cost the majority of nations
3% of their gross domestic product [1]. Not only this but
also a major problem is that the responsible authorities can’t
detect when and where the accident has happened and this
latency will lead to a delay in the help victims need. This
system has developed a solution that can be applied in the
actual world to all of the aforementioned difficulties, all of
which can be solved automatically. It has automatic accident
detection and, when an accident involves the vehicle, also
introduces automated accident alerting. The suggested system
will immediately transmit an alert to a registered database with
the necessary information, such as speed and location, after it
detects an accident involving a specific vehicle. In this case, we
have combined already-existing technology into the creation
of this solution. Lack of prompt medical attention for accident
victims at the scene accounts for half of the death rate. We
can learn the precise location of an accident vehicle using the
Global Positioning System (GPS) device’s location tracking
features. This is a special use of the technology that tackles
significant shortcomings in accident reporting and detection.
Impact sensors and GPS provide a quick, impartial, and maybe
acceptable way to discover and report accidents in a society
that is growing increasingly motorised. Additionally, since
this is not a very expensive form of accident detection and
reporting, the majority of people will be able to utilise it in
their cars.
On our roadways, there are several kinds of motor traffic.
These include cars, motorcycles, buses, vans, and trucks. All of
these vehicle categories have been in collisions at some point.
The following categories apply to motor vehicle collisions:-
•Vehicle collisions, in which two or more automobiles
are involved.
•Rollovers, or overturning.
•Collisions with stationary objects, such as rocks, solid
concrete, buildings, and trees, are all examples.
•Knocking against pedestrians and animals.
Crash sensors can detect impacts of varying magnitudes that
occur in every collision [2], [3]. Road Accidents are one
of the leading causes of death, disability, and hospitalization
of people worldwide in general and Indian particular. The
problem is considerably more concerning for low- and middle-
income countries because, despite the fact that these nations
have about 60% of the world’s vehicles, they account for 93%
of all traffic fatalities worldwide. India accounts for at least one
in ten of the global traffic fatalities. Given that India has one of
the greatest road networks in the world, the issue of road safety
is even more crucial. The issue has been made worse by the
unheard-of rate of motorization and the expanding urbanisation
brought on by a rapid rate of economic expansion [4].
Fig. 1 presents a million cities between 2015 and 2020
showed a slightly dropping trend in the percent share of
accidents and injuries. In 2020 the Covid-19 pandemic’s spread
and the ensuing restrictions on movement, notably during lock-
down, greatly decreased the overall number of traffic accidents
2022 IEEE 2nd International Symposium on Sustainable Energy, Signal Processing and Cyber Security (iSSSC) | 978-1-6654-9056-6/22/$31.00 ©2022 IEEE | DOI: 10.1109/iSSSC56467.2022.10051547
978-1-6654-9056-6/22/$31.00 @2022 IEEE
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
Fig. 1: Share of million-plus cities statistics on road accidents.
and fatalities in the country. However, the number of fatalities
in cities with a population of more than a million remained
largely consistent, showing a slight drop in 2020. From 2015 to
2020, the accident severity has shown an upward tendency in
the cities with a population of more than a million, rising from
14.9 percent in 2015 to 23.1 percent in 2020. The police and/or
other agencies receive information on traffic accidents through
a variety of ways, including direct reporting to the police
station, phone calls from unreliable sources, casual discovery
by officers on patrol, etc. This delay will prevent victims
from receiving the aid they require. Here, we have used the
MPU 9250 sensor, arduino nano, ESP8266, GPS module TTL
UART – external antenna, Global Positioning System (GPS)
and internet, programming, and Firebase Realtime Database
which is a cloud-hosted NoSQL database as the technology
involved for the development of the above-mentioned system.
The further discovery of the comprehensive accident data on
Firebase can yield significant information and knowledge that
can be put to good use in reducing the steadily rising number of
motor vehicle accidents on our roads. For instance, black spots
can be easily identified, making it easier to mark such spots
on the roads. The classification of vehicle kinds and classes
that are more likely to be involved in collisions can also be
aided by analysis.
The following is how the paper is set up: The proposed
system and methodology is included in Section II. The exper-
imental studies and result discussion are presented in Section
III and conclusions of this article are presented in Section IV.
II. PROPOSED SYSTEM AND METHODOLOGY
A. Motivation
Accidents are happening every day despite many efforts
made by various agencies. Also, during the post-accident
process, many lives could have been saved if the accident
victims were rescued in time and hospitalized. An efficient
method of accident detection and timely reporting of accidents
is essential to save many lives from danger [5]–[7]. Various
automated accident detection systems have been developed,
some systems are offline, and some systems are real-time
Internet of Things (IoT) systems [8]–[10]. A work by Li
et al. [11] proposed a system for detecting accidents using
accelerometers, airbag sensors and GPS with GSM-based
information. However, the non-availability of airbag sensors
in most cars imposes limitations. Also, smartphone based
crash systems are expensive and prone to false alarms [12].
While IMU sensors are based on MEMS technology, offer
high update rates, are cheap, and mostly contain accelerom-
eters, gyroscopes, and magnetometers for all three axes. It
provides ample information about sudden acceleration, angular
and heading change [13]. In real-time testing, noise is often
embedded with the signal during the data acquisition process
[14]. It would be a difficult task to detect the error, and thus
it is necessary to remove the noise to detect errors in signal
acquisition.
B. Work implementation methodology
1) Methodology: The methodology considered to achieve
the system design is shown in Fig. 2. System hardware,
software and system evaluation follows the normal process
of job implementation work. The hardware components used
are MEMS 9-axis IMU sensor (MPU9250), a GPS module
with external antenna TTL UART, arduino Nano, ESP8266:
a Wi-Fi enabled microcontroller and DC power supply. The
Arduino IDE was used to program the microcontrollers. A
Firebase cloud platform is used for user friendly mobile or
web application creation. The performance of the system is
evaluated based on a threshold-based algorithm.
Fig. 2: Proposed process of work implementation
IMU calibration
Connecting to WIFI and
sign in to the firebase
Accident
happened
?
Acquiring GPS Data
Send data to the firebase
along with IMU Data
Yes No
Send GPS data to the firebase
repeatedly for fixed time interval
End
Start
Fig. 3: Algorithm steps for implementation
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
In the system design phase, an algorithmic flow chart was
designed for the software code as shown in Fig. 3. Based on the
flowchart the actual coding of the software was Implemented
in programming languages. The coding part was done for
ESP8266 and Arduino Nano. Proper execution of the previous
steps ensured smooth and easy implementation of this phase.
In the testing phase the complete coding part was verified.
In the deployment step, the verified code was uploaded to
the board (implementation on the breadboard), and it was
tested by moving the entire circuit. This information will be
given to the fire base. In the final stage proper execution
of all the preceding steps ensured a working system as per
the requirements provided and most importantly also ensured
satisfactory utilization.
2) Implementation Overview: This section outlines the
software, procedures, personnel, and information that were
immediately committed to the successful implementation of
a GPS-based motor vehicle accident detection and reporting
solution. The following software products were used during
the implementation process.
•Arduino IDE: An open-source electronics platform
called Arduino is built on simple hardware and soft-
ware. This software was used to program the ESP8266
and Arduino Nano in this project.
•Firebase: Google has created a platform called Fire-
base for building mobile and web applications. In
this project, Firebase received information about the
accident’s position, location (latitude and longitude),
and speed. The GPS module and MPU9250 sensor
were used to collect those data. The firmware of the
ESP8266 Wi-Fi module was used to send such data.
The following hardware, gadgets and tools were used for
implementation process.
•ESP8266: A family of systems on chip microcon-
trollers that have integrated Wi-Fi and dual-mode
Bluetooth and are inexpensive and low-power.
•Ardunio Nano:The Nano is a small, complete, and
breadboard-friendly board based on the ATmega328
(Arduino Nano). It has more or less the same function-
ality as the Arduino Uno but in a different package. It
lacks only a DC power jack and works with a Mini-B
USB cable instead of a standard one.
•MPU9250: A system in a package (SiP) that combines
two chips: the MPU-6500, which contains a 3-axis
gyroscope, a 3-axis accelerometer, and an onboard
digital motion processor (DMP) that processes com-
plex motion fusion algorithms.
•GPS Module: A good GPS TTL UART module with
external antenna option provides 3.3V - 5V DC.
•Breadboard: Sometimes called a plug block, this is
used to build temporary circuits. This is useful to
designers as it allows components to be removed and
replaced easily.
•Jumper wire: This is an electrical wire that connects
remote electric circuits used for printed circuit boards.
3) Testing: Here all the hardware parts were checked one
by one. First, the microcontroller ESP8266 and Arduino Nano
were checked. Then, the MPU9250 was checked by sending
data to the serial monitor of Arduino Nano and then using
the GPS module with the ESP8266. In this work, GPS and
MPU9250 data will be sent to Firebase by making accounts
of Firebase. This is achieved by adding libraries and some
coding parts to the Arduino code. The hardware components
are assembled with the required details. The entire circuit is
tested with an IoT environment on a breadboard to observe
laboratory-based work performance. In-house testing process
of data acquisition, data transfer to cloud database, decision
making and reporting on accidents are done. The details of
the test results will be discussed in the following sections.
III. EXPERIMENTAL STUDIES AND RESULTS DISCUSSION
This section provides experimental evaluation of the system
and results discussion. After the implementation and testing
part, the prototype was evaluated for the purpose of determin-
ing whether the developed system was giving the expected
results. The expression for absolute linear acceleration (ALA)
is defined as
ALA =(Accx2+Accy2+Accz2)(1)
Where, Accx = acceleration in Xdirection, Accy = accelera-
tion in Ydirection and Accz = acceleration in Zdirection.
A. Assumptions and laboratory setup for system evaluation
•We have reduced the ALA limit from 3.5 to 1.5 for
testing purposes.
•If the user chooses to stop auto monitoring through
the app, we assume that the driver is safe, and the
device will not send any accident-related information
to the cloud.
•We assume that this device is intended to be used on
a normal vehicle such as Maruti 800.
•The user must have a stable internet connection.
•Must have a reliable GPS connection.
•The device must be mounted on a flat surface relative
to the vehicle.
The experimental setup for laboraty based model testing
can be seen from the Fig. 4. The hardware components are
assembled in breadboard to develop the complete circuit.
The DC power to the circuit can be given from external
power bank or from car battery supply. The hardware module
implementation of GPS, MPU9250, ESP8266 and Arduino
Nano is shown in Fig. 5.
MPU9250 processing is implemented in Arduino nano
board and GPS & cloud Firebase is processed in the ESP8266
board. Optimizing the code until expected results come. The
hardware connection circuit board is developed to test in real
time with vehicle and fine tuning threshold values for required
operations. The process of data acquisition and cloud base data
storage can be seen from the Fig. 6.
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
Fig. 4: Experimental study setup in laboratory implementation
Fig. 5: The breadboard circuit implementation.
Fig. 6: A screenshot of the data acquisition and storage process.
B. Real-time experimental result analysis
In normal driving scenario the ALA threshold was set to
3.5g as the real world ALA threshold should be between 3.5g
to 5g to detect fatal accidents. In uneven road condition the
ALA threshold was set knowingly to 2g for testing purposes.
we customize the ALA threshold value to know that how
many times ALA value go beyond 2g during our testing.
Our setup was able to identify it.In a real-world scenario,
the ALA threshold was set to a maximum value of 5g to
detect any abnormal event, which occurs when it crosses the
5g level, regardless of vehicle or road conditions. As accident
is an unpredictable event and it occurs some times but in our
testing for safety purpose we verified the IMU sensor impact
for various normal test cases under control environment with
safe drive. Various real time test cases are conducted in a real
road environment with controlled non-accident scenario. Fig.
7 shows the hardware made for in-vehicle use in real-world
testing.
Fig. 7: Designed hardware module for in-vehicle testing.
While testing our device, we used the Timestamp Camera
Application version 1.212. That app displays a watermark of
time with millisecond accuracy, a GPS coordinates, a location
address, a location with a map, the vehicle’s speed, heading,
and altitude on top of the video in real-time. It made it much
easier for us to document the road conditions and obstacles we
encountered during the test drive. A screen-shot of the Time-
stamp Camera App based result is shown in Fig. 8.
Fig. 8: Road scenario with time-stamp application result.
C. Comparison with existing works
The proposed system will offer much more features than
what is provided by state-of-the-art systems. Table I gives the
comparative studies with existing systems.
IV. CONCLUSIONS AND FUTURE WORKS
This study has designed and created an IoT-based automo-
tive test-bed environment that automatically detects collision
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
TABLE I: Comparative studies on IoT-based system for accident detection
Reference Advantages Limitations
Sarker et al. [15]
•Developed an intelligent accident detection, location tracking and
notification system using GPS and GSM.
•provides a low-cost portable system for accident detection and
location tracking.
•The use of ultrasonic sensor imposes limitations as it operates for
short distances. Sometimes it predicts the collision wrong.
Nasr et al. [16]
•A smart and reliable IoT system with shock sensors detects road
accidents and notifies immediately with their geographic coordinates.
•It provides basic medical information to rescue teams, identifies
precise accident locations and facilitates the routing process.
•A minimum of three alarms must be sent to confirm an accident. If
one is sent, it is considered a fault alarm.
•The load on the server is not considered as the number of transactions
is limited by the number of crashes during a period of time.
White et al. [12]
•Shows how smartphone sensors and their processing capabilities
can be used to address the challenges of detecting traffic accidents
without direct interaction with the vehicle’s onboard sensors.
•Reduced software maintenance complexity through smartphone ap-
plication upgrade mechanism.
•The destruction of the smartphone can prevent the distribution of
accident information.
•The proposed system consumes a significant amount of battery
power.
•Low speed traffic can cause WreckWatch to deactivate.
Navidi et al. [17]
•Provides GPS/INS-based navigation algorithms, calibration of navi-
gational sensors, and a de-noshing method for vehicle accidents.
•Accident detection using GPS/INS sensors and providing emergency
information using internet along with latitude and longitude of the
accident location.
S. Soma [18]
•The developed system sends short messages to WhatsApp of a
mobile number via Wi-Fi over the Internet in case of a crash.
•It monitors vehicle speed, detects vehicle location, sends alert
messages to mobile phones for remote information and provides for
changing mobile numbers at any time.
•System development is expensive and data transfer is not secure.
•It requires a strong internet network connection.
Chaudhariet al. [19]
•Provides a threshold-based crash detection mechanism for measuring
vibration followed by automatic alerts using GPS and GSM.
•In case of false alerts, the vehicle occupant can stop emergency
assistance requests by pressing a button.
•System development uses motion sensor module and GSM/GPRS
module. However, real-time performance requires faster processing.
Lakshmyet al. [20]
•Developed an accident prevention mechanism through automatic
engine locking and abnormal vibration detection followed by alcohol
detection.
•Deep-learning-based accident prediction using the accident scene
image captured by the vehicle’s front-facing camera.
•Information sharing to the nearest emergency center using GPS and
GSM module.
•System development uses multiple hardware devices, however, fast
processing is required for real-time performance.
•Keras is the neural network API used here with TensorFlow as the
back-end which requires high-end processors.
Proposed
•Development of an IoT-based in-vehicle embedded system for acci-
dent detection for reduction in accident fatality rate.
•A low-cost Wi-Fi-based communication system implementation us-
ing IMU sensors, Arduino Nano, ESP8266 microcontroller, and GPS
to perform processing and location tracking.
•The ESP8266 is used as a dedicated device that sends the data
received from the sensor units to the Google Firebase cloud server,
therefore reducing latency.
•The performance of the developed device was also evaluated in a
laboratory setup and in real-time driving scenarios.
•It can be used in vehicles to save many lives in the event of an
accident.
•A crash detection signal needs to be transmitted via the Internet
to a cloud server application. Internet connectivity issues can affect
performance.
and roll-over and notifies the appropriate organizations right
away. But in here, we are testing accidents causing above 3.5g
force, roll over, and free fall. Also this device won’t detect
small accidents such as scratches, shattering, etc. and it only
supports a WIFI bandwidth of 2.4GHz. To get the most accu-
rate results this device should place away from strong magnets
such as speakers and all other electromagnetic interference.
Not only that but also the GPS antenna should place where we
can get a strong signal and the supply power must be within
the given rating. The relevant authorities can readily obtain
the precise location by using this prior information, can help
in saving many lives in this process. Additionally, by using
the data in the database and the stored data in the cloud, they
can identify important information such as the locations and
times where many accidents occur, etc. in future, which may
enable the relevant authorities to make policies and put them
into practice effective measures to lower the number of injuries
and fatalities brought on by traffic accidents.
REFERENCES
[1] W. H. O. D. of Violence, I. Prevention, W. H. O. Violence, I. Prevention,
and W. H. Organization, Global status report on road safety: time for
action. World Health Organization, 2009.
[2] M. S. Shaheed and K. Gkritza, “A latent class analysis of single-vehicle
motorcycle crash severity outcomes,” Analytic Methods in Accident
Research, vol. 2, pp. 30–38, 2014.
[3] S. Komarizadehasl, B. Mobaraki, J. A. Lozano-Galant, and J. Turmo,
“Evaluation of low-cost angular measuring sensors,” in Proc. Int.
Conf. Recent Trends in Geotech. Geo-Environmental Eng. and
Edu.(RTCEE/RTGEE), 2020, pp. 23–25.
[4] ROAD ACCIDENTS IN INDIA 2020 lR., ROAD ACCIDENTS IN
INDIA 2020. [Online]. Available:~https://morth.nic.in.
[5] G. K. Sahoo, S. A. Patro, P. K. Pradhan, S. K. Das, and P. Singh, “An
IoT-based intimation and path tracing of a vehicle involved in road
traffic crashes,” in Proc. IEEE-HYDCON, 2020, pp. 1–5.
[6] G. K. Sahoo, P. K. Pradhan, S. K. Das, and P. Singh, “A user specific
APDS for smart city applications,” in Research in Intelligent and
Computing in Engineering. Springer, 2021, pp. 267–277.
[7] K. Parasana, G. K. Sahoo, S. K. Das, and P. Singh, “A health perspective
smartphone application for the safety of road accident victims,” in 2021
Advanced Communication Technologies and Signal Processing (ACTS).
IEEE, 2021, pp. 1–6.
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
[8] N. Kumar, D. Acharya, and D. Lohani, “An IoT-based vehicle accident
detection and classification system using sensor fusion,” IEEE Internet
of Things Journal, vol. 8, no. 2, pp. 869–880, 2020.
[9] N. Pathik, R. K. Gupta, Y. Sahu, A. Sharma, M. Masud, and M. Baz,
“AI enabled accident detection and alert system using IoT and deep
learning for smart cities,” Sustainability, vol. 14, no. 13, p. 7701, 2022.
[10] E. Fantin Irudaya Raj and M. Appadurai, “Internet of things-based smart
transportation system for smart cities,” in Intelligent Systems for Social
Good. Springer, 2022, pp. 39–50.
[11] C.-z. Li, R.-f. Hu, and H.-w. Ye, “Method of freeway incident detection
using wireless positioning,” in Proc. IEEE International Conference on
Automation and Logistics, 2008, pp. 2801–2804.
[12] J. White, C. Thompson, H. Turner, B. Dougherty, and D. C. Schmidt,
“Wreckwatch: Automatic traffic accident detection and notification with
smartphones,” Mobile Networks and Applications, vol. 16, no. 3, pp.
285–303, 2011.
[13] M. S. Amin, M. B. I. Reaz, S. S. Nasir, and M. A. S. Bhuiyan, “Low
cost gps/imu integrated accident detection and location system,” Indian
J. Sci. Technol, vol. 9, no. 10, pp. 1–9, 2016.
[14] B. B. Pradhan, S. Ari, G. K. Sahoo, D. K. Jena, S. K. Patra, and
R. Appavuraj, “Wavelet transform based error detection in signal
acquired from artillery unit,” in Proc. IEEE 1st Int. Conf. Condition
Assessment Techn. Electrical Syst. (CATCON), 2013, pp. 243–248.
[15] S. Sarker, M. S. Rahman, and M. N. Sakib, “An approach towards
intelligent accident detection, location tracking and notification system,”
in Proc. IEEE Int. Conf. Telecommun. Photonics (ICTP), 2019, pp. 1–4.
[16] E. Nasr, E. Kfoury, and D. Khoury, “An iot approach to vehicle accident
detection, reporting, and navigation,” in Proc. IEEE Int. Multidisci-
plinary Conf. Eng. Technol. (IMCET), 2016, pp. 231–236.
[17] N. Navidi et al., “Collision vehicle detection system based on gps/ins
integration,” Journal of Computer and Communications, vol. 5, no. 2,
pp. 48–70, 2017.
[18] S. Soma, “Iot based vehicle accident detection and tracking system
using gps modem,” International Journal of Innovative Science and
Research Technology, 2017.
[19] A. Chaudhari, H. Agrawal, S. Poddar, K. Talele, and M. Bansode,
“Smart accident detection and alert system,” in Proc. IEEE India
Council Int. Subsec. Conf. (INDISCON), 2021, pp. 1–4.
[20] S. Lakshmy, R. Gopan, M. Meenakshi, V. Adithya, and M. R. Elizabeth,
“Vehicle accident detection and prevention using iot and deep learning,”
in Proc. IEEE Int. Conf. Signal Process., Informatics, Commun. Energy
Syst. (SPICES), vol. 1, 2022, pp. 22–27.
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA. Downloaded on April 14,2023 at 06:55:11 UTC from IEEE Xplore. Restrictions apply.
