Received 11 March 2024, accepted 21 March 2024, date of publication 28 March 2024, date of current version 3 April 2024.

Digital Object Identifier 10.1109/ACCESS.2024.3382508

IoT-Enhanced Transport and Monitoring of

Medicine Using Sensors, MQTT, and
Secure Short Message Service
DAVID SAMUEL BHATTI 1 , MUHAMMAD MUEED HUSSAIN2 , BEOMKYU SUH3 ,
ZULFIQAR ALI 4 , ISMATOV AKOBIR5 , AND KI-IL KIM 6
1 Facultyof Information Technology, University of Central Punjab, Lahore 54000, Pakistan
2 Department of Computer Science, Institute of Management Sciences, Lahore 54000, Pakistan
3 Department of Computer Science and Engineering, Chungnam National University, Daejeon 34134, Republic of Korea
4 Department of Computer Science, National University of Technology, Islamabad 44000, Pakistan
5 Department of Computer Science and Engineering, Chungnam National University, Daejeon 34134, Republic of Korea
6 Department of Computer Science and Engineering, Chungnam National University, Daejeon 34134, Republic of Korea

Corresponding author: Ki-Il Kim ([email protected])

This work was supported by Basic Science Research Program through the National Research Foundation of Korea(NRF) funded by the
Ministry of Education (RS-2023-00237300) and Institute for Information & communications Technology Planning & Evaluation(IITP)
grant funded by the Korea Government, [Ministry of Science and ICT (MSIT)], (No. 2022-0-01200, Convergence security core talent
training business (Chungnam National University)).

ABSTRACT Since its inception more than a decade ago, Internet of Things (IoT) technology has been
guiding people in the development of a world full of smart solutions in which all devices and physical
objects, represented as ‘‘things,’’ are interlinked with sensors using the Internet. In some areas, the delivery
of medications to patients or receivers at their destinations remains highly outdated and informal. In smart
medicine delivery, the medicine needs to maintain its original state while facing multiple environmental
factors, such as temperature fluctuations, humidity, etc. This paper presents an effective implementation of
IoT (Internet of Things) for monitoring the transportation of medicines and vaccines, along with temperature
control facilitated through mobile applications and sensor networks. The system employs mobile applications
as the user interface, utilizes Arduino, MQTT (Message Queuing Telemetry Transport) for communication,
incorporates a temperature sensor, and employs a mini portable cooling box. Designed for the generalized
delivery of medicines/vaccines from sender to receiver, the system also suggests CRC-32 as an optimal
algorithm for error detection instead of complex hash functions such as MD5 and SHA, ensuring better
performance, smooth operation, and data integrity. In addition, elliptic curve-based shared keys are used for
protected data transmission. The accuracy of the proposed system is 89.88%, and the value of the F1-score is
0.686, which is greater than the threshold of 0.5, hence conclusively validating the authenticity and reliability
of the proposed system.

INDEX TERMS Arduino, Internet of Things, web server, medicine, vaccine, temperature sensor.

I. INTRODUCTION to a virtual or physical environment [2]. IoT solutions are

IoT and sensor communication play a vital role in daily investing significantly in propelling the worldwide IoT in
life [1]. Internet of Things refers to everything connected smart transportation and medical solutions [3]. In order to
to everything else through the internet, in which different function properly in IoT-based smart transportation systems,
nodes are connected to the network and each other in numerous mechanisms and algorithms are being used [4].
order to process information, make decisions, and respond These systems play a crucial role in the operation of
railway systems [5], roads [6], flights [7], and ships [8],
The associate editor coordinating the review of this manuscript and significantly enhancing the overall customer experience by
approving it for publication was Barbara Masini . improving the efficiency, tracking, and forwarding of goods
2024 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
46690 VOLUME 12, 2024
D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

TABLE 1. A case study of cost analysis of medicine waste. This paper presents a robust and user-friendly system
developed for monitoring the transportation of medicine,
ranging from distributors, pharmacies, or individuals as
senders to recipients, whether individuals or pharmacies.
Additionally, the system includes a temperature control
feature for refrigerating the medicines during transit. The
proposed solution encompasses both hardware and software
components, providing a straightforward user interface for
ease of use. Furthermore, to ensure message integrity and
facilitate error detection during communication, this paper
will implement a seamless mechanism aimed at enhancing
in transportation. The Internet of Things (IoT) is radically system efficiency and optimizing battery usage. A careful
altering the transportation industry [9]. Smart transportation selection of the most suitable checksum mechanism for the
systems are enhancing the mobility of both people and proposed solution will be made among CRC-32, MD5, and
goods, contributing to the overall improvement of the SHA-1 [15].
economy, general well-being, security, and environmental This manuscript is organized in different sections, which
sustainability. IoT and sensor networks collaboratively exert are I-Introduction, II-Related Work, III-Proposed Model, VI-
significant influence on the healthcare system, offering Limitations and Challenges, V-New Temperature Detection
solutions that facilitate patient care through continuous Methods, IV-Results and Discussion, VII-Conclusion and
monitoring of physiological changes and physical activity. Future Work.
Because of these technologies, vital data such as heartbeat,
blood pressure, and sugar levels are consistently recorded A. CONTRIBUTIONS
and transmitted directly to relevant healthcare professionals, The chosen issue, addressing the secure transportation
even when the patient is not physically present in the doctor’s of medicines, is unique, and the proposed framework is
office [10]. genuinely innovative with the following key contributions:
Moreover, the healthcare industry operates not only across
1) The solution utilizes advanced technologies like sen-
different cities but also spans international borders, employ-
sors, microcontrollers, resource efficient MQTT proto-
ing aeroplanes and ships to facilitate the swift transportation
col, and cellular networks to ensure the safe delivery of
of medicines [11]. In the healthcare sector, mere trans-
medicines.
portation alone is insufficient to meet the intended purpose;
2) Elliptic curve-based shared secret keys are generated
the safety of medicines during travel is equally critical.
for multi-party protected communication.
In contrast to the large fridges and deep freezers traditionally
3) The system ensures equal access to monitoring data for
utilized in pharmacies for medicine management, there are
all stakeholders involved in medicine transportation,
now compact and intelligent fridges/cooling containers that
preventing deception due to adverse environmental
can be conveniently carried along with the medicines during
conditions like temperature variations.
a journey [12]. GPS trackers play a crucial role in monitoring
4) The proposed system has the potential to save lives
the movement of delivery vehicles. The industry has adeptly
by preventing expired medicines and reducing phar-
leveraged these innovative technologies, integrating them
maceutical losses by incorporating sensing and backup
seamlessly to achieve their objectives promptly and securely.
hardware.
However, a significant gap persists in fully monitoring
the transportation of medicines and remotely managing the
temperature of the system. This gap has led to a dramatic II. RELATED WORK
increase in the wastage of medicines, posing a severe risk A preliminary system for a drone-based shipment was devel-
to people’s lives through the consumption of expired drugs. oped by Gatteschi et al. [16]. In this study, the architectural
The utilization of expired medicines or vaccines creates a framework of the program is initially explored, focusing on
substantial risk of fatalities among consumers. Table 1 shows hardware specifications. Subsequently, a software framework
the impact of medicine expiry and its wastage on human is outlined to manage and respond to shipment requests.
lives and the economy. Visually assessing whether a medicine The study particularly addresses challenges associated with
or vaccine has been exposed to detrimental temperature fully autonomous parcel deliveries, proposing potential
extremes is impractical. In many cases, there is no significant solutions to preemptively mitigate issues within this specific
change, even after harmful exposure. Stakeholders involved application scope. Given the urgency often associated with
in medicine and vaccine transportation remain unaware of medication deliveries and the feasibility of drones for small
the medicine’s condition. Key factors such as temperature, and lightweight packages, the system is customized for
humidity, the transporting vehicle, the container or box drug consignments. However, it is crucial to acknowledge
housing the medicine, and the associated hardware have a that the system model could potentially adapt to various
direct impact on the medicine’s state. other scenarios. While proficient for short-range deliveries,

VOLUME 12, 2024 46691

 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

shipment. Additionally, the solution advocates for effective

communication between stakeholders, emphasizing the need
for mutual agreement on delivery timelines. It asserts that
both the sender and receiver should coordinate and confirm
the delivery time frame based on the distance between them,
enabling proper arrangements for a successful and timely
delivery process.
Konovalenko and Ludwig [21] given a drone-mediated
medicine distribution system to remote locations has wit-
nessed increased attention in recent years. However, there is
a notable scarcity of information pertaining to the feasibility
and impacts of drone delivery for medications, underscoring
the significance of this study. Employing pharmacopoeial
methodologies, the study delves into the effects of vibration
and temperature on insulin, simulating crucial parameters
during drone flight. The research demonstrates the preser-
FIGURE 1. 30-day electronic refrigerator temperature logger.
vation of insulin quality through human insulin testing,
even after exposure to conditions mimicking a 30-minute
drone delivery (temperatures ranging from −20 to +40
it is important to note that the system is not designed to ◦C, vibration frequencies between 0 and 40 Hz). With
accommodate large parcels. the exception of a minor bubble formation post-takeoff,
By Nanda et al. [17], Blockchain technology is seam- no adverse effects were observed on Act fast following
lessly integrated with IoT to enhance health supply chain drone transportation. To the best of the authors’ knowledge,
management. The proposed solution introduces decentral- this marks the inaugural instance of utilizing a drone to
ized tracking and tracing of medical products, effectively assess insulin quality, providing compelling evidence for the
eliminating the risks associated with counterfeit medications. feasibility of drone transportation in maintaining pharma-
It ensures real-time product status updates throughout the ceutical stability and medication quality. Consequently, this
distribution process from the manufacturer to the end- study establishes a foundation for future research ventures
user. Smart contracts, designed using the solidity program- exploring diverse drug classes, environmental influences,
ming language and implemented on the public permission or alternative drone models. Loisel et al. [22] underscore the
Ethereum blockchain, play a pivotal role in ensuring security, critical importance of temperature monitoring in medicine
transparency, and trust in the system. This strategy offers a to ensure patient safety and maintain product quality using
means to reduce errors and eliminate fraud, minimize delays, machine learning approaches. Despite significant advance-
limit human interaction, enhance efficiency and accuracy, ments in temperature monitoring devices driven by the
improve inventory management, and ultimately reduce costs. Internet of Things (IoT), challenges persist, particularly in the
WHO pre-qualified and re-validated in 2009 system is Cold Chain of medicine. Balachandar and Chinnaiyan [23]
featured, equipped with the capability to store temperature introduced a unique conceptual model that comprehensively
readings for 30 days [18]. The system is designed to addresses temperature monitoring challenges within the
generate alarming alerts in the event of a deviation from the medical cold chain, considering perspectives from both the
required temperature range. These recorded readings can be pharmaceutical and healthcare industries. This model is
retrieved on the receiving end, providing a means to validate versatile, and applicable not only to pharmaceutical logistics
the reliability of the stored medicines. While proficient in but also to temperature-controlled laboratory medicine and
recording temperatures and alarms within a 30-day time vaccine cold chains. Its innovative approach paves the
frame, it’s important to note that this system does not have the way for the development of an Internet of Things-based
functionality to control the fridge temperature. The 30-day intelligent Cold Chain temperature monitoring system,
electronic refrigerator temperature logger is shown in the aiming to enhance responsiveness by identifying areas for
Figure 1. Lloyd et al. [19] proposed a mechanism that aims improvement and analyzing flaws in the existing Cold Chain
to address the issue of vaccine wastage due to accidental processes.
freezing by implementing a systematic temperature monitor- Liu et al. [24] proposed a system that utilizes the Raft con-
ing approach. This solution necessitates ongoing and vigilant sensus algorithm to enhance throughput and the Blockchain
temperature surveillance, integrated into hand-carry boxes to Distributed Network (BDN) with bloXroute servers for
ensure continuous monitoring and prevention of temperature- improved network scalability. The hybrid approach incor-
related issues. Similarly, the solution proposed by Sykes [20] porates multiple cryptographic techniques for privacy and
underscores the importance of ensuring the high-quality security, offering adaptability to system requirements. The
packaging and delivery of medicines, incorporating measures model allows for the addition or removal of components, and
such as the use of ice to maintain optimal temperatures during the use of alternative consensus algorithms for higher fault

46692 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

tolerance, addressing network bottleneck challenges through CRC codes are ubiquitously employed for detecting data
bloXroute implementation. transmission and errors. The common examples of CRCs
Stringent international regulations govern shipping and include CRC-16-CDMA2000 deployed in 3G mobile net-
storage temperatures in the cold chain, crucial for works, CRC-CCITT utilized for Bluetooth, CRC-24 applied
ensuring the quality and safety of goods, particularly in Long-Term Evolution (LTE), CRC-32 utilized in Ethernet
temperature-sensitive items like medications and food. and High-Level Data Link Control (HDLC) protocols, and
Incorrect handling of these items poses risks to public CRC-40-GSM employed in the GSM control channel. If the
health and the economy. Vivaldi et al. [25] proposed a probability of undetected error is given by the expression 1
novel solution to address temperature-related challenges with n as the degree of the CRC polynomial (CRC length)
during medical supplies transportation, a modified RFID and k as the number of errors, then the probability of error
tag with a copper-doped ionic liquid was employed to detection of CRC-40>CRC-32>CRC-24>CRC-16, which
detect temperature thresholds, unaffected by humidity implies an increase in computing complexity in the same
changes. This tag permanently changed upon surpassing order.
the ionic liquid’s melting point, adjustable through dopant
1 k
 
concentration. The introduction of copper ions facilitated PunDetectErr(CRC-n) ≈ 1 − 1 − n (1)
melting and freezing point adjustments, providing a 2
cost-effective solution for identifying cold chain breakdowns CRC-32 is selected due to its significant error detection
in transporting temperature-sensitive goods. The modified property and lower computational complexity than CRC-40.
RFID tags enable non-contact acquisition of information It is energy efficient compared with MD5 and the family of
from multiple packages simultaneously, offering a versatile SHA message authentication protocols, and it is also highly
and economical means of monitoring the cold chain. efficient in terms of hardware and software implementation.
As far as CRC encoding is concerned, it is done by
multiplying a message polynomial M (x) with the term x n ,
A. CONCLUSION OF LITERATURE REVIEW
followed by division by a generator polynomial g(x) of degree
Examining the existing literature reveals a significant
n. The resulting remainder, represented as r(x) = M (x) · x n
technology gap in the secure transportation of medicines,
mod g(x), constitutes the CRC check bits. The addition of
with minimal attention dedicated to this critical area. Our
these check bits to M (x) · x n generates a codeword: M (x) ·
evaluation underscores the opportunity to guarantee the safe
x n + r(x). CRC decoding conventionally involves dividing
and secure transit of medicines by harnessing state-of-the-
the received message by the shared generator polynomial
art sensing and information and communication technologies,
g(x) and then comparing the coefficients of the obtained
including sensors, micro-controllers, MQTT, 5G, etc. that are
remainder with CRC check bits. A mismatch in bits indicates
neglected in prevailing solutions.
an error [27]. Algorithm 1 shows the simplicity and ease of
the code to be implemented on the hardware. Similarly, it is
III. PROPOSED MODEL
The proposed model is given in Figure 2, which clearly Algorithm 1 CRC32 Calculation Algorithm
depicts the flow of the system, overall, the designed archi- 1: Input: data: Input data for CRC32 calculation
tecture is capable of delivering medicine/vaccine safely and 2: procedure GenerateCRCTable
securely to the end users with use of the latest tools and tech- 3: for i ← 0to255 do
nologies such as IoT devices, sensors, smart fridge, MQTT, 4: crc ← i
client application, database and remote server. MQTT is a 5: for j ← 0to7 do
lightweight, open-standard protocol designed for efficient 6: crc ← (crc ≫ 1) ⊕
machine-to-machine communication in resource-constrained (CRC32_POLYNOMIAL & (−(crc & 1)))
IoT environments over ordered, lossless, and bi-directional 7: crc32_table[i] ← crc
connections, typically utilizing TCP/IP or potentially QUIC
8: function CalculateCRC32(data, length)
(Quick UDP Internet Connections) [26]. MQTT QoS (Qual-
9: crc ← 0xFFFFFFFF ▷ Initial CRC32 value
ity of Service) levels 0, 1, and 2 refer to the different levels
10: for i ← 0tolength − 1 do
of message delivery guarantees between the MQTT broker
11: crc ← (crc ≫ 8) ⊕ crc32_table[(crc ⊕
and the client. QoS 0 provides the lowest assurance of
data[i]) & 0xFF]
delivery, as the message is sent with the expectation that
it may be lost or delivered multiple times. QoS 1 requires 12: return crc ⊕ 0xFFFFFFFF ▷ Final XOR operation
the sender to receive an acknowledgment (ACK) from the
receiver, which may lead to potential duplicate deliveries. believed that symmetric key cryptography is more efficient
QoS 2 uses a four-step handshake process to ensure the in terms of large or frequent data encryption as compared
message is delivered exactly once, eliminating the possibility to public key cryptography, which requires more resources
of duplicates. QoS 0 is preferred due to the low-resource and infrastructure and consumes more energy. In the realm
environment. of IoT, traditional asymmetric cryptography methods like

VOLUME 12, 2024 46693

 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

FIGURE 2. Proposed model.

RSA pose challenges due to resource-intensive computations. facilitating secret key exchange between two parties, A and
To address this, we propose lightweight alternatives, such B, based on an agreed-upon group G and a fixed group
as Elliptic Curve Diffie-Hellman (ECDH) with AES-256. element g. A group is a mathematical set of elements with
ECDH outperforms DH and RSA in terms of key length and defined operations that combine two elements to produce a
performance. In fact, the ECC 160-bit key security is equal third element within the group. For secret key sharing in the
to the 1024-bit key security of RSA and DH. The Diffie- DH protocol, each party generates a private random number.
Hellman (DH) key exchange is a mathematical protocol If A generates a and broadcasts ga , and B generates b and

46694 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

broadcasts gb , the common secret is gab . The security of

this protocol relies on the discrete log problem’s difficulty
in the given group, which is to find a given ga and g. ECDH
key exchange protocol, which is based on DH, is considered
for establishing a shared secret key between the sender (A),
the recipient (B), and the medicine container (C) through bi-
linear pairing-based cryptography, as shown in Figure 4 [28],
[29]. The primary difference between DH and ECDH is that
the former operates on numbers while the latter uses curve
points [30].
An elliptic curve defined over real numbers in its ‘‘standard
form,’’ is represented by the equation y2 = x 3 + ax + b,
where a and b are fixed values, and 4a3 + 27b2 ̸ = 0. This
elliptic curve forms an abelian group, where the addition
operation is a group operator. Consequently, the curve follows
the properties of closure, associativity, identity, inverse, and FIGURE 3. BN256 elliptic curve.

commutativity with respect to the addition of points on the

elliptic curve. For adding two points P(x1 , y1 ) and Q(x2 , y2 )
on the curve to get the third point R(x3 , y3 ), the following
rules are followed:
1- If points P and Q have distinct x and y coordinates, the
addition of P and Q (i.e., R) involves reflecting the result
about the x-axis, forming the point −R. This point is the
intersection of the straight line connecting P and Q with
the curve. The coordinates x3 and y3 are determined by
calculating the slope of the line PQ as m = yx22 −y 1
−x1 and then
2
computing x3 = m − x1 − x2 and y3 = m(x1 − x3 ) − y1 .
2- In the case of two overlapping points, a tangent is drawn
at point P, intersecting the curve at −R. The addition of the
two points results in R = P + P. The coordinates x3 and
y3 are found by first calculating the slope of the tangent line
3x12 +a
as m = 2y 1
and then determining x3 = m2 − x1 − x2 and
y3 = m(x1 − x3 ) − y1 .
3- When adding the points P(x1 , y1 ) and −P(x1 , −y1 ), the
line connecting these two points does not intersect the curve
at a third point. The intercepting point is at infinity, denoted
as the point of infinity O, serving as the additive identity of
the group (P + (−P) = O). FIGURE 4. Group key generation.

4- To find kP, where k is a scalar number, point P is added k

times, expressed as kP = P + P + . . . + P k times.
We recommend the Barreto-Naehrig (BN) curve, which participant, User A, B, and Container C, selects a secret key
is commonly expressed in the Weierstrass form, defined by value: a, b, and c respectively. The public keys, represented as
the equation y2 = x 3 + b, where x, y are coordinates and b points on the elliptic curve, are denoted as (Pa, Qa), (Pb, Qb),
is a constant term. The BN curve parameters vary based on and (Pc, Qc), where Pi = di ·P and Qi = di ·Q. Users share the
the chosen values, with BN256 being a standardized instance public part of the shared key to be generated with one another.
characterized by the equation y2 = x 3 + 3. BN256 operates Shared secret key computation is shown in Algorithm 2
over a 256-bit prime field and finds extensive use in pairing- where each participant (at A, B, and Container C) computes
based cryptography [31], [32]. BN256 elliptic with addition their shared secret using the Weil pairing [33]. The shared
of two points P and Q as R and its image as -R image is shown secret is a point of an elliptic curve that is passed to the
in Figure 3. HMAC-based Extract-and-Expand Key Derivation Function
For the purpose of group key sharing, we propose a (HKDF) to generate a 256-bit symmetric key [34], [35]. This
tripartite Diffie-Hellman protocol that utilizes the Weil symmetric is used for encrypted communication between A,
pairing, as introduced by Joux in 2000. This protocol operates B, and C using block cipher AES-256.
on an elliptic curve denoted as E, featuring two independent The shared secret computed using the Weil pairing
base points, P and Q implying that no integer d exists such should be hard to drive given the public keys (PA, QA),
that P = d · Q or Q = d · P. In the key generation phase, each (PB, QB), and (PC, QC) are available to adversaries.
VOLUME 12, 2024 46695
 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

Algorithm 2 Tripartite Diffie-Hellman With Weil Pairing application. Additionally, two GPS sensors are integrated to
1: Input: Elliptic curve E, independent base points P and Q provide real-time location updates to both the sender and
2: Initialize: Secret keys a, b, and c for User-A, User-B, and User- receiver. Similar to the redundancy in temperature sensors,
C
3: Compute and broadcast public keys:
dual GPS sensors are employed to bolster system efficiency
4: User-A: (Pa , Qa ) = (a · P, a · Q) and reliability. This comprehensive sensing and notification
5: User-B: (Pb , Qb ) = (b · P, b · Q) system ensures the secure and well-monitored transport of
6: User-C (Container): (Pc , Qc ) = (c · P, c · Q) medicines or vaccines.
7: Compute shared secret and symmetric key using the Every piece of data throughout the entire process will
HMAC-based Extract-and-Expand Key Derivation Func-
tion (HKDF) [34]
be meticulously recorded. The Arduino, equipped with
8: User-A: both temperature and GPS sensors, will transmit this data
9: SharedSecretABC = F(a, Pb , Qc ) = F(a, Qb , Pc ) = to a remote server via an LTE module using the secure
e(Pb , Qc )a HTTP protocol, functioning as a gateway. Subsequently,
10: SymmetricKey = HKDF(SharedSecretABC ) the module forwards the data to an MQTT broker, which,
11: User-B:
12: SharedSecretABC = F(b, Pa , Qc ) = F(b, Qa , Pc ) =
in turn, pushes the information to the remote server via
e(Pa , Qc )b the HTTP protocol; the basic operations of MQTT are
13: SymmetricKey = KDF(SharedSecretABC ) given in Algorithm 3. Both the sender and the receiver will
14: Container-C: have access to relevant data through their respective mobile
15: SharedSecretABC = F(c, PA , QB ) = F(c, Qa , Pb ) = applications. For instance, the receiver’s application will
e(Pa , Qb )c
16: SymmetricKey = HKDF(SharedSecretABC )
display comprehensive details, including GPS coordinates
and temperature records. To maintain a historical record and
preserve all shipment details, the data will be stored in a
database.
But such security solutions, while computationally secure,
may be vulnerable to breaches with powerful adversaries, Algorithm 3 MQTT Algorithm
while information-theoretic cryptography solutions will offer 1: Initialize: Broker:=ServerIP, Publisher:=Fridge,
resilience in the future [36]. Subscriber:=Sender, Receiver, Topic:=Medicine Fridge
To initiate the process, the system involves two key actors: State
2: Connect to Broker:
the sender and the receiver. This process revolves around Connect
3: Publisher −−−−→ Broker
the reliable shipment of medicines or vaccines. The sender, Connect
4: Subscriber −−−−→ Broker
typically a service provider, initiates the shipment by creating 5: Publish a Message:
an order and placing the medicine into a smart fridge or a Publish(topic, message)
6: Publisher −−−−−−−−−−−−−→ Broker
container designed to maintain a specific temperature range. 7: Broker stores and forward messages to subscribers
The screen layouts used for start shipment and user selection 8: Subscribe to a Topic:
are shown in Figures 5 and Figure 6 respectively. The smart Subscribe(topic)
9: Subscriber −−−−−−−−−→ Broker
fridge, secured with a password, requires a PIN code for 10: The broker adds a subscriber to the list for a specified topic
access; this code is securely delivered to the customer via 11: Receive Published Messages:
email upon order creation. The smart fridge, managed by 12: Broker forwards message to subscribers in the topic
an Arduino, is equipped with temperature and GPS sensors 13: Unsubscribe from a Topic:
Unsubscribe(topic)
to ensure optimal conditions during transportation. Despite 14: Subscriber −−−−−−−−−−−→ Broker
these precautions, the challenge to medicine’s reliability 15: The broker removes subscribers for a specified topic
16: Disconnect:
arises during transit, where the fridge encounters fluctuating Disconnect
17: Publisher −−−−−−→ Broker
temperatures due to variations throughout the day, differing Disconnect
18: Subscriber −−−−−−→ Broker
climates in various areas, and the immediate environmental
conditions impacting the fridge’s performance.
During the transportation of medicine, temperature sensors Upon the completion of the shipment, the end user can
play a pivotal role in overseeing and broadcasting the tem- issue a command to unlock the fridge through the application,
perature levels, promptly alerting the Arduino system to any utilizing the secure PIN code sent to it via email at the
fluctuations. If the temperature falls below the required range initiation of the shipment. The screen shown in Figure 7 is
for the medicine, the Arduino instructs the fridge to initiate used for authentication purposes that demand this PIN. This
cooling, safeguarding the medicine from expiration due to meticulous data recording and secure access system ensures
adverse conditions. This dual-sensor approach enhances the transparency, accountability, and reliability throughout the
safety measures, with a primary temperature sensor taking entire shipment process. Mobile security of different mobiles
the lead and a secondary sensor serving as a reliable backup. and business applications can also be ensured through
In the event of a failure in the primary sensor, an immediate different suggestions and recommendations, such as MVP,
notification is sent to the service provider, who can remotely app right management, and inspection (anti-virus McAfee,
activate the backup temperature sensor through the client CLAMAV, Bit Defender) [37]. Algorithm 4 lays out the

46696 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

detailed functionality of the proposed model that clearly

describes the functions of different elements of the solution.

Algorithm 4 Medicine Transportation Monitoring

1: Initialize medicine fridge with Arduino module and GSM chip.
2: Connect sensors to the Arduino for temperature and status
monitoring.
3: Set up an MQTT server for messaging.
4: For the Diffie-Hellman key exchange
5: Make use of ECDH Tri-Partite Algorithm 2
6: Encrypted Data Transmission:
7: procedure Data_Transfer()
8: while Medicine Transportation is ongoing do
9: if One cooling fan fails then
10: Activate the backup cooling fan.
11: if Temperature is outside the specified range then
12: Adjust cooling fan speed to regulate temperature.
13: Read data from fridge sensors.
14: CRC ← CalculateCRC32(data) Algorithm 1
CRC || AES_Encryption(Data, Symmetric Key)
15: User-C −→ User B,
A
16: Transmit encrypted data and CRC to sender and receiver
mobile screens using MQTT.
17: Store data securely in the database.
18: procedure MedicineContainerUnlocking()
19: Sender:
20: Generate a unique code.
21: Encrypt the code using a 256-bit ECDH symmetric key. FIGURE 5. Start shipment.
22: Compute CRC for the encrypted code.
23: Transmit encrypted code and CRC via MQTT Algorithm 3
to the receiver.
24: Receiver:
25: Receive encrypted code and CRC via MQTT Algorithm 3.
26: Decrypt the code using the ECDH symmetric key.
27: Verify the integrity of the code using the CRC.
28: User Interaction:
29: Enter the code to open the container.
30: Container Access Control:
31: if Entered code is correct then
32: Open the medicine container.
33: else
34: Do not open the container.

A. IMPLEMENTATION
The system is implemented using different hardware and soft-
ware components. A brief description of these components is
given in Table 2.

IV. RESULTS AND DISCUSSIONS

A. TEMPERATURE RECORDING AND REPORTING TEST
In the testing phase of the monitoring system, every
feature integrated into the application demonstrated optimal
performance, consistently yielding the anticipated results.
To assess the efficacy of the proposed solution in mitigating FIGURE 6. Select user.
the environmental impact on medicine during transporta-
tion, a rigorous 20-day testing period was conducted on
insulin. Throughout this duration, the transportation box 8 degrees Celsius to simulate real-world conditions. The
was personally carried from home to the workplace, with choice of insulin, a commonly used and temperature-sensitive
daily temperature averages recorded. The testing parameters medication, served as a stringent test case, providing valuable
maintained a scale range between 2 degrees Celsius and insights into the system’s reliability under challenging

VOLUME 12, 2024 46697

 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

FIGURE 8. Temperature history of 20 days.

TABLE 3. Analysis of CRC, MD5, and SHA on 1 GB of data.

FIGURE 7. Shipment authentication.

TABLE 2. Hardware and software components.

FIGURE 9. Energy drainage analysis [20-day trial].

components. Its use not only saves battery power but also
improves the system’s overall performance and durability.
circumstances. According to Figure 8 taken from the data Furthermore, taking into account the amount of data produced
generated by the system, the temperature remained within by Arduino, it is around 100,000 bytes per day. In order to
range for the majority of days. In 20 days, only once did it finalize the best algorithm for the given system, Figure 9
go below 2 degrees Celsius and once slightly above 8 degrees depicts the performance of CRC-32, MD5, and SHA-1
Celsius. over 20 days. Hence, it is evident that CRC exhibited
significantly faster execution times with minimal impact on
B. INTEGRITY TEST resources such as the battery. In contrast, running the other
After running MD5, CRC32, and SHA-1 using 1GB of data two algorithms resulted in higher resource consumption and
produced by the system on the fully charged Arduino, we got slightly increased hardware demands.
the results shown in Table 3. In the context of hashing for low-level devices, the
The findings indisputably show that CRC32 is the evidence presented above strongly supports the conclusion
most efficient solution for low-level devices and hardware that CRC32 is the most advantageous choice. Its efficiency

46698 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

TABLE 4. Confusion matrix table. To measure the performance of the system, another
important coefficient is used. The Matthews Correlation
Coefficient (MCC) is a statistical measure designed to
evaluate the effectiveness of a binary categorization system.
It accounts for true positives, true negatives, false positives,
and false negatives and gives an equitable metric even when
the groups are of varying sizes [41]. The Equation 2 formula
in terms of speed, resource utilization, and minimal impact defines the MCC:
on hardware makes it a standout solution for applications (TP ∗ TN − FP ∗ FN )
where these factors are paramount. Choosing CRC32 not MCC = √
(TP + FP)(TP + FN )(TN + FP)(TN + FN )
only ensures effective hashing but also enhances the overall
performance and resource management of low-level devices. (2)
Where TP is the number of true positives, TN is the number
C. DATA COMMUNICATION TEST of true negatives, FP is the number of false positives, and FN
The test is conducted to verify the bidirectional commu- is the number of false negatives.
nication between the server application and the device By putting in the values we got from Table 5 in MCC
microcontroller via Cloud MQTT. Over 20 days, we initiated Equation 2, we get a value of 0.65, as shown in the
20 shipments to assess the system’s performance, subjecting mathematical expression 3.
it to diverse situations and circumstances. Throughout the (93 ∗ 662 − 15 ∗ 70)
20-day trial, not a single packet was missed when the network MCC = √ = 0.65
(93 + 15)(93 + 70)(662 + 15)(662 + 70)
was accessible, with successful encoding and decoding
(3)
occurring on both ends. Whenever there is network unavail-
ability, data packets are stored locally and subsequently The MCC score falls between −1 and +1, with +1
pushed to the server once connectivity is restored. This representing excellent prediction, 0 representing no better
process ensures data integrity and resilience. As depicted than random prediction, and −1 representing complete
in Figure 10, the data is then transmitted to Cloud MQTT, disagreement between forecast and observation. Overall,
becoming accessible to the devices. a higher MCC value indicates greater classifier performance.
Based on this, our system produced 0.65, which is more than
D. NOTIFICATION ALERT TEST half way better than 0 and closer to +1.
Over the 20-day shipment period, the testing focus was The purpose of the notification alert test is to check the
primarily on the crucial aspects of the application, with a validity of the proposed framework when the container is
particular emphasis on notification alerts. The significance of moved from one location to another. We could not test for
these alerts lies in their pivotal role in ensuring the medicine’s longer distances but for a short distance that is from lab to
reliability. They serve as a real-time update mechanism, home and home to lab; on average, the distance is 35-40
keeping users informed about the current state of the medicine kilometers. It is tested in hot summers where temperatures
during transit. Throughout the trial of these 20 shipments, may vary from 40Co - 47Co and cellular service may
a comprehensive assessment of 840 temperature records also be disrupted in some areas or may become weak in
was conducted. While the system was expected to generate overcrowded passages due to rush hours. Close investigations
108 temperature alerts, it exceeded expectations by producing reveal that moisture and humidity, the aging of sensors, and
a total of 163 alerts. This robust performance underscores the inappropriate sensor placement may cause FNR and FPR.
effectiveness and responsiveness of the notification system in So, the following actions are recorded for reducing FNR and
providing timely and accurate information. FPR:
Now for data analysis, we will use the confusion matrix
(Table 4). It is a useful tool for determining the variables that 1) ADVANCED FILTERING ALGORITHMS
impact accuracy and precision. Advanced filtering algorithms, such as Kalman filters or low-
Now we need to calculate accuracy [38], error rate, pass filters, can be explored to smooth temperature data and
precision [38], true positive rate [39], false positive rate [39], reduce the impact of short-term fluctuations [42], [43].
true negative rate, prevalence, and F-score [40] from Table 4,
and the results are given in Table 5. 2) REDUNDANCY AND CROSS-VALIDATION
The aforementioned calculations were conducted to com- Redundancy and cross-validation for temperature sensors are
prehensively assess various aspects of the system through an also suggested, with redundant sensors deployed and consen-
analysis of the data using a confusion matrix, widely regarded sus algorithms developed to cross-validate readings [44].
as one of the best methods for evaluating system performance.
Upon running the confusion matrix, it was determined that the 3) PERIODIC REVIEW, AND MAINTENANCE
accuracy of the notification alerts is 0.89, indicating that this Periodically reviewing and updating the error detection
classifier is accurate 89% of the time. algorithm based on system performance and evolving

VOLUME 12, 2024 46699

 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

FIGURE 10. Data packets on cloud mqtt.

TABLE 5. Confusion matrix calculations.

requirements, regularly updating shared keys for secure data pin. It has good accuracy, which is +/-0.5◦C (at -10 to
transmission, employing best practices for key management, 85◦C) [45], [46].
and implementing protocols for detecting and mitigating 2) TMP1826 is a OneWire sensor, that is also compatible
security breaches to ensure the authenticity of transmitted with Arduino libraries. It offers high precision and a
data may also reduce FNR and FPR. OneWire interface for simplified wiring [47].
3) Infrared thermopile sensors such as the
MLX90616ESF-HCA use the Seebeck effect to mea-
V. USING NEW TEMPERATURE DETECTION METHODS
sure temperature and can be interfaced with Arduino
The proposed solution does not introduce or invent any novel using protocols like I2C. They offer non-contact
methods for temperature detection. Instead, it leverages state- temperature measurements and are suitable for mea-
of-the-art temperature sensing technology to conduct tem- suring surface temperatures. Some models offer digital
perature measurements. This concise explanation assists in
outputs [48].
making informed decisions when selecting the most suitable
These sensors are not traditional semiconductor-based but
approach for the given scenario. For instance, the LM35
provide digital output and can be easily interfaced with
temperature sensor is conventionally semiconductor-based
Arduino microcontrollers. It is important to consider the
and can be interfaced with Arduino; there are others that
specific requirements of an application, such as temperature
are not semiconductor-based but provide high precision and
range, accuracy, and interfacing preferences, along with
can be interfaced with Arduino. The most relevant are given
checking datasheets and Arduino libraries for compatibility
below.
and ease of integration.
1) The DS18B20 is a OneWire temperature sensor
with a wide temperature range, high accuracy, and VI. LIMITATIONS AND CHALLENGES
programmable resolution. It supports multiple sensors 1) The proposed solution is based on cellular networks;
on a single bus and communicates over a single digital communications may be disrupted if the service is

46700 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

not available in any region, and information about [4] D. Oladimeji, K. Gupta, N. A. Kose, K. Gundogan, L. Ge, and F. Liang,
medicines exposure to hostile environmental factors ‘‘Smart transportation: An overview of technologies and applications,’’
Sensors, vol. 23, no. 8, p. 3880, Apr. 2023.
cannot be obtained. Multiple service providers may [5] World Summit on the Information Society Outcome Documents, Int.
be required to ensure the robustness of the proposed Telecommun. Union, Geneva, Switzerland, 2005.
system. [6] G. Mao, Y. Hui, X. Ren, C. Li, and Y. Shao, ‘‘The Internet of Things for
2) sensors may behave poorly and may report inaccu- smart roads: A road map from present to future road infrastructure,’’ IEEE
Intell. Transp. Syst. Mag., vol. 14, no. 6, pp. 66–76, Nov. 2022.
rate readings due to moisture, aging of the sensor [7] S. Chakraborty, T. Chakravorty, and V. Bhatt, ‘‘IoT and AI driven
hardware, electromagnetic interference, voltage fluctu- sustainable practices in airlines as enabler of passenger confidence,
ations, direct contact with the heating surface, and an satisfaction and positive WOM : AI and IoT driven sustainable practice
in airline,’’ in Proc. Int. Conf. Artif. Intell. Smart Syst. (ICAIS), Mar. 2021,
abrupt change in temperature. pp. 1421–1425.
3) MQTT supports three QoS levels (0, 1, 2), but higher [8] S. Aslam, M. P. Michaelides, and H. Herodotou, ‘‘Internet of Ships:
QoS levels come with increased overhead and may A survey on architectures, emerging applications, and challenges,’’ IEEE
impact performance, especially in low-bandwidth or Internet Things J., vol. 7, no. 10, pp. 9714–9727, Oct. 2020.
[9] W. D. Hoyer, M. Kroschke, B. Schmitt, K. Kraume, and V. Shankar,
high-latency environments. ‘‘Transforming the customer experience through new technologies,’’
J. Interact. Marketing, vol. 51, pp. 57–71, Aug. 2020.
VII. CONCLUSION AND FUTURE WORK [10] A. El Attaoui, S. Largo, A. Jilbab, and A. Bourouhou, ‘‘Wireless medical
An innovative system has been introduced to monitor medi- sensor network for blood pressure monitoring based on machine learning
for real-time data classification,’’ J. Ambient Intell. Humanized Comput.,
cation distribution and regulate the temperature of medicines vol. 12, no. 9, pp. 8777–8792, Sep. 2021.
during transportation, addressing the issue of medicine [11] A. Musamih, K. Salah, R. Jayaraman, J. Arshad, M. Debe, Y. Al-Hammadi,
waste and potential harm caused by environmental factors. and S. Ellahham, ‘‘A blockchain-based approach for drug traceability in
The system uses electronic sensors, GPS, and temperature healthcare supply chain,’’ IEEE Access, vol. 9, pp. 9728–9743, 2021.
[12] E. N. Mambou, S. M. Nlom, T. G. Swart, K. Ouahada, A. R. Ndjiongue,
sensors to reduce errors and cater to those who may not and H. C. Ferreira, ‘‘Monitoring of the medication distribution and the
be computer professionals. The cloud-based system allows refrigeration temperature in a pharmacy based on Internet of Things
supervisors to remotely connect and monitor data, ensuring (IoT) technology,’’ in Proc. 18th Medit. Electrotech. Conf. (MELECON),
Apr. 2016, pp. 1–5.
medicines are consistently maintained within a reasonable
[13] BBC news. (2017). South Sudan Vaccination Error Kills 15 Children.
temperature range. The system also proactively notifies users [Online]. Available: https://www.bbc.com/news/world-africa-40135814
of temperature fluctuations, providing real-time awareness [14] AirCargo News. (2019). Failures in Temperature-Controlled Logistics
of the medicine’s status. This technology demonstrates Cost Biopharma Industry Billions. [Online]. Available: https://www.
aircargonews.net/sectors/pharma-logistics/failures-in-temperature-
the effectiveness of IoT in integrating digital devices and controlled-logistics-cost-biopharma-industry-
transforming the global landscape into a real-time display. billions/
Despite the security mechanisms used by MD5 and SHA, [15] H. Zhuoyu and L. Yongzhen, ‘‘Design and implementation of efficient hash
CRC excels in speed, making it a recommended choice for functions,’’ in Proc. IEEE 2nd Int. Conf. Power, Electron. Comput. Appl.
(ICPECA), Jan. 2022, pp. 1240–1243.
enhanced performance. Statistical measures like accuracy [16] V. Gatteschi, F. Lamberti, G. Paravati, A. Sanna, C. Demartini, A.
and F1-score confirm the credibility and effectiveness of the Lisanti, and G. Venezia, ‘‘New frontiers of delivery services using
proposed solution. drones: A prototype system exploiting a quadcopter for autonomous drug
shipments,’’ in Proc. IEEE 39th Annu. Comput. Softw. Appl. Conf., vol. 2,
In the future, we want to enhance the proposed solu- Jul. 2015, pp. 920–927.
tion with blockchain-based technology that can make the [17] S. K. Nanda, S. K. Panda, and M. Dash, ‘‘Medical supply chain integrated
real-time readings of sensors and the state of the medicine with blockchain and IoT to track the logistics of medical products,’’
container immutable. Blockchain technology offers a promis- Multimedia Tools Appl., vol. 82, no. 21, pp. 32917–32939, Sep. 2023.
[18] Ü. Kartoğlu, E. Nelaj, and D. Maire, ‘‘Improving temperature monitoring
ing solution for data protection and distributed sharing as in the vaccine cold chain at the periphery: An intervention study using a
it is a decentralized and distributed ledger technology that 30-day electronic refrigerator temperature logger (fridge-tag® ),’’ Vaccine,
ensures transparency, security, and immutability of data. vol. 28, no. 24, pp. 4065–4072, May 2010.
Blockchain’s immutability makes it difficult to alter data, [19] J. Lloyd, P. Lydon, R. Ouhichi, and M. Zaffran, ‘‘Reducing the loss of
vaccines from accidental freezing in the cold chain: The experience of
providing a robust layer of protection against unauthorized continuous temperature monitoring in Tunisia,’’ Vaccine, vol. 33, no. 7,
modifications or tampering. pp. 902–907, Feb. 2015.
[20] C. Sykes, ‘‘Time- and temperature-controlled transport: Supply chain
REFERENCES challenges and solutions,’’ Pharmacy Therapeutics, vol. 43, no. 3, p. 154,
2018.
[1] M. Majid, S. Habib, A. R. Javed, M. Rizwan, G. Srivastava,
T. R. Gadekallu, and J. C.-W. Lin, ‘‘Applications of wireless sensor [21] I. Konovalenko and A. Ludwig, ‘‘Comparison of machine learning
networks and Internet of Things frameworks in the industry revolution classifiers: A case study of temperature alarms in a pharmaceutical supply
4.0: A systematic literature review,’’ Sensors, vol. 22, no. 6, p. 2087, chain,’’ Inf. Syst., vol. 100, Sep. 2021, Art. no. 101759.
Mar. 2022. [22] J. Loisel, S. Duret, A. Cornuéjols, D. Cagnon, M. Tardet,
[2] I. Azimi, A. M. Rahmani, P. Liljeberg, and H. Tenhunen, ‘‘Internet E. Derens-Bertheau, and O. Laguerre, ‘‘Cold chain break detection
of Things for remote elderly monitoring: A study from user-centered and analysis: Can machine learning help?’’ Trends Food Sci. Technol.,
perspective,’’ J. Ambient Intell. Humanized Comput., vol. 8, no. 2, vol. 112, pp. 391–399, Jun. 2021.
pp. 273–289, Apr. 2017. [23] S. Balachandar and R. Chinnaiyan, ‘‘Reliable pharma cold chain
[3] V. S. Naresh, S. S. Pericherla, P. S. R. Murty, and S. Reddi, ‘‘Internet monitoring and analytics through Internet of Things edge,’’ in Emergence
of Things in healthcare: Architecture, applications, challenges, and of Pharmaceutical Industry Growth With Industrial IoT Approach.
solutions,’’ Comput. Syst. Sci. Eng., vol. 35, no. 6, pp. 411–421, 2020. Amsterdam, The Netherlands: Elsevier, 2020, pp. 133–161.

VOLUME 12, 2024 46701

 D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

[24] X. Liu, A. V. Barenji, Z. Li, B. Montreuil, and G. Q. Huang, ‘‘Blockchain- [46] ANSpecified. (2024). DHT11 Vs DHT22 Vs LM35 vs DS18B20 Vs
based smart tracking and tracing platform for drug supply chain,’’ Comput. BME280 Vs BMP180. Accessed: Mar. 7, 2024. [Online]. Available:
Ind. Eng., vol. 161, Nov. 2021, Art. no. 107669. https://randomnerdtutorials.com/dht11-vs-dht22-vs-lm35-vs-ds18b20-vs-
[25] F. Vivaldi, B. Melai, A. Bonini, N. Poma, P. Salvo, A. Kirchhain, bme280-vs-bmp180/
S. Tintori, A. Bigongiari, F. Bertuccelli, G. Isola, and F. Di Francesco, [47] T Instruments. (May 2023). TMP1826 1-Wire® , 4±0.2◦ C Accurate
‘‘A temperature-sensitive RFID tag for the identification of cold Temperature Sensor With 2Kb EEPROM. Accessed: Mar. 7, 2024.
chain failures,’’ Sens. Actuators A, Phys., vol. 313, Oct. 2020, [Online]. Available: https://www.ti.com/lit/ds/symlink/tmp1826.pdf?ts=
Art. no. 112182. [Online]. Available: https://www.sciencedirect.com/ 1709754722978&ref_url=https%253A%252F%252Fwww.ti.
science/article/pii/S0924424720301606 com%252Fproduct%252FTMP1826
[26] B. Mishra and A. Kertesz, ‘‘The use of MQTT in M2M and IoT systems: [48] Melexis. (Aug. 2012). MLX90616ESF-HCA Infra Red Thermometer in to-
A survey,’’ IEEE Access, vol. 8, pp. 201071–201086, 2020. 39 for High Temperature Thermometer Guns. Data Sheet. Accessed:
[27] E. Dubrova, M. Näslund, and G. Selander, ‘‘CRC-based message Mar. 7, 2024. [Online]. Available: https://www.melexis.com/en/
authentication for 5G mobile technology,’’ in Proc. IEEE Trust- product/MLX90616/Thermal-Infrared-Thermopile-Sensor-High-
com/BigDataSE/ISPA, vol. 1, Aug. 2015, pp. 1186–1191. Temperature-Measurements
[28] S. Ricci, P. Dzurenda, J. Hajny, and L. Malina, ‘‘Privacy-enhancing group
signcryption scheme,’’ IEEE Access, vol. 9, pp. 136529–136551, 2021.
[Online]. Available: https://api.semanticscholar.org/CorpusID
[29] A. Menezes. (2005). An Introduction to Pairing-Based Cryptography.
[Online]. Available: https://api.semanticscholar.org/CorpusID
[30] B. Buchanan. (Apr. 30, 2023). Elliptic Curve Integrated Encryp-
tion Scheme (ECIES): Encrypting Using Elliptic Curves, ASecuri-
tySite: When Bob Met Alice. Accessed: Mar. 10, 2024. [Online]. DAVID SAMUEL BHATTI received the Ph.D.
Available: https://medium.com/asecuritysite-when-bob-met-alice/elliptic- degree in computer science (information security)
curve-integrated-encryption-scheme-ecies-encrypting- from the School of Electrical Engineering and
using-elliptic-curves-dc8d0b87eaa Computer Science (SEECS), National University
[31] W. J. Buchanan. (2024). Tripartite Diffie–Hellman Algorithm With
of Sciences and Technology (NUST), Islamabad,
BN256. Accessed: Feb. 7, 2024. [Online]. Available: https://asecuritysite.
Pakistan, in 2020. He is currently an Assistant
com/keyexchange/go_bn256
[32] P. Dzurenda, S. Ricci, J. Hajny, and L. Malina, ‘‘Performance analysis and Professor with the University of Central Punjab
comparison of different elliptic curves on smart cards,’’ in Proc. 15th Annu. (UCP), Lahore, Pakistan. His research interests
Conf. Privacy, Secur. Trust (PST), Aug. 2017, pp. 365–36509. include networks, mobiles, and smartphones secu-
[33] A. Joux, ‘‘A one round protocol for tripartite Diffie–Hellma,’’ in Algorith- rity. His expertise extends to secure routing
mic Number Theory. Berlin, Germany: Springer, 2000, pp. 385–393. protocols, secret key establishment, and device authentication, particularly
[34] H. Krawczyk and P. Eronen, ‘‘HMAC-based extract-and-expand key in low-resource devices like wearable, body-worn, and wireless body area
derivation function (HKDF),’’ IBM Res. Nokia, Internet Eng. Task network (WBAN) devices. His current research interests include the design
Force (IETF), NY, USA, Tech. Rep. 5869, 2010. [Online]. Available: of security protocols using probabilistic data structures. The aim is to
https://www.rfc-editor.org/rfc/rfc5869 optimize time and space complexity in low-resource devices, enhancing their
[35] W. J. Buchanan. (2024). Ecdh With Golang and Kryptology. Accessed: efficiency, and robustness in the realm of information security.
Mar. 10, 2024. [Online]. Available: https://asecuritysite.com/ecdh/go_ecdh
[36] D. S. Bhatti and S. Saleem, ‘‘Ephemeral secrets: Multi-party secret key
acquisition for secure IEEE 802.11 mobile ad hoc communication,’’ IEEE
Access, vol. 8, pp. 24242–24257, 2020.
[37] D. S. Bhatti, N. A. Saqib, and Z. Anwar, ‘‘SCEAMS: Secure corporate
environment adhered to mobile & smartphones,’’ in Proc. 6th Int. Conf.
Innov. Comput. Technol. (INTECH), Aug. 2016, pp. 561–566.
[38] R. Yacouby and D. Axman, ‘‘Probabilistic extension of precision, MUHAMMAD MUEED HUSSAIN received the
recall, and F1 score for more thorough evaluation of classification B.S.C.S. degree from Virtual University and the
models,’’ in Proc. 1st Workshop Eval. Comparison NLP Syst., 2020, M.Phil. degree in computer science from the Insti-
pp. 79–91. tute of Management Sciences, Pakistan. He is
[39] D. van Ravenzwaaij and J. P. A. Ioannidis, ‘‘True and false positive rates
currently a Renowned AI and IoT Researcher
for different criteria of evaluating statistical evidence from clinical trials,’’
with expertise in information security, wireless
BMC Med. Res. Methodol., vol. 19, no. 1, pp. 1–10, Dec. 2019.
[40] C. Goutte and E. Gaussier, ‘‘A probabilistic interpretation of precision,
networks, and software programming. With over
recall and F-score, with implication for evaluation,’’ in Proc. Eur. Conf. Inf. ten years of experience, he is also a Senior
Retr. Santiago de Compostela, Spain: Springer, Mar. 2005, pp. 345–359. Software Engineer with Invensify Ltd. He focuses
[41] D. Chicco and G. Jurman, ‘‘The advantages of the Matthews correlation on securing communication protocols, optimizing
coefficient (MCC) over F1 score and accuracy in binary classification wireless network performance, and developing robust software solutions.
evaluation,’’ BMC Genomics, vol. 21, no. 1, pp. 1–13, Dec. 2020.
[42] X. Zhang, H. Liang, J. Feng, and H. Tan, ‘‘Kalman filter
based high precision temperature data processing method,’’
Frontiers Energy Res., vol. 10, Apr. 2022. [Online]. Available:
https://www.frontiersin.org/articles/10.3389/fenrg.2022.832346, doi:
10.3389/fenrg.2022.832346.
[43] Y. Zhang, R. Wang, S. Li, and S. Qi, ‘‘Temperature sensor denoising
algorithm based on curve fitting and compound Kalman filtering,’’ BEOMKYU SUH received the B.S. and M.S.
Sensors, vol. 20, no. 7, p. 1959, Mar. 2020. [Online]. Available: degrees in computer engineering from Chungnam
https://www.mdpi.com/1424-8220/20/7/1959 National University, Daejeon, South Korea, in
[44] R. H. Dodier, ‘‘Unified prediction and diagnosis in engineering systems 2022 and 2024, where he is currently pursuing
by means of distributed belief networks,’’ Ph.D. dissertation, Dept. Civil, the Ph.D. degree. His research interests include
Environ., Architectural Eng., Univ. Colorado, Boulder, CO, USA, 1999. wireless sensor networks and deep reinforcement
[45] D. Yulizar, S. Soekirno, N. Ananda, M. A. Prabowo, I. F. P. Perdana, and learning.
D. Aofany, ‘‘Performance analysis comparison of DHT11, DHT22 and
DS18B20 as temperature measurement,’’ in Proc. 2nd Int. Conf. Sci. Educ.
Sci., 2022, pp. 37–45, doi: 10.2991/978-94-6463-232-3_5.

46702 VOLUME 12, 2024

D. S. Bhatti et al.: IoT-Enhanced Transport and Monitoring of Medicine

ZULFIQAR ALI received the Ph.D. degree in KI-IL KIM received the M.S. and Ph.D. degrees in
computer science from the National University computer science from Chungnam National Uni-
of Computer and Emerging Sciences, Islamabad. versity, Daejeon, South Korea, in 2002 and 2005,
He is currently an accomplished Assistant Pro- respectively. He is currently with the Department
fessor. With extensive teaching experience, he is of Computer Science and Engineering, Chung-
also with the National University of Technology. nam National University. He has been with the
His research interests include AI, optimization, Department of Informatics, Gyeongsang National
and networks, particularly in computational intel- University, since 2006. His research interests
ligence and swarm intelligence. He received include machine learning for networks, wire-
a Research Exchange Fellowship with Arizona less/mobile networks, fog computing, MANET,
State University, in 2014. Apart from his teaching roles, he actively con- QoS for wireless, and wireless sensor networks.
tributes to conferences, including the Frontiers of Information Technology.
He has supervised research projects and published extensively in reputable
journals. His commitment to academic excellence is evident in his continuous
contributions to the field of computer science.

ISMATOV AKOBIR received the B.S. degree

from Woosong University, Daejeon, South Korea,
in 2023. He is currently pursuing the M.S. degree
with Chungnam National University. His research
interests include machine learning and wireless
networks.

VOLUME 12, 2024 46703