Security Aspects of Digital Twins in IoT

Vitomir Pavlov, Florian Hahn and Mohammed El-Hajj

Twente University, EEMCS (SCS), Enschede, The Netherlands

Keywords:

Digital Twin, IoT, Security, Authentication Arduino, Raspberry PI.

Abstract:

The number of Internet-connected devices are expected to reach almost 30 billion by 2030, and already to-

day the Internet of Things (IoT) technologies is a part of everyday life in sectors like public health, smart

cars, smart grids, smart cities, smart manufacturing and smart homes. An even tighter integration between

IoT technology and physical objects within these sectors has been made possible by the Digital Twin (DT)

technology providing better abilities for real-time monitoring, data-driven modeling and process optimization.

One integral aspect of this approach is the connection between IoT end-devices and their corresponding digital

twins for real-time data communication. Depending on the envisioned scenario, the involved data and derived

processes affect the safety of human lives, hence an authentic connection is of major importance. At the

same time, IoT devices have restrictions on the available power sources and provided computing resources.

In this work we report on our experiments with the Azure IoT Hub, the commercial platform that supports

digital twins offered by Microsoft. First, we set up a real-time connection between the cloud platform and two

different IoT devices and explore how an authentic connection is established between IoT devices and their

corresponding DTs. Based on a test bed consisting of widely used IoT devices we analyse the power con-

sumption and execution time of the offered authentication mechanisms that are based on general symmetric or

asymmetric encryption. While the authentication time for a Raspberry Pi is below 0.5 seconds, the same task

took above 4.5 seconds for an Arduino, highlighting the importance of lightweight authentication mechanisms

for real-time communication between IoT devices and DT platforms.

1 INTRODUCTION

Advancements in sensor, microprocessor, and battery

technology in recent years has given rise to small,

low-powered devices that can be used to collect infor-

mation about the physical world (El-hajj et al., 2017;

El-Hajj et al., 2019). Due to their relatively low cost

and small form factor, these sensing devices can be

embedded into objects ranging from buildings, ship-

ping containers, and infrastructure to airplanes and

automobiles (El-Haii et al., 2018). Connecting these

devices together over wireless communication creates

the Internet of Things (IoT), a term initially coined

by Kevin Ashton in 1999 when he proposed linked

RFID enabled devices to the internet for Proctor and

Gamble (Ashton et al., 2009). The technology has ap-

plications in various ﬁelds, such as logistics tracking,

environmental monitoring, theft prevention (a recent

example being the Apple AirTag, and patient moni-

toring in a medical context (Datta and Sharma, 2017).

As an emerging technology, IoT has to deal with a

https://orcid.org/0000-0002-4022-9999

number of growing Challenges (El-Hajj et al., 2019;

El-Hajj et al., 2021). Managing the vast amount of

data coming from myriad sensors poses new network-

ing and data science challenges, and optimizing soft-

ware for low-power, battery restrained devices is be-

ing actively worked on. Another challenge is the se-

curity of the IoT devices and the data transferred be-

tween them. Due to the low-power nature of many

IoT sensor nodes, traditional security techniques no

longer work efﬁciently (Elhajj et al., 2022), and litera-

ture reports the absence of proper lightweight encryp-

tion and authentication mechanisms (El-Hajj et al.,

2019). Recently, the new concept of Digital Twins

(DTs) has been introduced (Fuller et al., 2020) in

various sectors such as industrial production, build-

ing management, health care, and smart cities. With

the help of a completely digital representation of a

physical system, its full life-cycle can be monitored,

analysed, controlled, and optimized (Liu et al., 2021).

The integration of this new technology into existing

IoT systems is becoming more and more popular and

commercial platforms are offered, for example by Mi-

crosoft or Nvidia providing this comprehensive tool

560

Pavlov, V., Hahn, F. and El-Hajj, M.

Security Aspects of Digital Twins in IoT.

DOI: 10.5220/0011714500003405

In Proceedings of the 9th International Conference on Information Systems Security and Privacy (ICISSP 2023), pages 560-567

ISBN: 978-989-758-624-8; ISSN: 2184-4356

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

with small adoption overhead.

In this work we study the security of data that is

being gathered by IoT devices and then forwarded to

Azure Digital Twins, the popular digital twins plat-

form offered by Microsoft. Despite the ongoing re-

search activities in lightweight cryptography, the de-

fault authentication mechanisms supported by Azure

Digital Twins all rely on general symmetric and asym-

metric cryptography including full-ﬂedged X.509 cer-

tiﬁcates. Our practical experiments are based on the

simulation of a minimal real-life scenario where the

Digital Twin is provided with real-time data from

real IoT devices via a mutually authenticated chan-

nel. In this testing environment, we have analyzed

the power consumption and execution time for differ-

ent authentication schemes on two popular IoT end-

devices, that is, the Raspberry Pi 3 Model B and Ar-

duino MKR 1010 WiFi.

The rest of this paper is structured as follows: sec-

tion 2 provides general background information on

digital twins and then explore related work in the ﬁeld

of security and current use-cases for Digital Twins.

In section 3 we detail the setup of our experiment to

show how Digital Twins can be utilised to improve

the security of IoT devices. Section 4 goes into the

results of our experimentation, and section 5 is a dis-

cussion of the work done. Finally, section 6 concludes

the research and presents areas of interest for further

research.

2 BACKGROUND

In this section, we ﬁrst review the state of the art in

Digital Twins applications, focusing on application in

the context of IoT. Then, we combined these reviews

to argue that the current issues in the state of the art

motivate the creation of a framework to apply DTs

IoT-based applications.

2.1 What is a Digital Twin?

A digitally determined model that uniquely represents

a physical instance, process, system, or similar ab-

straction is known as a Digital Twin (DT) in broad

terms(Vrabi

c et al., 2018). The Digital Twin was

ﬁrst introduced by Michael Grieves in a 2003 pre-

sentation on Product Life-cycle Management (PLM),

where Grieves was working with John Vickers of

NASA (Grieves, 2014). Following that, in the PLM

courses, this conceptual model served as a ”Mir-

rored Spaces Model” by Grieves (Grieves and Vick-

ers, 2017). Grieves in (Grieves, 2014) also deﬁned

the architecture to be used while dealing with digital

twins. This architecture consisted of three main com-

ponents: The physical component, the virtual one and

a communication channel between the physical and

virtual component. The whole process is illustrated

in Figure 1.

Information

Virtual

Component

Data

Physical

Component

Figure 1: The Twinning between Physical and Virtual com-

ponents.

In the current state of the art, the only consen-

sus is that a Digital Twin is a virtual representation

of a physical object. The majority of reviewed litera-

ture also considers a bi-directional connection and the

ability to perform simulations on the digital represen-

tation as an integral part of the deﬁnition of Digital

Twins. The ability to persist across multiple phases in

the physical object’s life cycle is explicitly mentioned

in about one third of the reviewed literature according

to (Kuehner et al., 2021).

2.2 Digital Twin Use-Cases

Industrial applications, platforms for life-cycle man-

agement, preventive maintenance, and the automobile

industry are examples of where DTs are used more

broadly (Liu et al., 2021). Irrigation (Purcell and

Neubauer, 2022), medical (Ahmadi-Assalemi et al.,

2020), supply chain (Defraeye et al., 2021), infras-

tructure (Hou et al., 2020), education (Vikhman and

Romm, 2021), natural disaster detection (Fan et al.,

2021), telecommunication (Seilov et al., 2021), and

cybersecurity (Sousa et al., 2021; Saad et al., 2020;

Salvi et al., 2022; Schellenberger and Zhang, 2017)

are some of the most current and upcoming DT uses.

In (Ivanov et al., 2020) authors show that DTs

serve as the foundation for smart communities and

areas in a smart city, which is deﬁned as a strategic

approach to incorporate data and digital technologies

to assure sustainable development, population safety,

and economic development of the urban environment.

Smart cities and city planning are two ﬁelds that have

recently experienced a rise in the incorporation of the

DT idea. Along with the focus on smart digital build-

ings, their servicing, and asset management, there is a

shift toward creating digital twins of entire communi-

ties, or so-called smart Digital Twin cities. Recently,

Security Aspects of Digital Twins in IoT

561

the literature has begun to report on the use of DTs

in medicine(Ahmadi-Assalemi et al., 2020). Numer-

ous uses have been mentioned, including in the areas

of athletics, viral infection simulations, well-being in

smart cities, telemedicine, and healthcare monitoring.

Cybersecurity and social ethics issues are framed as

two of the major barriers facing the conceptual model

of a human Digital Twin (HDT) as mentioned by the

authors in (Shengli, 2021). Their main objective was

to create a virtual representation of the body of a per-

son by utilizing data from wearable devices, mobile

phone and medical records. Web services are used to

update the content of the virtual representation on a

regular basis. The HDT concept incorporates a cer-

tain index that is given to humans at birth, and as they

develop, their biological data is continuously sent into

the HDT as an input.

For modeling, simulating, and improving cyber-

physical systems, the DT idea is essential. Through

application services related to evaluation, model-

ing, tracking, optimizing, and predictive manage-

ment, it can offer a deeper understanding of com-

plicated physical processes (Hou et al., 2020). The

goal of DT engineering applications is to anticipate

future behavior and efﬁciency of physical systems

and to produce useful data that enable the devices

to behave autonomously and providing support in

decision-making process (Zhou et al., 2019). For ex-

ample a DT multi-dimensional model used in con-

struction as an illustration was created by authors in

(Liu et al., 2020) to monitor pre-stressed steel struc-

tures in real-time for safety estimations. Although the

usage of DT principles in the automobile industry has

increased, the majority of the current research is con-

centrated on the automotive assembly processes and

design implementation of automobiles (Sharma and

George, 2018). Even there was a great effort to use

DT in the production of electrical vehicles with the

introduction of IoT and network technologies that en-

ables the ability to convert ofﬂine digital model into

DT. Authors in (Van Mierlo et al., 2021) show that the

development of DT in electrical vehicles could lead to

the ability to prognosis and planning of future mainte-

nance events, monitoring system, fault prediction and

fault location, etc.

Based on this review we conclude that DT appli-

cations are increasingly being used in a wide range

of ﬁelds. This technology – if merged with enabling

technologies like big data, simulation, Information

Technologies (IT) and communication interfaces – be

a novel, useful tool for designers and operators at the

same time. It is clear that DT technology is still in

its early life-stages; overcoming profound challenges

that face a modern DT implementation, such as costs,

information complexity and maintenance, a lack of

rules and regulations, and problems with cybersecu-

rity and communications, will be necessary before DT

technology reaches its full potential.

3 METHOD AND EXPERIMENT

SETUP

In this section we detail our methodology for setting

up a test bed for the chosen DT platform and then

testing the authentication done between the physical

component and the virtual one before start communi-

cating in a real time manner. We ﬁrst list the com-

ponents of our test bed, then explain the setup of the

Digital Twin, and ﬁnally show how the authentica-

tion schemes occurred between the physical and vir-

tual device.

3.1 Framework Overview

For our study we ﬁrst prepared a working scenario

for the deployment of DTs and before we conduct our

analysis later. We used the following components in

our test bed:

1. IoT Hub: An IoT application and the connected

devices use the managed Azure IoT Hub ser-

vice, which is hosted in the cloud and serves as

a central messaging hub. Millions of devices and

their backend programs can be securely and reli-

ably connected. Several messaging patterns are

supported, including device-to-cloud telemetry,

uploading ﬁles from devices, and request-reply

methods to control your devices from the cloud.

IoT Hub offers tracking to assist you in keeping

track of the creation, connections, and failures of

your devices. To serve your IoT workloads, IoT

Hub expands to millions of devices connected at

once and millions of events per second.

2. Azure Digital Twin: Using digital models of

complete environments, such as buildings, facto-

ries, farms, energy networks, railways, stadiums,

and more, Azure Digital Twins is a platform as a

service (PaaS) solution that makes it possible to

create twin graphs. Twin graphs may even be cre-

ated for entire cities. These digital models can be

utilized to gather data that leads to breakthrough

consumer experiences, process optimization, bet-

ter products, and lower costs.

3. Physical Device: We deployed a Raspberry Pi

3B+ as the physical device. Boasting a 64-bit

quad core processor running at 1.4GHz, dual-

band 2.4GHz and 5GHz wireless LAN, Blue-

ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy

562

tooth 4.2/BLE, faster Ethernet, and Power over

Ethernet (PoE) capability via a separate PoE

HAT. And for the purpose of comparison and

benchmarking, we also used Arduino MKR 1010

WiFi which includes a 32-bit Cortex-M0+ ARM

micro-controller (programmable like a conven-

tional Arduino) running at a clock speed up to 48

MHz,providing up to 20 analog inputs and up to 8

digital I/O ports, Wi-Fi and Bluetooth connectiv-

ity, and a crypto-chip for secure communication

using SHA-256 hashing algorithm. In addition,

it has 32 KB of Static Random Access Memory

(SRAM) and 8 general purpose digital pins that

can be either outputs or inputs.

4. DHT11 Sensor: It is a sensor for collecting the

Temperature and Humidity data. It measures the

airﬂow using a sensitive humidity sensor and a

thermostat and outputs a digital signal on the data

pin (no analog input pins needed).

3.2 Experimental Setup

Our experimental setup implements the framework

described in section 3.1. The source code is avail-

able in Gitlab

. Figure 2 describes our prototypi-

cal implementation and documents technology and

data ﬂows. It shows all processes that are executed

during the communication between the IoT device

and the temperature sensors, all the applications and

the Azure services. Certiﬁcate based authentication

is build upon X.509 public key infrastructure (PKI)

standard between the physical device and the DTs (a

more detailed description about authentication pro-

cess is given in section 3.3). The two temperature

sensors are connected to the IoT device and each sen-

sor is authenticated via a separate certiﬁcate. When

the device application authenticates the sensors with

the X.509 certiﬁcates, there are two possible outputs:

ﬁrst, IoTHubDeviceClient indicating a working con-

nection or second, empty indicating a failed authen-

tication and hence no connection. If the clients are

authenticated, then the data from both of the sensors

is formatted as reported properties

and sent to the

Azure IoT Hub which will then update the Digital

Twins that are related to the physical sensors. We

activated a twin patch handler which listens for any

changes of the desired properties, which are usually

updated from the Azure or via backend program. In

our case, we have implemented a back-end program

Removed for blinded submission purposes. Available

upon request

In a JSON-like language format called Digital Twins

Deﬁnition Language (DTDL)

that authenticates with the connection string, which

contains the private key of the IoT Hub and con-

nects the sensors via their Digital Twin IDs. If au-

thentication is successfully achieved, then the desired

properties are updated and the twin patch handler re-

acts that was activated on the IoT device. However,

if authenticated failed, then the IoTHubRegistryMan-

ager is not initialized and an error will be shown.

Our test program also indicates where the sensors are

hosted and which are connected via WiFi. For the pur-

posed of comparison regarding the performance anal-

ysis we deployed our framework on two physical IoT

devices, that is, the Raspberry Pi 3B+ and the Arduino

MKR1010 WiFi.

3.3 Mutual Authentication Between

Physical and Digital Twin

Certiﬁcate based authentication is built by leveraging

the X.509 public key infrastructure (PKI) standard.

Stronger security is provided via certiﬁcate authen-

tication, which uses a Certiﬁcate Authority (CA) to

mutually authenticate the client and the server. De-

vices can be authenticated to an Azure IoT Hub using

X.509 certiﬁcates. A certiﬁcate is a digital record that

includes a device’s public key and can be used to au-

thenticate itself. X.509 certiﬁcates using ECDSA and

the RSA signature algorithms were used in our proof

of concept framework. We used OpenSSL to create

a certiﬁcation authority (CA), a subordinate CA, and

a device certiﬁcate. The example then signs the sub-

ordinate CA and the device certiﬁcate into a certiﬁ-

cate hierarchy. The X.509 digital certiﬁcate contains

various ﬁelds like version, serial number, signature

(hash) algorithm, issuer, valid from, valid to, subject,

the public key of the entity that is the certiﬁcate is is-

sued for and its parameters, enhanced key usage, sub-

ject alternative name, subject key identiﬁer, key usage,

basic constraints, and the thumbprint. Figure 3 shows

the list of ﬁelds created for each certiﬁcate.

The authentication process between the physical

device and the DT is illustrated in Figure 4.

Initially, the IoT device will instantiate the X.509

Certiﬁcate as an object with the pre-generated key

pair from the certiﬁcate ﬁles that we already moved

to the device. Then, it will send the public data of the

document to the Azure CA and the IoT Hub. The CA

and the hub will verify the device’s certiﬁcate and, if

everything is correct, it will instantiate the AzureIo-

THubClient

and establish a secure connection. The

IoTHubClient

will have access to the DT public key

The IoT Hub service client is used to communicate

with devices through an Azure IoT hub.

Is the primary interface for developers using the Azure

Security Aspects of Digital Twins in IoT

563

Figure 2: Framework Overview.

Figure 3: X.509 Certiﬁcate Fields.

which help the device to encrypt the messages. Af-

terwards, when the device is authenticated correctly,

it will start sending the sensor data using the IoTHub-

DeviceClient

which will encrypt this data and send

IotHub client library

A synchronous device client that connects to an Azure

IoT Hub instance.

it to the IoTHub

. This is veriﬁed by the IoT Hub

using the root certiﬁcate, then the data is decrypted

and the used to update the DT properties. Whenever

the Digital Twin make a change from its behalf, it

will ﬁre up patch event which will then be noticed

by the IoT Hub. Then, it will encrypt the modiﬁed

data and send the desired proper- ties to the physical

device. Finally, the physical device, by using the Io-

THubDeviceClient, will decrypt the data passed in the

event handler and will receive it successfully. All the

scripts used for this work can be seen in Github

. Ad-

ditionally, screenshots with the results from our ex-

periments are in the pics-results folder.

4 RESULTS

After we prepared the simulation and everything was

running smoothly, we decided to evaluate the sym-

Azure IoT Hub is a managed service hosted in the

cloud that acts as a central message hub for communication

between an IoT application and its attached devices.

https://github.com/Vitomir2/Digital-Twins-Azure-

IoT-Hub

ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy

564

Figure 4: Authentication Process.

metric key authentication and the X.509 certiﬁcates

by measuring the power consumption. We did this

by using a USB power consumption tester which has

the ability to monitor the energy consumption. Ad-

ditionally, we executed test scenarios to monitor the

execution time of the authentication which was our

second metric used to compare both authentications.

The results can be seen in Table 1. We created two

additional test scripts that are only authenticating the

devices, in order to have more precise results. You

can ﬁnd them on the repository, in folder test-scripts.

For the power consumption we ran the script ten times

and then took the average values that the USB tester

was giving in amperes and volts. In all authentication

types, the Raspberry Pi was working on 5.32 volts.

For the execution time, we decided to execute the run

the script ten times and calculate the average execu-

tion time. Then, we ran this additional ten times in

order to get the more accurate average results for the

execution time. Furthermore, we measured the gener-

ation time of the device certiﬁcates and the generation

time for the root certiﬁcates. The overall results are

presented in Table 2.

Table 1: Measurement Results.

IoT Hardware

Platform

Authentication

protocol

Current(mA) Power(W)

Execution

time(s)

Raspberry PI

Symmetric key 112 0.596 0.0452

X.509 with ECDSA 151 0.803 0.412

X.509 with RSA 163 0.867 0.396

Arduino MKR 1010

Symmetric key 12 0.061 4.683

X.509 with ECDSA 15 0.076 5.121

Table 2: Certiﬁcate Generation Time.

Certiﬁcate

Generation

time(s)

Root Generation

time(s)

X.509 with ECDSA 1.4 5.3

X.509 with RSA 1.8 6.5

5 DISCUSSION

Our environment used a bidirectional communication

which enables both, the physical devices and the DTs,

to communicate to each other. They provide the new

data via the reported and desired properties which are

in a JSON object format. One can add any property

and value and if there is a change of the value of an

already existing property it will then trigger an event

and will update the data accordingly. However, if the

physical twin passes the same data twice, it will up-

date it only once. The same goes for the updates from

Security Aspects of Digital Twins in IoT

565

the DT to the physical device. From the measure-

ment results, for both boards, we see that the execu-

tion time of the connection authenticated by a string

derived from symmetric cryptography is faster than

the both certiﬁcate options. For the Raspberry Pi, it is

approximately 0.0452 seconds, whereas for the cer-

tiﬁcates it is 0.412 seconds and 0.396 seconds for the

ECDSA and RSA certiﬁcates, respectively. For the

Arduino board symmetric key authentication requires

4.683 seconds for the connection string versus 5.121

for the ECDSA certiﬁcate. Additionally, we can see

that it uses approximately 112 mA for the current and

0.596 watts for the power, whereas the certiﬁcates use

approximately 151 mA and 0.803 watts, and 163 mA

and 0.867 watts, respectively. These power measure-

ments show that the digital certiﬁcates are less efﬁ-

cient, in the manner of the energy consumption, than

the symmetric key authentication. Furthermore, we

can see that both certiﬁcates have slight differences in

both the power consumption and the execution time

measurements. For the power consumption on the

Arduino MKR 1010 board, we can see for symmet-

ric and X.509 it consumes a less power but for the

execution time Raspberry PI is faster. Lastly, from

the certiﬁcate generation times, we can see that both,

the generation time for the root certiﬁcates and the

devices’ certiﬁcates, are faster for the ECDSA. We

measured approximately 5.3 seconds for the genera-

tion of the Root ECC certiﬁcate and 1.4 seconds for

the generation of the certiﬁcates for the devices with

ECC. However, for the RSA, we evaluated the genera-

tion time and it was a bit higher of 6.5 seconds for the

root certiﬁcate and 1.8 seconds for the devices’ certiﬁ-

cates. This execution time might be higher, because

of the larger keys for the RSA 2048 bits, whereas in

the ECC, the key was 256 bits. We conclude that the

symmetric key encryption is much faster than the dig-

ital certiﬁcates, and requires less energy. However,

the certiﬁcates give us the opportunity to have better

security, to keep the private keys securely only on the

devices and to have a relation with a speciﬁc identity,

e.g, the IoT devices or individual sensors.

6 CONCLUSION

We study how mutual authentication between phys-

ical devices and Digital Twins is currently imple-

mented at Azure IoT Hub - the commercial plat-

form by Microsoft. Our experiments based on a

proof of concept successfully established an authen-

ticated communication between IoT end-devices and

its mapped Digital Twins. To our surprise, no authen-

tication mechanisms offered by this platform are par-

ticularly lightweight and hence speciﬁcally tailored

towards low-power consumption - a requirement of-

ten formulated for IoT applications. Our performance

analysis measuring the execution time and the power

consumption of two hardware platforms show differ-

ence in execution time that are of one order of mag-

nitude and a similar difference is measured for the

power consumption necessary for the authentication

process between IoT devices and DTs. Speciﬁcally,

the authentication process for the Arduino MKR 1010

required more than 4.5 seconds and hence are of lim-

ited use for real-time applications. From our results

we conclude that further studies are necessary when

combining IoT solutions with the recent Digital Twins

technology. Especially the security challenges for

real-time monitoring are of high relevance. In fu-

ture, we are going to work with a lightweight authen-

tication schemes to ensure the mutual authentication

between the DT and its physical device and then we

will compare the performance analysis with the tradi-

tional authentication schemes like the ones provided

by Azure.

REFERENCES

Ahmadi-Assalemi, G., Al-Khateeb, H., Maple, C.,

Epiphaniou, G., Alhaboby, Z. A., Alkaabi, S., and Al-

haboby, D. (2020). Digital twins for precision health-

care. In Cyber defence in the age of AI, Smart societies

and augmented humanity, pages 133–158. Springer.

Ashton, K. et al. (2009). That ‘internet of things’ thing.

RFID journal, 22(7):97–114.

Datta, P. and Sharma, B. (2017). A survey on iot architec-

tures, protocols, security and smart city based applica-

tions. In 2017 8th International Conference on Com-

puting, Communication and Networking Technologies

(ICCCNT), pages 1–5.

Defraeye, T., Shrivastava, C., Berry, T., Verboven, P., On-

wude, D., Schudel, S., B

uhlmann, A., Cronje, P., and

Rossi, R. M. (2021). Digital twins are coming: Will

we need them in supply chains of fresh horticultural

produce? Trends in Food Science & Technology,

109:245–258.

El-Haii, M., Chamoun, M., Fadlallah, A., and Serhrouchni,

A. (2018). Analysis of cryptographic algorithms on

iot hardware platforms. In 2018 2nd Cyber Security

in Networking Conference (CSNet), pages 1–5. IEEE.

El-hajj, M., Chamoun, M., Fadlallah, A., and Serhrouchni,

A. (2017). Analysis of authentication techniques in in-

ternet of things (iot). In Cyber Security in Networking

Conference (CSNet), 2017 1st, pages 1–3. IEEE.

El-Hajj, M., Fadlallah, A., Chamoun, M., and Serhrouchni,

A. (2019). Ethereum for secure authentication of iot

using pre-shared keys (psks). In 2019 International

Conference on Wireless Networks and Mobile Com-

munications (WINCOM), pages 1–7.

ICISSP 2023 - 9th International Conference on Information Systems Security and Privacy

566

El-Hajj, M., Fadlallah, A., Chamoun, M., and Serhrouchni,

A. (2019). A survey of internet of things (iot) authen-

tication schemes. Sensors, 19(5):1141.

El-Hajj, M., Fadlallah, A., Chamoun, M., and Serhrouchni,

A. (2021). A taxonomy of puf schemes with a novel

arbiter-based puf resisting machine learning attacks.

Computer Networks, 194:108133.

Elhajj, M., Jradi, H., Chamoun, M., and Fadlallah, A.

(2022). Lasii: Lightweight authentication scheme us-

ing iota in iot platforms. In 2022 20th Mediterranean

Communication and Computer Networking Confer-

ence (MedComNet), pages 74–83. IEEE.

Fan, C., Zhang, C., Yahja, A., and Mostafavi, A. (2021).

Disaster city digital twin: A vision for integrating ar-

tiﬁcial and human intelligence for disaster manage-

ment. International Journal of Information Manage-

ment, 56:102049.

Fuller, A., Fan, Z., Day, C., and Barlow, C. (2020). Digi-

tal twin: Enabling technologies, challenges and open

research. IEEE access, 8:108952–108971.

Grieves, M. (2014). Digital twin: manufacturing excel-

lence through virtual factory replication. White paper,

1(2014):1–7.

Grieves, M. and Vickers, J. (2017). Digital twin: Mitigat-

ing unpredictable, undesirable emergent behavior in

complex systems. In Transdisciplinary perspectives

on complex systems, pages 85–113. Springer.

Hou, L., Wu, S., Zhang, G., Tan, Y., and Wang, X.

(2020). Literature review of digital twins applications

in construction workforce safety. Applied Sciences,

11(1):339.

Ivanov, S., Nikolskaya, K., Radchenko, G., Sokolinsky, L.,

and Zymbler, M. (2020). Digital twin of city: Concept

overview. In 2020 Global Smart Industry Conference

(GloSIC), pages 178–186. IEEE.

Kuehner, K. J., Scheer, R., and Strassburger, S. (2021). Dig-

ital twin: Finding common ground – a meta-review.

Procedia CIRP, 104:1227–1232. 54th CIRP CMS

2021 - Towards Digitalized Manufacturing 4.0.

Liu, M., Fang, S., Dong, H., and Xu, C. (2021). Review of

digital twin about concepts, technologies, and indus-

trial applications. Journal of Manufacturing Systems,

58:346–361.

Liu, Z., Bai, W., Du, X., Zhang, A., Xing, Z., and Jiang, A.

(2020). Digital twin-based safety evaluation of pre-

stressed steel structure. Advances in Civil Engineer-

ing, 2020.

Purcell, W. and Neubauer, T. (2022). Digital twins in agri-

culture: A state-of-the-art review. Smart Agricultural

Technology, page 100094.

Saad, A., Faddel, S., Youssef, T., and Mohammed, O. A.

(2020). On the implementation of iot-based digi-

tal twin for networked microgrids resiliency against

cyber attacks. IEEE transactions on smart grid,

11(6):5138–5150.

Salvi, A., Spagnoletti, P., and Noori, N. S. (2022). Cyber-

resilience of critical cyber infrastructures: Integrating

digital twins in the electric power ecosystem. Com-

puters & Security, 112:102507.

Schellenberger, C. and Zhang, P. (2017). Detection of

covert attacks on cyber-physical systems by extend-

ing the system dynamics with an auxiliary system. In

2017 IEEE 56th Annual Conference on Decision and

Control (CDC), pages 1374–1379. IEEE.

Seilov, S., Kuzbayev, A., Seilov, A., Shyngisov, D.,

Goikhman, V. Y., Levakov, A., Sokolov, N., Zhursin-

bek, Y. S., et al. (2021). The concept of building a net-

work of digital twins to increase the efﬁciency of com-

plex telecommunication systems. Complexity, 2021.

Sharma, M. and George, J. (2018). Digital twin in the

automotive industry: Driving physical-digital conver-

gence. Tata Consultancy Services White Paper.

Shengli, W. (2021). Is human digital twin possible? Com-

puter Methods and Programs in Biomedicine Update,

1:100014.

Sousa, B., Arieiro, M., Pereira, V., Correia, J., Lourenc¸o,

N., and Cruz, T. (2021). Elegant: security of criti-

cal infrastructures with digital twins. IEEE Access,

9:107574–107588.

Van Mierlo, J., Berecibar, M., El Baghdadi, M., De Cauwer,

C., Messagie, M., Coosemans, T., Jacobs, V. A., and

Hegazy, O. (2021). Beyond the state of the art of elec-

tric vehicles: A fact-based paper of the current and

prospective electric vehicle technologies. World Elec-

tric Vehicle Journal, 12(1):20.

Vikhman, V. and Romm, M. (2021). “digital twins” in edu-

cation: Prospects and reality. Vysshee Obrazovanie v

Rossii= Higher Education in Russia, 30(2):22–32.

Vrabi

c, R., Erkoyuncu, J. A., Butala, P., and Roy, R. (2018).

Digital twins: Understanding the added value of in-

tegrated models for through-life engineering services.

Procedia Manufacturing, 16:139–146. Proceedings of

the 7th International Conference on Through-life En-

gineering Services.

Zhou, M., Yan, J., and Feng, D. (2019). Digital twin frame-

work and its application to power grid online anal-

ysis. CSEE Journal of Power and Energy Systems,

5(3):391–398.

Security Aspects of Digital Twins in IoT

567