Vulnerabilities of internet of things, for healthcare devices and applications

Vulnerabilities of Internet of Things, for Healthcare Devices and Applications Vulnerabilities of Internet of Things, for Healthcare Devices and Applications Kyriaki Tsantikidou SCYTALE Group, Compute[.]

Vulnerabilities of Internet of Things, for Healthcare Devices and Applications Kyriaki Tsantikidou

SCYTALE Group, Computer Engineering and Informatics Department (CEID)

University of Patras, Hellas

k.tsantidikidou@upatras.gr

Nicolas Sklavos

SCYTALE Group Computer Engineering and Informatics Department (CEID)

University of Patras, Hellas

nsklavos@upatras.gr

Abstract—As the Internet of Things (IoT) is progressively

integrated into people's life, the security aspect of Healthcare

applications is drastically becoming a major concern for the

research community The IoT technology comes with many

vulnerabilities, which can threaten the personal and public

safety Many new technologies are being introduced as solutions

to IoT issues Three of the most promising are Machine learning,

Deep Learning and Blockchain However, these solutions are

not impeccable In this paper, various IoT-based hardware

implementations for security in Healthcare applications are

demonstrated Afterwards, the major vulnerabilities as well as

threats and few common attacks of the utilized IoT components

in Healthcare are discussed; that includes, the challenges and

usual attack methods of the mentioned three security techniques

(Machine learning, Deep Learning and Blockchain) Finally,

efficient and flexible solutions for Healthcare, are presented

Keywords—IoT, Vulnerabilities, Healthcare, Hardware

Implementations, Machine Learning (ML), Deep Learning (DL)

I INTRODUCTION

The Internet of Things (IoT) is a promising technology

that mainly consists of a plethora of computationally

constrained devices that are interconnected to each other and

further connected to the Internet IoT can have profound

effects in smart applications and thus in humans' daily routine

and well-being Specifically, it can integrate the Internet and

digital world with the physical one [1], hence offering new

services and possibilities Many domains, such as smart cities,

smart homes, agriculture, environmental monitoring, etc., are

adopting IoT-based technologies Healthcare is also one of the

main domains that employs the IoT technology, deeming it

easily accessible, user-friendly and expanding its capabilities

However, as more heterogeneous devices are introduced

into the interconnected network of IoT and the application

requirements are increasing, the scale of security problems is

expanding and the maintenance of the main security concepts,

namely confidentiality, availability and integrity, is becoming

a major concern The continuously evolving IoT technology

presents new vulnerabilities that can be exploited by attackers

and the existing security policies are inadequate to prevent [2]

These drawbacks can be devastating for e-healthcare

applications First, the collected data are private and must be

protected because of ethical and legal implications Moreover,

a malfunction, such as Denial of Service (DoS), caused by an

attack can negatively affect the users and possibly threaten

their lives Conclusively, the implementation of IoT in

Healthcare must be carefully examined

New security methods, with possibly fewer

vulnerabilities, are being investigated for IoT-based

healthcare applications Two of them are Machine Learning

(ML) and Deep Learning (DL), whose advances have increased the systems’ intelligent In recent years, ML/DL algorithms have also been applied in IoT security [3] The results have been positive and many ML/DL mechanisms have been proposed for securing one or multiple of the IoT layers Prime examples are physical-layer authentication [4], attack detection [5] and malware detection [6] Another security method, whose compatibility with IoT network is being tested, is Blockchain The anticipated results of the integration of Blockchain and IoT technologies are mechanisms for providing potential solutions to the limitations of both, as presented in [7] Furthermore, with the addition of 6G, the security as well as the bandwidth can be increased However, with new opportunities even more unexpected challenges constantly emerge providing attackers with additional vulnerabilities to explore

In the past years, many scientific papers have been concerned with the rapid utilization of IoT technology and its continuously increasing number of vulnerabilities and threats Most papers discuss the vulnerabilities, threats and attacks for each layer of IoT and for diverse IoT applications [2], [8-10] Moreover, they present solutions, guidelines and future challenges for IoT security Nevertheless, studies that present hardware implementations of existing or novel security methods, such as [11], are few

In this paper, a number of hardware implementations of IoT-based Healthcare security applications are discussed, including Machine learning- and Deep learning-based architectures and Blockchain integration designs After their further inspection, the primary security vulnerabilities of IoT technology in smart Healthcare are presented in detail Some efficient and flexible approaches for securing IoT systems are introduced Essentially, the contributions of this work are:

• This work presents a variety of IoT-based hardware implementations by related papers that are utilized for securing Healthcare systems or have great potential of being applied as secure mechanisms in this domain

• It assembles and analyses the most well-known, to the best of our knowledge, vulnerabilities of the major IoT components, viz hardware, software and network, and lists some of their succeeding attacks

• The research includes additional information about the main challenges of three promising techniques for IoT security (Machine learning, Deep learning and Blockchain technology)

• Finally, exemplary solutions for appropriately securing IoT-based Healthcare systems are displayed

II HARDWARE IMPLEMENTATIONS

A IoT-based Healthcare security architectures

In this section, a variety of IoT-based hardware

implementations for security, which can be utilized in

Healthcare applications are presented First, exclusively smart

health architectures are analysed Afterwards, more general

implementations, such as Wireless Sensor Networks (WSNs)

security mechanisms, ML/DL and Blockchain designs, etc.,

are demonstrated These designs are employed for securing

various features and IoT components Nevertheless, all these

IoT elements are also being utilized in Healthcare systems,

hence the discussed ideas and techniques are beneficial

E-health implementations: E-health applications in IoT

technology improve quality of life, monitor and record vital

functionalities and contribute to the recognition and

prevention of serious medical emergencies A general

architecture for an IoT-based healthcare system is depicted in

Fig 1 In [12] a reliable healthcare IoT system for monitoring

diabetic patients, storing the appropriate data and controlling

an insulin pump is designed The implementation utilizes the

IoT-cloud, an Alaris 8100 infusion pump and the Keil

LPC-1768 board with embedded Ethernet port and Cortex-M3

micro-controller The experimental results prove that the

proposed implementation is reliable, secure, authentic and

achieves a 99.3% availability probability The authors propose

the development of a more generalized reliability model as a

future research for this design In [13] the SecureData scheme

is proposed This secure data collection scheme attends to

privacy and security concerns for IoT-based healthcare The

presented security algorithm is optimized on a FPGA, but an

evaluation of the design's security and privacy has not been

displayed [14] develops a real-time authentication and

random signatures generation system It utilizes three sources

of entropy, two of which are unique for each person, namely

Electroencephalogram (EEG) and heart rate variability

(HRV), and the last one is unique for each device, namely

SRAM-based physical unclonable function (PUF) The

developed system was fabricated in a 65-nm LP CMOS

technology and is efficient for real-time authentication and

production of random secret keys Finally, [15] presents a

design framework which can assist Internet of Medical Things

(IoMT) designers to achieve a more secure implementation

The framework can provide early evaluations and remarks of

a design under Side Channel Analysis (SCA) attacks

General implementations: In [16], an FPGA-based

implementation for monitoring the embedded operating

system is presented The developed hardware-based security

design can detect any attacks which alter the original

behaviour without utilizing a maximum amount of additional

hardware resources [17] proposes a System-level Mutual

Authentication (SMA) scheme securing both hardware and

firmware from various familiar attacks Specifically, the

proposed approach enables hardware and firmware to

authenticate each other with the utilization of a unique system

ID and a firmware obfuscation methodology In [18] a library

of emerging family of lightweight elliptic curves is designed

for IoT systems It is optimized resulting in a high-speed

version and a memory-efficient version, both resilient against

timing and simple power analysis (SPA) attacks The

implementation is evaluated in an 8-bit ATmega128 and a

MSP430 processor with 16-bit multiplier

Fig 1 IoT-based Healthcare architecture (Adapted from [19])

Security implementations for WSNs: Wireless Sensor

Network (WSN) is utilized by a wide range of application areas and is being integrated into IoT [20] Therefore, it is important to study WSNs in the interest of further securing Healthcare systems In [21], an enhanced lightweight version

of Advanced Encryption Standard (AES) is proposed This Lightweight AES (LAES) replaces the MixColumns function with a bitwise permutation The algorithm is implemented in

a Virtex-7 FPGA and is simulated with Xilinx ISE tools The resulted architecture utilizes fewer resources than previous implementations and achieves higher percentages on plaintext bit flip and key bit flip Hence, the LAES is deemed more secure in terms of avalanche effect

Implementations of ML and DL security algorithms: In

[22], an online learning approach for protecting a custom many-core architecture from unexpected attacks is proposed The authors present the training process of the Trojan detection model that can perceive even unexpected attacks The average Trojan detection accuracy is 93% and requires 5.6 uS to be executed Furthermore, the architecture is implemented on Xilinx Virtex-7 FPGA with a low area overhead and latency Another FPGA-based implementation

of a deep learning algorithm for anomaly detection is introduced in [23] The proposed model is a three-layer Deep Belief Network (DBN) which is trained with the MNIST dataset and tested with two different datasets while using the FPGA The purpose of the model is the improvement of anomaly detection in network attacks Overall, the resource-efficient design has a slight decrease in accuracy but an efficient increase in detection speed compared to a C/C++ implementation Finally, [24] implements a machine learning-based Trojan detection method in SAKURA-G circuit board with Xilinx SPARTAN-6 The primary machine learning method utilized is Support Vector Machine (SVM) and the developed design produces high accuracy detection of Trojan

in the circuit

Implementations of the Blockchain concept: In [25], an

FPGA-based implementation of SHA-256, which provides security and privacy in the Blockchain architecture, is proposed The presented parameterizable SHA-256 hardware design is simulated with ModelSim RTL Simulator and implemented with Xilinx Vivado 18.2 Design Suite The results are positive with the authors proposing more future utilizations of their architecture Lastly, [26] demonstrates a blockchain architecture called PUFChain which utilizes physical unclonable functions (PUFs) and can be integrated into the IoT technology The authors also propose a consensus algorithm named Proof of PUF-Enabled Authentication (PoP)

which combines a PUF and hashing module, provides device

and data security and reduces the computational load and

transaction time For the experimental segment, the PoP

algorithm was implemented in the Altera DE2 FPGA module

and for its evaluation, supplementary Raspberry pis were

employed as the network nodes Overall, the developed

architecture achieves appropriate speedup, resource utilization

and scalability

B Security mechanisms in other domains

Smart Home: A Smart Home consists of various

IoT-based home automation devices which can be remotely

controlled by a user The design in [27] can be a significant

example for also securing smart health applications In the

reference paper, three different detection methods for home

IoT devices are displayed The system's architecture consists

of three components, the SPIDAR home Wi-Fi router, the

SPIDAR Raspberry Pi and the SPIDAR web application The

three methods' processing time, CPU utilization and detection

accuracy of primarily Brute force password attacks,

DoS/DDoS, SQL injection, Cross-site scripting (XSS) and

Evil twin attacks are calculated through various

implementation experiments

Smart Agriculture: Smart Agriculture consists of a smart

sensors network that can assist in managing and monitoring

fields and farms The architecture presented in [28] can also

be referenced as a fine example for securing remote health

systems Specifically, [28] proposes a method for securing a

WSN for remote farm monitoring The WSN is developed

using Atmega and MSP Microcontroller and Raspberry pi

These components were selected based on their ability to

execute the designed security scheme Finally, this security

efficiency is evaluated through the resistance to attacks and

the bit number used for the symmetrical and asymmetric keys

III SECURITY VULNERABILITIES, CHALLENGES AND

SOLUTIONS

In this section, the vulnerabilities and threats in every layer

of an IoT-based Healthcare system are presented and

analysed The Healthcare architecture depicting all IoT layers

is presented in Fig 2 Following the example of paper [2], the

vulnerabilities are categorized into two sections, the

embedded vulnerabilities, namely hardware and software, and

the network vulnerabilities, as illustrated in Fig 3 Moreover,

the three most common security methods of recent years are

investigated while their challenges and common attacks are

introduced

A Embedded vulnerabilities

First, the majority of IoT applications require their devices

to execute functions autonomously or remotely in an

unattended and questionable environment [29] Therefore, the

devices can be subjected to a variety of hardware attacks, such

as reverse-engineering, side-channel attack, etc An

undisturbed attacker can also inflict physical damage or even

remove completely the device, resulting in Denial of Service

[30] Furthermore, because the IoT devices mostly implement

weak or no security algorithms for prevention of physical

tampering and even have unnecessary open ports accessible

without authentication, the attackers can obtain unauthorized

access and control of the device by extracting hardware

credentials and thereafter corrupting the entire system [8]

Finally, natural disasters that are unpredictable can cause

some physical damage to poorly designed and exposed

devices

Fig 2 IoT layers for a Healthcare system (Adapted from [19], [31])

Second, the definition of IoT states the use of mostly computational and memory constrained devices with low energy efficiency The result is a device that is defenceless to attacks because of the inability to implement heavyweight security algorithms, which are more reliable than many of the proposed lightweight solutions [32] Moreover, the device cannot execute multiple security protocols offering protection from all possible attacks in a variety of user cases that threaten the IoT technology [10] Additional hardware may be needed for a more universal security, but in many cases, such as the design of e-health IoT devices, it is not ideal [15] The mentioned constraints can also affect the implementation of secure authentication mechanisms [2], [8], [29] Finally, the limited memory and insufficient energy consumption can be exploited leading to Denial of Service because the device has been corrupted or simply shut down [9]

In addition, most corporations focus on the functionality and the implementation cost of the commercial devices rather than the security and the hardware integrity [10] This can lead

to adoption of inept programming practices, inefficient access control, improper patch management capabilities and non-existent security patches/updates [8] In some cases, the inaccessibility of the devices, because of the deployment area, can hinder the timely installation of security patches/updates and even raise the cost [33]

Fig 3 Security Vulnerabilities of IoT-based Healthcare systems

Other vulnerabilities emerge from the IoT supply chain A supply chain is described as the required steps for a product or service to be completed and delivered to the final costumer The primarily threats are hardware counterfeiting attacks that vary from simple relabelling attacks to complete reverse engineering The result is loss of revenue and reliability and

even more damage to the system by malware injections and

hardware trojan insertions [34] Current commercialised

products are vulnerable to these attacks because of the

advances in reverse engineering and the high cost required to

design and maintain an anti-counterfeiting technique

Lastly, the heterogeneous nature of IoT systems requires a

more generic approach which is not always feasible because

of the limitations in memory and power Moreover, the lack

of authentication and enforcement of the least privilege

principle [2] can lead to security breaches because of the

inclusion of computationally weak and easily corrupted

devices that can access primary control ones This can be

extended to user unawareness of security attacks and proper

protection methods For example, a user can bestow a

wearable device to an attacker without really realizing the

possible security damage which can be caused either through

malware injection, corruption and control of the entire system

or data exposure Furthermore, an attacker can execute many

easy attacks because the user has unintentionally revealed the

security credentials, granted access to an adversary, or

downloaded a malicious code [10-11]

B Network vulnerabilities

The majority of IoT applications consist of wearable or

mobile devices that constantly connect or disconnect from a

variety of familiar or unknown and public or private networks

Hence, the devices must implement dynamic algorithms

which can sustain their proper functionality and provide total

protection from this range of networks In addition, as more

devices are invented and introduced, more malicious

presences can easily pass undetected, and more vulnerabilities

of communication protocols can emerge and be exploited

[35] Also, because of the sheer number of interconnected IoT

devices, their management and servicing to the global network

requires more complex security schemes [9], [36] However,

as already mentioned, the IoT devices have memory, power

and computational constraints which render the

implementation difficult

Moreover, wireless communication already has a variety

of unsolved challenges The unique vulnerabilities of WSNs

will also be a concern in IoT networks Many papers, such as

[36] and [37], demonstrate the necessity of studying WSNs

and present their vulnerabilities and unresolved challenges

IoT applications, like WSNs, use wired and wireless

communications, such as Wi-Fi, Bluetooth, ZigBee, Radio

Frequency Identification (RFID), 6LoWPAN and LoRaWAN

technologies, which all have notorious vulnerabilities and

threats As revealed in [38], Wi-Fi, Bluetooth, ZigBee and

RFID technologies have a major vulnerability in their

authentication protocol that the attacker can exploit and then

compromise the system [11] also touches upon the existing

threats to ZigBee, Bluetooth, 6LoWPAN and LoPaWAN and

introduces their possible attack vectors Finally, large-scale

IoT applications can utilize Fifth Generation (5G)

communication systems that add more security requirements

and manners for potential attacks, as described in [39]

Another network vulnerability relates to routing and key

management First, in [40], the vulnerabilities and potential

attacks of the routing protocol RPL are presented

Furthermore, the authors emphases the need for more research

in that field because RPL-specific attacks can have a large

impact on the IoT networks Second, the implementation of

weak key management schemes, because of the IoT device

constraints, or the utilization of common keys, which originate

from the commercial devices due to bad security practises, cannot prevent attackers, from easily extracting the communication keys with methods, such as brute force [2] Lastly, the main threat that can exploit a variety of vulnerabilities from all IoT layers and in particular the network layer is the Denial-of-Service (DoS) attack and the Distributed Denial of Service (DDoS) attack Specifically, all the previously mentioned vulnerabilities, which result in partially or totally unauthorized control of devices, can be utilized by forcing the compromised devices to send huge number of requests or messages to another crucial device or even the server The result is the unavailability of the system's services for the simple user because the huge number of requests cannot be handled or the device has run out of battery Many studies [2], [8-11], [41] have outlined the vulnerabilities which lead to DoS and DDoS attacks and the importance of designing mitigation mechanisms

C Security challenges, Machine learning and Deep learning

First and foremost, as ML and DL techniques evolve, they can be applied for both securing and exposing a system Specifically, neural networks can be used for attacking security systems that were previously hard to break [42] For example, an attacker can use a neural network to copy the mathematical contains of security functions in various systems, e.g., Physical Unclonable Functions (PUFs)

In addition, due to IoT devices' hardware limitations and ML/DL algorithms' computational complexity, a completely secure hardware implementation of these mechanisms, especially for real-time applications, is becoming increasingly hard for future researchers [31] Microprocessors and circuits that may be befitting for executing ML/DL are difficult to design, expensive and have high-energy consumption ratio

As most IoT manufacturers utilize low-cost and low-power components for their application [1], such solutions are discarded Moreover, the lack of established methods that explain the execution of DL algorithms depending on their architecture, renders problematic their understanding Therefore, the design of lightweight DL mechanisms that can

be appropriate to resource-constrained devices is deemed challenging [31]

Another vulnerability in machine learning algorithms is associated with the training data set The massive amount of heterogeneous data received by a single device requires high efforts of pre-processing for their utilization as inputs to a specific model In many cases, ML-based networks do not implement data pre-processing or cleaning methods Instead, they assume the integrity and availability of the training data [43] Furthermore, an attacker can render useless a ML security algorithm by injecting false data and incorrect labels

to the training set or by modifying the features of their attack scheme to differentiate from the training samples The result

is the decreasing of ML's accuracy and performance [31] displays the various attacks that can exploit the training data vulnerability and even expose the classifier's training parameters Thus, the training phase of the ML/DL algorithms must first be protected and secured from incorrect data injection and information extraction to provide the security benefits such a mechanism can offer

D Security challenges with Blockchain technology

Blockchain and IoT integration is applied to various sections, as demonstrated in Fig 4, and offers many benefits

However, many issues also arise The primary challenge is

related with the resource constraints in IoT devices

Specifically, for the implementation of Blockchain, all IoT

devices of the network must be capable of executing the

resource-intensive process of reaching consensus by voting

[7] Thus, even though a decentralized architecture can be

cost-efficient compared to other centralized methods, the

overall resource wasting can be deterrent for many IoT

designers Moreover, the Blockchain mechanism alone does

not provide total security Instead, supplementary

cryptography algorithms must be implemented which further

burdens the resource constrained IoT device

Fig 4 Sections of Blockchain integrated with IoT applications

A second challenge stems from the sheer number of

interconnected devices in an IoT network Specifically, as

more IoT devices are introduced into the network, the size of

the Blockchain is increasing because more transactions, which

need to be stored, arise This can cause problems for IoT's

massive number of interconnected devices and transmitted

data because some Blockchain architectures can only manage

a few transactions per second [7] Blockchain technology is

not constructed to keep up with the scalability of IoT

Lastly, for the integration to succeed, both Blockchain and

IoT security and functionality must be successfully certified

independently [7] For example, IoT's susceptibility to data

alteration attacks can cause problems when these corrupted

data remain in the Blockchain Furthermore, the probabilistic

nature of the transaction latency in Blockchain can hinder and

degrade the availability of the IoT system, which can be

dangerous in critical applications such as healthcare

E Efficient and Flexible Solutions

In recent years, many papers have attempted to present

encouraging approaches towards securing IoT-based systems

[44] and [45] describe an exemplary framework for

Healthcare applications with the intention to limit the

vulnerabilities and possible threats of Healthcare designs

Moreover, the authors demonstrate security characteristics

and future directions for efficient architectures The latter also

investigates various proposed solutions and models with their

advantages and drawbacks listed

Some appropriate architectures that can offer

implementational solutions to mentioned vulnerabilities of

IoT systems are presented In [46], a cryptographic system is

efficiently implemented in an IoT device while providing

flexible security schemes The architecture can be applied to

various scenarios, based on encryption/decryption needs and

available hardware resources, and effectively offer security,

authentication and confidentiality Lastly, [47] proposes an

interactive lightweight architecture for data streaming that is

implemented in a FPGA Even though it employs three

different stream ciphers for three modes, the design achieves

to allocate fewer resources than conventional means while maintaining the security level offered by the three ciphers

IV CONCLUSIONS AND OUTLOOK

The analysis of IoT systems exposes numerous vulnerabilities that can easily be exploited resulting in serious concerns Paradoxically, the newly integrated security methods come with additional threats that render the systems even more hazardous As the presented hardware implementations demonstrate, it is difficult to design a hardware-based lightweight security mechanism that is both resource/energy efficient and completely secure Nevertheless, the careful analysis of the mentioned approaches and the consideration of the topics discussed in this paper will direct future research aiming to develop solutions for IoT security in Healthcare applications

References

[1] H Fei, Security and Privacy in Internet of Things (IoTs): Models, Algorithms, and Implementations, CRC Press, 2016, pp 237-266, 355-380

[2] I Stellios, P Kotzanikolaou, M Psarakis, C Alcaraz and J Lopez, “A Survey of IoT-Enabled Cyberattacks: Assessing Attack Paths to

Critical Infrastructures and Services,” IEEE Communications Surveys

Tutorials, vol 20, no 4, pp 3453-3495, 2018

[3] Y E Sagduyu, Y Shi and T Erpek, “IoT Network Security from the

Perspective of Adversarial Deep Learning,” in 2019 16th Annual

IEEE International Conference on Sensing, Communication, and Networking (SECON), 2019, pp 1-9

[4] N Wang, T Jiang, S Lv and L Xiao, “Physical-Layer Authentication

Based on Extreme Learning Machine,” IEEE Communications

Letters, vol 21, no 7, pp 1557-1560, 2017

[5] S S Swarna Sugi and S R Ratna, “Investigation of Machine Learning Techniques in Intrusion Detection System for IoT Network,”

in 2020 3rd International Conference on Intelligent Sustainable

Systems (ICISS), 2020, pp 1164-1167

[6] Niall McLaughlin, Jesus Martinez del Rincon, BooJoong Kang, Suleiman Yerima, Paul Miller, Sakir Sezer, Yeganeh Safaei, Erik Trickel, Ziming Zhao, Adam Doupé, and Gail Joon Ahn, “Deep

Android Malware Detection,” in Proceedings of the Seventh ACM on

Conference on Data and Application Security and Privacy, New York,

NY, USA, Association for Computing Machinery, 2017, p 301–308 [7] Mohammad Maroufi, Reza Abdolee, Behzad Mozaffari Tazekand,

“On the Convergence of Blockchain and Internet of Things (IoT)

Technologies,” CoRR, vol abs/1904.01936, 2019

[8] N Neshenko, E Bou-Harb, J Crichigno, G Kaddoum and N Ghani,

“Demystifying IoT Security: An Exhaustive Survey on IoT Vulnerabilities and a First Empirical Look on Internet-Scale IoT

Exploitations,” IEEE Communications Surveys Tutorials, vol 21, no

3, pp 2702-2733, 2019

[9] S Khanam, I B Ahmedy, M Y Idna Idris, M H Jaward and A Q Bin Md Sabri, “A Survey of Security Challenges, Attacks Taxonomy

and Advanced Countermeasures in the Internet of Things,” IEEE

Access, vol 8, pp 219709-21974, 2020

[10] I Makhdoom, M Abolhasan, J Lipman, R P Liu and W Ni,

“Anatomy of Threats to the Internet of Things,” IEEE

Communications Surveys Tutorials, vol 21, no 2, pp 1636-1675,

2019

[11] F Meneghello, M Calore, D Zucchetto, M Polese and A Zanella,

“IoT: Internet of Threats? A Survey of Practical Security

Vulnerabilities in Real IoT Devices,” IEEE Internet of Things Journal,

vol 6, no 5, pp 8182-8201, 2019

[12] Zeyad A Al-Odat, Sudarshan K Srinivasan, Eman M Al-Qtiemat, Sana Shuja, “A Reliable IoT-Based Embedded Health Care System

for Diabetic Patients,” CoRR, vol abs/1908.06086, 2019

[13] H Tao, M Z A Bhuiyan, A N Abdalla, M M Hassan, J M Zain and T Hayajneh, “Secured Data Collection With Hardware-Based

Ciphers for IoT-Based Healthcare,” IEEE Internet of Things Journal,

vol 6, no 1, pp 410-420, 2019

[14] S K Cherupally, S Yin, D Kadetotad, C Bae, S J Kim and J Seo,

“A Smart Hardware Security Engine Combining Entropy Sources of

ECG, HRV, and SRAM PUF for Authentication and Secret Key

Generation,” in 2019 IEEE Asian Solid-State Circuits Conference

(A-SSCC), 2019, pp 145-148

[15] K Nomikos, A Papadimitriou, G Stergiopoulos, D Koutras, M

Psarakis and P Kotzanikolaou, “On a Security-oriented Design

Framework for Medical IoT Devices: The Hardware Security

Perspective,” in 2020 23rd Euromicro Conference on Digital System

Design (DSD), 2020, pp 301-308

[16] A Pouraghily, T Wolf and R Tessier, “Hardware support for

embedded operating system security,” in 2017 IEEE 28th

International Conference on Application-specific Systems,

Architectures and Processors (ASAP), 2017, pp 61-66

[17] U Guin, S Bhunia, D Forte and M M Tehranipoor, “SMA: A

System-Level Mutual Authentication for Protecting Electronic

Hardware and Firmware,” IEEE Transactions on Dependable and

Secure Computing, vol 14, no 3, pp 265-278, 2017

[18] Z Liu, X Huang, Z Hu, M K Khan, H Seo and L Zhou, “On

Emerging Family of Elliptic Curves to Secure Internet of Things: ECC

Comes of Age,” IEEE Transactions on Dependable and Secure

Computing, vol 14, no 3, pp 237-248, 2017

[19] J J P C Rodrigues et al., “Enabling Technologies for the Internet of

Health Things,” IEEE Access, vol 6, pp 13129-13141, 2018

[20] M Kocakulak and I Butun, “An overview of Wireless Sensor

Networks towards internet of things,” in 2017 IEEE 7th Annual

Computing and Communication Workshop and Conference (CCWC),

2017, pp 1-6

[21] H B Acla and B D Gerardo, “Security Analysis of Lightweight

Encryption based on Advanced Encryption Standard for Wireless

Sensor Networks,” in 2019 IEEE 6th International Conference on

Engineering Technologies and Applied Sciences (ICETAS)}, 2019, pp

1-6

[22] A Kulkarni, Y Pino and T Mohsenin, “Adaptive real-time Trojan

detection framework through machine learning,” in 2016 IEEE

International Symposium on Hardware Oriented Security and Trust

(HOST), 2016, pp 120-123

[23] K Alrawashdeh and C Purdy, “Reducing calculation requirements in

FPGA implementation of deep learning algorithms for online anomaly

intrusion detection,” in 2017 IEEE National Aerospace and

Electronics Conference (NAECON), 2017, pp 57-62

[24] T Hu, L Wu, X Zhang, Y Yin and Y Yang, “Hardware Trojan

Detection Combine with Machine Learning: an SVM-based Detection

Approach,” in 2019 IEEE 13th International Conference on

Anti-counterfeiting, Security, and Identification (ASID), 2019, pp 202-206

[25] D K.N and R Bhakthavatchalu, “Parameterizable FPGA

Implementation of SHA-256 using Blockchain Concept,” in 2019

International Conference on Communication and Signal Processing

(ICCSP), 2019, pp 0370-0374

[26] S P Mohanty, V P Yanambaka, E Kougianos and D Puthal,

“PUFchain: A Hardware-Assisted Blockchain for Sustainable

Simultaneous Device and Data Security in the Internet of Everything

(IoE),” IEEE Consumer Electronics Magazine, vol 9, no 2, pp 8-16,

2020

[27] V Visoottiviseth, P Sakarin, J Thongwilai and T Choobanjong,

“Signature-based and Behavior-based Attack Detection with Machine

Learning for Home IoT Devices,” in 2020 IEEE REGION 10

CONFERENCE (TENCON), 2020, pp 829-834

[28] A F Ametepe, S A R M Ahouandjinou and E C Ezin, “Secure

Encryption by Combining Asymmetric and Symmetric Cryptographic

Method for Data Collection WSN in smart Agriculture,” in 2019 IEEE

International Smart Cities Conference (ISC2), 2019, pp 93-99

[29] W Zhou et al., “Reviewing IoT Security via Logic Bugs in IoT

Platforms and Systems,” IEEE Internet of Things Journal, vol 8, no

14, pp 11621-11639, 2021

[30] J Deogirikar and A Vidhate, “Security attacks in IoT: A survey,” in

2017 International Conference on I-SMAC (IoT in Social, Mobile,

Analytics and Cloud) (I-SMAC), 2017, pp 32-37

[31] M A Al-Garadi, A Mohamed, A K Al-Ali, X Du, I Ali and M Guizani, “A Survey of Machine and Deep Learning Methods for

Internet of Things (IoT) Security,” IEEE Communications Surveys

Tutorials, vol 22, no 3, pp 1646-1685, 2020

[32] Marcos A Simplicio Jr., Marcos V.M Silva, Renan C.A Alves, Tiago K.C Shibata, “Lightweight and escrow-less authenticated key

agreement for the internet of things,” Computer Communications, vol

98, no 0140-3664, pp 43-51, 2017

[33] Shancang Li, Theo Tryfonas, Honglei Li, “The Internet of Things: a

security point of view,” Internet Research, vol 26, no 2, pp 337-359,

2016

[34] M T Arafin, A Stanley and P Sharma, “Hardware-based anti-counterfeiting techniques for safeguarding supply chain integrity,” in

2017 IEEE International Symposium on Circuits and Systems (ISCAS), 2017, pp 1-4

[35] A Burg, A Chattopadhyay and K -Y Lam, “Wireless Communication and Security Issues for Cyber–Physical Systems and

the Internet-of-Things,” Proceedings of the IEEE, vol 106, no 1, pp

38-60, 2018

[36] I Butun, P Österberg and H Song, “Security of the Internet of

Things: Vulnerabilities, Attacks, and Countermeasures,” IEEE

Communications Surveys Tutorials, vol 22, no 1, pp 616-644, 2020

[37] J Yu, E Lee, S Oh, Y Seo and Y Kim, “A Survey on Security Requirements for WSNs: Focusing on the Characteristics Related to

Security,” IEEE Access, vol 8, pp 45304-45324, 2020

[38] K Lounis and M Zulkernine, “Attacks and Defenses in Short-Range

Wireless Technologies for IoT,” IEEE Access, vol 8, pp

88892-88932, 2020

[39] M Wazid, A K Das, S Shetty, P Gope and J J P C Rodrigues,

“Security in 5G-Enabled Internet of Things Communication: Issues,

Challenges, and Future Research Roadmap,” IEEE Access, vol 9, pp

4466-4489, 2021

[40] A Raoof, A Matrawy and C Lung, “Routing Attacks and Mitigation

Methods for RPL-Based Internet of Things,” IEEE Communications

Surveys Tutorials, vol 21, no 2, pp 1582-1606, 2019

[41] A Murzaeva, B Kepçeoğlu and S Demirci, “Survey of Network

Security Issues and Solutions for the IoT,” in 2019 3rd International

Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), 2019, pp 1-6

[42] M A Alamro, Y Zhuang, A O Aseeri and M S Alkatheiri,

“Examination of Double Arbiter PUFs on Security against Machine

Learning Attacks,” in 2019 IEEE International Conference on Big

Data (Big Data), 2019, pp 3165-3171

[43] F Hussain, R Hussain, S A Hassan and E Hossain, “Machine Learning in IoT Security: Current Solutions and Future Challenges,”

IEEE Communications Surveys Tutorials, vol 22, no 3, pp

1686-1721, 2020

[44] K Tsantikidou, N Sklavos, “Hardware Security for IoT-Based, Healthcare Applications”, New England Hardware Security Day 2021 Workshop, New England, USA, April 9, 2021

[45] A N Bikos, N Sklavos, “The Future of Privacy and Trust on the Internet of Things (IoT) for Healthcare: Concepts, Challenges, and Security Threat Mitigations”, Chapter in the Book: Recent Advances

in Security, Privacy, and Trust for Internet of Things (IoT) and Cyber-Physical Systems (CPS), editors Kuan-Ching Li, Brij B Gupta, Dharma P Agrawal, CRC Press, ISBN: 9780367220655, 2020 [46] P Panagiotou, N Sklavos, I D Zaharakis, “Design and Implementation of a Privacy Framework, for the Internet of Things (IoT)”, proceedings of 21th EUROMICRO Conference on Digital System Design, Architectures, Methods, Tools (DSD'18), Prague, Czech Republic, August 29–31, 2018

[47] M Tsavos, N Sklavos, G Ph Alexiou, “Lightweight Security Data Streaming, Based on Reconfigurable Logic, for FPGA Platform”, proceedings of 23th EUROMICRO Conference on Digital System Design, Architectures, Methods, Tools (DSD'20), Portoroz, Slovenia, August 26–28, 2020