Evaluation criteria

Your presentation is expected to last for max 15 minutes, you will be stopped at 17 min. That means you need to identify the key messages of the selected paper. There is penalty for being too long. You will get questions about your paper and your presentation. One block of presentation and questions will approximately take 30 min.

Your presentation will be evaluated with respect to the following criteria:

Knowledge and Overview over topics in the paper

what is the paper about, which specific aspects are addressed?
- 1-2 slides on terms and how they are related
- 1-2 slides on "about", how do these topics relate and what comes out

Identification of key findings of the paper:

why are the findings key findings, to what extend are they different from the state-of-knowledge?
- 1 slides: does it look like that there is something new, or just "bla, bla"
- 1-2 slides: compared to our lectures, which security challenges does the paper address, and to what degree
what are the strengths/weaknesses of the approaches being mentioned
- 1 slide: does the paper mention strength (S) and weaknesses (W)? what is your analysis of S/W

Evaluation of the paper

what is your take on the paper? Criticism, own thoughts,...
- 1-slide how do you judge the paper? own thoughts, what would you have done?
- 1 slide conclusion (short summary)

Scientifically precise presentation

Be precise in your formulations, avoid story telling.
Use numbers where appropriate, e.g. 20% more instead of much more

Style:

clear structure, clear and concise slides
Provide examples, and respect the time

The presentation does not contribute to the final grade, but is a training for you to (i) analyse scientific knowledge and (ii) present scientific results.

List of papers for TEK5530

Attached is a list of papers, from whom you may select one to be presented later in the course. Alternatively, select an academic paper of your choice. We suggest to start with SemanticScholar.org, Google Scholar or Microsoft Academics. Read more on: Guide on how to search for Literature

List of Papers 2026

We used Elicit to define search phrases and get a list of 10 selected papers. Please select one paper, and add that link with your name to the Announcement on Canvas. Presentation of papers is on 5 Mar 2026.

Note: using +oa:true adds only open access papers

List of Topics and Elicit report

What are the in-network security challenges in IoT home networks?- https://elicit.com/find-papers/f797a399-64b9-42b6-b1a9-0e576b7865c3

Which parameters do you need to consider when establishing a digital twin monitoring system of the power grid? - https://elicit.com/find-papers/f3dc3165-094f-4135-afe2-9a2b91bb59cd

What are the promising methods and practices for secure IoT and OT system developments? - https://elicit.com/find-papers/fc67c26b-c6fd-4606-b177-244635ce9a83

What are specific security challenges in IoT and OT industrial systems? - https://elicit.com/find-papers/f2e29d57-afac-4754-ac24-c493194e0db3

What are the in-network security challenges in IoT home networks? - https://elicit.com/find-papers/f797a399-64b9-42b6-b1a9-0e576b7865c3

What are pathways or frameworks for finding zero-day vulnerabilities of IoT or OT systems? - https://elicit.com/find-papers/f7f148a6-a4bb-4ebc-8184-1624f3e0bf96

Security Evaluation of Smart Home Access - https://elicit.com/review/5a5d31ab-01f4-4086-8550-c2782d6fd18a

Security Challenges in Smart Grids - https://elicit.com/review/5aca7010-1053-4ad0-a8a7-789b12868814

List of Papers 2025

IoT Security Metrics

a set of measurable security indicators for assessing the security of IoT devices and systems.

Elicit results on: What are security metrices for the internet of things?

Security metrics for IoT devices are crucial for assessing and improving their security posture. Several studies have proposed various metrics and frameworks to quantify IoT security parameters. These include network-related metrics like throughput, packet loss rate, and jitter (Ebad, 2023), as well as device-specific metrics such as password strength, security transmission rate, and energy consumption (Setzler & Mountrouidou, 2021; Ebad, 2023). Researchers have developed hierarchical metric systems that integrate different security properties into a single score (Doynikova et al., 2021; Fedorchenko et al., 2022). Some approaches focus on code quality metrics, linking them to general quality characteristics from ISO/IEC 25010:2011 (Klíma et al., 2022). Others propose methods for specifying IoT security-quality metrics on a globally accepted scale (Ito et al., 2021). Security graphs and metrics have also been used to identify critical components and vulnerabilities in IoT networks (Bodei et al., 2020). These metrics aim to provide clear and objective measures for both users and manufacturers to assess and compare IoT device security (Bonilla et al., 2017).

Elicit papers (shared as .csv)

0) The paper proposes using security metrics to analyze the security level of IoT systems by modeling data flows and dependencies.

C. Bodei, P. Degano, G. Ferrari, Letterio Galletta, 2020, Security Metrics at Work on the Things in IoT Systems, 10.1007/978-3-030-41103-9_9

1) "Firmware is a crucial component of an electronic system.", "Based on the review at the end we proposed a combination of static and dynamic analysis methods to detect Authentication Bypass Firmware Vulnerability in IoT and Embedded systems."

Saimon Hemram et al. (2022), "Firmware Vulerability Detection in Embedded Systesm and the Internet of Things" https://www.semanticscholar.org/paper/Firmware-Vulnerability-Detection-in-Embedded-and-of-Hemram-Kathrine/e335c9d772157c3b493edbcd34835b7d58d62218

2) "The results indicate how the security vulnerability identification helps in prioritizing business decisions by vulnerabilities quantification.", "The research work deals with the identification and mitigation ofthesecurity vulnerabilities by an intelligent and smart software vendor, which enumerates common vulnerabilities in its database and provides the possible solution for mitigating the same."

Vinita Malik & Sukhdip Singh (2019), Security risk management in IoT environment, https://doi.org/10.1080/09720529.2019.1642628

3) "We propose augmenting secure boot with a mechanism to protect against compromises to field-upgradeable devices.", "In particular, secure boot standards should verify the firmware of all devices in the computer, not just devices that are accessible by the host CPU."

James Hendricks & L. V. Doorn (2004), Secure bootstrap is not enough: shoring up the trusted computing base, https://doi.org/10.1145/1133572.1133600

4) "Not exclusively is the proposed framework model taking out conceivable security weaknesses, it additionally can be utilized alongside the best security strategies to countermeasure the cyber security dangers confronting every single unit of IOT.", "The proposed protocol actualized security certificates to permit information move between nodes of the proposed IOT model."

Sanaa. S. Abd El dayem (2021), SECURITY For INTERNET OF THINGS-IOT, https://doi.org/10.1109/iemcon53756.2021.9623184 - applicability?

5) "In this work, we aimed to investigate security of IoT devices through performing automatic penetration test on IoT devices.", "We have (1) investigated how such models can be applied for performing autonomous IoT penetration testing."

Fartein Lemjan Færøy et al. (2023), Automatic Verification and Execution of Cyber Attack on IoT Devices, https://doi.org/10.3390/s23020733

6) "Penetration testing is generally used to identify the vulnerabilities/ possible attacks on traditional systems periodically.", "A timely fix of these vulnerabilities can avoid future attacks."

Geeta Yadav et al. (2020), IoT-PEN: A Penetration Testing Framework for IoT, https://doi.org/10.1109/ICOIN48656.2020.9016445

Machine learning algorithm performance in IoT security

evaluation of machine learning algorithms, including Random Forest, for securing IoT.

Machine learning (ML) techniques are increasingly being applied to enhance security in Internet of Things (IoT) systems. ML algorithms can detect anomalies, intrusions, and other security threats in IoT networks by analyzing network traffic patterns and device behaviors (Bagaa et al., 2020; Shahid et al., 2019). Various ML approaches, including supervised, unsupervised, and deep learning methods, have been explored for IoT security (Kikissagbe & Adda, 2024). These techniques can improve intrusion detection systems, authentication mechanisms, and overall security management in IoT environments (Androcec & Vrček, 2018; Hussain et al., 2023). ML-based approaches have shown promising results in detecting attacks with high accuracy and low cost (Bagaa et al., 2020). However, challenges remain, such as addressing the heterogeneity of IoT devices and ensuring the scalability of ML solutions (Canedo & Skjellum, 2016). Despite these challenges, ML offers a proactive and adaptive approach to securing IoT systems, capable of responding to evolving cyber threats in real-time (Kumar & Singh, 2024; Almajed, 2023).

7) Further the utilization of machine learning algorithms like Decision Tree, Random Forest, and ANN tests have resulted in an accuracy of 99.4%.", "Accuracy, precision, recall, f1 score, and area under the Receiver Operating Characteristic Curve are the assessment metrics utilized to compare the performance of the existing techniques.

Joel et al. (2023) https://www.semanticscholar.org/paper/An-Analysis-of-Security-Challenges-in-Internet-of-Joel-Manikandan/50041a6d870091c6901a7ff306235d16f00f2947

IoT Security Measures, a collection of security measures and mechanisms for ensuring the security of the Internet of Things ecosystem

low quality paper - 8) "Protecting data during network transit is also very important concern in the measures.", "This study defines security requirements and challenges that are common in IoT and their solutions, ranging from simple things like authentication to trust management."

S. Lee et al. (2018), Security and Privacy Measures on Data Mining for Internet of Things, https://www.semanticscholar.org/paper/Security-and-Privacy-Measures-on-Data-Mining-for-of-Lee-Vallent/32a7c88fbfda4e6daaf1f1d4200f8eef10ba6069

not recommended paper, very specific task - 9) "Therefore, secured IoT applies equally to: (i) IoT device-to-device communication, (ii) IoT sensing/actuating, and (iii) IoT information exchange.", "In this context, ensuring security during ‘thing’ operation and information exchange becomes a critical task."

IOAN TUDOSA et al. (2019), Hardware Security in IoT era: the Role of Measurements and Instrumentation, https://doi.org/10.1109/METROI4.2019.8792895

10) "In addition to these, we extensively analyze the intrusion detection and mitigation mechanisms proposed for IoT and evaluate them from various points of view.", "We review the characteristics of IoT environments, cryptography-based security mechanisms and D/DoS attacks targeting IoT networks."

Ahmet Aris et al. (2018), Security of Internet of Things for a Reliable Internet of Services, https://doi.org/10.1007/978-3-319-90415-3_13

Area: Domain specific security assessment or metrices for IoT systems

Summary of top 8 papers Recent research has focused on developing security metrics and assessment frameworks for IoT systems. Several studies propose automated approaches for security evaluation, including machine learning and natural language processing to analyze vulnerabilities (Duan et al., 2021; Payne et al., 2019). IoT-specific metrics have been introduced to quantify security parameters, such as throughput, packet loss rate, and energy consumption (Ebad, 2023; Bonilla et al., 2017). Researchers have also developed frameworks for classifying and comparing these metrics (Ebad, 2023; Doynikova et al., 2021). Some studies emphasize the importance of considering both device-level and network-level security aspects (Bodei et al., 2020; Setzler & Mountrouidou, 2021). Additionally, researchers have explored the use of attack graphs and circuits to model potential threats and assess network vulnerabilities (Payne et al., 2019; Evangelou, 2020). These advancements aim to provide more comprehensive and accurate security assessments for IoT ecosystems, addressing the unique challenges posed by the rapidly growing number of connected devices.