Mehmet Demirci
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mehmet Demirci has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Altuğ Çil
Informatics Institute Gazi University Ankara Türkiye
2 shared publications - Banu Gunel
Middle East Technical University
2 shared publications - Burak Gulbay
Department of Information Security Engineering, Graduate School of Natural and Applied Sciences, Gazi University, Ankara 06500, Turkey, Turkey (Türkiye)
2 shared publications - Mevlut Serkan Tok
GAZI UNIVERSITY, INSTITUTE OF SCIENCE, DEPARTMENT OF INFORMATION SECURITY (INTERDISCIPLINARY), INFORMATION SECURITY (INTERDISCIPLINARY) (DR)
2 shared publications - Sedef Demirci
Gazi Üniversitesi
2 shared publications - Alper Sarıkaya
Informatics Institute Middle East Technical University Ankara Turkey
1 shared publication
Latest Publications
- 18 June 2026
- Published by Taylor & Francis in Leisure/Loisir
- p. 1-32
Digitalizing campus recreation: action research on developing a mobile application for leisure engagement
- 18 June 2026
- Published by Taylor & Francis in Leisure/Loisir
- p. 1-32
Digitalizing campus recreation: action research on developing a mobile application for leisure engagement
- 12 February 2026
- Published by Elsevier in Ad Hoc Networks
- Vol. 185
DeepSpect: An RF spectrogram-based deep learning approach for near-real-time attack detection in FANETs
- 12 February 2026
- Published by Elsevier in Ad Hoc Networks
- Vol. 185
DeepSpect: An RF spectrogram-based deep learning approach for near-real-time attack detection in FANETs
- 14 November 2025
- Published by Wiley in International Journal of Communication Systems
- Vol. 39 (1)
SARE: Secure, Automated, and Regulatory‐Compliant Platform for Improving Forensic Readiness in Software‐Defined Networks
The significant increase in cybercrimes increases the importance of digital forensics processes. Designing digital forensics regulations before cyber incidents occur is important in terms of compliance with the legislation, speed, and quality. The lack of systems meeting forensic requirements leads to various administrative and judicial problems. In this article, a novel platform named SARE has been developed to automate network forensic processes in a software‐defined network (SDN) environment in a manner fully compatible with regulatory requirements. SARE was developed to meet the requirements arising from both Turkish legislation and certain international regulations. It automatically logs the required Internet access records specified in the legislation with the help of sFlow. It stores the log files in an electronically signed, backed‐up, secure, accessible, and integrity‐preserved manner and wipes them at the end of the specified period. Additionally, SARE can prevent attacks on the platform and prevent access to malicious targets. SARE was examined in experiments using real network traffic, and it successfully fulfilled all the requirements without the need for an extensive technical infrastructure. All necessary data can be stored with a file that is only 0.87% of the traffic size generated according to user habits. With 5.5% of the total number of packages, the regulation‐compliant logging process has been completed. In addition, it has been shown that the same data in different file types create a size difference of more than three times. The use of SARE will enable compliance with regulations automatically without any intervention and improve the forensic readiness of organizations.
- 14 November 2025
- Published by Wiley in International Journal of Communication Systems
- Vol. 39 (1)
SARE: Secure, Automated, and Regulatory‐Compliant Platform for Improving Forensic Readiness in Software‐Defined Networks
The significant increase in cybercrimes increases the importance of digital forensics processes. Designing digital forensics regulations before cyber incidents occur is important in terms of compliance with the legislation, speed, and quality. The lack of systems meeting forensic requirements leads to various administrative and judicial problems. In this article, a novel platform named SARE has been developed to automate network forensic processes in a software‐defined network (SDN) environment in a manner fully compatible with regulatory requirements. SARE was developed to meet the requirements arising from both Turkish legislation and certain international regulations. It automatically logs the required Internet access records specified in the legislation with the help of sFlow. It stores the log files in an electronically signed, backed‐up, secure, accessible, and integrity‐preserved manner and wipes them at the end of the specified period. Additionally, SARE can prevent attacks on the platform and prevent access to malicious targets. SARE was examined in experiments using real network traffic, and it successfully fulfilled all the requirements without the need for an extensive technical infrastructure. All necessary data can be stored with a file that is only 0.87% of the traffic size generated according to user habits. With 5.5% of the total number of packages, the regulation‐compliant logging process has been completed. In addition, it has been shown that the same data in different file types create a size difference of more than three times. The use of SARE will enable compliance with regulations automatically without any intervention and improve the forensic readiness of organizations.
- Open Access
- 17 July 2024
- Published by MDPI AG in Preprints.org
A Framework for Developing Strategic Cyber Threat Intelligence from Advanced Persistent Threat Analysis Reports Using Graph-Based Algorithms
Advanced persistent threat (APT) attacks are sophisticated and organized attacks commonly motivated by political, financial, and strategic objectives. In order to comprehend their tactics, techniques, and procedures (TTP) and indicators, APT reports are valuable sources. While blue teams typically rely on server logs, firewall rules and user authorizations managed in database tables, attackers have a graph-based mindset. In this work, we propose a framework for discovering and evaluating APTs using graph-based algorithms. Cyber threat intelligence (CTI) was extracted from 40,358 pages of APT reports and transformed into a graph. Centrality, community, and similarity analyses were executed on the graph. As a result, critical and influential APT groups and indicators of compromise (IoC) were discovered. Similar attacks and APT groups were revealed. Analysis results were interpreted to create new strategic CTI that can be utilized in future security operations.
- 17 July 2024
- Published by MDPI AG in Preprints.org
Open AccessA Framework for Developing Strategic Cyber Threat Intelligence from Advanced Persistent Threat Analysis Reports Using Graph-Based Algorithms
Advanced persistent threat (APT) attacks are sophisticated and organized attacks commonly motivated by political, financial, and strategic objectives. In order to comprehend their tactics, techniques, and procedures (TTP) and indicators, APT reports are valuable sources. While blue teams typically rely on server logs, firewall rules and user authorizations managed in database tables, attackers have a graph-based mindset. In this work, we propose a framework for discovering and evaluating APTs using graph-based algorithms. Cyber threat intelligence (CTI) was extracted from 40,358 pages of APT reports and transformed into a graph. Centrality, community, and similarity analyses were executed on the graph. As a result, critical and influential APT groups and indicators of compromise (IoC) were discovered. Similar attacks and APT groups were revealed. Analysis results were interpreted to create new strategic CTI that can be utilized in future security operations.
- Open Access
- 27 March 2024
- Published by Politeknik Dergisi in Journal of Polytechnic
- Vol. 27 (2) , 665-679
Ağ Adli Bilişimi Süreç Gereksinimlerinin Belirlenmesi ve Yazılım Tanımlı Ağlarda İncelenmesi
Türkiye’de internet üzerinden işlenen suçların aydınlatılması amacıyla adli, idari ve teknik süreçleri tanımlayan mevzuat ve uygulamalar mevcuttur. Ancak bunların ağ adli bilişimi odağında incelenmediği, sağlıklı ve mevzuata uygun ağ adli bilişimi süreçleri için gereksinimlerin belirlenmediği ve yazılım tanımlı ağlar (YTA) üzerinde incelenmediği görülmüştür. Bu çalışmada YTA adli bilişimi ve bilirkişiliği için gerekli delil kaynaklarının tespiti, Türkiye’de kullanılan mevzuat ve uygulamaların ağ adli bilişimi odağında temel gereksinimlerinin belirlenmesi, gereksinimlerin YTA ortamına uygulanabilirliğinin gösterilmesi ve bir YTA adli bilişim süreci oluşturulması amaçlanmıştır. Uluslararası çapta yapılan diğer teknik ve idari ağ adli bilişim çalışmalarının da incelenmesinin ardından YTA deneyleri adli bilişim süreçleri gözetilerek gerçekleştirilmiştir. Deney sonuçları YTA’nın belirlenen gereksinimleri sağladığını göstermiştir. Geliştirilebilir alanlar, uluslararası çapta yapılan diğer çalışmalar da ele alınarak tartışılmış olup, mevzuat ve uygulamaların YTA’ya tam uyum sağlaması için önerilerde bulunulmuştur. Türkiye’de adli bilişim alanında alınabilecek önlemlerin de değerlendirilmesinin ardından YTA adli bilişim yönetimi niteliğinde olan bir süreç önerisinde bulunulmuştur. YTA adli bilişiminin sadece güney arayüzü verileriyle yapılabilmesinin mümkün olduğu ancak bu konuda idari düzenlemelerin gerektiği, mevzuat kapsamında belirlenen verilerin en önemli veriler olmasına rağmen en az eforla elde edilebileceği ve sınır ötesi adli bilişim çalışmalarının gerekli olduğu sonuçları elde edilmiştir.
- 27 March 2024
- Published by Politeknik Dergisi in Journal of Polytechnic
- Vol. 27 (2) , 665-679
Open AccessAğ Adli Bilişimi Süreç Gereksinimlerinin Belirlenmesi ve Yazılım Tanımlı Ağlarda İncelenmesi
Türkiye’de internet üzerinden işlenen suçların aydınlatılması amacıyla adli, idari ve teknik süreçleri tanımlayan mevzuat ve uygulamalar mevcuttur. Ancak bunların ağ adli bilişimi odağında incelenmediği, sağlıklı ve mevzuata uygun ağ adli bilişimi süreçleri için gereksinimlerin belirlenmediği ve yazılım tanımlı ağlar (YTA) üzerinde incelenmediği görülmüştür. Bu çalışmada YTA adli bilişimi ve bilirkişiliği için gerekli delil kaynaklarının tespiti, Türkiye’de kullanılan mevzuat ve uygulamaların ağ adli bilişimi odağında temel gereksinimlerinin belirlenmesi, gereksinimlerin YTA ortamına uygulanabilirliğinin gösterilmesi ve bir YTA adli bilişim süreci oluşturulması amaçlanmıştır. Uluslararası çapta yapılan diğer teknik ve idari ağ adli bilişim çalışmalarının da incelenmesinin ardından YTA deneyleri adli bilişim süreçleri gözetilerek gerçekleştirilmiştir. Deney sonuçları YTA’nın belirlenen gereksinimleri sağladığını göstermiştir. Geliştirilebilir alanlar, uluslararası çapta yapılan diğer çalışmalar da ele alınarak tartışılmış olup, mevzuat ve uygulamaların YTA’ya tam uyum sağlaması için önerilerde bulunulmuştur. Türkiye’de adli bilişim alanında alınabilecek önlemlerin de değerlendirilmesinin ardından YTA adli bilişim yönetimi niteliğinde olan bir süreç önerisinde bulunulmuştur. YTA adli bilişiminin sadece güney arayüzü verileriyle yapılabilmesinin mümkün olduğu ancak bu konuda idari düzenlemelerin gerektiği, mevzuat kapsamında belirlenen verilerin en önemli veriler olmasına rağmen en az eforla elde edilebileceği ve sınır ötesi adli bilişim çalışmalarının gerekli olduğu sonuçları elde edilmiştir.
- 29 December 2023
- Published by IGI Global Scientific Publishing
- p. 73-101
Adaptive Learning in IoT-Based Smart City Applications
Internet of things (IoT) based smart city applications rely on constant data collection and accurate data analytics, yet the fast-changing nature of such data often causes the performance of machine learning models to deteriorate over time. Adaptive learning has been increasingly utilized in these applications in recent years as a viable solution to this problem. Moreover, IoT applications are vulnerable to various security threats due to their large-scale deployment, resource-constrained devices, and diverse protocols. This has led to an increased interest in efficient security and intrusion detection mechanisms tailored for IoT environments. In this chapter, the authors first focus on methods to address the issue of concept drift in time series streaming data for IoT-based smart city applications, such as weather, flood, and energy consumption forecasting, through adaptive learning. Furthermore, the authors examine adaptive learning-based security solutions to various attacks in different domains of the dynamic smart city landscape. Request access from your librarian to read this chapter's full text.
- 29 December 2023
- Published by IGI Global Scientific Publishing
- p. 73-101
Adaptive Learning in IoT-Based Smart City Applications
Internet of things (IoT) based smart city applications rely on constant data collection and accurate data analytics, yet the fast-changing nature of such data often causes the performance of machine learning models to deteriorate over time. Adaptive learning has been increasingly utilized in these applications in recent years as a viable solution to this problem. Moreover, IoT applications are vulnerable to various security threats due to their large-scale deployment, resource-constrained devices, and diverse protocols. This has led to an increased interest in efficient security and intrusion detection mechanisms tailored for IoT environments. In this chapter, the authors first focus on methods to address the issue of concept drift in time series streaming data for IoT-based smart city applications, such as weather, flood, and energy consumption forecasting, through adaptive learning. Furthermore, the authors examine adaptive learning-based security solutions to various attacks in different domains of the dynamic smart city landscape. Request access from your librarian to read this chapter's full text.
- 16 September 2023
- Published by Elsevier in Computers & Security
- Vol. 135
RAIDS: Robust autoencoder-based intrusion detection system model against adversarial attacks
- 16 September 2023
- Published by Elsevier in Computers & Security
- Vol. 135
RAIDS: Robust autoencoder-based intrusion detection system model against adversarial attacks
- 6 May 2023
- Published by Elsevier in Computer Networks
- Vol. 231
ConPoolUBF: Connection pooling and updatable Bloom filter based SYN flood defense in programmable data planes
- 6 May 2023
- Published by Elsevier in Computer Networks
- Vol. 231
ConPoolUBF: Connection pooling and updatable Bloom filter based SYN flood defense in programmable data planes
- 19 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 49-54
Processing Honeypot Logs with Big Data and Data Visualization via Hadoop- Power BI Integration
The increase in the amount of assets and services in cyberspace encourages attackers to create new threats based on more advanced techniques. To protect cyber assets from cyber threats, honeypots are also utilized to detect attack methods and tools within the scope of cyber threat intelligence efforts besides reactive measures such as firewalls and intrusion prevention systems. In this study, we obtained 2,161,988 lines of raw data of IP addresses, connection requests, access credentials, executed commands, connected URL addresses, and downloaded malicious files through Cowrie honeypot set up in the cloud, which was specifically configured to listen SSH and TELNET connection requests. The raw data including honeypot logs was processed with auxiliary tools, the data volume was reduced by 82%, and the raw data was transformed into a structured dataset. Then the dataset was imported to the Hadoop environment and analyzed. Possible botnet campaigns were determined by performing basic and comparative analysis, and findings were visualized as charts and heatmaps through the Hadoop- Power BI integration. The results showed that four different possible botnet campaigns are active on Internet and one of these botnets has cryptocurrency mining features.
- 19 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 49-54
Processing Honeypot Logs with Big Data and Data Visualization via Hadoop- Power BI Integration
The increase in the amount of assets and services in cyberspace encourages attackers to create new threats based on more advanced techniques. To protect cyber assets from cyber threats, honeypots are also utilized to detect attack methods and tools within the scope of cyber threat intelligence efforts besides reactive measures such as firewalls and intrusion prevention systems. In this study, we obtained 2,161,988 lines of raw data of IP addresses, connection requests, access credentials, executed commands, connected URL addresses, and downloaded malicious files through Cowrie honeypot set up in the cloud, which was specifically configured to listen SSH and TELNET connection requests. The raw data including honeypot logs was processed with auxiliary tools, the data volume was reduced by 82%, and the raw data was transformed into a structured dataset. Then the dataset was imported to the Hadoop environment and analyzed. Possible botnet campaigns were determined by performing basic and comparative analysis, and findings were visualized as charts and heatmaps through the Hadoop- Power BI integration. The results showed that four different possible botnet campaigns are active on Internet and one of these botnets has cryptocurrency mining features.