YAPAY SİNİR AĞLARI İLE KLASİK KRİPTOGRAFİ ALGORİTMALARININ ŞİFRELİ VERİLER ÜZERİNDEN SINIFLANDIRILMASI

Year 2020, Volume: 25 Issue: 2, 651 - 664, 31.08.2020

Sevtap Türk , Rüya Şamlı

https://doi.org/10.17482/uumfd.695128

Cited By: 1

Abstract

Şifreleme bilimi olarak ifade edilebilecek olan kriptoloji için kullanılan algoritmaların temel amacı bir noktadan bir noktaya iletilen, ya da herhangi bir ortamda saklanan verilere izinsiz kişilerin erişmesini engellemek ve bu veriler ele geçirilse dahi verilerin anlaşılmasını imkânsız hale getirmektir. Günümüzde, birçok farklı tipteki şifreleme algoritmalarının temeli klasik simetrik şifreleme yöntemlerine dayanmaktadır. Gelişen teknolojiyle ortaya çıkan veri güvenliği sorununu çözmek için daha karmaşık matematiksel altyapıya sahip yöntemler denense de donanımsal gerçekleme zorlukları araştırmacıları farklı arayışlara yöneltmiştir. Bunlardan biri de YSA (Yapay Sinir Ağları – Artificial Neural Networks)’dır. Kriptoloji ve YSA’nın birleşimi ile oluşan ve “Nöral Kriptografi” olarak adlandırılan çalışma alanında hem şifreleme hem de kriptanaliz aşamalarında YSA modellerinden faydalanılmaktadır. Bu çalışmada, bir Nöral Kriptografi uygulaması ile klasik simetrik şifreleme yöntemlerinden birkaçıyla şifrelenen verilerin, YSA yöntemi ile klasik şifreleme algoritmalarından hangisine ait olduğu tahmin edilmeye çalışılmıştır.

Keywords

Klasik Şifreleme, Simetrik Şifreleme Algoritmaları, Yapay Sinir Ağları

Supporting Institution

TÜBİTAK

Project Number

118E682

Thanks

Bu çalışma Doç. Dr. Rüya ŞAMLI danışmanlığında gerçekleştirilen ve 2017 yılında tamamlanan yüksek lisans tez çalışmasının (Türk, 2017) bir bölümünü oluşturmaktadır.

Classification of Classical Cryptography Algorithms Through Encrypted Data With Using Artificial Neural Networks

Abstract

The main aim of cryptography algorithms is to prevent unauthorized people from attaining data that transmitted from one node to another or stored in any environment, even if it is captured, making it impossible to decrypt. Today, basis of many different types of encryption methods is based on classical encryption algorithms. Although many methods which have more complex mathematical infrastructure are tried to solve the data security problem become important by the advancement of technology. The hardware implementation difficulties of these complex methods have led the researchers to the different areas. One of these areas is ANN (Artificial Neural Networks). In the study area called "Neural Cryptography" which is formed by the combination of cryptology and ANN, ANN models are used in both encryption and cryptanalysis phase. In this study, we prepared a Neural Cryptography application and have tried to determine which data is encrypted by which classical method with using ANN. 

Keywords

Classical Encryption, Symmetric Cipher Algorithms, Artificial Neural Networks

Project Number

118E682

References

• 1. Abd, A.J., Al-Janabi, S.T.F. (2019) Classification and Identification of Classical Cipher Type Using Artificial Neural Networks, Journal of Engineering and Applied Sciences, 14(11), 3549-3556. doi: 10.36478/jeasci.2019.3549.3556
• 2. Chandra, B., Varghese, P.P., Saxena, P.K., Kant, S. (2007) Neural Networks for Identification of Crypto Systems, The 3rd Indian International Conference on Artificial Intelligence (IICAI-07), 402-411.
• 3. Dileep, A.D., Sekhar, C.C. (2006) Identification of Block Ciphers Using Support Vector Machines, The IEEE International Joint Conference on Neural Network Proceedings, 2696-2701. doi: 10.1109/IJCNN.2006.247172
• 4. Dunham, J.G., Sun, M.T., Tseng, J.C.R. (2005) Classifying File Type of Stream Ciphers in Depth Using Neural Networks, The 3rd ACS/IEEE International Conference onComputer Systems and Applications. doi: 10.1109/AICCSA.2005.1387088
• 5. Fausett, L.V. (1994) Fundamentals of Neural Networks: Architectures, Algorithms and Applications, Prentice Hall, United States.
• 6. Gallant, S.I. (1993) Neural Network Learning and Expert Systems, MIT Press, London, England.
• 7. Hagan, M.T., Demuth, H.B., Beale, M.H., De Jesus, O. (2014) Neural Network Design, Martin Hagan.
• 8. Haykin, S. (1998) Neural Networks: A Comprehensive Foundation, Prentice Hall, Delhi, India.
• 9. Kara, O. (2009) Kriptografinin Yapıtaşları Kriptografik Algoritmalar ve Protokoller, Bilim ve Teknik, 500, 34-41.
• 10. Khadivi, P., Momtazpour, M. (2010) Cipher-Text Classification With Data Mining, The 4th IEEE International Symposium on Advanced Networks and Telecommunication Systems, 64-66. doi: 10.1109/ANTS.2010.5983530
• 11. Öztemel, E. (2006) Yapay Sinir Ağları, Papatya Yayıncılık, İstanbul, Türkiye.
• 12. Piper, F. (1997) Introduction to Cryptology, Information Security Technical Report, 10-13.
• 13. Sharif, S.O., Kuncheva, L.I., Mansoor, S.P. (2010) Classifying Encryption Algorithms Using Pattern Recognition Techniques, The IEEE International Conference on Information Theory and Information Security, 1168-1172. doi: 10.1109/ICITIS.2010.5689769
• 14. Shihab, K. (2006) A Cryptographic Scheme Based on Neural Networks, The 10th WSEAS International Conference on Communications, 7-12.
• 15. Soto, J., Melin, P., Castillo, O. (2018) Ensembles of Type 2 Fuzzy Neural Models and Their Optimization with Bio-Inspired Algorithms for Time Series Prediction, Springer.
• 16. Stinson, D.R. (2002) Cryptography: Theory and Practice, Chapman & Hall/CRC Press.
• 17. Tan, C., Ji, Q. (2016) An Approach to Identifying Cryptographic Algorithm from Ciphertext, The 8th IEEE International Conference on Communication Software and Networks (ICCSN), 19-23. doi: 10.1109/ICCSN.2016.7586649
• 18. Trappe, W., Washington, L.C. (2006) Introduction to Cryptography with Coding Theory, Pearson Prentice Hall.
• 19. Türk, S. (2017). Yapay Sinir Ağları Kullanılarak Şifreleme Yöntemlerinin Performans Analizlerinin Gerçekleştirilmesi, Yüksek Lisans Tezi, İstanbul Üniversitesi Fen Bilimleri Enstitüsü, İstanbul.
• 20. Türk, S., Samli, R., Orman, Z. (2019) A Sample Substitution Cipher Data Processing Using Neural Networks, The 4th International Conference on Theoretical and Applied Computer Science and Engineering (ICTACSE), 18-22.
• 21. Volna, E., Kotyrba, M., Kocian, V., Janosek, M. (2012) Cryptography Based On Neural Network, The 26th European Conference on Modelling and Simulation (ECMS), 386-391.
• 22. Yee, L.P., Silva, L.C.D. (2002) Application of Multilayer Perceptron Networks in Public Key Cryptography, The International Joint Conference on Neural Networks (IJCNN), 1439-1443. doi: 10.1109/IJCNN.2002.1007728
• 23. Yu, W., Cao, J. (2006) Cryptography Based on Delayed Chaotic Neural Networks, Physics Letters A, 356 (4-5), 333-338. doi: 10.1016/j.physleta.2006.03.069