+7 (495) 987 43 74 ext. 3304
Join us -              
Рус   |   Eng

Authors

Bayanov B.

Degree
Student, Kazan national technical research university named after A. N. Tupolev — KAI
E-mail
bayanov_bulat@mail.ru
Location
Kazan
Articles

Classification of DDoS-attacks based on the neural network model

This article describes a trend of DDoS-attacks development, gives classification of such attacks and specifies some of the most popular types of DDoS-attacks. Also, this work looks statistical methodologies of analysis of malicious traffic, which can be used in modern security provision systems. Methodology, which based on multiple simulations with neurolink models, was chosen as a main in this article. The major problem of this study is to develop an architecture of such models and choose the most perspective with an assessment of the adequacy of this model. The purpose of this article is to gaining results of use neurolink model in case of classification of DDoS-attacks. Thereby, based on the raw data that were obtained through multiple simulations of different types of traffic, we found the most perspective architecture of neurolink model that has both: the least complex learning sample and at the same time provides high quality of classifying traffic. On the basis of results, which were gained as a result of work with developed mathematic model and assessment of its adequacy, we can approve that considered methodology carrying out its purpose, classification of DDoS-attacks, completely, and could be used in security algorithms or that kind of software.
Read more...