DETECTION OF UNAUTHORISED ACTIONS AND ATTACKS IN NETWORKS OF THE METHOD OF WAVELET ANALYSIS

Abstract

Behavioural methods based on the models of "normal" functioning of an information network. The principle of work is in identifying differences between the current state of the information system functioning and the one that is thought to be examplary for this network. Any discrepancy is seen as an intrusion or anomaly. Perspective method for detecting unauthorised invasions must have high accuracy of detection known and unknown types of cyberattacks, perfectly own accuracy of decision-making, that is low number of false positives. The detecting cyberattacks method based on wavelet analysis (WA) proved itself well in struggling against non stationary signals. The wavelet transform (WT) has a self-adjusting moving frequency-time window, but equally well reveals both low-frequency (LF) and high-frequency (HF) characteristics of the signal on different time scales. Wavelet filters allow not only to fight against noise, but also to extract the necessary components of the signal. Today’s society cannot do without information technology. IT plays a major role in and is an integral part of every sphere of our life. Current development of the information society is inextricably linked with collecting, processing and passing huge amounts of data, and converting data into a product that has significant value. This is the main reason behind the global shift from the industrial society to the information society. The invention of the Internet has led to a significant increase in international communication in various spheres of human life. On the other hand, along with a large number of benefits, there has been a significant number of threats related to the current technology. There are currently multiple threats now on both national and international level. Therefore, different mechanisms for protecting cyberspace are now part of many countries’ national strategy Network anomalies are so varied that they cannot be categorized using one simple classification. The number of attacks, their power and complexity are increasing. Malicious users are looking for radically new ways of network intrusion and the existing methods of cyber protection often turn out to be inefficient. [1].