DocumentCode
251619
Title
Discovering new indicators for botnet traffic detection
Author
Adamov, Alexander ; Hahanov, Vladimir ; Carlsson, Anders
Author_Institution
Kharkov Nat. Univ. of Radioelectron., Kharkov, Ukraine
fYear
2014
fDate
26-29 Sept. 2014
Firstpage
1
Lastpage
5
Abstract
Botnets became the powerful cyber weapon that involves tens of millions of infected computers - “cyber zombies” - all over the world. The security industry makes efforts to prevent spreading botnets and compromising an Individual Cyberspace (IC)[1] of users in such way. However, botnets continue existing despite numerous takedowns initiated by antivirus companies, Microsoft, FBI, Europol and others. In this paper we investigate existed methods of traffic detection represented mostly by IDS system and discover new indicators that can be utilized for improving botnet traffic detection. To do this we analyse the most prevalent backdoors communication protocols that stay behind of the popular botnets. As a result, we extracted new data that might be used in detection routines of IDS (Intrusion Detection System). An objective of the study is mining new indicators of compromise from botnet traffic and using them to identify cyber-attacks on IC. The analysis method assumes analysis of a communication protocol of the top botnet backdoors. The discovered results that can be used to improve detection of infected hosts in a local network are presented in this paper.
Keywords
invasive software; IDS system; backdoor communication protocol; botnet traffic detection; cyber weapon; cyber zombies; cyber-attack; individual cyberspace; infected computer; intrusion detection system; security industry; Encryption; IP networks; Protocols; Servers; Trojan horses; IDS; Indicator-of-Compromise; Individual Cyberspace; botnet; detection; encryption; signature; traffic;
fLanguage
English
Publisher
ieee
Conference_Titel
Design & Test Symposium (EWDTS), 2014 East-West
Conference_Location
Kiev
Type
conf
DOI
10.1109/EWDTS.2014.7027100
Filename
7027100
Link To Document