作者: Wei Wei , Yabo Dong , Dongming Lu , Guang Jin
DOI: 10.1007/11758549_8
关键词: Fuzzy logic 、 Correlation function (quantum field theory) 、 Computer network 、 Fuzzy classification 、 Network traffic control 、 Denial-of-service attack 、 Network packet 、 Traffic generation model 、 Computer science 、 Traffic shaping
摘要: In legitimate traffic the correlation exists between outgoing and incoming of a server network because request-reply actions in most protocols. When DDoS attacks occur, attackers send packets with faked source addresses. As result, to addresses does not induce any related traffic. Our main idea is find changes caused by DDoS. We sample traffics using Extended First Connection Density (EFCD), express cross-correlation function. Because DDoS-initiating stage much similar traffic, we use fuzzy classification order guarantee accuracy. Experiments show that can be identified accurately our algorithm.