Research On Streaming Media Detection Methods Against DoS\DDoS Attack Based On Analysis Of Self-similarity

With the rapid development of Internet technology, audio and video information in theform of streaming media is becoming increasingly popular, and the user base is increasinglylarge. Streaming media technology has been widely used in multimedia news releases, onlinelive, video on demand, E-commerce, online advertising, telemedicine, remote education,Internet radio, real-time video conferencing and other aspects of Internet services withincreasingly attention, especially the importance of service quality.Streaming media technology has the characteristic of high time-sensitive which directly impactson people's experice degrees, so once it is subjected to malicious attack flows, theconsequences will be much more serious by response delays and even paralysis. Form the point,this article specifies the security vulnerabilities which may lead to DoS, DDoS attacks on thebasis of systematic study of the traditional DoS, DDoS attacks and in-depth analysis of thestreaming protocol interaction RTSP. Then we design and implement a streaming media defensesystem against DoS, DDoS attacks, and put forward effective strategies for attack detection andattack prevention based on the self-similarity of network traffic. At last, by doing a lot ofdefense test, the test results prove that this defense system has a significant effect of defenseagainst streaming media applications'DoS, DDoS attacks.