Flexible Approach to Perceive DDoS Attack From Flash Crowds in Dynamic Miscellaneous Traffic
The Denial of Service and Distributed Denial of Service attacks is highly threatening the World Wide Web. Retribution of these attacks and to preserve the authors' server a new novel technique to detect the application layer-based DDoS attacks is proposed in this paper. This Paper presents an effective technique named Zombi detection technique. In this technique Pearson's Correlation Coefficient is proposed to detect the repeated features of Packet's Arrival. The Time Stamp and Changing Rate of Port Numbers are considered for the feature extraction from the successive requests from the clients [Legitimate/Illegitimate]. The generated patterns is directed to the proxy server. Consequently, the decision is left to the proxy server whether the requested client can be denied or provided the service.