Detection of DDoS using Artificial Neural Network

Cyber attacks are one of the most deadliest and lethal attacks in today’s world. One of these attacks being the Distributed Denial of Service attack. A DDoS attack is a malicious way to disturb the ongoing traffic of the server. It overflows the victim or its assets by flooding the internet traffic. DDoS attacks attain effectiveness by using compromised computers or IoT devices as sources of internet traffic. In a DDoS attack, a malicious actor attempts to cause troubles in the normal functioning of a network or server by buffer overflowing it with a flood of internet requests. In simple words Distributed Denial of Service attacks can be defined as a traffic stop in which cars are stopped on the highway due to unknown circumstances and are prevented from reaching their destination. In this attack the incoming logs of the network traffic suffer from many incoming network logs which takes the server down and causes the loss of time and resources to the organization. The intruders use the bots to buffer the traffic to the victim or organization, overwhelming its infrastructure and causing it to crash or become inaccessible. In this paper we have presented a way to detect the Denial of Service attacks using neural networks. We compared our model with the other models in the field and got a higher accuracy rate and consistency of the model. It has been trained on over 5000 sites which are malicious and brought out an accuracy of about 98.