Real Time DDos Attack Detection System Using Apache Flink and Gradient Boosting Algorithm

Abstract

When large amounts of traffic from hundreds, thousands, or even millions of other computers are routed to a network or server to crash the system and disrupt its function to make the system crash.A DDoS attack aims to overwhelm the devices, services, and network of its intended target with fake internet traffic, rendering them inaccessible to or useless for legitimate users.DDoS attacks are designed to look like a flood of calls, or requests, made by browsers asking a web page to load.Ddos attacks are very crucial as they make the website or webapp down which leads to losses to the company.Early DDoS detection is critical for businesses because it can help protect the functioning and security of a network.Networks without a robust DDoS defense strategy may have trouble defending against the wide range of DDoS attacks, which can be difficult to trace.The problems need to be addressed with models that can manage the time information contained in network traffic flows.We can detect the attack while the initial requests are being made to the server and block the requests made by such IP addresses or if they are false dods attack warnings we can scale our app to handle the traffic.The system to be proposed should be as real time as it can be as it will help in preventing system crashes. In this System we will be using apache flink to give real time detection and for better classification of attack the gradient boosting is proposed. Keywords: Distributed Denial of Service attack detection; machine learning; network security .