AcLGB: A Lightweight DDoS Attack Detection Method

With the development of Internet technology, distributed denial of service(DDoS) attack has always been a hot and difficult point in network security.Protecting network infrastructure and information security is also becoming more and more important.However, cyber security is an arms race, as attacks develop and network traffic surges, intelligent solutions face the challenge of detecting sensitive changes in traffic characteristics.In this paper, we propose a lightweight Adaptive Clustering-based LightGBM(AcLGB) detection method.This is a new DDoS traffic classification method and an effective lightweight detection method.We introduce a new clustering technique to learn the clustering centers that can be used to extend the characteristics of a given dataset.It solves the challenge of difficult detection when traffic characteristics change sensitively.The model separates the samples of different categories in the best way, and outperforms the current detection method with 99.98% detection accuracy. In the CIC-DDoS2019 data set, the detection time of 802s is better than other detection methods.