作者: Nobuaki Furutani , Jun Kitazono , Seiichi Ozawa , Tao Ban , Junji Nakazato
DOI: 10.1007/978-3-319-26561-2_45
关键词:
摘要: This paper presents an adaptive large-scale monitoring system to detect Distributed Denial of Service (DDoS) attacks whose backscatter packets are observed on the darknet (i.e., unused IP space). To classify DDoS backscatter, 17 features traffic defined from IPs/ports information for source and destination hosts. adapt change attacks, we newly implement online learning function in proposed system, where SVM classifier is continuously trained with transformed during a certain period. In performance evaluation, use MWS Dataset 2014 that consists collected 1st January 28th February (8 weeks). We demonstrate keeps good test detection (0.98 F-measure).