【Author】
Essaid, Meryam; Kim, DaeYong; Maeng, Soo Hoon; Park, Sejin; Ju, Hong Taek
【Source】2019 20TH ASIA-PACIFIC NETWORK OPERATIONS AND MANAGEMENT SYMPOSIUM (APNOMS)
【Abstract】Recently Distributed Denial-of-Service (DDoS) are becoming more and more sophisticated, which makes the existing defence systems not capable of tolerating by themselves against wide-ranging attacks. Thus, collaborative protection mitigation has become a needed alternative to extend defence mechanisms. However, the existing coordinated DDoS mitigation approaches either they require a complex configuration or are highly-priced. Blockchain technology offers a solution that reduces the complexity of signalling DDoS system, as well as a platform where many autonomous systems (Ass) can share hardware resources and defence capabilities for an effective DDoS defence. In this work, we also used a Deep learning DDoS detection system; we identify individual DDoS attack class and also define whether the incoming traffic is legitimate or attack. By classifying the attack traffic flow separately, our proposed mitigation technique could deny only the specific traffic causing the attack, instead of blocking all the traffic coming towards the victim(s).
【Keywords】DDoS Mitigation; Blockchain; Smart Contracts; Ethereum; Deep Learning; RNN-LSTM
评论