Design of Comprehensive Experimental Scheme for Cyberspace Security Teaching Based on SDN and Machine Learning

Nowadays, the cyberspace security industry has been booming, and there is an urgent need for excellent technical talents in related fields. Thus, it is necessary to design a comprehensive and practical simulation experimental scheme oriented to important cyberspace security issues, and apply it to the teaching process of core courses in cyberspace security discipline. Currently, DDoS attacks in cloud environment have become one of the thorny issues faced by cyberspace security. In order to solve this issue, a novel collaborative DDoS detection and mitigation system based on software defined network (SDN) and machine learning in the cloud (DDMS) is designed, in which a detection module combines CBF method with DNN method to handle both flooding attacks and low-rate attacks, and a mitigation module establishes a distributed multi-controller architecture to make multiple SDN controllers work collaboratively. Compared with related schemes, DDMS has faster data processing speed, higher detection accuracy, and lower computational overhead. Consequently, we use the system as a comprehensive experimental scheme for the core courses of cyberspace security, which can equip students with technologies of advanced network security attack and defense based on SDN, develop the students’ innovation and practice ability, and then cultivate excellent talents for the industry.