Cyber-Physical System Intrusion Detection Model Based on Software-Defined Network
Yunting Xiao, Jingyong Liu, Lichen Zhang · 2021
Due to the large scale, increasing complex, wide distribution, and heterogeneous network environment of Cyber Physial System(CPS), CPS becomes more vulnerable to network attacks. Concerning this, we introduce the software defined network paradigm into CPS architecture to simply CPS management and provide a solution against network security problems. We have also proposed a detection approach based on extreme learning machine(ELM) to protect CPS. The proposed approach is tested using the generated dataset and the seven features subset of NSL-KDD dataset. And it is proved to be effective with accuracy rate of 99% and 75.4% respectively. Experiments confirm that our proposed extreme learning appraoch can be a feasible countermeasure for flow-based intrusion detection in CPS environment.