Intelligent IoT Device Abnormal Traffic Detection Method Based on Secure Multi-Party Computation

The traditional Intelligent Internet of Things (IIoT) system has the problems of low detection efficiency, high cost, and difficulty in ensuring accuracy in detecting Abnormal Traffic Detection (ATD) of equipment, making it unable to cope with the growing large power grid network. A ATD method for IIoT devices based on Secure Multi-Party Computing (SM-PC) is proposed to address the above issues. Firstly, the basic architecture of the IIoT system and the principle of the device ATD were analyzed. Then, an analysis was conducted on the necessity of multi-party participation in ATD, and an SM-PC model was designed based on this. Finally, an ATD model was constructed based on SM-PC and the Isolation Forest (IsFo) algorithm: SMPC-IsFo. A comparative analysis was conducted between the proposed method and the other three comparative methods under the same conditions through simulation experiments. The results indicate that the proposed method has the highest accuracy and Fl score when using the CICIDS2017 and UNSW-NB15 datasets. On the two datasets, it reached 98.24%, 97.33%, and 89.31%, 88.19% respectively. Its performance is superior to the other three comparison methods.