Indoor security method based on WiFi for privacy environment

Abstract:
With the development of Internet and the wireless devices, we propose a privacy-oriented indoor security method based on WiFi under the background of 6G communication and perception integration. Firstly, we propose the compiled wireless network to collect CSI in different situation. Next, the CSI is pre-processed by combining the median filter algorithm and the Hampel filter algorithm. Then, neural network is used for feature extraction. Finally, machine learning algorithm is used to classify and detect the intrusion state. The results of the experiment show that the performance in different environments can reach 95%. Besides, the universality and robustness of the algorithm both meet the needs of practical scenarios and have good practical application value.