Research On Indoor Intrusion Detection Based On WiFi-CSI

With the widespread application of WiFi wireless network technology,the distinctive features of wireless perception,such as sensorless,wireless,and non-contact,closely link human behavior detection and communication systems,and wireless perception technology for passive intrusion detection without equipment has also been more and more popular.The more attention.Traditional intrusion detection methods usually use sensors,infrared rays and cameras,but these methods have certain problems in accuracy or cost.Intrusion detection ability under wireless perception has the advantages of wide coverage,little environmental impact and privacy protection.In addition,WiFi sensing technology does not require personnel to actively participate in the entire wireless sensing process.However,although some research results have been achieved in passive intrusion detection accorded to wireless sensing technology,there are still problems such as single research scenario,limited detection distance and insufficient detection rate.In response to the above problems,this literature first explain the key knowledge points of CSI,expounds the relevanted applications of wireless sensing means,and discusss the existing CSI-based personnel intrusion detection ability in detail.Combining with important research in the past,this research explain a novel indoor occupant intrusion discussion meanings based on WiFi-CSI.First,use the fine-grained CSI on the WiFi device to capture the small changes caused by the movement of the human body,and then preprocess the collected data,and then use the MUSIC algorithm to process any array to obtain the characteristics of the covariance matrix,according to the correlation of the type To get the noise subspace,and then obtain the AOA,and judge whether there is an intrusion according to the manifestation of the spectral peak diagram.At the same time,by calculating the phase distinction changes of different paths caused by the behavior of the human body,it is judged whether there is an invasion.The difference between this method and the traditional method is that it combines spectral peak search and phase difference.The advantages of the two are complementary,which resistes the interference of scene and noise,and solves the influence of multipath functions on the results.Secondly,two prominent indoor scenes,meeting room and darkroom,are selected in the article to test the effectiveness of the method.The experimental conclusion show that the precision of the put forwarded method in the conference room and the dark room are 93.2%and 94.9%,respectively.In addition,this works also detects the detection performance of the proposed method in three motion modes of small-amplitude,medium-amplitude and bigamplitude.The detection rates are respectively 93%,94% and 94.7%.The method is verified by comparison with existing methods.The strong and stability of the system are in line with the expected results.