Intrusion Detection And Posture Recognition In Sensitive Environment Based On Radio Frequency Sensing

In recent years,with the rapid development of radio frequency sensing technology and the wide deployment of wireless LAN,many scholars found that wireless signal can not only realize basic data communication,but also be applied to intrusion detection and attitude recognition in sensitive environment.Therefore,radio frequency sensing technology based on wireless signal has been widely concerned and deeply studied by scholars at home and abroad.In the face of some sensitive environments that are not allowed or easy to enter,such as polluted environment,dangerous areas of chemical plants and cultural relic sites that need to be carefully protected,the traditional intrusion detection and attitude recognition technology is inadequate.The technology based on radio frequency sensing has the advantages of wide coverage,not affected by light,not exposing privacy and strong penetration performance,which can well complete the task of intrusion detection and attitude recognition in sensitive environment.Aiming at the shortcomings of traditional intrusion detection technology,such as high hardware requirements and harsh detection conditions,this paper proposes an intrusion detection technology based on radio frequency sensing.Firstly,according to the amplitude and phase characteristics of channel state information,median filtering,deconvolution,linear transformation and other methods are used for preprocessing;secondly,self-organizing competitive neural network algorithm is selected for feature extraction and intrusion signal fingerprint database is established;then,Softmax regression algorithm is used to establish the statistics of nonlinear dependence between intrusion behavior and signal in off-line phase Finally,in the online phase,the trained statistical model is used for classification and recognition,and the voting mechanism is used to complete the final decision.The results show that the intrusion detection rate of the technology is as high as 98%,and the misjudgment rate is only 2%.In view of the shortcomings of traditional attitude recognition technologies such as cameras and sensors,such as high cost,blind area,complex deployment and invasion of user privacy,this paper also proposes a kind of attitude recognition technology based on radio frequency sensing.In the process of feature vector extraction,not only the conventional maximum value,minimum value,difference and standard deviation are selected as features,but also the quartile range representing the dispersion of channel state amplitude information and the information entropy describing the ordering degree of phase information are introduced The improved linear discriminant analysis algorithm and Softmax regression algorithm are combined to generate the human posture model.The experimental results show that the method has high accuracy and good robustness,the average recognition rate can reach95.87%,and the recognition performance is good.At the same time,for practical engineering applications,this paper designs and develops an intrusion detection system based on radio frequency sensing.In order to facilitate the design and development of the system,this paper adopts the modular design idea,and divides the system into data acquisition module,preprocessing module,feature extraction module,classification recognition module and interactive display module.In addition,this paper also introduces the design and development process in detail from four aspects of system requirements analysis,overall architecture design,software and hardware platform system construction and module development,focuses on the specific implementation methods of each module,and tests the function and performance of the system,and the test results meet the expected requirements.