Campus Crowd Monitoring And Implementation Based On WiFi Signal

It is important to understand the law of campus population,improv the level of campus integration,and protect campus security.At present,the main monitoring means in the campus include video monitoring and security patrol.Video monitoring often sometimes ductile,have no way to guarantee the control room all the time someone staring at the monitor,and video monitor will have dead Angle,in a bad light or a mask effect is poor,the security patrol to achieve 24 hours monitoring on campus,both methods have certain limitations,the campus crowd activity monitoring and vulnerability.In order to clearly understand the overall activity rules of students in the campus,and solve some accidents caused by crowd gathering and intrusion monitoring events in important places on campus,this paper proposes a campus crowd monitoring system based on the status of WiFi channel.The main contents and contributions are as follows.(1)WiFi probe is used to collect data from multiple nodes,and we can more researchs such as MAC(Media Access Control Address)statistics,AP(Wireless Access Point)analysis and channel analysis and RSSI(Received Signal Strength Indication)according to the collected data.(2)Time series prediction.I used a variety of prediction models to predict the flow of people,and predicted the flow of people at the next moment according to the historical data of the flow of people at nodes,and set the number of people on the line limit.Calculate the number of people of the next time,and compare with the predicted value,trigger alarm outside the threshold.(3)Intrusion monitoring.Google mobile phone Nexus 5 is used to collect indoor manned and unmanned CSI(Channel State Information)Channel State Information,and the convolutional neural network classification algorithm is used to train and save the obtained data.When someone enters the room,CSI Channel State will change and alarm will be issued.This article adopts the method of machine learning and deep learning developed campus crowd monitoring system based on WiFi channel state and deployed to the cloud,the system proposed in this thesis is very convenient to deploy and has high monitoring accuracy,dug up some have important value of campus personnel gathered and activity state information,for the study of campus or other sensitive area provides the mobile population behavior experiment platform,provides a new technical means for management department and the solution.