Research On Emotional Perception And Behavior State Based On Channel State Information

With the advent of the mobile Internet era and the development of 5G,smart home and intelligent wearable equipment industry ushered in a new air vent,more and more people choose to use a variety of sensor equipment to track and monitor their daily behavior,control the surrounding household appliances,and even finally assess their recent emotional state.Unlike these traditional ways,more and more researchers are trying to use the channel state information(CSI)of Wi-Fi signals as an invisible sensor for human behavior monitoring.It has the advantages of wide coverage,no contact,no privacy infringement problems,and the prospect of exploration and application in the later stage is very extensive.In this paper,an emotional perception system based on channel state information is proposed for the first time by using the phenomenon that emotional expression has corresponding external limb expression,and we correspond emotion,action and signal one by one.By establishing a lookup table method based on Fresnel zone theory,the antenna is set up reasonably to capture as many detailed features of emotional expression as possible,and then four common emotions(joy,sadness,anger,fear)are identified by KNN algorithm.We then try to monitor some of the behavior states with finer action granularity,such as operating smartphones.We divide mobile phone operation gestures into clicks,long press,horizontal sliding,vertical sliding,doubleclick,press the Power key,and static,i.e.seven categories.By monitoring the frequency and order of these seven types of actions over a period of time,using HMM algorithm to identify app usage of the smartphone in this time,chatting,playing games,watching movies or browsing the web.The contributions of this thesis are as follows:First,we describe the correspondence between action and Wi-Fi signal information,and a table checking method based on Fresnel zone theory is proposed to set up antennas.Second,we establish the mapping between emotional and channel state information,and four different emotions are identified by the classification method.Third,we further monitor the user's fine-grained action to operate the phone,and through monitoring results over a period of time,distinguish the user's mobile phone usage status during that time period.