Research On WiFi-based Passive Target Motion Detection Algorithm In Complex Scenes

The development of science and technology has brought great convenience to people's lives.Applications such as smart homes have begun to appear in daily life.By detecting whether there are moving targets in the indoor environment and providing corresponding services,applications such as smart home can be more intelligent.As wireless fidelity(WiFi)technology is widely distributed in indoor environments,and it has the advantages of low cost and easy deployment,the use of WiFi for device-free target motion detection has received increasing attention from researchers.The signals between the transceivers in the indoor environment include Line-of-Sight(LOS)signals and through-the-wall signals.However,the existing WiFi-based device-free target motion detection technology is only for a single scene,and there are not many applications in complex scenarios containing two types of signals,such as indoor home environments.In view of this situation,this paper conducts research on device-free target motion detection algorithms for complex scenes based on Channel State Information(CSI).The main research contents of this article are as follows:First,the research on Line-of-Sight(LOS)/through-the-wall signal recognition is carried out.Because there are differences between the carriers of the received signal,a suitable carrier is selected by calculating the maximum variance,and then calculate the cumulative probability distribution of the selected carrier data.Finally,the dynamic time warping is used to calculate the distance of different set of data,and the LOS/ through-the-wall signal recognition is realized by comparing with the threshold.Second,the research on the passive moving target detection under the condition of the through-the-wall signal is carried out.First,the data processing method of conjugate multiplication is used to eliminate the inherent phase offset in the received data and then perform linear interpolation.Next,Doppler-MUSIC is used to estimate the Doppler frequency shift of the received data and its corresponding spectrum power,and further extract the data characteristic values such as the average value of the spectrum power and the mean of the first order difference.Finally,Support Vector Machine(SVM)is used to realize passive target motion detection under the condition of through-the-wall signals.Again,the area detection of the moving target is performed.Use dynamic time regularization to determine the link status of multiple indoor links online.Using linear chain conditional random field to parameterize the target motion area information,and the parameters are updated with the improved iterative scaling method.Finally,the Viterbi algorithm is used to estimate the target motion area.The algorithm verification is performed in indoor complex scenes.The experimental results show that it is possible to detect whether the link state has changed with high accuracy.At the same time,the moving target detection algorithm can maintain a detection success rate of more than 93% under the condition of through-the-wall signal.For different moving areas of the target,the algorithm can accurately estimate the moving area of the target.