Research On Passive Target Tracking Algorithm Based On WLAN Channel State Information

With the advent of the artificial intelligence era,intelligent applications such as smart security and smart home have become more and more extensive.Passive target tracking has become a research hotspot at home and abroad as a key technology for these applications.Passive target tracking technology based on special equipment such as radar and infrared is costly and difficult to popularize in indoor environment.And with the widespread deployment of Wireless Local Area Network(WLAN),WLAN-based passive target tracking technology has attracted much attention from researchers because of its low cost,ease of deployment,and high versatility.But existing WLAN-based passive target tracking technologies mostly require offline data acquisition such as fingerprint positioning methods,which have great limitations in practical applications.In order to solve these problems,this paper carries out research based on wireless target tracking algorithm.The main research contents are as follows:First,for the target motion detection,according to that the mobile target has the Doppler shift effect,the target motion detection parameters are extracted in the frequency domain.Firstly,the influence of target motion on CSI phase is analyzed,and then the time-frequency analysis of CSI power is performed by short-time Fourier transform in order to eliminate the influence of phase error on Doppler shift estimation.Then,based on the magnitude of the spectral intensity in both the target and non-target motion states,the target motion detection is achieved using the sum of the spectral intensities on the frequency sequence.Compared with the detection method based on CSI amplitude,the proposed algorithm can effectively improve the detection accuracy.Second,for target tracking,the Doppler shift is first extracted from the time-frequency spectrograms to calculate the length change rate of the target reflection path.Secondly,the influence of the moving direction of the target relative to the transceiver link on the received signal power is analyzed by the ray tracing model.Based on this,the spectral intensity at the time of extraction is used as an estimation parameter of the target motion direction,which can reduce the influence of multipath effect,and then use the relationship between the trend of the spectral intensity in time series and the direction of motion to achieve direction estimation.Finally,the joint reflection path length change rate and the target motion direction are used to estimate the target moving speed,thereby achieving target tracking.Compared with the existing passive target tracking method,this paper can achieve higher tracking accuracy without the need of offline data training.Third,this paper implements the proposed algorithm on Wi-Fi devices and verifies its performance in real indoor environment.The experimental results show that the proposed algorithm has a motion detection accuracy of 97.35% and a confidence of 60% with a tracking error of 1.3 meters.