Research On CSI-based WiFi Indoor Positioning Algorithm

With the development of the information age,the application of indoor positioning technology is becoming more and more extensive,especially in medical monitoring and disaster relief,and accuracy requirements of positioning services are becoming higher and higher.The existing Wi Fi indoor positioning technology mainly uses the received signal strength indicator(RSSI)as a measurement indicator,but because the strength of the Wi Fi signal is easily affected by the complex indoor environment,the measurement accuracy is usually not high,and then the positioning accuracy is limited.In contrast,the measurement accuracy of Channel State Information(CSI)is higher,and there is a strict signal model as a support,so it has the potential to achieve high positioning accuracy.However,the current indoor positioning systems based on CSI,such as the well-known Spot Fi positioning system,still suffer from poor noise immunity and weak resolution of coherent sources,and cannot meet the need of high-precision positioning services.In order to achieve a stable and high-precision indoor positioning system,this paper proposes a CSI-based Wi Fi indoor positioning improved algorithm(compared to Spot Fi)for complex indoor scenes,and builds an actual test system to verify and evaluate the performance of the proposed positioning algorithm.Our main work is as follows:(1)To solve the problems of poor noise-resistance ability,poor signal-decoherence ability and low accuracy of indoor positioning algorithms,this paper has optimized and improved the positioning algorithm.Specifically,a.Based on the algorithm framework of Spot Fi,on the basis of the classic Multiple Signal Classification(MUSIC)algorithm,this paper proposes an improved MUSIC algorithm based on the forward and backward subcarrier smooth expansion array.This algorithm combines forward spatial smoothing and conjugate backward spatial smoothing to optimize the subcarrier array of the CSI matrix,which can effectively reduce the aperture loss of antenna array and increases the number of equivalent antennas.In addition,the noise subspace of the covariance matrix is modified to enhance the orthogonality between the noise and the signal steering vector,thereby improving the performances of signaldecoherence and the accuracy of the angle of arrival(Ao A)estimation.b.In order to solve the problem that the MUSIC algorithm cannot accurately estimate the number of signal sources under a low signal-to-noise ratio,this paper proposes an algorithm based on the eigenvalue gradient(CSI covariance matrix)to improve the noise resistance of the MUSIC algorithm.In addition,in order to solve the problem of direct path recognition after the MUSIC algorithm calculating the multipath parameters,an improved clustering method is proposed to accurately classify similar multipaths of multiple data packets,thereby improving the direct path recognition by the likelihood probability.(2)Combining the above positioning algorithms,this paper builds a complete measured system on the basis of existing commercial Wi Fi devices to verify the effectiveness of the proposed algorithm and evaluate performance.First,to provide a reliable data basis for high-precision indoor positioning,a multi-packet linear fitting method is used to eliminate CSI errors and improve CSI data accuracy.Then based on the corrected CSI,the improved positioning algorithm in(1)is used to estimate the Ao A,and combined with the RSSI for positioning.Experiments show that the positioning error of this method is reduced by 17% compared with the Spot Fi algorithms.The positioning algorithm proposed in this paper has great guiding significance and practical application value for actual indoor Wi Fi positioning.