Algorithm Improvement And System Performance Analysis Of Multi-user Single-node Indoor Positioning

With the construction of large shopping malls,stations,and underground parking lots and the increasingly complex indoor environment,people's demand for indoor positioning is becoming more and more intense.Since the Wi Fi for network communication is generally laid in the indoor environment,it can be used for indoor positioning while the network communication is being performed.Based on the single node indoor positioning system of Wi Fi,this paper studies the performance of multi-user positioning system in different situations from the perspective of network transmission.First of all,aiming at the lack of traditional indoor positioning in positioning accuracy and deployment,this paper proposes a single node positioning model based on channel state information(CSI).Then,a phase error model of CSI is proposed for the phase information characteristics of the received CSI.The linear error related to the subcarriers and the nonlinear error caused by noise are eliminated by the interpolation method and the least squares method.Then,this paper introduces the basic principle of MUSIC-based indoor positioning algorithm,constructs the spatial spectrum by using the phase value of CSI after eliminating the error,estimates the transmission delay and angle of the multipath signal by searching for the peak value of the spectrum,and calculates the location of the source.Aiming at the problem of positioning accuracy and computational complexity of MUSIC algorithm,a frequency band splicing algorithm and a direct search algorithm are proposed.They improve the positioning accuracy of the algorithm and reduce the computational complexity by extending frequency band and reducing the search dimension.On the basis of direct search,the Root-MUSIC algorithm and ESPRIT algorithm are used to avoid one-dimensional peak search,which further reduces the computational complexity of the positioning algorithm.The positioning error and running time of different positioning algorithms are actually measured.Finally,this paper constructs a multi-user indoor positioning system,which analyzes the influencing factors of multi-user positioning system performance from three aspects: type of access users,single node positioning system and multi-useraccess protocol.The effects of different positioning algorithms,different number of transmitted data packets,different communication methods and different maximum initial contention windows on user positioning delay are described.After analyzing the different influencing factors,this paper calculates the maximum number of times that a multi-user positioning system can handle in the case of a different number of Internet users.