Research On Indoor Localization Technology Of Fusion Multidimensional Spatial Spectrum Information

With the rapid growth of information data services and Internet services,people's demand for localization and navigation services in different environments has gradually increased.Especially in indoor environments with complex environments,such as large shopping malls,museums,and underground parking lots,it is often necessary to determine the specific location of persons or objects.However,currently the most widely used GPS technology can not be used for indoor localization because its signal is a line-of-sight propagation.Therefore,research on indoor localization technology has become a hot topic in the current research field.The main work and innovation of this paper are as follows:Firstly,signal subspace fingerprints,covariance fingerprints and RSSI fingerprints are proposed,and multidimensional fingerprints library is constructed.The latter two fingerprints can be obtained directly from the calculation of the signal subspace,greatly reducing the burden of database construction.Compared with the traditional RSSI single fingerprint,they contain the intensity and arrival angle information of the incident signal to the array antenna,which can more effectively combat the time-varying characteristics of the channel caused by multipath propagation and environmental changes.Secondly,a fusion localization algorithm based on the double threshold assisted mode is proposed to calculate the variance between the prediction grids of multiple classifiers.The predicted grid points corresponding to the collected samples and classifiers is removed when the environmental fluctuations are large,and the majority voting criterion is adopted to get the final forecast grid.The classifier is trained by using GBDT.Finally,it was verified by experimental simulation.Thirdly,in the actual verification work,an indoor localization system based on the open source software radio GNU Radio was designed.This system uses the USRP to connect multiple antennas to form an array signal receiving system to collect signals at various indoor points.Finally,the algorithm introduced and improved by the multidimensional fingerprints collected in the actual indoor environment was verified.