| The development of wireless communication promotes the advent of the era of "Internet of Everything",and the services based on location information have a very important role and significance in the construction of Intelligence Interconnection.Satellite positioning and navigation is widely used outdoors,but indoors,due to its environmental constraints,wireless signals are susceptible to many unfavorable factors such as multipath and non-visual distance propagation,so accurate and reliable positioning faces a great challenge.The processing of multipath and non-line-of-sight errors has been the focus of research and hotspot.In this paper,based on indoor multipath environment information,we study the improvement and optimization methods for handling multipath in two directions:error compensation and path assistance.Firstly,the thesis reviewed the relevant literature of many domestic and foreign research scholars,described the current development status of wireless localization technology in multipath environment at home and abroad,introduced the basic theory of wireless localization,and the problems in indoor localization are briefly outlined and analyzed.Then,for the multipath error in indoor line-of-sight environment as well as mixed environment,on the basis of the indoor floor plan,a dynamic error compensation method based on error model is proposed for the geographical and time-varying characteristics of multipath error.The method is divided into two stages.Offline modeling:Firstly,grid the location area,the deviation data of each grid time of arrival(TOA)are collected,and then the deviation model is fitted by a learning method,solved and the parameters of the model are saved as a deviation model library.Online localization:After matching the initial positioning to the corresponding grid and using the deviation model to generate each vertex error,the ratio of the angle of arrival(AOA)parameter values deviating from the angle range of the grid vertices is used as the weight of each vertex compensation value,and the compensated data is used to correct the estimated position and reduce the positioning error by repositioning.Both simulation and experimental results show that the proposed method has significantly improved the positioning accuracy compared with other methods,and the system performance is stable under different levels of interference noise.Next,for the pure NLOS environment without line-of-sight paths,this paper improves the traditional method based on scattering paths and deviation factors for the disadvantage that its localization error increases or even fails to locate when the possible position lines(LPMDs)are parallel to each other or to the region.The algorithm divides the LPMD distribution into three categories and constructs the possible position regions corresponding to each case separately.A new step-by-step dichotomous search method is proposed to find the optimal estimated position within the region by back-projecting the position of the scattering point based on known information,dichotomously finding the point with the minimum error with the measured data on the central axis of the possible position region,then making a vertical line on the central axis of the region at this point,and then using the same dichotomous finding method to determine the estimated position on the vertical line,which reduces the complexity of the algorithm while improving the positioning accuracy.And the localization accuracy of the original method and the proposed method are compared by simulation and experiment respectively,and the localization performance of the algorithm is verified.Finally,the research work of this paper is reviewed and summarized,but there are still areas of improvement in the scheme,and future research work is planned based on these shortcomings. |
PDF Full Text Request