Research And Application Of Wireless Signal Localization:an Approach Of Network Propagation Influence Analysis

In recent years,the rapid development of wireless communication networks leads to increasingly complex wireless signal propagation environments,and various types of interference generated within and outside networks seriously affect the quality of communication networks.Localization of radio interference sources outside networks in dense urban areas becomes a major difficult work in current network interference troubleshooting.Due to the multi-path and non-line-of-sight propagation of wireless signals in complex urban environments,the geometric positioning-based methods cannot achieve higher localization accuracy.In practical processes of interference troubleshooting,manual drive-testing is often required to gradually find interference sources in a segment-bysegment manner and at higher labor and time costs.Therefore,how to quickly and accurately determine the locations of interference sources outside networks is an urgent problem to be solved.Aiming at LTE electric wireless private networks,this thesis proposes an interference source localization method based on network propagation influence analysis.This method takes into account the complex propagation behavior of wireless signal in urban environments and uses drive-testing data and 3D ray tracing to simulate the interference signal propagation trajectories.Depending on the characteristic linkage of spatial locations in the interference region,the 3D space and the ray distribution within the region are modeled as an influential network,and then the network influence propagation model is designed for determining the interference source locations.The study described in this thesis consists of three parts as follows.(1)Generating of interference signal propagation trajectory based on reverse ray tracing.Firstly,the wireless signal propagation simulation environment is established by interference detection in localization areas and scene modeling.The strong signal drive-testing points are then filtered from the drive-testing data and used as the starting points to generate the possible propagation trajectory of interference signal by 3D ray tracing.(2)Interference localization via network propagation influence analysis.Firstly,the spatial raster index is obtained by analyzing the coverage of the propagation trajectory of the interference signal in the spatial raster.Then the network structure of the influence model is determined and simplified by taking the raster as nodes and the trajectory links between neighboring rasters as edges.After that,network node attributes such as node influence and network edge attributes such as edge propagation probability are calculated based on the spatial raster index.In this way,the network influence propagation model for interference localization is established.This model is taken to analyze the propagation influence of the interference signal and determine the interference source location through three steps,i.e.propagation trajectory de-duplication,influence propagation,and interference source location evaluation.Finally,the local search based second-round positioning is employed to further improve the accuracy of the localization results from the first-round modelbased positioning.(3)Design and implementation of the interference localization software system.Based on the above two parts of work,the interference localization software system based on network propagation influence analysis is developed.Large-scale localization test data are generated by simulating influence propagation in the dense urban areas in Nanjing.The accuracy of first-round positioning based on influence analysis and secondround positioning based on local search is evaluated.The causes of localization errors are also analyzed.It is shown that the first-round positioning can guarantee 67%of the localization error within 116m,and the second-round positioning can guarantee 67%of the localization error within 83m.In summary,the proposed interference localization method takes into account the non-line-of-sight propagation of interference signals in dense urban environments and has high localization accuracy.Therefore,it can effectively solve the problem of locating extra-network interference sources in electric wireless private networks,and strongly supports the work of network optimization and interference detection and localization in practice.