A Reverse Ray Tracing Algorithm Based on Cluster Analysis
Introduction of ray tracing can effectively solve problems of multi-path propagation,non-line-of-sight (NLOS)noise and hearability faced by localization algorithms in complex circumstances.However,the environment models are often too ideal and thus may affect the actual application of ray tracing;meanwhile,measurement errors of localization parameters can severely undermine the tracing results of signalspropagation paths.To address these problems,a reverse ray tracing algorithm is proposed based on cluster analysis.First,a two-dimensional propagation environment model is built,with the building surface assumed as the quasi-mirror to simulate the actual propagation environment of rays;second,the position orientating path and the area orientating path are defined based on features of the propagation path of multi-path localization signals.The algorithm proposed uses the branch tracing and search algorithm to calculate feasible paths according to the estimated TOA and AOA values of multi-path localization signals,and then employs cluster analysis to recognize position orientating paths and trace their propagations,based on which the sub-section weighting method is used to trace the maximum recognizable path among area orientating paths.Finally,simulation results demonstrate the rationality and effectiveness of the algorithm.
Ray Tracing Quasi-mirror Geographic Information System(GIS) Localization
Kong Fanzeng Chen Guojun Zheng Nae Chen Song
Zhengzhou Institute of Information Science and Technology Zhengzhou,China
国际会议
重庆
英文
529-537
2016-03-20（万方平台首次上网日期，不代表论文的发表时间）