一种非视距信号环境下智能手机速度平滑定位方法CSCD

随着智能手机的普及,基于智能手机的位置服务深刻地影响了人们的生活。通常,智能手机内嵌有全球导航卫星系统(GNSS)芯片,通过跟踪导航卫星信号获得载波频率、载波相位和码相位信息,从而解算位置和速度信息。对于单点定位(SPP)技术而言,位置主要由码相位计算确定,而速度信息主要由载波多普勒频移信息计算确定。在复杂城市环境下,GNSS信号可能会被周围的建筑物反射,形成非视距(NLOS)信号。NLOS信号会导致额外的传输路径,并引起数十米的位置误差。由于速度比位置更精确,通常使用卡尔曼滤波器(KF)进行基于位置-速度(P-V)动态模型的位置平滑。提出了一种因子图优化(FGO)方法来减少NLOS引起的位置误差,将时间相关历史测量值添加到FGO中以优化当前位置估计,并采用两种鲁棒策略来减小NLOS条件下的位置偏差。最后,使用智能手机采集实际场景数据集进行了实验,结果表明FGO方法在速度辅助下可以获得更好的定位结果。

A Robust Positioning Method for Smartphone GNSS Under NLOS Environments

With the popularity of smartphones, location services based on smartphones have deeply affected people''s lives. Usually, a smartphone is embedded with a Global Navigation Satellite System (GNSS) chip, which provides three-dimensional position information. The GNSS chip obtains the carrier frequency, carrier phase and code phase information by tracking the navigation satellite signal to get the position and speed information. Typically, for Single Point Positioning (SPP) techniques, the position is primarily determined by code phase calculations, and the velocity information is primarily determined by carrier Doppler shift information. In urban environments, GNSS signals may be reflected by the surrounded buildings, forming a Non-Line-Of-Sight (NLOS) signal. NLOS signals cause additional transmission paths and cause position errors of tens of meters. Since velocity is more reliable than position, Kalman Filter (KF) is usually utilized for position smoothing based on Position-Velocity (P-V) dynamic model. In this paper, we propose a Factor Graph Optimization (FGO) method to reduce the position error caused by NLOS, the time-correlated historical measurements are added to FGO to optimize the current position estimate, and two robust strategies are adopted in FGO to mitigate position bias under NLOS conditions. Finally, experiments are carried out using a smartphone to collect real scene datasets. The experimental results show that the FGO method can obtain better localization results with the aid of speed.