Program

Abstract:

In recent years, wireless positioning technologies have been increasingly used in both civilian and military applications, with Global Navigation Satellite System (GNSS) being the most widely used positioning technology. However, in the complex environment of urban canyons, some satellite signals are obscured and reflected by surrounding buildings before they can be received. These signals that do not propagate in a straight line to the receiver are called non-line-of-sight signals (NLOS). NLOS signals contain additional errors caused by reflections, which can lead to positioning multipath errors of tens of meters or more. To address this problem, this paper proposes a NLOS signal identification method based on fuzzy C-mean clustering for precise positioning of GNSS in urban canyon environment. In this paper, three signal features, namely carrier-to-noise ratio (C/N0), pseudorange residual and elevation angle, are extracted and used to classify the signal of NLOS by combined clustering. And the NLOS/MP is reasonably excluded and suppressed based on the PDOP changes after NLOS identification, and finally the localization algorithm is solved. The experiments show that the localization accuracy in the three directions of east, north and sky is significantly improved after NLOS processing in the deep urban canyon environment. Compared with mathematical statistics and supervised classification methods, fuzzy C-mean clustering does not require a priori information and avoids tedious GNSS data preprocessing, so the method reduces computational load and equipment cost.