Anomaly Detection Method of BDS Signal-in-Space Based on Autoregressive Distributed Lag Model

Abstract

The signal-in-space (SIS) anomaly of BeiDou navigation satellite system (BDS) is an important factor affecting its high accuracy SIS quality assessment. Detecting and eliminating SIS anomaly is not only an important method to build SIS fault model of BDS, but also helps to guarantee the integrity of BDS navigation and positioning. Based on the problem that the traditional empirical threshold method cannot accurately identify the start and end times of anomalies in anomaly detection, which leads to anomaly detection leakage, a combined detection method based on autoregressive distributed lag model and empirical threshold is proposed in this paper. Before the calculation, the spurious anomalies of SIS are removed by data cleaning. The high-precision SIS ranging error (SISRE) is recovered by Space State Representation (SSR) correction number, and then projected to the user’s line of sight direction, and the anomaly detection threshold is determined by using the combined threshold of empirical threshold and autoregressive distributed lag (ARDL) model. The feasibility and effectiveness of the method were analyzed by using the data collected in 2021. The test results show that, compared with the traditional threshold method, the proposed method can more accurately detect the start and end points of SIS anomalies caused by clock anomalies, which is also the detection accuracy is improved. In addition, the anomaly detection method proposed in this paper is used to count the anomalies throughout the year, and the results show that the highest frequency of anomalies is found in geostationary orbit (GEO) and inclined geosynchronous orbit (IGSO), and these anomalies are mainly caused by satellite clocks.