Urban city robust GNSS positioning: a hybrid RTK model driven by theories of frequency-adaptive switching and solution separation

Abstract Real-time kinematic (RTK), the most commonly used Global Navigation Satellite System (GNSS)-based positioning technology at present, adopts a double-differenced (DD) combination to eliminate observing errors, which can offer high-accuracy positioning solutions for users. However, while under a complex user environment (i.e. urban city), frequent partial or complete GNSS outages bring frequent observation availability changing among single-frequency, dual-frequency, and multi-frequency observations. This appearance will lead to the available GNSS observations cannot be fully utilized in the conventional RTK model that is based on single-frequency, dual-frequency, or triple-frequency observations. To overcome this problem, a mixed-frequency hybrid RTK model based on theories of the frequency adaptive switching (FAW) and the solution separation (SS) test is provided in this work. To present the performance of the proposed method, a set of unmanned aerial vehicle-borne dual-frequency multi-constellation GNSS data is processed and also used to simulate the frequently changed GNSS availability. The experimental results indicate that (1) the FAW-based RTK model can offer 12.5%, 39.5%, and 30.0% positioning accuracy improvements and 13.6% fixed rate improvements on average compared to the traditional dual-frequency RTK mo (2) the SS test can detect the faultily fixed ambiguities and brings about 17.2%, 29.4%, and 23.1% positioning accuracy improvements in three directions compared to the FAW-based method while using partially ambiguity fixed method.