An enhanced sensitivity analysis framework for residual ionospheric error in network RTK geometry-free three-carrier ambiguity resolution
摘要
With the rapid development of multi-frequency Global Navigation Satellite Systems (GNSS), Geometry-Free (GF) Three-Carrier Ambiguity Resolution (TCAR) has become an increasingly viable solution for high-precision positioning. The GF TCAR model enables satellite-by-satellite ambiguity resolution, offering robustness independent of satellite geometry. However, its performance in network Real-Time Kinematic (RTK) applications is highly sensitive to residual atmospheric errors. To accurately quantify these impacts and optimize ambiguity resolution strategies, this study proposes an enhanced analysis framework and presents a comprehensive sensitivity analysis of the GF TCAR model under different atmospheric and noise conditions. Theoretical derivations and simulations confirm that non-dispersive errors, such as tropospheric delays, are effectively mitigated within the GF framework. In addition, the study establishes quantitative accuracy thresholds for ionospheric modeling across constellations. While Extra-Wide-Lane (EWL) and Wide-Lane (WL) ambiguities remain robust against ionospheric errors of up to dozens of Total Electron Content Units (TECU) and several TECU, respectively, Narrow-Lane (NL) ambiguity resolution is the most challenging step, requiring residual errors to remain about within ± 0.2 TECU. Furthermore, the analysis shows that EWL ambiguity resolution is relatively insensitive to measurement noise, whereas WL ambiguity resolution is predominantly noise-limited, especially as the EWL wavelength increases. These findings provide valuable error-tolerance references for the quality control of ionospheric corrections in network RTK services.