Enhancing PPP-RTK resilience: a comprehensive evaluation of atmospheric product extrapolation methods for handling transmission interruptions

Precise Point Positioning Real-Time Kinematic (PPP-RTK) represents a groundbreaking advancement in Global Navigation Satellite System (GNSS) positioning technology, enabling rapid centimeter-level accuracy without dependence on proximate reference stations. A critical yet underexplored dimension in current scholarly discourse involves the management of atmospheric product transmission interruptions prior to successful ambiguity resolution in PPP-RTK user terminals. This study methodically confronts the pivotal issue of atmospheric product transmission disruptions in PPP-RTK user terminals. We deploy and scrutinize four distinct time series extrapolation techniques for atmospheric products Trend Moving Average (TMA), Double Exponential Smoothing (DES), Adaptive Filtering (AF), and Naive Forecasting (NF). The results reveal that atmospheric products extrapolated via the DES method demonstrate markedly superior accuracy relative to alternative approaches, particularly in Hong Kong, China, characterized by intense ionospheric activity. The precision of products generated by the DES method surpassed those of the AF, TMA, and NF methods by 12.2%, 1.5%, and 12.3%, respectively. Over a week-long assessment in Hong Kong, China, the DES extrapolation model enhanced the average three-dimensional RMS of positioning errors by 25.7%, 9.7%, 31.0%, and 38.5% compared to the AF, TMA, NF methods, and the non-extrapolation approach, respectively. Furthermore, the initial convergence times were reduced by 6.7%, 2.6%, 7.3%, and 25.8%, respectively. The insights derived from this research provide a robust solution for PPP-RTK user terminals during atmospheric product data transmission interruptions, significantly enhancing terminal resilience, particularly in low-latitude regions experiencing vigorous ionospheric fluctuations.