Effect of stochastic ionospheric delay modeling for GPS ambiguity resolution

Abstract

The network-derived ionospheric delay can improve the fast and accurate determination of the long baseline in both the rapid-static and kinematic Global Positioning System (GPS) positioning mode. In this study, an interpolation of the undifferenced (UD) ionospheric delays is performed on a satellite-by-satellite and epoch-by-epoch basis, respectively, using the least-squares collocation (LSC) to provide not only ionospheric delays but also their variances. The developed method retains the simplicity of the two-dimensional (2-D) model, but it does not introduce errors due to the thin-shell assumption made in the single-layered model. Our method also provides flexibility in forming the predicted double-differenced (DD) ionospheric delays. Faster and more reliable positioning solutions can be obtained when the developed method is used to predict DD ionospheric delays. The numerical test applying the method to the Ohio Continuously Operating Reference Station network shows a 23% improvement in mean time-to-fix with the network-derived ionospheric delays.