Performance Assessment of a Long Range Reference Station Ambiguity Resolution Algorithm for Network RTK GPS Positioning

Abstract

The authors firstly propose a new ambiguity resolution (AR) method for processing long range reference station baselines in Network Real-Time Kinematic (NRTK) GPS positioning. This method has the following advantages: no need for observational equation linearization and resolution, satellite by satellite based ambiguity resolution and long range baseline ambiguity resolution of up to a few hundred kilometres. In order to assess the performance of this method, several sub-networks of Ordnance Survey's OSNetTM in which the baseline lengths vary from 200km to 400km are formed. The corresponding initialization times and Double-Differenced (DD) biases are estimated. The results show that even if the inter-station distances are up to 400km the proposed long range AR algorithm can quickly and accurately resolve the ambiguities. According to the performance of ambiguity resolution and the character of DD biases between reference stations, the authors conclude that it is possible to implement highly accurate NRTK GPS positioning based on much sparse OSNetTM with inter-station distances up to 400km and this will significantly reduce the investment in the establishment and maintenance of a network of reference stations.