Algorithms for Sparse Network-based RTK GPS Positioning and Performance Assessment

The average inter-station distances in most established network Real Time Kinematic (RTK) systems are constrained to around 50 km. A sparse network RTK system with an average inter-station distance of up to 300 km would have many appealing advantages over a conventional one, including a significant reduction in the development and maintenance costs. The first part of this paper introduces the key approaches for sparse network RTK positioning technology. These include long-range reference baseline ambiguity resolution and real-time kinematic ambiguity resolution for the rover receivers. The proposed method for long-range kinematic ambiguity resolution can overcome the network weaknesses through three procedures: application of the interpolated corrections from the sparse network only to wide-lane combination; searching the ambiguities of wide-lane combination; and searching L1 ambiguities with wide-lane combination and ionosphere-free observables. To test these techniques, a network including ten reference stations was created from the Ordnance Survey's Network (OS NetTM) that covers the whole territory of the United Kingdom (UK). The average baseline length of this sparse network is about 300 km. To assess the positioning performance, nine rover stations situated inside and outside the network were also selected from the OS Net™. Finally, the accuracy of interpolated corrections, the positioning accuracy and the initialization time required for precise positioning were estimated and analysed. From the observed performance of each rover receiver, and the accuracy of interpolated network corrections, it can be concluded that it is feasible to use a sparse reference station network with an average inter-station distance up to 300 km for achieving similar performance to traditional network RTK positioning. The proposed approach can provide more cost-efficient use of network RTK (NRTK) positioning for engineering and environmental applications that are currently being delivered by traditional network RTK positioning technology.