Temporal and Spatial Decorrelation Error Reduction by a Compact Network RTK

Byungwoon Park and Changdon Kee

Abstract:	The use of a Real Time Kinematic (RTK) enables position accuracy increase up to within several centimeters. The GNSS common errors are highly correlated in a small time-latency and from a short distance, so an RTK correction message generated from a reference station can easily cancel it out. A decorrelation error, however, becomes larger when the correction age gets older or the distance between the reference and a rover grows, so the single reference station (RS) should broadcast a RTK correction message every second within a 10-20km baseline. To reduce the temporal decorrelation error, SNUR-2000 v2.2 has been suggested as a RTK Compact Correction in RTCM Message, and it has shown that the accuracy degradation is negligible even under a low-rate (500bps) data-link. On the other hand, Network RTK is a maturing technology that can compensate for the spatial decorrelation error. Currently, there are several Network RTK method in the market including the Master- Auxiliary concept (MAC), Virtual Reference Stations (VRS), and Fl„chenkorrekturparameter(FKP). Each of Compact RTK or Network RTK provides a proper solution to either temporal- or spatial-related issue, but none of them deals with both decorrelation errors. This paper suggests a combination method of the Compact and Network RTK and shows it can reduce the temporal and spatial decorrelation error at a time. Network-based RTK methods generally use conventional observation-based messages for RTK correction such as RTCM version 3 Message Type 1004. These messages are so susceptible to the latency that the master reference station should broadcast them every second. The SNUR- 2000 protocol and its efficient transmission technique, which were designed to reduce the latency effect of a single RS RTK, are still helpful to the network RTK. In this paper, we prove its compatibility with all the generalized Network RTK methods theoretically. To verify this compatibility, we processed 1 hr rinex data from the CORS reference stations in North Carolina State. The system construction is the same as that of the normal network RTK, which consists of one master and several auxiliary stations. Using the GNSS raw observables and meteorological data transmitted from all the stations, the phase ambiguities of master station can be estimated and those from all stations in the network are reduced to a common level. The master station broadcasts Compact RTK correction message instead of the Type 1004 message, and network RTK correction messages are regularly transmitted in the RTCM-recommended format. No matter what network RTK method may be applied, our Compact Network RTK resolves the ambiguity of the carrier phase in 10-40 seconds and determines the position with 6~7cm horizontal and 7~8cm vertical error (95%) in the 50-60km radius-region. The Compact RTK Correction can make the MAC technique be available in even a 700 bps data-link, while conventional one should use at least 1900bps. FKP method using RTCM message type 1004 needs 2000bps data rate, but 600 bps is enough for it to be operated with the Compact RTK. In the case of VRS, moreover, we can reduce the baud-rate from 1400bps to 500bps without a severe degradation of accuracy. The Compact RTK Correction is not only compatible with general RTK network solutions, but also beneficial to them in the sense of data rate. We expect this compatibility and contribution of Compact Network RTK correction enable future RTK services to be wide, general, and efficient.
Published in:	Proceedings of the 2009 International Technical Meeting of The Institute of Navigation January 26 - 28, 2009 Disney's Paradise Pier Hotel Anaheim, CA
Pages:	341 - 352
Cite this article:	Park, Byungwoon, Kee, Changdon, "Temporal and Spatial Decorrelation Error Reduction by a Compact Network RTK," Proceedings of the 2009 International Technical Meeting of The Institute of Navigation, Anaheim, CA, January 2009, pp. 341-352.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In