Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #1

RAPID STATIC AND KINEMATIC POSITIONING WITH THE HONG KONG GPS ACTIVE NETWORK:
W. Chen, C. Hu, Y. Chen, X. Ding, Hong Kong Polytechnic University; Simon, C. Kowk, Geodetic Survey Section, Lands Department, Hong Kong

The Lands Department of Hong Kong is establishing a permanent GPS array of reference stations for various applications. It collects GPS data continuously from multiple reference stations and delivers quality-checked data to the users. The well-positioned and dense array of the reference stations enables the users to achieve cm-level accuracy within a short period of time even using only one low-cost single frequency GPS receiver. It also supports meter-level navigation and transportation management applications using DGPS correction. The array of permanent GPS stations can also be used for scientific research such as deformation monitoring and weather forecasting. The station spacing for the active network is about 10 to 15 km and there will be 13 to 14 stations covering the whole territory.

This paper describe an algorithm for GPS rapid static positioning and kinematic positioning based on multiple reference GPS network. First the phase integer ambiguities in the active network are solved epoch by epoch by making using of the known coordinates of the reference station. Then a Kalman filter method is used to process the rover (or stationary) data together with its adjacent reference stations with both carrier phase and code data. It is based on the network adjustment (multi-baseline) and the correlation among baselines is taken into account. Moreover the double difference residuals among the reference stations are also used as a quality control factor to check the observation errors for each satellite at each epoch. It is demonstrated that with the Hong Kong GPS Active Network the observation time for ambiguity resolution can be reduced significantly. More importantly, improvement has gained on the liability of the integer ambiguity and the precision of the baseline.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #2

RTK-GPS POSITIONING IN JAPAN BY VIRTUAL REFERENCE STATION (VRS) SYSTEM WITH GPS-BASED CONTROL STATIONS:
H. Namie, N. Hagiwara, National Defense Academy of Japan; H. Kim, FuiYoSe Co., Ltd; N. Nitta, AD NET INC.; Y. Shibahara, Mitsubishi Electric Co., Japan; T. Imakiire, Geographical Survey Institute of Japan; A. Yasuda, Tokyo University of Mercantile Marine, Japan

The accuracy of a few centimeters can be easily obtained in 2drms by RTK-GPS positioning, as the range from a user GPS antenna to the satellite is determined by the phase measurement of the carrier waves.

It is necessary, however, to provide the carrier phase data from a reference station located at a known position, to a user receiver. It is hard to prepare expensive GPS receiver for RTK-GPS. RTK-GPS users in Japan have desired the development of a flexible, wider and less expensive carrier phase data dissemination service.

While, Geographical Survey Institute (GSI) of Japan has established the network of as many as 947 GPS-based control stations covering all over the Japanese Islands, in order to observe crustal movement for seismic prediction and research.

The service for RTK-GPS carrier phase data dissemination has been operated experimentally since October in 1997 from several GPS-based stations. And since December in 2000 has been operated experimentally by Virtual Reference Station (VRS) system using a several stations, known as FKP (flachen-korrectur-parameter; Germany) system in Tokyo area.

The authors have investigated RTK-GPS positioning at a fixed point in National Defense Academy (NDA) of Japan and in Tokyo University of Mercantile Marine, both out of the VRS network area. First, the carrier phase data from several GPS-based stations are transmitted to the data control center. And they are converted to FKP data there. And the data are transmitted to user receiver via a cell phone with data rate of 9,600 bps. The authors evaluated their validity by applying the data to RTK-GPS positioning.

First, the authors measured the data delay of the VRS system before positioning. The system fixes the position from a combination of the data transmitted by the reference station and those by user receiver collected at the same time.
The delay is measured to be about 1 s with 0.08 s in S.D. The data delay of portable phone is about 0.25 s that is included in the total delay. Sometimes the delay exceeds 8 s by the trouble of the cell phone. It is confirmed 2drms and S.D. in altitude are smaller than 2 cm respectively.

The nominal positioning accuracy of ordinal RTK-GPS is 2 cm plus the length of baselinetimes 2ppm, within a radius of about 10 km from the reference station. Although the sites to which base lines length from the nearest GPS-based station exceeds 30 km and locate out of GPS-based station network area for VRS, the remarkable accuracy is obtained comparable to the conventional RTK-GPS positioning.

Next, the authors measured the time to fix the ambiguity. They put off the GPS antenna cable from GPS receiver and then put it to the receiver again. The time is measured to be about 2 minutes to recover for RTK-GPS fix.

Further, the authors tried positioning onboard a ship under navigation. They measured the time to fix the ambiguity after the ship passed below the Rainbow Bridge. The time is measured to be about 7 to 12 minutes to recover RTK-GPS fix positioning.

These results show that the present VRS experiment system operates satisfactorily even outside of the area.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #3

SUBSIDENCE MONITORING SYSTEM USING REAL-TIME GPS SENSORS:
K. Borre, Aalborg University, Denmark; K. de Jong, Delft University of Technology, Netherlands; C. Pichot, THALES-Navigation, France

Today GPS observables are so accurate that it makes sense to investigate if they can be used for subsidence monitoring at the mm-level.

This challenge, of course, asks for using sophisticated filters and data handling. Most systematic errors are eliminated to some extent by using relative positioning techniques.

The difficulties in ionospheric and tropospheric modeling set certain limits for the extension of the system: all baselines should be short and the height differences small.

The ultimate difficulty is to cope with multipath effects. We exploit the well-known repetition of the satellite constellation every 24 hours. Efficient filters model this error source down to sub-millimeter level. Hence we have a system that works at a mm-level both within a 24 hours period and over weeks.

The software has both an initialization and a monitoring mode. The initialization period takes 24 hour, monitoring takes place in the 24 hour periods following the initialization by comparing the results from these periods with those from the initialization periods. Since the orbital repeatability is not exactly 24 hours, the software determines the optimal repetition period. Also, it takes care that the same satellite configuration is used when comparing corresponding epochs.

The software runs under MS Windows (tm). The number of receivers for which data can be processed simultaneously is limited only by the capacity of the data links. The developed software builds on the Scorpio receiver from THALES-Navigation. We believe the patented method makes a break-through for a new precise application of GPS.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #4

RTK NETWORKS BASED ON GEO++R GNSMART - CONCEPTS, IMPLEMENTATION, RESULTS:
G. Wuebbena, A. Bagge, Geo++ GmbH, Germany

The accuracy of todays RTK is limited by the distance dependent errors from orbit, ionosphere and troposhere. The basic idea of GN-SMART (Geodetic Navstar - State Monitoring And Representation Technique) is to analyse the data from a reference station network and to estimate the state of the GPS error budjet components in real time. All stations of a network are processed simultaneously for best estimation of global paramteres and to increase the reliability of the results.

The complete state is normally too complex to be used by the rover directly. Therefore GNSMART can derive several types of representations of the complete model, adequate for special transmission or rover requirements to reduce the distance dependent errors significantly. The implementation was operable before the current solar activity maximum, and is currently installed in many reference stations around the world under different ionospheric conditions. Recent results show the capabilities of GNSMART. Horizontal accuracy of 1 centimeter can be achieved with initialization times of 30 seconds, often even within 10 seconds.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #5

THE ISSUES OF PRACTICAL IMPLEMENTATION OF THE COMMERCIAL RTK NETWORK SERVICE:
I. Petrovski, S. Kawaguchi, H. Torimoto, DX Antenna Co.Ltd., Japan; K. Fuji, K. Ebine, Hitachi Ltd.; G. Lachapelle, M.E. Cannon, The University of Calgary, Canada

The paper presents a commercial RTK VRS Network service in Japan and issues related to its implementation. The infrastructure for commercial RTK correction service was introduced for public service this year. We discuss the user segments the system aims, the segment's size, user requirements in each segment and potential competitive systems for each segment in Japan. The concept for future development of the system is presented regarding Internet based globalization, GALILEO and GLONASS utilization, use of the predicted ephemeris.

The approach for choosing an algorithm for RTK corrections calculations for the RTK Network service is described. The guidelines for infrastructure development, number and location of reference stations, considerations for the different data link utilization, including broadcast service, cellular phone and the Internet are presented.

In conclusion, the demands that this system and others alike bring are discussed including those to the RTCM standards, and Internet protocols.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #6

STUDY OF A SIMPLIFIED APPROACH IN UTILISING INFORMATION FROM PERMANENT REFERENCE STATION ARRAYS:
H.J. Euler, Leica Geosystems AG, Switzerland; G. Wuebbena, Geo++ GmbH, Germany

Permanent reference station networks are being world wide under development for surveying type applications targeting cm-accuracies. While the research is already being carried out for quite some time, researchers have identified the obstacles in using single permanent reference stations with rovers in the traditional baseline modes. The advantages emerging by utilising the information out of reference station arrays such as modelling remaining tropospheric, ionospheric and orbit biases have been discussed. Methods and concepts show the improvements in performance and reliability in some kind of closed system approaches. Activities such as standardisation discussions within RTCM are under way to target the interoperability between these reference station systems and roving receivers from various manufacturers. One obstacle in the discussion and therefore in later interoperability are the opening and proper description of models used for deriving the models for the above-noted biases. This difficulty has to be mitigated and will vanish with time, but interoperability is needed urgently. The paper studies a different approach to utilise and distribute the information out of permanent reference station arrays in compact messages. The separation of different calculation tasks provides an easier way to describe an efficient and easy to standardise way of transferring and distributing the information. The proposed way may solve the current dilemma for interoperability standards.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #7

A STUDY OF GPS/GLONASS MULTIPLE REFERENCE STATION TECHNIQUES FOR PRECISE REAL-TIME CARRIER PHASE-BASED POSITIONING:
L. Dai, S. Han, J. Wang, C. Rizos, University of New South Wales, Australia

Real-time high precision GPS/GLONASS surveying and navigation applications have been constrained to the short-range case due to the presence of distance-dependent biases in the between-receiver single-differenced observables. Over the past few years, the use of a GPS reference station network approach, to extend the inter-receiver distances (user-to-reference receiver separation), has shown great promise. In order to model the distance-dependent residual biases, such as the atmospheric biases and orbit errors, several multiple reference station techniques -- including Partial Derivative algorithm, NetAdjust method, Linear Combination method, Distance-Based Linear Interpolation algorithm and Virtual Reference Station method -- have been developed. All of these methods aim to model the distance-dependent biases between the base station(s) and the user receiver with the support of a multiple reference station network.

In this paper these techniques are reviewed in detail. The advantages and disadvantages of each of these technique are described. The relationships among these methods will also be presented. Based on these analyses, a new method is proposed. The distance-dependent biases have been separated into the frequency-dependent errors (ionospheric bias) and frequency-independent errors (e.g. troposphere bias and orbit bias). The separate estimates of the two types of errors which are generated from the carrier-phase measurements using the multiple reference stations can be used to model the user distance-dependent biases for L1, L2 carrier phase and pseudorange measurements in different ways. Therefore, pseudorange correction terms don't need to be generated from the network station pseudorange measurements. As a result, the accuracy of the distance-dependent bias estimate for the pseudorange measurements can be significantly improved, and the volume of correction term data to be broadcast can also decreased.

Test data from GPS (and GLONASS) reference stations was used to evaluate the performance of this proposed method, compared to existing methods. The numerical results show that the proposed methodology can significantly reduce the distance-dependent biases of carrier-phase and pseudo-range measurements at the GPS user station. The efficiency of single-epoch ambiguity resolution is also improved.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Paper #8

IMPROVING A MULTI-REFERENCE GPS STATION NETWORK METHOD FOR OTF POSITIONING IN THE ST. LAWRENCE SEAWAY:
L.P. Fortes, M.E. Cannon, S. Skone, G. Lachapelle, University of Calgary, Canada

Differential GPS has the capability to provide cm-level positioning accuracy, as long as the carrier phase ambiguities are resolved on-the-fly (OTF) to integer values. Former existing methods were based on the use of a single fixed reference station located in the vicinity of the rover. The maximum distance allowed between the reference station and user in this case is generally limited to within 50 km due to effects from the atmosphere and orbit. The number of reference stations can be increased to extend the coverage. However, the installation and maintenance of such stations and their data transmission links is operationally complex and expensive. The optimal use of existing stations is therefore of utmost importance to maximize cost effectiveness. A novel and unique method was developed at the University of Calgary, which uses all available reference stations to generate regional code and carrier phase corrections, which can be transmitted to the user in order to resolve integer ambiguities OTF over the region. One of the major advantages of this method is to increase the coverage under which successful OTF ambiguity resolution is possible. This method has been tested using several data sets collected under various atmospheric conditions in the world. The improvement brought by the method, very good in practically all cases, depended on the geometry of the network coupled with the atmospheric conditions. Further research is being developed at the University of Calgary towards optimizing the method in order to maximize the improvement obtained in the observation, position and ambiguity domains. This new approach, also using least square collocation, separately models the errors into ionospheric, tropospheric and satellite orbital components. A major effort has been carried out in terms of modeling the ionosphere in directional components, which has shown to be relevant under high ionospheric conditions. Results of the enhanced method are presented using data collected in the St. Lawrence Seaway region, Canada, under such high differential ionospheric effects.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Alternate #1

ACCURACY CONTROL OF DGPS IN RTK POSITIONING WITH RTCM STANDARD TROUGH INTERNET PROTOCOL AND GPRS:
M. Caprioli, A. Scognamiglio, Politecnico di BARI, Italy

A great number of permanent station GPS are installing by many private operators and institutional subjects in Italy and all over the world, realizing a very interesting network that can be used as base stations in RTK positioning where the user can reach accuracy of differential GPS with only one receiver.

Many studies are realized on transmission of data whit radio waves and with GSM, and now we'll test the RTCM transmission with Internet protocol.

The permanent station on "Politecnico di Bari" will be connected to the Internet and RTCM data transmitted with IP protocol using a Linux server trough the 2101 port assigned by IANA (Internet Assigned Numbers Authority) to this function.

The Rover station (GPS receiver Trimble 4700) connected to Internet receives data with a client program on Pc or PDA linked to the GPS receiver, that can operate the necessary correction in RTK positioning.

The paper presents results from practical tests realized in Puglia (Italy) in order to verify accuracy of this system of positioning and transmission comparing with the previous experiences of data transmission with GSM link at different distances from permanent station.

When the telephone companies will be ready, a new test with GPRS protocol of GSM standard we shall realizing an Internet link "always on" with a very interesting saving of money.

Finally we will want to verify until which distance can be used this method maintaining a sufficient accuracy and how can be used many permanent stations of GPS Network in order to realize one Virtual Reference Station.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Alternate #2

THE EVALUATION OF THE ACCURACY AND MEASUREMENT OF LATENCY OF INTERNET-BASED DGPS AND RTK-GPS POSITIONING:
Z. Liu, A. Yasuda, Tokyo University of Mercantile Marine, Japan; C. Fan, Tokyo University of Fisheries

The differential GPS techniques, including DGPS and RTK-GPS, are indispensable when highly accurate service is needed, although the accuracy of GPS positioning has been improved substantially by turning off the Selective Availability which was the largest source of error for GPS receivers before. For the DGPS and RTK-GPS positioning system, how to transmit the correction data is very important. We use Internet to transmit correction data. Internet has lots of merits such as convenience, global covering and so on. However, one of the critical characteristics of Internet is the unexpected transfer latency, which will have a great affect to the DGPS and RTK-GPS positioning accuracy. This paper describes the transfer latency of GPS correction data via Internet and the results of Internet-based DGPS and RTK-GPS positioning.

We measure the transfer latency via Internet of GPS correction data. TCP protocol and a client-server request-answer model are used. We propose a model to measure the data transfer time corresponding to this reliable connection-oriented protocol. In this model, the client computer, which assumes the data transmissions to the rover, is connected to Internet via a 56 kbps modem and telephone line in the pertinent experiments. The server computer is assigned with global IP address 192.244.150.144. The software employed in the experiments is written in the C programming language and use the Winsock API functions.

The precision of the measured results is 10ms. The length of DGPS correction data to be transmitted is 520 bits, and that of RTK-GPS is 3824 bits. The average transfer latency, including the time lag caused by modem and telephone line, is about 310 ms; the average transfer latency of RTK-GPS correction data is about 425 ms.

In order to exclude the errors caused by other elements which have the affects to positioning accuracy, two RT-2 receivers are used for the reference and the rover with zero base line. The twice drms of Internet-based DGPS positioning is within one meter and that of Internet-based RTK-GPS positioning is within one centimeter. They are of the same order of accuracy compared with the conventional DGPS/RTK-GPS positioning results.
[Back to Program]

Session D1: NETWORK-BASED TECHNIQUES FOR RTK APPLICATIONS
Alternate #3

NETWORK AIDED RTK SYSTEM:
M.P. Hyun, G.J. Soo, Y. Cho, P. Sungjin, ELEXTech, Inc., Korea

RTK system have provided ambiguity resolution on the fly. This paper has purpose to improve performance of RTK system using Network assistance.

In GPS carrier phase integer ambiguity search method, the problem is that the time to fix ambiguity resolution is long. The receiver being search integer ambiguity, the Network could have used as assister. In this paper, The Navtron G2000 receiver of ELEX Tech, using VS_GPS of VLSI Solution Oy have been used.
[Back to Program]