Results of testing the prototype of high precision GPS attitude determination system and relative position determination of close approaching/moving apart objects

For carrying out the experimental researches* (testing) of the developed prototype of high precision GPS attitude determination system (ADS) of moving objects it has been created the angular instrument prototype (fig. 1). In the given case it has been realized the version of ADS with use of independent GPS receivers with free-running oscillators (clocks).


layout angle measuring system based on GPS

Figure 1. Prototype of GPS attitude determination system (ADS), baselines: 1.546 m (2-1), 1.542 m (4-1), ~2.166 m (3-1); observation interval duration while testing: ~5 hours


Structure of ADS prototype:

  • support-rotating mechanism;
  • 4 receivers – two dual frequency receivers NovAtel DL-V3 receiving antennas – NovAtel-702GG (1) and Antcom G5Ant-52AT1 (3)) and two single frequency receivers NovAtel OEM6 with the appropriate receiving antennas Tallysman TW3430 (2, 4);
  • GPS observations recorder (laptop);
  • power supply;
  • NovAtel receivers support software;
  • domestic software «OCTAVA» for GPS observations post-processing and analysis;
  • the developed special software for ADS of moving objects.

In addition to the task of attitude determination within the framework of the performed works it has been considered and solved the task of relative high precision positioning/navigation (using carrier-phase GPS observations) of close approaching/moving apart objects (so called «moving base» mode) on baselines between the objects up to ~200 m (observation interval made up ~40 minutes).

While testing, for comparison and quality control positioning in the kinematic mode for the purpose of solving of both given tasks there have been also used the observations of SURE permanent reference GPS station (Kharkov, KhNURE).

The main tasks of the experimental researches (testing):

  • determination (in single frequency and dual frequency measuring modes) of minimal intervals of observations (so called TTFF – Time-To-First-Fix), at which it is provided the reliable integer carrier-phase ambiguity resolution (fixation) (CPAR) at kinematic determinations – both in the task of attitude determination and also in the task of relative navigation while close approaching/moving apart of the objects;
  • estimation of attitude determination accuracy (in conditions of testing) and determination of relative coordinates of moving objects.

Main results of testing


1. According to the researches, in case of use of dual frequency observations the task of carrier-phase ambiguity resolution is reliably (with frequency of 100%) solved for every separate epoch on determining the vectors of bases both in the task of ADS and in the task of precision relative navigation in «moving base» mode. Along with that there is no need to use the constraints in the task of ADS, when there are known (or previously determined with millimeter accuracy) the baselines between the phase centers (PC) of ADS receiving antennas.

2. In case of using only single frequency observations for the reliable solution of the task of carrier-phase ambiguity resolution it is required not less than 7 minutes, since in conditions of testing there prevailed a strongly time correlated multipath error component of code and carrier phase GPS observations. For improving the reliability of single-frequency observations carrier-phase ambiguity resolution there were used the constraints – the information about precision baselines between PC of the antennas. The baselines were determined experimentally in the process of the preliminary calibration. Table 1 presents the probabilities (frequency) of correct fixation of integer carrie-phase ambiguities by use and without using the constraints.


Table 1 – Success rate (reliability) of single frequency observations CPAR

Sample length T (sec/min) Probability (frequency) of correct fixation of integer carrier-phase ambiguities, %
Without constraints (L1 antennas Tallysman) With constraints (L1 antennas Tallysman) Without constraints (L1L2 antenna Antcom) With constraints (L1L2 antenna Antcom)
5 сек - - 82 91
30 sec - - 90 95
60 sec 47 74 91 100
180 sec (3 min) 74 83 95 100
300 sec (5 min) 74 100 95 100
420 sec (7 min) 85 100 100 100
600 sec (10 min) 100 100 100 100

The features of currently used receiving antennas considerably influence the carrier-phase ambiguity resolution reliability. So, in table 1 there are given the results for two different antennas: single frequency antenna Tallysman TW3430 and dual frequency antenna Antcom G5Ant-52AT1 (as an antenna of the reference/first point of the ADS prototype there was used the dual frequency antenna NovAtel-702GG). It is clear that in the second case the carrier-phase ambiguity resolution reliability is appreciably higher even without using the constraints. This is caused by the difference of the receiving antennas «Antcom» and Tallysman in part of the level of multipath mitigation.

3. Table 2 gives the accuracy characteristics (RMS) of ADS prototype which were obtained in the course of experiments.


Table 2 – Accuracy (RMS) of attitude determination

Yaw Roll Pitch
RMS, angular degrees 0,06 0,3 0,28

4. When solving the task of relative kinematic determination/navigation of close approaching objects («moving base» mode) the positioning errors (RMS) did not exceed 10 mm for all the three coordinates on baselines ~200 m during all the session of measuring (~40 min.).

5. Further researches.

In the given case there was implemented and researched the variant of ADS by use of separate GPS receivers with free-running (nonsynchronized) clocks. Such variant of creation of the system by GPS/GNSS observations requires the joint estimation of the current vectors of bases and relative receivers’ clock devergence. This approach has got both the drawbacks (extended state vector) and certain advantages. This fact is discussed in the appropriate scientific literature.

Another approach which is expedient to be realized and researched further for comparison, implies the creation of the attitude determination specialized system where it is formed one time scale for separate receivers at hardware level (usually several OEM receiver modules are united on one board). In this case there are estimated only current vectors of the measuring bases, and carrier-phase ambiguity resolution is carried out for single but not for dual differences of carrier-phase observations. In case of such implementation of the system it may be expected the improvement of reliability characteristics of carrier-phase solutions and increase of accuracy of attitude parameter determination. But along with that there arises the problem of high precision calibration of carrier-phase delays differences in measuring channels of the system. This problem does not occur in the first variant of ADS.

There are also another approaches of improving the carrier-phase solutions and increasing the accuracy of attitude determination. There is of interest the implementation of the approach by which carrier-phase ambiguity resolution is carried out simultaneously for all the baselines of the system, but at that there are taken into account all apriori linear-angular constraints – information about baselines and angles between the vectors of the bases.


Download the express-releas [En]