Testing was performed at Ft. Hood Texas to quantify the performance of Geodetics Epoch-by-Epoch® software and products as used in an Army operational test environment.

Testing of RTD-Vector and the Epoch-by-Epoch® Attitude Determination System (EADS) at Ft. Hood consisted of mounting two sets of three antennas on the Ft. Hood pan-tilt unit. The test configuration is illustrated in the following figure:



The accuracy obtainable from both RTD-Vector and the EADS is a function of the antenna separation. The pan-tilt unit was designed such that the antenna separation can be precisely adjusted to 1, 2 and 3 meters allowing us to test all three baseline lengths under identical conditions (sans satellite geometry). The image below illustrates a pair of antennas (one feeding the EADS and the other feeding RTD-Vector) mounted on the pan-tilt unit when for the 3 meter baseline distance.


The pan-tilt unit was programmed to execute a sequence of pans and tilts with a short pause between each change to gather static data (note that the unit did not provide a roll capability). As show in the picture above, the unit started with a heading of approximately 128 degrees. With the unit fixed at this heading, the pan-tilt controller script affected the shifts in pan and tilt. There were three headings that the unit tested. The headings were approximately 128°, 218° and 7°. At each heading, the tilt was set to the following angles: 0°, 15°, 30°, 45°, 50°, 55°, 60°, 75°, 45°, -15°, -30°, -40° and back to 0°. There was a short pause between each change in tilt to record data.

Tests show that tilt angles beyond 55° are problematic. This is a result of loss of visibility to enough satellites to compute reliable solutions. When the antennas are angled at beyond 55°, a large section of the sky is not visible to the antennas, thus not enough satellites are visible.

If we look at the average of the standard deviations, excluding angles beyond 55°, we get a good idea of how the RTD-Vector and EADS products performed at a 3 meter baseline separation. Results are 0.12, 0.11 and 0.18 degrees for heading, pitch and roll, respectively.