Inside GNSS Media & Research

JUL-AUG 2019

Issue link: https://insidegnss.epubxp.com/i/1148308

Contents of this Issue

Navigation

Page 59 of 67

60 Inside GNSS J U L Y / A U G U S T 2 0 1 9 www.insidegnss.com FIGURE 12 UAV simulation environment with the Globalstar, Orbcomm, and Iridium LEO constellations. (a) LEO satellites' trajectories. (b) UAV trajectory and GPS cutoff location. Map data: Google Earth. FIGURE 13 UAV simulation results with the Globalstar, Orbcomm, and Iridium LEO constellations. (a)-(b) UAV simulated and estimated trajectories. (c) Simulated and estimated trajectories and the fi nal 95-th percentile uncertainty ellipsoid for one of the simulated LEO satellites. Map data: Google Earth. FIGURE 14 UAV simulation environment with the Starlink LEO constellation. (a) LEO satellites' trajectories. The elevation mask was set to 35 degrees. (b) UAV trajectory and GPS cut off location. Map data: Google Earth. FIGURE 15 UAV simulation results with the Starlink LEO constellation. (a)-(b) UAV simulated and estimated trajectories. (c) Simulated and estimated trajectories and the fi nal 95-th percentile uncertainty ellipsoid for one of the simulated LEO satellites. Map data: Google Earth. navigates over Santa Monica, California, USA, for about 25 kilometers in 200 sec- onds, during which it had access to GPS signals only for the fi rst 100 seconds. A er li -off , the UAV makes 4 banking turns. A total of 10 LEO satellite trajec- tories were simulated. e LEO satellite orbits corresponded to the Globalstar, Orbcomm, and Iridium constellations. e UAV made pseudorange and pseu- dorange rate measurements to all 10 LEO satellites throughout the entire trajectory. e LEO satellites' positions and velocities were initialized using TLE fi les and SGP4 propagation. Figure 12 shows the trajectories of the simulated LEO satellites and the UAV along with the location at which GPS signals were cut off . To estimate the UAV's trajectory, 2 navigation frameworks were imple- mented: (i) the LEO-aided INS STAN framework and (ii) a traditional GPS- aided INS for comparative analysis. Each framework had access to GPS for only the fi rst 100 seconds. Figure 13(a)-(b ) illustrate the UAV's true trajectory and those estimated by each of the 2 frame- works while Figure 13(c) illustrates the simulated and estimated trajectories of one of the LEO satellites, as well as the fi nal 95-th percentile uncertainty ellip- soid (the axes denote the radial (ra) and along-track (at) directions). Table 2 sum- marizes the fi nal error and position root mean squared error (RMSE) achieved by each framework a er GPS cutoff . B. UAV SIMULATION WITH THE STARLINK LEO CONSTELLATION WITH PERIODICALLY TRANSMITTED LEO SATELLITE POSITIONS A UAV was equipped with (i) a tactical- grade IMU and (ii) GPS and LEO satel- lite receivers. e UAV navigates over Santa Monica, California, USA, for about 82 kilometers in 10 minutes, dur- ing which it had access to GPS signals only for the fi rst 100 seconds. A er li - off , the UAV makes 10 banking turns. e simulated LEO satellite trajectories corresponded to the upcoming Starlink constellation. It was assumed that the LEO satellites were equipped with GPS receivers and were periodically trans- mitting their estimated position. ere was a total of 78 LEO SVs that passed within a preset 35° elevation mask set, with an average of 27 SVs available at any point in time. The UAV made pseudorange and pseudorange rate measurements to all LEO satellites. e LEO satellites' positions in the STAN framework were initialized using the fi rst transmitted LEO satellite positions, which were produced by the GPS receiv- ers onboard the LEO satellites. Figure 14 shows the trajectories of the simulated LEO satellites and the UAV along with STAN WITH LEO

Articles in this issue

Links on this page

view archives of Inside GNSS Media & Research - JUL-AUG 2019