Absolute and relative positioning accuracy of static receivers using NavIC

Abstract

Our objective in this thesis is to assess the navigation accuracy of stationary re ceivers using NavIC (Navigation with Indian Constellation). All errors in NavIC’s pseudoranges are investigated so that they can be mitigated or accounted for ap propriately to achieve high-accuracy navigation. Multipath error in the L5 and S1 pseudoranges are different; hence, oscillatory multipath is removed using sidereal re peatability at these frequencies separately for accurate iono delay estimation. Due to multipath, the iono delay estimates using uncompensated pseudoranges have as much as 7 m difference from the nominal night time delay of 1-3 m. When oscillations are removed, the difference reduces to ≤ 3 m. Dual-frequency iono delay estimates using code phase are within 2σ of the SBAS (Space Based Augmented System) grid based estimates. A comparison of Global Neustrelitz TEC Model (NTCM-GL) and Klobuchar is presented for a low latitude region. The time-varying frequencies and periods of multipath errors caused by a single NavIC signal path reflected from the ground are shown as functions of signal elevation angle and 24 hour orbital period. For the GEO satellites the single reflected path frequencies are as low as 0.1 milliHz, and for the GSO (geosynchronous orbit) satellites, the frequencies vary from -1.6 to 1.6 milliHz. The corresponding non-sinusoidal and biased short-delay multipath oscillation periods thus vary from less than an hour to several hours. The multipath error caused by reflections from the neighboring walls, however, depend on both elevation angles and relative azimuth angles of the reflections, and therefore their frequencies are higher and periods shorter than those of the ground reflections. Furthermore, since the elevation and azimuth traverses of the GSO satel lites are far wider than those of the GEO satellites, the errors of the GSO signals have much higher frequencies and variation in amplitudes and are less orderly than the errors of the GEO signals. In addition, since the wavelength of the S1 signals (0.12 m) is about half of the L5 signals (0.25 m), the S1 multipath frequencies are two times those of the L5 multipath frequencies. Also, the real multipath errors in the seven pseudoranges are caused potentially by several surrounding reflectors, so the periods of the real multipath errors are relatively shorter than what is stated above, have a spectrum, and are less predictable.