Abstract:
Ambiguity Resolution (AR) is a non-trivial procedure; it becomes more important if it is aimed
at characteristics like robustness, and high performance over varying conditions such as
observation span, baseline length, or elevation mask. It is because of this, lots of research has
been done in this field in the 1990's and still going on. This is due to the vast applications of
global positioning system (GPS). India is trying to get itself into the group of nations having
indigenous satellite based "Air Traffic Control". For such application it is prerequisite to have
satellite telemetry data processing modules which can give flexibility to access satellite based
positioning system data processing procedure at every step so that intermediate results can be
analyzed and thus the positioning accuracy of ground station can be enhanced. In this study, with
a vision of developing an indigenous GPS data processing module, a step is taken by performing
ambiguity resolution in GPS data using different AR schemes to test their performance,
especially under challenging scenarios like very low observation span, low elevation mask etc. In
this study, ROUND, Least Squares Ambiguity Decorrelation Adjustment (LAMBDA), and Fast
Ambiguity Resolution Approach (FARA) are implemented on Matlab. Raw GPS data was
acquired and was brought to Matlab environment and processed. The above mentioned methods
are employed on this processed data and then the results are compared. Interpretation of results
revealed that as observation span reduces, performance of these methods decreases drastically
for desired accuracy. Comparison is also made on the basis of other predefined scenarios. The
study reveals that LAMBDA method is best on data recorded under low elevations because it
removes the correlation between ambiguities hence isolating the ambiguity affected by multipath
whereas ROUND and FARA are unable to mitigate this problem.
