MLM Technique for Measuring Airborne GPS Accuracy with Ranging Instruments

Theodore A. Miller

Abstract:	GPS/INS systems, especially with modern RTK and post-processing methods, are the most accurate systems available to measure the location and attitude of airborne vehicles. While this is a good circumstance for the GPS industry, it becomes problematic when someone asks “How accurate is it”? It is problematic because there is no good truth source to measure the accuracy of this most-accurate instrument. Currently, GPS accuracy is typically best estimated using external truth sources while the device is on the ground, or with side-by-side comparisons to other GPS systems flying alongside the test unit. Neither of these methods is satisfactory. Consequently, in this paper, a method is described which will measure airborne GPS accuracy using a set of reasonably accurate external systems (ranging laser and radar) and a statistical methodology to maximize their potential. The ranging instruments at ATR are good candidates for the external instruments and an extension of the Maximum Likelihood Method (MLM) provides a statistical method. Using these instruments in this manner exploits the difference in the native geometry of ranging systems vs. GPS. This combination is probably the best available option for airborne testing of GPS accuracy. MLM is a method that is commonly used to define the best estimate for the µ and s parameters of a normal (Gaussian) distribution, i.e. the arithmetic mean and standard deviation respectively. In this GPS context, MLM can take advantage of the geometrical differences of GPS vs. ranging instruments by explicitly referring to the geometry in the probability density function that initiates MLM. The inclusion of geometry eliminates the possibility of an analytic solution, but a computerized minimization routine can readily extract the accuracy parameters. Additionally, statistical sampling methods can extract a quantitative factor so that the measurement engineer can assess the quality of the MLM estimate. This paper briefly describes MLM applied to a test flight that had a track that, while not ideal for MLM, was good enough for an initial experimental verification of the method. The result demonstrates a significant systematic error that could not be seen with typical measurement techniques.
Published in:	Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011) September 20 - 23, 2011 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	2004 - 2008
Cite this article:	Miller, Theodore A., "MLM Technique for Measuring Airborne GPS Accuracy with Ranging Instruments," Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011), Portland, OR, September 2011, pp. 2004-2008.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In