GPS filtering to minimize ionosphere divergence error for safe aircraft landing

Aircraft use navigation guides to help pilots land especially in low visibility conditions. The commonly used Instrument Landing System is very effectively yet expensive and inflexible. The Global Positioning System (GPS) can be used to design potentially cheaper and more versatile navigation systems. The Local Area Augmented System (LAAS) is one of the proposed systems for navigation. However it is yet to be certified by the Federal Aviation Administration (FAA) owing to its susceptibility to certain rare events which can potentially cause a system failure in LAAS.; LAAS implements the idea of using differential corrections from a nearby reference station to increase the accuracy of the position estimate of the user. One of the biggest threats to LAAS, which undermines the concept of differential corrections, is an ionospheric storm caused by high solar activity. The ionosphere is a spherical shell of ionized gases over earth and affects radiowave communication passing through it. Although ionospheric storms are rare events, under certain worst cases they can be undetectable and might cause a missed landing. Therefore to ensure the safety of the usage of LAAS, this problem is addressed in this research.; Errors due to the ionospheric storm are magnified by the standard real-time smoothing filter in LAAS, the Hatch Filter. The Hatch Filter leverages the two components of GPS signals, namely the code and carrier measurement, and provides excellent attenuation of noise under nominal conditions. However under the influence of the ionospheric storm, the Hatch Filter suffers a delay in its position estimate and hence produces a systematic bias.; This thesis designs two alternative real-time smoothing filters with the motivation of increasing insensitivity towards ionospheric storms without sacrificing the "nominal day" noise attenuation. The first design uses mathematical rigor to optimize a general linear digital filter for the least maximum transient error under the influence of an ionospheric storm. The optimized linear filter is found to reduce the maximum transient error by a maximum of 20% during the offline simulation test conducted on GPS data. The second filter is designed with the motivation of reducing steady state errors caused by the ionospheric storm. This filter is nonlinear and is able to overcome the limitations of linear filters (which suffer degraded transient error while reducing steady state errors). In the offline experiments the nonlinear filter attained zero steady state error in each case and also reduced the maximum transient error as compared to the Hatch Filter.