Gravity-Aided Navigation Technology for Reducing Ballistic Missile Submarines’ (SSBN) Dependence on the Global Positioning System (GPS)

Gravity aided navigation has been explored previously using Kalman filter methods by various researchers. A Kalman filter based implementation provides best estimates under the assumption of Gaussian noise properties and linearity of the system dynamics. Typically, the geophysical aided navigation applications are non-Gaussian and non-linear. We propose to develop a real-time particle filter based mechanization for terrain matching of gravitational data with platform INS data in order to provide a cost-effective high-accuracy navigation solution applicable over extended, perhaps indefinite, endurance intervals. Such a mechanization is not limited to the requirements of Gaussian noise properties, as the particle filter based approach generates and updates an effective probability density function (PDF) from the input data. Also, since the PDF is not assumed Gaussian and in fact is generated and updated in the course of processing, this particular approach is applicable to both the legacy ESGN system and future alternative to ESGN (possibly FOGN) system whose errors will predominantly be driven by gyro angle random walk (ARW) errors.