Navigation filters for use in the Global Positioning System

This dissertation explores the nonlinear noise properties of pseudorange measurements in the Global Positioning System (GPS) while presenting several new approaches to pseudorange-only navigation.; Nonlinear noise associated with pseudorange multipath effects in the GPS is complex and typically poorly modeled in classical approaches. In this dissertation, real-world multipath test cases are presented and new navigation filters are developed to reduce noise and improve navigation accuracy.; The first navigation technique involves optimal linear modeling which determines the system equations at the receiver's operating point in space; as opposed to traditional approaches that linearize about equilibrium points. The result is a new approach to navigation that uses an adaptive Kalman filter to produce superior solutions.; The second approach builds the same optimal linear model and adds measurement differencing to achieve improved accuracy. The result is a navigation solution that is less susceptible to nonlinear noise.; Finally, rational model estimation theory is applied to pseudorange measurements to identify the pseudorange measurement structure. A single pseudorange model is constructed such that it is dependent upon additional pseudorange measurement inputs from other observed GPS satellites. The generated pseudorange is then used for navigation instead of the actual pseudorange measurement. The result is a navigation solution that is less susceptible to nonlinear noises.