Processing GPS carrier-phase measurements can provide high positioning accuracy for several navigation applications. However, if not detected, cycle slips in the measured phase can strongly deteriorate the positioning accuracy. Cycle slips frequently occur in areas surrounded by trees, buildings, and other obstacles. The dynamics experienced by the GPS receiver in kinematic mode of navigation also increases the possibility of cycle slips. Detection and correction of these cycle-slips is essential for reliable navigation. One way of detecting and correcting for cycle slips is to use another system to be integrated with GPS. Inertial Navigation Systems (INS), using three-axis accelerometers and three-axis gyroscopes, is integrated with GPS to provide more reliable navigation solution. Moreover, INS was utilized in the past for GPS cycle slip detection and correction. For low cost applications, Micro-Electro-Mechanical-Systems (MEMS) accelerometers and gyroscopes are used inside INS. For land navigation, reduced inertial sensor system (RISS) utilizing two accelerometers, one gyroscope, and the vehicle odometer was suggested. MEMS-based RISS has the advantage of using less number of MEMS-based gyroscopes and accelerometers thus reducing the overall cost and avoiding the complex error characteristics associated with MEMS sensors. In this thesis, we investigate the use of MEMS – based RISS to aid GPS and detect and correct for cycle slips. The Kalman filter was employed in centralized fashion to integrate the measurements from both GPS and RISS. This thesis research also offers a new threshold selection criterion resulting in a more robust cycle slip detection and correction. The proposed method was tested in different scenarios of road tests in land vehicle. Results show accuracy iii improvement over the conventional double differenced pseudoranges-based integrated system. Moreover, the adaptive selection criterion of the detection threshold proposed in this thesis improves the detection rate, especially in the case of small-sized cycle slips. / Thesis (Master, Electrical & Computer Engineering) -- Queen's University, 2013-05-06 18:11:57.076

Improving and predicting the accuracy of positioning estimates derived from the global positioning system (GPS) continues to be a problem of great interest. Dependable and accurate positioning is especially important for navigation applications such as the landing of commercial aircraft. This subject gives rise to many interesting and challenging mathematical problems. This dissertation investigates two such problems. The first problem involves the study of the relationship between positioning accuracy and satellite geometry configurations relative to a user's position. In this work, accuracy is measured by so-called dilution of precision (DOP) terms. The DOP terms arise from the linear regression model used to estimate user position from GPS observables, and are directly related to user position errors. An analysis of the statistical properties explaining the behavior of the DOP terms is presented. The most accurate satellite geometries and worst configurations are given for some cases. The second problem involves finding methods for detecting and repairing cycle-slips in range delay data between a satellite and a receiver. The distance between a satellite and a receiver can be estimated by measuring the difference in the carrier frequency phase shift experienced between the satellite and receiver oscillators. Cycle-slips are discontinuities in the integer number of complete cycles in these data, and are caused by interruptions or degradations in the signal such as low signal to noise ratio, software failures, or physical obstruction of the signals. These slips propagate to errors in user positioning. Cycle-slip detection and repair are crucial to maintaining accurate positioning. Linear regression models and sequential hypothesis testing are used to model, detect, and repair cycle-slips. The effectiveness of these methods is studied using data obtained from ground-station receivers.