GNSS Signal Tracking Using a Bank of Correlators

A combined PLL/DLL algorithm is developed for tracking GNSS carrier phase and code phase using the output from a large number of correlators. This approach has advantages for limb-scanning applications, in which useful meteorological information, available only at the initial rising time of a GPS satellite, is desired. The technique uses the bank of correlators to span wide ranges of uncertainty in code phase and carrier phase, thereby avoiding the need for a separate acquisition and the associated loss of an initial span of data. A fusion of optimal estimation methods processes the output from these correlators. A batch optimization of a signal model’s fit to the many accumulations provides the Kalman filter with “measurements” of the most likely signal parameters, and the Kalman filter utilizes a signal dynamics model to provide estimates that drive the PLL and DLL. The effectiveness of this algorithm is demonstrated by using a truth-model simulation of a limb scan. With this method and 50 Hz accumulations, pull-ins of at minimum 122 Hz and 5 C/A code chips, for the PLL and DLL respectively, are achieved. Moreover, uncertainties and errors for all signal parameters are rapidly driven close to zero. The rising signal is tracked successfully from approximately 0.03 s since emergence.