Differential integer ambiguity resolution with Gaussian a priori knowledge and Kalman filtering

In this paper, a new maximum a posteriori probability estimation of ambiguities and baselines is proposed for differential carrier phase positioning. It performs a recursive least-squares estimation with an extended Kalman filter, that uses double difference code and carrier phase measurements and Gaussian a priori knowledge about the baseline length, elevation/ pitch angle and azimuth/ heading. The maximum a posteriori probability estimator finds the optimum trade-off between a solution that minimizes the range residuals and one which is close to the priori knowledge. It is shown that the Gaussian a priori knowledge enables a ten times faster convergence of the float solution, and it substantially suppresses multipath and, thereby, prevents divergence of float ambiguities and baselines. Moreover, the Gaussian a priori knowledge allows some errors in the a priori information, i.e. it is more robust than deterministic a priori knowledge.