Filtering Problem for Ambiguous Phase Measurements

摘要

The problem of recurrent parameter estimation for a linear discrete dynamic system is considered under the conditions that some of the measurements are the phase measurements and are read ambiguously. The cycle slips are assumed to occur between arbitrary measurement time moments. The solution is obtained for the phase measurements distributed in compliance with the truncated convolute Gaussian law. The developed algorithm is a multimodal estimation algorithm, which involves a procedure for resolving ambiguity and transforms to Kalman's filter in the absence of ambiguous measurements.