IEEE Transactions on Geoscience and Remote Sensing · 2004 · 202 citations · 10 references
Natural SciencesGeophysicsGeodesyGlobal Navigation Satellite SystemsRemote SensingCarrier Phase Ambiguity ResolutionInterferometryParameter EstimationParameter Estimation Under UncertaintyRadar Signal ProcessingGlobal Positioning SystemInteger Least-squares EstimationNonlinear InversionNonlinear Parameter EstimationRadar InterferometrySynthetic Aperture RadarSynthetic Aperture Radar InterferometryVariance-covariance Matrix Estimation
In the permanent scatterer technique of synthetic aperture radar interferometry, there is a need for an efficient and reliable nonlinear parameter inversion algorithm that includes estimation of the phase cycle ambiguities. Present techniques make use of a direct search of the solution space, treating the observations as deterministic and equally weighted, and which do not yield an exact solution. Moreover, they do not describe the quality of the estimated parameters. Here, we use the integer least squares estimator, which has the highest probability of correct integer estimation for problems with a multivariate normal distribution. With this estimator, the propagated variance-covariance matrix of the estimated parameters can be obtained. We have adapted the LAMBDA method, part of an integer least squares estimator developed for the ambiguity resolution of carrier phase observations in global positioning systems, to the problem of permanent scatterers. Key elements of the proposed method are the introduction of pseudo-observations to regularize the system of equations, decorrelation of the ambiguities for an efficient estimation, and the combination of a bootstrap estimator with an integer least squares search to obtain the final integer estimates. The performance of the proposed algorithm is demonstrated using simulated and real data.
10
Least-squares estimation of the integer GPS ambiguities
P. J. G. Teunissen · Medical Entomology and Zoology · 1993 · 343 citations