Variational Bayes Estimation of Off-grid Migrating Targets with a Delta Method

This paper describes a variational Bayes sparse recovery to estimate off-grid range-migrating targets unambiguously. Off-grid parameters depend nonlinearly on the observation and are thus challenging to estimate. Though, if not estimated, targets may be erroneously recovered in the velocity-ambiguity sidelobes of the true ones. A so-called Delta method is used and results in a fast and efficient algorithm able to leverage velocity ambiguity on synthetic and experimental data. The proposed approach is less complex than state-of-the-art algorithms, a desirable property for real-time applications.