Range-Doppler Imaging via Forward-Backward Sparse Bayesian Learning

We consider range-Doppler imaging via transmitting a train of probing pulses as in radar and active sonar. We show that range-Doppler imaging can be formulated as a sparse signal recovery problem and that we can use an expectation maximization based sparse Bayesian learning (EM-SBL) algorithm to achieve high resolution imaging. We also reduce the complexity of EM-SBL significantly by using an efficient forward-backward algorithm in the E step of the EM algorithm.