ABSTRACT
The compressed sensing (CS) paradigm is invoked to allow the target scene to be unambiguously reconstructed, while assuming reasonably that the signal of interest is sparse or at least compressible in a given dictionary. This chapter investigates the use of a wideband waveform jointly with sparse recovery techniques to unambiguously reconstruct the target signal. For conventional narrowband radars, a target remains confined in its range resolution cell during the whole coherent processing interval (CPI). The radar ambiguity function is an important analysis tool that allows to characterize a waveform jointly with the filter matched to the received signal. SSR techniques are known to have good deconvolution properties that enable to increase the dynamic range of the estimated sparse signal. According to the hierarchical Bayesian model, classical Bayesian estimates of the target amplitude vector can be derived. The Bayesian community is actively working on decreasing the complexity of hierarchical inference while first attempts at designing CS radar detector are suggested.
