Super-resolution method for coherent DOA estimation of multiple wideband sources

We focus on coherent Direction Of Arrival (DOA) estimation of wideband sources based spatial sparsity. By development super-resolution algorithms, DOA narrowband has already been addressed gridless sparse algorithms and atomic norm minimization. In this paper, a method is proposed for multiple sources. To do so, we first develop new focusing to map the subband with largest center frequency other ones no requirement any initial estimates. Then, using latter method, introduce an problem by defining set atoms exploiting signal joint sparsity different subbands in continuous domain. This then cast as semidefinite program, which leads implementing higher resolution more robustness noise. Our only needs one single snapshot each subband, although it can also be used snapshots. Numerical simulations show outperformance compared conventional ones.