Underdetermined DOA Estimation for Wideband Signals Using Robust Sparse Covariance Fitting
Zhen-Qing He, Zhiping Shi, Lei Huang, Hing Cheung So · IEEE Signal Processing Letters · 2014
From the co-array perspective, sparse spatial sampling can significantly increase the degrees-of-freedom (DOFs), enabling us to perform underdetermined direction-of-arrival (DOA) estimation. By leveraging the increased DOFs from the sparse spatial sampling, we develop a new underdetermined DOA estimation method for wideband signals, named wideband sparse spectrum fitting (W-SpSF) estimator. In W-SpSF, we formulate a sparse reconstruction problem that includes a quadratic$({\ell_2})$weighted covariance fitting term added to a sparsity-promoting$({\ell _{2, 1}})$regularizer. Meanwhile, the optimal regularization parameter of W-SpSF is studied to ensure robust sparse recovery. Numerical results enabled nested arrays demonstrate that the W-SpSF estimator outperforms the spatial smoothing based MUSIC algorithm and works well in nonuniform noise environment.