Low Complexity Robust Direction Finding Method for Impulsive Noise in Lp-Space

Rui Lu,Shitao Zhu,Binke Huang,Ming Zhang,Xiaobo Liu,Anxue Zhang
DOI: https://doi.org/10.1016/j.procs.2017.03.085
2017-01-01
Procedia Computer Science
Abstract:A robust low complexity direction of arrival (DOA) estimation methodforimpulsive noise is proposed in this paper. The presence of outliers makes it difficult to estimate the subspace accurately, and as a result leads to serious estimation errors. In our method, robust signal subspace is first obtained by iterative re-weight singular value decomposition (IR-SVD) of data matrix. Then subspace rotation operator is calculated via matrix lp-norm minimization procedureinstead of least squares (LS). Compared to traditional ESPRIT, our proposed method performs more robust in the presence of impulsive noise. Besides, this ESPRIT like method provides close-form solution of DOA, which saves computational load by avoiding grid searching. Simulation results illustrate that proposed method outperforms than several outliers-resistant algorithms in scenario of impulsive noise.
What problem does this paper attempt to address?