Low Complexity Robust Direction Finding Method for Impulsive Noise in l p -Space
Rui Lu, Shitao Zhu, Binke Huang, Ming Zhang, Xiaobo Liu, Anxue Zhang · Procedia Computer Science · 2017
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 l p -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.