Direction-of-Arrival Estimation Using an Array Covariance Vector and a Reweighted l₁ Norm

抄録

We propose a novel sparse representation-based direction-of-arrival (DOA) estimation method. In contrast to those that approximate l₀-norm minimization by l₁-norm minimization, our method designs a reweighted l₁ norm to substitute the l₀ norm. The capability of the reweighted l₁ norm to bridge the gap between the l₀- and l₁-norm minimization is then justified. In addition, an array covariance vector without redundancy is utilized to extend the aperture. It is proved that the degree of freedom is increased as such. The simulation results show that the proposed method performs much better than l₁-type methods when the signal-to-noise ratio (SNR) is low and when the number of snapshots is small.