Fig. 3.
Mean square error (20 log10 ||x̄ − x||2) as a function of ℓ and SNR, for ν = 1/2 and ε(5) = 10−4. This figure highlights the sparsifying properties of fν(·)-minimization by IRLS for ν < 1. As in the previous figure, only a few number of IRLS iterations are required to reach a satisfactory value of the MSE.