Compensation of signal with erasures via sparse representation into its significant subspace

A novel method to the compensation of a signal with erasures is proposed in this paper by considering the sparsity of the signal of interest. First, a measurement is present in order to describe the sparsity of a signal under a given error. Second, an iterating method is proposed to yield the compensation of an erased signal from its significant subspace which can be estimated by using the Orthogonal Match Pursuit algorithm. Finally, simulation results show that the proposed method is well suitable to the compensation for erasures.