Sparse signal recovery with measurement matrix uncertainty

Tiedan Wang,Ronghui Tao,Ping Wei
2012-01-01
Journal of Information and Computational Science
Abstract:Traditional sparse reconstruction algorithms require the accurately known dictionary, and the sparse recovery performance would degrade significantly with the measurement matrix uncertainty. In order to eliminate the negative effects of measurement matrix uncertainty, this paper proposed a robust reconstruction algorithm. Through the variance of additive errors, this algorithm considered the sparse constraint and the effect of the additive errors from the statistical point at the same time. Numerical simulations demonstrated that the proposed method has a better reconstructing performance than the traditional methods. © 2012 Binary Information Press.
What problem does this paper attempt to address?