In this paper, the clipping noise estimation problem based on compressed sensing (CS) is investigated when the subcarriers used for creating measurement vector are on the edge of the frequency band. Although selection of such subcarriers would not degrade data rate and is suitable for systems requiring high rates, it still weakens the performance of CS recovery algorithms especially in the presence of noise. To address this issue, a new approach based on the orthogonal matching pursuit (OMP) algorithm is proposed for improving the clipping noise estimation. Simulation results indicate that in the chosen scenario under unfavorable conditions, the proposed algorithm is still able to estimate clipping noise and according to its structure, only slightly increases the complexity of the original OMP algorithm.
