Improved sparse channel estimation for multicarrier systems with compressive sensing

In this paper, we propose an effective channel estimator by exploiting compressive sensing (CS) for multicarrier systems. Conventional linear channel estimators are considered optimal under the assumption of rich multipath, while the practical physical multipath channels tend to exhibit sparse structure. Exploiting the inherent sparsity, CS-based channel estimation can achieve higher spectral efficiency through reducing the number of pilots compared with the linear estimation methods. For the channel sparsity is usually unavailable, a modified compressive sampling matching pursuit sparse channel estimation (mCoSaMP-SCE) method is proposed. Simulations confirm the proposed method with respect to the MSE performance and the computational complexity.