Partial sparse multi-path channel estimation using ℓ1-regularized LS algorithm

Sparse multi-path channels (MPC) are often encountered in several wireless communications. Exploiting the sparse property of MPC, several estimation methods have been proposed in recent years. Unlike the previous methods based on either dense MPC or sparse MPC models, we propose a novel MPC estimation method based on a generalized hybrid MPC model. The existed least squares (LS) and sparse component analysis (SCA) methods are also extended to partial SCA method by using a partially lscr₁-regularized LS algorithm. Simulation results confirm that the proposed method performs better than both the dense MPC and sparse MPC estimation methods.