Blind Sparse Channel Identification Using Subspace-Based Algorithm

This paper addresses the problem of blind channel identification under sparse channel condition. Our approach is an extension of the subspace blind channel identification methods. Unlike previous approaches for blind channel identification where the optimization is in least square sense, i.e., the L-2 norm, the proposed extension includes the identification of sparse channels and uses the L-1 norm. By doing so, we show that the performance of the proposed method outperforms previous approach, under sparse channel conditions. Numerical examples are included in order to demonstrate the effectiveness of the proposed approach. Bit Error Rate and normalized error performances of our approach are also included.