Sparsity analysis using a mixed approach with greedy and LS algorithms on channel estimation

Various channels can be denoted by sparse channels and many algorithms have been proposed to exploit their sparsity. In this paper, we propose a mixed algorithm based on Greedy and LS algorithms for sparse channel estimation. Analyses of the proposed and commonly used algorithms in terms of performance and complexity are performed considering the channel's sparsity, the length of training sequence and the stopping criterion. Our results show that a suitable trade-off can be found and effective channel estimations can be obtained with a low-cost algorithm.