An Improved SAMP Channel Estimation Method via Compressed Sensing for MIMO-OFDM Systems

To address the problem that in the Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM) system, the Compressed Sensing (CS) greedy algorithm uses the inner product criterion in the process of selecting atoms for channel estimation will lose part of the original signal information and affect the accuracy of reconstruction, this paper uses the Dice criterion instead of the inner product criterion for atom screening, the Dice criterion can better preserve the original state of the atoms. Due to the problem that the fixed step size of the Sparsity Adaptive Matching Pursuit (SAMP) algorithm affects the accuracy of channel estimation, this paper proposes an improved SAMP algorithm by using the idea of power function variable step size. The simulation results show that the improved SAMP algorithm has better Mean Square Error (MSE) and Bit Error Rate (BER) estimation performance than the traditional compressed sensing greedy algorithm in MIMO-OFDM system.