Improved Channel Estimation Algorithm Based on Compressed Sensing

Aiming at the characteristics of OFDM (Orthogonal Frequency Division Multiplexing) system that the channel is sparse and the sparsity is unknown, a kind of new sparsity adaptive compressed sensing channel estimation algorithm is proposed, namely MSAMP (Modified Sparsity Adaptive Matching Pursuit). The algorithm obtains channel information of time domain through inverse Fourier transform on the basis of traditional LS (Least Square) channel estimation. Then the point with the highest energy in the time domain response is replaced by the mean value of the noise. After iteration, the actual channel sparsity estimation value is obtained, which is used as the initial index set of MSAMP algorithm to improve the reconstruction speed. And the idea of atom pre-screening and minimum constant step size are combined. That is, in the initial selection stage of each iteration, a threshold is set to perform preliminary screening of atoms and discard non-ideal atoms with less correlation. On this basis, a secondary screening of atoms is performed to further remove inappropriate atoms; At the same time, the minimum constant step size 1 is used as the size of the change to prevent under-estimation or over-estimation to improve system accuracy. The simulation results show that the reconstruction speed of MSAMP algorithm is better than SAMP algorithm under the same conditions. Moreover, MSAMP algorithm is better than the traditional least square estimation algorithm and SAMP (Sparsity Adaptive Matching Pursuit) algorithm in the mean square error and bit error rate of the channel with unknown sparsity.