부산대학교 도서관

Focused on the poor estimation performance of traditional algorithms for massive MIMO systems with time division duplex, an improved sparse adaptive matching pursuit (SAMP) algorithm based on weak selection and variable stepsize is proposed. The algorithm combines the atomic selection characteristics of the stagewise weak orthogonal matching pursuit (SWOMP) algorithm and the variable step-size of the power function. The weak selection rule of the SWOMP algorithm is taken as the preprocessing of the SAMP algorithm. By adding fixed atoms through a reasonable threshold, improving the estimation accuracy and accelerating the convergence of the SAMP algorithm is achieved. The simulation results show that compared with the SAMP algorithm, the channel estimation accuracy is significantly improved, and the estimation accuracy is improved by 2 to 3 dB.