Sparse channel estimation with regularization methods in massive mimo systems

Massive multiple-input multiple-output (MIMO) technology has recently gained a lot of attention as a candidate technology for the next generation wireless systems. With a higher number of antennas, pilot-based channel estimation faces a limitation in the number of orthogonal pilots to be used among users in all cells. Sparse channel estimation by using regularization methods can reduce the pilots compared to pilot-based channel estimation. In this paper, we study two regularization methods: least absolute shrinkage and selection operator (lasso) and elastic net. We investigate the performance of least squares (LS), lasso, and elastic net when the sparsity of the channel changes over time. We study the optimum tuning parameters for lasso and elastic net based channel estimators to achieve the best performance with the different number of pilots and values of signal-to-noise ratio (SNR). Finally, we present the asymptotic analysis of LS, lasso, and elastic net based channel estimators.