Adaptive semi-blind channel estimation for massive MIMO systems

Traditional semi-blind channel estimation schemes for massive multiple-input multiple-output systems are based on eigenvalue decomposition (EVD) or singular value decomposition (SVD). However, EVD based or SVD based estimators are too complex for practical implementation. To reduce the complexity, a new adaptive semi-blind channel estimator for massive MIMO systems is developed in this paper. The new estimator is based on the fast single compensation approximated power iteration (FSCAPI) subspace tracking algorithm, which converges fast, possess good orthogonality, and has low computational complexity. Furthermore, the closed form expression of the channel capacity under inaccurate CSI is derived to study the effect of channel estimation error on channel capacity. Simulation results show that the proposed FSCAPI-based semi-blind channel estimator achieves nearly the same estimation performance with the SVD-based estimator, and superior to the EVD-based estimator in terms of mean square error and channel capacity.