TCGTR-based channel estimation for cooperative multi-cell large scale antenna systems

In this paper, a cooperative channel estimation strategy is proposed in multi-cell large scale antenna systems (so-called “massive MIMO”). Based on time-shifted uplink pilot sequences, the strategy can alleviate the performance degradation caused by the pilot contamination. In addition, the truncated conjugate gradient trust region (TCGTR) algorithm is proposed to estimate the channel vectors. The TCGTR algorithm, as an iterative regularization method, is widely used for large scale optimization because of its remarkable performance and fast convergence. The novelty of the TCGTR algorithm lies in the fact that it has never been used for channel estimation before. The complexity of the TCGTR algorithm is analyzed in this work. The simulation results show that this proposed algorithm significantly outperforms the noncooperative channel estimation.