Uplink Channel Estimation and ICI Elimination for Cell-Free Massive MIMO High-Speed Trains Communications

This article proposes an uplink transmission scheme for cell-free (CF) massive multiple-input multiple-output (MIMO) systems in high-speed railway communications to mitigate the negative effects of fast time-varying multipath channels with fewer antennas and insufficient angular resolution. To enhance the estimation performance by utilizing the possible similarity of the relatively slow time-varying channel parameters, the Dirichlet process (DP) is introduced to construct the probabilistic model of the doubly-selective multiparameter channel estimation problem, which is solved by adopting the expectation maximization (EM) framework. Furthermore, by introducing successive approximation, an enhanced channel estimation scheme is proposed. Finally, for the data transmission utilizing the orthogonal frequency division multiplexing (OFDM) modulation, based on estimated channel parameters, a distributed inter-carrier interference (ICI) elimination algorithm with local channel quality-based combining is proposed, and the performance of the system under different architectures and ICI elimination algorithms is analyzed. Simulation results show that DP can improve estimation performance by utilizing implicit similarity, the combination of first-order Taylor expansion and successive approximation can effectively improve estimation performance at higher speeds under relatively higher pilot power, and compared to centralized processing, the proposed distributed ICI elimination algorithm with local channel quality-based combining has a relatively small performance loss.