Robust Multi-User Channel Tracking Scheme for 5G New Radio

Recently, massive multiple input multiple output (MIMO) channel tracking in the fifth generation (5G) new radio (NR) systems has attracted intensive interest. By exploiting the dynamic sparsity of massive MIMO channels, it is possible to design a high-accuracy channel tracking scheme. However, the existing channel estimation/tracking algorithms often ignore the practical imperfections in real systems, such as the channel aging effect due to the hopping Sounding Reference Signal (SRS) pattern, the time offset, phase noise, and multi-user SRS interference. In this paper, we propose a robust multi-user uplink channel tracking scheme, which is compatible with 5G NR systems and robust against various system imperfections. Specifically, we propose a sparse Markov channel model to capture the dynamic sparsity of massive MIMO-OFDM channels under the consideration of imperfect factors. Then, we propose a robust multi-user channel tracking scheme, which iterates between two components until convergence. Particularly, in the channel estimation component, we employ the Turbo-CS method to exploit the channel dynamic sparsity to perform efficient multi-user channel estimation under non-orthogonal SRSs, where a multi-stage successive interference cancellation (MSIC) scheme is proposed to mitigate the multi-user SRS interference. Then, in the imperfection parameter estimation component, we further estimate the unknown imperfection parameters based on the updated Bayesian channel estimates from the channel estimation component. Simulation results verify the superior channel tracking performance of our proposed scheme over the baselines.