Near-Field Channel Estimation for Extremely Large Antenna Arrays With Uniform Circular Arrays

Extremely large-scale antenna array (ELAA) is viewed as one of the most promising technologies for future 6G communication systems. To fully exploit the performance of ELAA, accurate channel state information (CSI) is very important. Recently, uniform circular array (UCA) has attracted a lot of attention thanks to its constant beamforming gain at all directions. However, with a large antenna array, the users are very likely to be located in the near-field range of the ELAA. In the near field, the electromagnetic cannot be approximated as the planar wave to analyze as the case in far-field scenarios. Currently, most CSI acquisition schemes are based on the far-field assumption where the electromagnetic waves propagate as planar waves. To improve the channel estimation accuracy in ELAA systems with UCA, we study the near-field channel estimation in this paper. Specifically, we first analyze the beamforming gain of UCA and reveal the multi-value property in UCA systems, based on which we propose a criterion of a proper codebook for UCA near-field channel estimation. Based on this criterion, we further design an improved concentric-ring codebook. Then, based on the designed codebook, we propose an improved concentric-ring codebook based orthogonal matching pursuit (C-OMP) scheme to recover the near-field channel of UCA effectively. We also analyze the computational complexity of the proposed scheme. Furthermore, we discuss how to adapt the proposed scheme to multiple other scenarios such as wideband scenarios and spatial non-stationary scenarios. Finally, we conduct the simulations to reveal that the proposed scheme can realize more accurate channel estimation than existing schemes and can be easily adapted to different scenarios.