A Low-Complexity 2-D Angle of Arrival Estimation in Massive MIMO Systems

Massive multiple-input multiple-output (MIMO) systems are promising technology to greatly increase the spectral efficiency for the 5G cellular system. However, the implementation is practically a challenge due to the limitation of cost, space, and complexity. Though the millimeter-wave (mm-wave) transmission can greatly save the space for deploying numerous antennas, the demand on the numerous RF chains increases the implementation cost significantly. The hybrid structures of sub-array antennas are then developed to alleviate the cost, where the entire array is grouped into several sub-arrays. All antennas in a subarray share a common RF chain, which greatly reduces the complexity. Furthermore, the downlink channel state information (CSI) is crucial for several pre-processing technologies such as precoding. Nevertheless, the CSI estimation is difficult due to the large dimension of a channel matrix. Accordingly, CSI estimation by the structured channel matrix is attractive since only few unknown angle-of-arrivals (AoA) and deterministic signatures can model the CSI. Estimation of CSI is equivalent to estimating the AoA. In this paper, we propose a new AoA estimation by using estimating signal parameters via rotational invariance technique (ESPRIT) for the massive MIMO system with two kinds of hybrid subarrays, referred to side-by-side and interleave sub-arrays. Numerical results validate the proposed AoA estimation and show that the proposed AoA estimation with side-by-side sub-arrays can approach to the fully-digitized arrays while keeping a lower computational complexity.