Directional Spatial Channel Estimation For Massive FD-MIMO in Next Generation 5G Networks

TL;DR

This paper introduces a novel FD-directional spatial channel estimation algorithm for massive FD-MIMO systems in 5G, leveraging uplink-downlink correlation to improve spectral efficiency without feedback overhead.

Contribution

It presents a new algorithm that estimates FD channel state information using uplink data and spatial correlation, reducing feedback needs in 5G massive MIMO systems.

Findings

Approaches ideal perfect-CSIT performance

Reduces uplink feedback overhead

Effective in massive FD-MIMO systems

Abstract

Full-dimensional (FD) channel state information at transmitter (CSIT) has always been a major limitation of the spectral efficiency of cellular multi-input multi-output (MIMO) networks. This letter proposes an FD-directional spatial channel estimation algorithm for frequency division duplex massive FD-MIMO systems. The proposed algorithm uses the statistical spatial correlation between the uplink (UL) and downlink (DL) channels of each user equipment. It spatially decomposes the UL channel into azimuthal and elevation dimensions to estimate the array principal receive responses. An FD spatial rotation matrix is constructed to estimate the corresponding transmit responses of the DL channel, in terms of the frequency band gap between the UL and DL channels. The proposed algorithm shows significantly promising performance, approaching the ideal perfect-CSIT case without UL feedback overhead.