Channel reciprocity calibration in TDD hybrid beamforming massive MIMO systems

TL;DR

This paper proposes a channel reciprocity calibration scheme for TDD hybrid beamforming massive MIMO systems that does not rely on channel models and can achieve highly accurate CSIT estimation.

Contribution

A novel reciprocity calibration method for TDD hybrid beamforming massive MIMO that does not depend on channel models and improves CSIT accuracy.

Findings

The proposed scheme can estimate CSIT with arbitrarily small error.

It does not require assumptions on the propagation channel model.

The method enhances the potential of hybrid beamforming systems.

Abstract

Hybrid analog-digital (AD) beamforming structure is a very attractive solution to build low cost massive multiple-input multiple-output (MIMO) systems. Typically these systems use a set of fixed beams for transmission and reception to avoid the need to obtain channel state information at transmitter (CSIT) for each antenna element individually. However, such a method can not fully exploit the potential of hybrid AD beamforming systems. Alternatively, CSIT can be estimated by assuming a model for the propagation channel, whereas this model is only validated in millimeter-wave (mmWave) band thanks to its poor scattering nature. In this paper, we focus on time division duplex (TDD) systems with hybrid beamforming structure and propose a reciprocity calibration scheme that allows to acquire full CSIT. Different to existing CSIT acquisition methods, our approach does not require any assumption on the channel model and can, in theory, estimate the CSIT up to an arbitrary small error.