# Channel Extrapolation for FDD Massive MIMO: Procedure and Experimental Results

## TL;DR

This paper investigates channel extrapolation in FDD massive MIMO systems using wideband measurements and the SAGE algorithm to reconstruct downlink channels from uplink data, highlighting challenges and environment-dependent performance.

## Contribution

It introduces a measurement-based approach for channel extrapolation in FDD massive MIMO using the SAGE algorithm and evaluates its effectiveness in real-world conditions.

## Key findings

- Extrapolation performance varies with calibration accuracy.
- LOS channels yield better extrapolation results than NLOS.
- Propagation environment significantly impacts extrapolation accuracy.

## Abstract

Application of massive multiple-input multiple-output (MIMO) systems to frequency division duplex (FDD) is challenging mainly due to the considerable overhead required for downlink training and feedback. Channel extrapolation, i.e., estimating the channel response at the downlink frequency band based on measurements in the disjoint uplink band, is a promising solution to overcome this bottleneck. This paper presents measurement campaigns obtained by using a wideband (350 MHz) channel sounder at 3.5 GHz composed of a calibrated 64 element antenna array, in both an anechoic chamber and outdoor environment. The Space Alternating Generalized Expectation-Maximization (SAGE) algorithm was used to extract the parameters (amplitude, delay, and angular information) of the multipath components from the attained channel data within the training (uplink) band. The channel in the downlink band is then reconstructed based on these path parameters. The performance of the extrapolated channel is evaluated in terms of mean squared error (MSE) and reduction of beamforming gain (RBG) in comparison to the ground truth, i.e., the measured channel at the downlink frequency. We find strong sensitivity to calibration errors and model mismatch, and also find that performance depends on propagation conditions: LOS performs significantly better than NLOS.