Location and Map-Assisted Wideband Phase and Time Calibration Between Distributed Antennas

TL;DR

This paper presents a novel wideband phase and time calibration method for distributed antenna networks, leveraging location and map information to improve coherence and localization accuracy.

Contribution

It extends a recent narrowband LoS calibration model to wideband scenarios, incorporating reflection paths and joint phase-time calibration with practical estimators.

Findings

The proposed estimators closely approach the CRLB in simulations.

Inclusion of map and location info improves calibration accuracy.

The method supports scenarios with and without prior knowledge.

Abstract

Distributed massive multiple-input multiple-output networks utilize a large number of distributed access points (APs) to serve multiple user equipments (UEs), offering significant potential for both communication and localization. However, these networks require frequent phase and time calibration between distributed antennas due to oscillator phase drifts, crucial for reciprocity-based coherent beamforming and accurate localization. While this calibration is typically performed through bi-directional measurements between antennas, it can be simplified to unidirectional measurement under perfect knowledge of antenna locations. This paper extends a recent phase calibration narrowband line-of-sight (LoS) model to a phase and time calibration wideband orthogonal frequency division multiplexing model, including both LoS and reflection paths and allowing for joint phase and time calibrations. We explore different scenarios, considering whether or not prior knowledge of antenna locations and the map is available. For each case, we introduce a practical maximum likelihood estimator and conduct Cramer-Rao lower bound (CRLB) analyses to benchmark performance. Simulations validate our estimators against the CRLB in these scenarios.