Experimental Performance of Bidirectional Phase Coherent Transmission and Sensing for mmWave Cell-free Massive MIMO Systems with Reciprocity Calibration

TL;DR

This work tackles the critical challenge of phase synchronization across distributed TRPs in mmWave cell-free MIMO by introducing a bidirectional reciprocity calibration framework that enables coherent joint transmission and high-precision sensing. It develops a two-step ML TLS method for initial calibration, coupled with phase-tracking using unilateral uplink/downlink CSI, and enhances robustness in dynamic environments through sensing-assisted calibration and sensing-assisted calibration strategies. The proposed methods achieve reduced calibration overhead, transferable calibration coefficients across UEs, and validated coherent transmission gains and high-precision sensing on a mmWave prototype with HAD-BF; experiments confirm practical viability and channel-reciprocity benefits for long-time coherent operation. The results demonstrate that integrated calibration and sensing can significantly extend coherence times and improve localization and target detection, offering a scalable path toward robust, large-scale CF-mMIMO deployments in 6G mmWave networks.

Abstract

Phase synchronization among distributed transmission reception points (TRPs) is a prerequisite for enabling coherent joint transmission and high-precision sensing in millimeter wave (mmWave) cell-free massive multiple-input and multiple-output (MIMO) systems. This paper proposes a bidirectional calibration scheme and a calibration coefficient estimation method for phase synchronization, and presents a calibration coefficient phase tracking method using unilateral uplink/downlink channel state information (CSI). Furthermore, this paper introduces the use of reciprocity calibration to eliminate non-ideal factors in sensing and leverages sensing results to achieve calibration coefficient phase tracking in dynamic scenarios, thus enabling bidirectional empowerment of both communication and sensing. Simulation results demonstrate that the proposed method can effectively implement reciprocal calibration with lower overhead, enabling coherent collaborative transmission, and resolving non-ideal factors to acquire lower sensing error in sensing applications. Experimental results show that, in the mmWave band, over-the-air (OTA) bidirectional calibration enables coherent collaborative transmission for both collaborative TRPs and collaborative user equipments (UEs), achieving beamforming gain and long-time coherent sensing capabilities.

Figures (27)

• Figure 1: Diagram for mmWave cell-free massive MIMO system.
• Figure 2: HAD-BF system where a UE is transmtting to a BS. It can also be seen as an analog phased array when $N_{RF}^{BS} = 1$ and $N_{RF}^{UE} = 1$.
• Figure 3: The bidirectional calibration process in a quasi-static scenario.
• Figure 4: Pilot mapping method and transmission diagram.
• Figure 5: Quasi-static scene spectral efficiency result for different algorithms with the transmission power of 30 dBm.