A System Concept for Online Calibration of Massive MIMO Transceiver Arrays for Communication and Localization

Massive multiple-input multiple-output (MIMO) techniques are being considered for the fifth generation (5G) mobile communication systems in order to deliver high multiplexing gain. However, hardware impairments like quadrature imbalance in mixers violate the requirement for channel reciprocity and may change, e.g., with temperature or while aging. In addition, advanced wireless localization techniques and the generation of predefined beam patterns require knowledge about all antenna phase center positions and the time and phase delay of all transmit and receive channels. Thus, an efficient online compensation method is needed that scales well for very large numbers of transceiver modules. We propose to extend the transmitter with a small measurement feature at the transmitter output based on one uncalibrated power detector per module as well as a single, external four-element backscatter array for the entire matrix. These enhancements facilitate a fast and efficient iterative calibration, which recognizes and mitigates all major error sources. Beside optimal communication throughput and energy efficiency, it thereby brings localization capabilities to mobile networks as an additional major benefit. For verification, a system of multiple cost-efficient 5.8-GHz massive MIMO transceivers with 150-MHz bandwidth and a backscatter array has been implemented. Measurement results demonstrate the capability of the proposed concept to efficiently compensate major error sources as well as its robustness.