Calibration of Nonlinear Crosstalk in MIMO Transmitter

Multiple-input multiple-output (MIMO) techniques have been extensively used in modern wireless communications. The front-end nonlinearity coupled with crosstalk in a MIMO transmitter is much more complex than that in a single-input single-output (SISO) system. In this paper, we propose a nonlinear crosstalk calibration algorithm with the corresponding reference signal design. We start with the nonlinear crosstalk model of the MIMO transmitter and derive its corresponding inverse compensation model. The proposed model characterizes the nonlinearity and the crosstalk in separated blocks. We then propose specifically designed reference signals that utilize the model structure. The nonlinear crosstalk model can be reorganized and simplified with the reference signal. Thus, the complexity of the model coefficients estimation is greatly reduced without compromising the model accuracy. In addition, with the proposed algorithm, the compensation of the nonlinearity and crosstalk can be done separately. Simulations show a satisfactory calibration results with low computational complexity.