Sensor-Aided Predictive Beam Tracking for mmWave Phased Array Antennas

This paper studies the beam tracking problem in a millimeter wave (mmWave) based system with phased array antennas. The key challenge is to jointly optimize the receive analog beamformer and estimate the angle of arrival (AoA) with consideration of mobile terminal (MT) rotation and environment changes over time. We propose an inertial sensor aided predictive beam tracking (PBT) scheme to exploit the attitude information available at the MT. Specifically, the AoA changes due to the local rotation which is on the time-scale order of a few 100 ms, and thus can be well predicted with prediction error modeled by an additive Gaussian random variable. Then, by taking the prediction error into account, the receive beamformer is opti- mized based on the Bayesian Cramér-Rao bound. Subsequently, the maximum a posteriori (MAP) estimate is employed for AoA estimation. The resultant estimator is equivalent to achieve a balance between the pilot-based AoA estimation and sensor-aided AoA prediction. The numerical simulation results validate that the proposed low- complexity PBT scheme can accurately track the beam even for dynamic scenarios. Meanwhile, the rotation information provided by the local sensors is of critical importance for beam tracking.