Adaptive channel prediction based on polynomial phase signals

TLDR: A new adaptive channel prediction based on non-stationary polynomial phase signals is proposed, which outperforms the classical Linear Prediction and previous prediction methods based on sinusoidal modeling.

Abstract: Motivated by recently published physics based scattering SISO and MIMO channel models in mobile communications [1, 2], a new adaptive channel prediction based on non-stationary polynomial phase signals is proposed. To mitigate the influence of the time-varying amplitudes and to reduce the computation complexity, an iterative estimation of the polynomial phase parameters using the Non-linear instantaneous LS criterion is proposed. Given the polynomial phase parameters, the time-varying amplitudes are estimated using the Kalman filter. The performance of the new predictor is evaluated by Monte Carlo simulations in SISO scenarios with multiple scattering clusters. The new predictor outperforms the classical Linear Prediction and previous prediction methods based on sinusoidal modeling.