Phase Noise Estimation Under the Assumption of Autoregressive Process Model in OFDM System

The phase noise (PN) is a challenge for Orthogonal Frequency Division Multiplexing (OFDM) system, which causes the nonlinear distortion and degrades the performance, especially in 5th generation New Radio (5G-NR). In this letter, the PN is described by the autoregressive (AR) process model and the corresponding maximum a posteriori (MAP) criterion is presented, which is named MAP-AR method. First of all, the sparse banded structure of inverse matrix of covariance matrix for AR process model is proven. Besides, the PN is described by the AR process model and the corresponding MAP criterion is converted to a sparse banded linear equation with the consideration of linear approximation. In addition, the corresponding Gaussian elimination (GE) method is given to solve such sparse banded linear equation with low computational complexity. Finally, the superiorities of our proposed MAP-AR method is confirmed by the numerical results.