Peak-to-Average Power Ratio Reduction of Orthogonal Frequency Division Multiplexing using the Autoregressive Model

In this paper, a novel distortionless peak-to-average power ratio (PAPR) reduction method using the autoregressive (AR) model is proposed. This method transforms parallel input data sequences into quasi-white noise signals by utilizing the AR model, which can reduce the autocorrelation of parallel input data sequences greatly, so PAPR can be reduced effectively. Based on simulation, this method can achieve about 2.2 dB PAPR reduction effect by contrast with original signals. Furthermore, when the number of subcarriers is large (e.g., 1024), to achieve the same PAPR reduction, the complexity of the proposed method that uses only one inverse fast Fourier transform (IFFT) is much lower than that of partial transmit sequences (PTS)