Abstract:

In view of poor performance of traditional blind estimation methods for zero prefix orthogonal frequency division multiplexing (ZP-OFDM) time-parameters, especially in low signal-to-noise ratio (SNR) and multipath channels, a novel blind estimation method is proposed. Firstly, the power autocorrelation function is calculated and then the pretreatments of wavelet de-noising and smoothing filtering are operated, the symbol duration is estimated by detecting of the distance between two peaks. Secondly, the amplitude square values for the received data length of unit symbol duration are calculated and then the wavelet decomposition and reconstruction are operated. The adaptive threshold is set to extract the estimated zero prefix (ZP) length. Finally, the final ZP length is decided by joint estimation of multiple segment data. Simulation results show that the proposed methods have the superior performance in low SNR and multipath channel conditions.