An Improved Model for Symbol Rate Estimation Based on Wavelet Transform

In non-cooperative communication system, symbol rate is a crucial part of parameter estimation. Based on the wavelet transform, an improved symbol rate estimation model is proposed. Without a prior knowledge, we raised a better symbol rate pre-estimation method in view of corrected average periodogram’s smoothing power spectrum. By choosing the optimal wavelet scale and fast Fourier transform, we adopted a multi-synchronous reference algorithm to extract the symbol rate. Simulation demonstrates that the performance of accuracy and anti-noise is improved than traditional algorithms. The normalized mean square is lower than 10 at SNR=0dB and reached 10 when SNR=10dB, which is approaching more closely to CRLB.