Abstract: Aiming at the problem that the performance of traditional MQAM signals symbol rate estimation method is deteriorating or even completely invalid under Alpha stable noise environment. According to the combination of the advantages of Myriad class filter and wavelet transform, a high-order vanishing-moment quadratic wavelet transform method based on Recursive Hybrid Myriad filter (RHMy) is proposed in this paper. Firstly, the RHMy filter is used to suppress the impulse noise of the receiver signals sufficiently, and then the high-order evanescent moment wavelet is used to estimate the symbol rate. Compared with the the multiscale Haar wavelet estimation method, the fractional low-order cyclic spectrum estimation method and the Weighted Myriad-based wavelet transform method, The simulation results show that the proposed method can effectively improve the estimation accuracy under low generalized signal-to-noise ratio in the Alpha stable distributed noise environment and even under different Alpha values the proposed method also has good robustness.