In order to solve the problem that existing algorithms for the symbol rate estimation of phase shift keying (PSK) signals will undergo performance degradation or even become invalid in the Alpha stable noise environment, a novel cyclic correntropy based method under the cyclostationary framework for symbol rate estimation of PSK signals is proposed. The paper presents the cyclic correntropy function, and deduces the cyclic correntropy function expression of binary phase shift keying (BPSK) signals theoretically. This method is easy to be realized owing to the adoption of fast fourier transform (FFT) in calculating the one-dimensional slices of the cyclic correntropy function, and can estimate the symbol rate by detecting the discrete spectral lines of the PSK signal cyclic correntropy function. Moreover, it does not require the prior information of the impulsive noise. The simulation results show that the proposed method can suppress the Alpha stable noise efficiently, and has excellent symbol rate estimation performance of PSK signals, especially in the strong impulsive noise environment.