The blind symbol rate estimation algorithms for autonomous radio in deep space are presented. In view of the unknown symbol starting point, an improved symbol rate estimation algorithm based on split-symbol moments estimator (SSME) is proposed according to the theoretical analysis of the estimation error. This new algorithm estimates the signal-to-noise ratio (SNR) in different ranges and selects the maximum value, which ensures the symbol rate estimation accuracy. In addition, the computational complexity of the new algorithm is compared with the conventional delay and multiply algorithm, improved delay and multiply algorithm with accumulations. The simulation results show that the proposed algorithm can effectively complete the task of symbol rate estimation at low SNR. Its symbol rate error estimation probability is lower than that of delay and multiply algorithm, and is close to that of delay and multiply algorithm with accumulations. However, the computational complexity of the new algorithm is less than 10% of that of delay and multiply algorithm with accumulations.