The advantages of Wavelet Transform and Fast Fourier Transform (FFT) are obtained by STransform. It has excellent time frequency aggregation performance. The complexity of signal distribution was quantified by normalized information entropy. The original vibration signal of the rolling bearing was transformed by STransform. The uniformity of frequency distribution was quantified by normalized information entropy at each time. An Stime entropy characteristic index was proposed to reflect the degradation process of rolling bearings. Based on the rolling bearing inner ring and the outer ring of the bearing, bearing three faults were analyzed in 3 different damage degree mathematical model simulation, and extract their Stime entropy index were compared to verify the feasibility of the method. The data of the accelerated fatigue life cycle of the rolling bearing are analyzed and compared with the effective values of the time domain indexes commonly used in the engineering. The results show the effectiveness of the method.