Satellites have excellent features of high control integration, various working modes, and complex telemetry data. To quantitatively describe the satellite performance degradation, a novel data mining analysis method is proposed to analyze the satellite’s telemetry data. The median filter method is used to reduce the noise data. Then, sample entropy is calculated to characterize state and support vector data description is utilized to analyze the satellite performance degradation process. Experimental results show that the proposed method could quantitatively describe the performance degradation process of satellite. It is an important approach for the ground-station-monitor to analyze the performance of satellite.