Table tennis forehand fast attack and loop technology is the core technology of the project, and table tennis footwork is directly related to the technical level and development prospects of athletes. However, it is very difficult to study technical issues such as the footwork of table tennis. This paper mainly studies the table tennis movement analysis and strength monitoring system based on the posture sensor. This article introduces the improved BP neural network algorithm of Secure Internet of Things. The dynamic error compensation of the sensor is verified by experiments. In the prediction stage, the system uses the estimated value of the previous state as the input value to estimate the current state. In the calibration stage, the output data of the current state can be used as the prediction data of the next state to get more accurate output value. MATLAB is used to read the data, the sampling frequency is 50 Hz, and the acceleration and angular velocity data are output at the same frequency. Finally, the flight trajectory of table tennis is predicted. The difference between the actual velocity and the carrier is about 0.07 m/s, and the actual displacement is about 0.45 m. The results show that the attitude sensor plays a good role in monitoring the movement track and serving strength of table tennis.