The invention discloses a non-contact user identity authentication method based on RFID and a convolutional neural network, and the method comprises the steps: simulating the unlocking of a sudoku of a mobile phone, deploying an RFID tag matrix, collecting the phase information during the movement of a hand, and achieving the identity authentication of a user through the convolutional neural network. The method comprises the following implementation steps of: 1, acquiring phase information of an action for identity authentication; 2, preprocessing the acquired phase information; 3, extracting starting point information of each action sample from the preprocessed phase information; 4, making the obtained phase information into a data set, and training the data by adopting a deep learning convolutional neural network to obtain a convolutional neural network model for authentication; 5, identifying the action of the user by using the trained model when the identity of the user is authenticated; and 6, obtaining an authentication result, and judging whether authentication succeeds or not. According to the invention, the invention employs the CNN-based convolutional neural network to extract the features in the phase, carries out the motion recognition, carries out the identity authentication, and is better in accuracy, safety and robustness.