移动终端身份认证的深度信念网络模型

Deep Belief Network Model for Mobile Terminal Identity Authentication

文章针对移动终端面临的信息安全问题,建立了移动终端身份认证的深度信念网络模型。利用触摸屏传感器采集用户原始触摸手势数据序列,经数据预处理后提取手势特征并传入深度信念网络模型;选用逐层贪婪算法进行无监督的预训练,再经反向传播算法进行有监督微调后固定模型参数;将测试手势特征数据作为模型输入层数据,经模型计算后得到输出层数据,由Softmax分类器对输出数据分类,认证用户身份。仿真实验结果表明,与连续隐马尔可夫模型、反向传播算法相比,文中深度信念网络模型能达到较低的错误率,明显提高认证的准确性。

Aiming at the information security problem faced by mobile terminal, a deep belief network model for mobile terminal identity authentication is established. The touch screen sensor is used to collect the user’s original touch gesture data sequence. After the data is pre-processed, the gesture features are extracted and passed into the deep belief network model. The layer-by-layer greedy algorithm is used for unsupervised pre-training, and then the back-propagation algorithm is used to supervise and fine-tune the fixed model parameters. The gesture feature data is used as the model input layer, and the output layer data is obtained after the model is calculated, and the output data is classified and authenticated by the Softmax classifier. Compared with the continuous hidden Markov model and back propagation algorithm, the simulation results show that the deep belief network method can achieve a lower error rate and significantly improve the accuracy of authentication.