为在低复杂度约束条件下提升电磁信号调制识别的性能,提出了一种基于稀疏深度神经网络(Sparse Deep Neural Network,SDNN)的电磁信号调制识别方法。首先,通过提取电磁信号同相和正交两路数据绘制出信号的星座图,作为信号的浅层特征表达...为在低复杂度约束条件下提升电磁信号调制识别的性能,提出了一种基于稀疏深度神经网络(Sparse Deep Neural Network,SDNN)的电磁信号调制识别方法。首先,通过提取电磁信号同相和正交两路数据绘制出信号的星座图,作为信号的浅层特征表达;然后,基于星座图中各信号点密度大小对星座图进行上色,增强星座图中信号特征;最后,通过SDNN对增强后的星座图进行识别分类。实验结果表明,SDNN模型选取合适的剪枝率后,能够有效降低模型存储规模和计算量,其中模型参数压缩了72%,浮点运算量压缩了45%,与原模型97%的综合识别率相比,稀疏化处理后模型的综合识别率为96.8%,在小幅度识别精度损失范围内大幅降低了模型复杂度。展开更多
In order to allow the guardians to monitor the physiological parameters of the infant more intuitively and to be able to respond to sudden irregularities in the pulse rate,abnormal blood oxygen,high or low body temper...In order to allow the guardians to monitor the physiological parameters of the infant more intuitively and to be able to respond to sudden irregularities in the pulse rate,abnormal blood oxygen,high or low body temperature and other conditions,and to facilitate communication with the medical staff or to request assistance in treatment,an STM32 microcontroller-based infant health monitoring system is designed.The digital signal acquisition module for pulse,blood oxygen and body temperature acquire the raw data,and the microcontroller performs algorithmic processing to display the physiological parameters such as pulse,blood oxygen and body temperature of the infant,and configures the threshold alarms for the physiological parameters by means of a keypad module.Finally,the test results are compared and tested against the standard physiological parameters of infants and children to verify that the system meets the requirements of medical precision and accuracy.展开更多
文摘为在低复杂度约束条件下提升电磁信号调制识别的性能,提出了一种基于稀疏深度神经网络(Sparse Deep Neural Network,SDNN)的电磁信号调制识别方法。首先,通过提取电磁信号同相和正交两路数据绘制出信号的星座图,作为信号的浅层特征表达;然后,基于星座图中各信号点密度大小对星座图进行上色,增强星座图中信号特征;最后,通过SDNN对增强后的星座图进行识别分类。实验结果表明,SDNN模型选取合适的剪枝率后,能够有效降低模型存储规模和计算量,其中模型参数压缩了72%,浮点运算量压缩了45%,与原模型97%的综合识别率相比,稀疏化处理后模型的综合识别率为96.8%,在小幅度识别精度损失范围内大幅降低了模型复杂度。
文摘In order to allow the guardians to monitor the physiological parameters of the infant more intuitively and to be able to respond to sudden irregularities in the pulse rate,abnormal blood oxygen,high or low body temperature and other conditions,and to facilitate communication with the medical staff or to request assistance in treatment,an STM32 microcontroller-based infant health monitoring system is designed.The digital signal acquisition module for pulse,blood oxygen and body temperature acquire the raw data,and the microcontroller performs algorithmic processing to display the physiological parameters such as pulse,blood oxygen and body temperature of the infant,and configures the threshold alarms for the physiological parameters by means of a keypad module.Finally,the test results are compared and tested against the standard physiological parameters of infants and children to verify that the system meets the requirements of medical precision and accuracy.
