CAJ | 학술논문

针对多种信号在低信噪比条件下识别率低的问题，利用高阶累积量良好的抑制噪声特性，通过构造高阶累积量作为特征参数之一，并联合其他特征参数，采用优化的径向基神经网络对模拟数字信号进行自动调制识别。Matlab仿真表明，该种方法能够有效提高低信噪比条件下的信号识别率。
침대다충신호재저신조비조건하식별솔저적문제，이용고계루적량량호적억제조성특성，통과구조고계루적량작위특정삼수지일，병연합기타특정삼수，채용우화적경향기신경망락대모의수자신호진행자동조제식별。Matlab방진표명，해충방법능구유효제고저신조비조건하적신호식별솔。
For a variety of signals recognition is low in the low SNR conditions, using iligh order cumulants good noise suppression characteristics, by building higher order cumulants as one of the characteristic parameters, and joining other characteristic parameters, using the optimized radial basis function neural networks for automatic modulation recognition analog and digital signal. Matlab simulation shows that the methods can effectively improve the low SNR of the signal recognition rate.