基于级联SSA-SOFM神经网络的信号分选方法

A Method of Signal Sorting Based on Cascaded SSA-SOFM Neural Network

针对应用传统SOFM神经网络进行辐射源信号分选的不足,提出了一种结构自调整算法以解决必须事先确定SOFM网络规模的问题,进而提出了一种基于级联结构自调整SOFM(SSA-SOFM)神经网络模型的信号分选方法,采用实际数据进行了实验仿真,结果表明了该方法的有效性.

Aimed at the drawback of the traditional SOFM neural network for radiation signal sorting, a structure self-adjusting （SSA） algorithm was discussed, which solves the problem of deciding the network scale in advance. Then, a signal sorting method based on the cascaded SSA-SOFM neural network model, was put forward and simulations made by using available data. The results show that the method dealt with is feasible.