基于BP神经网络的GSM-R抗同频干扰方法研究

在GSM-R通信网络中,同频干扰是影响通信质量的一个因素。由于被训练的BP神经网络可以不断校正误差,使输出误差不断减小。为解决GSM-R同频干扰问题,提出了用BP神经网络及其算法对GSM-R通信信号进行校正,通过构建GSM-R通信系统和BP神经网络系统的模型,阐述了GSM-R通信系统抗同频干扰的原理,并对GSM-R进行了数据采集和仿真。

GSM-R Resisting Same-Frequency Interference based on BP Neural Network

In GSM-R communication network, the same-frequency interference is an important factor in affecting the communication quality. As the trained BP neural network could constantly correct the errors and thus make the output error constantly decrease. For solving the problem of the same-frequency interference in GSM-R, BP neural network and its algorithm are proposed to correct GSM-R communication signals. The model for GSM-R communication system and BP neural network is constructed. The principle for resisting the same-frequency interference in GSM-R communication system is discussed and the data of GSM-R is simulated and collected.