基于神经网络的传播预测模型浅析

传播预测模型是网络规划和频谱资源合理利用的基础。在分析了现有模型不足的基础上,介绍了人工神经网络的结构及其传播预测模型的构建,并对该模型进行了改进,提出了应用反向传播(Back-Propagation,BP)神经网络的混合传播模型。人工神经网络具有良好的非线性逼近能力和泛化能力,非常适用于特定地区传播损耗的预测。通过对试验数据的分析处理,验证了此方法能够更真实地反映该区域的无线电波传播环境,得到更高的预测准确度。

Propagation Prediction Models Based on Artificial Neural Network

Propagation Prediction Model is the basis of network planning and spectrum utilization.Based on an analysis of the disadvantages of existing models,the paper introduces the basic structure of artificial neural network(ANN) and its prediction model,and proposes a new hybrid propagation model using BP neural network,which improves the artificial neural network model.Artificial neural network has strong capabilities of nonlinear approaching and generalization,which makes it very suitable for application in pro...