基于神经网络的UTRAN网络质量综合评价  

3G UTRAN Quality Evaluation Based on Neural Network

作　　者：徐海东[1] 李冶文[1] 江峰[2] 宋俊德[2] 

机构地区：[1]中国移动通信集团公司网管中心,北京100032 [2]北京邮电大学电子工程学院,北京100876

出　　处：《北京邮电大学学报》2005年第4期41-44,49,共5页Journal of Beijing University of Posts and Telecommunications

基　　金：国家自然科学基金项目(60372098)

摘　　要：为了评价3G无线网络UTRAN的运行质量,给出了3G无线网络UTRAN质量用户感知关键指标,提出了利用基于前向单隐层BP神经网络方法的UTRAN网络质量综合评价模型.该模型采用模糊综合评价算法构造训练样本集.对比模糊综合评价算法对3G试验测试数据和该模型的评价结果可以看出,该模型能稳定、一致、可重复地评价UTRAN网络质量.A set of key performance indicator of UMTS terrestrial radio access network （UTRAN） running quality was presented to evaluate the running quality of 3G UTRAN. Meanwhile, an evaluating model based on BP neural network was proposed to evaluate the 3G UTRAN running quality. The fuzzy evaluating method was used to construct training data. It is resulted by the fuzzy evaluating method, that the model will be able to evaluate the UTRAN running quality very well.

关 键 词：3G无线网络质量 BP神经网络 评价模型 

分 类 号：TN915.07[电子电信—通信与信息系统]