摘要
传统的5G Massive MIMO(multiple input multiple output)3维信道模型复杂度高、计算量大,无法满足网络需求.针对此问题,提出一种基于神经网络的3维无线信道特征预测及评估模型.该模型只依赖于高精地图产生的射线追踪数据,无须搭建测试网络.仿真结果表明:该模型能降低网络优化成本及时间开销、快速预测信道特征和评估网络性能.
The traditional 5G Massive MIMO(multiple input multiple output)three-dimensional channel model has high complexity and large computation complexity,which cannot meet the network requirements.To solve this problem,a three-dimensional wireless channel feature prediction and evaluation model based on neural network was proposed.The model only relied on ray tracing data from high-precision map,and did not need to build a test network.The simulation results showed that this model could reduce network optimization costs and time overhead.This model could quickly predict channel feature and evaluate network performance.
作者
朱军
蒋一鸣
李凯
王写
成博
ZHU Jun;JIANG Yiming;LI Kai;WANG Xie;CHENG Bo(Institute of Electronic Information Engineering,Anhui University,Hefei 230601,China;Institute of Creativity and Art,Shanghai University of Science and Technology,Shanghai 201210,China;Huawei Technical Service Co.,Ltd.,Shanghai 201206,China)
出处
《安徽大学学报（自然科学版）》
CAS
北大核心
2020年第6期36-42,共7页
Journal of Anhui University(Natural Science Edition)
基金
安徽省科技重大专项(18030901010)。
关键词
5G
大规模多输入多输出
3维信道模型
信道特征
神经网络
5G
Massive multiple input multiple output
three-dimensional channel model
channel feature
neural network
