摘要
针对5G无线传播模型的构建,使用Pearson系数量化特征与目标值之间的相关性,以此构造出新的特征。将这些特征送入到BP神经网络、决策树、随机森林中来建立无线传播模型,并且能够预测新环境下无线信号覆盖的强度。该模型为建立精准的无线网络提供技术支持,使网络建设成本降低,并提高建设效率。
For building 5G wireless propagation model,the Pearson coefficient was used to measure the correlation between the target and characteristics,and new features were constructed.Then these features were input to the BP neural network,decision tree,random forests to build wireless propagation model,and they could predict the strength of the wireless signal under the new environment.Our method can provide technical support for building accurate wireless networks,and it greatly reduces the network construction cost and improves the construction efficiency.
作者
谭海军
朱世宇
单欲立
陈善雄
Tan Haijun;Zhu Shiyu;Shan Yuli;Chen Shanxiong(Information Office,Yangtze Normal University,Chongqing 408100,China;College of Computer and Internet of Things,Chongqing Institute of Engineering,Chongqing 400056,China;College of Computer and Information Science,Southwest University,Chongqing 400715,China)
出处
《计算机应用与软件》
北大核心
2022年第2期120-127,共8页
Computer Applications and Software
基金
国家自然科学基金项目(61303227)。
关键词
无线网络
决策树
神经网络
传播模型
Wireless network
Decision tree
Neural network
Propagation model
