摘要
针对长期演进LTE网络上行干扰分类模型中噪声敏感、训练时间长的问题,建立了结合堆栈降噪自编码器与极限学习机的LTE网络上行干扰分析模型。使用上行干扰原始数据无监督地预训练堆栈降噪自编码(SDAE)提取高层抽象特征,并为极限学习机(ELM)分类器提供初始参数。该模型发挥了ELM收敛快和SDAE抑制噪声的优势,同时克服了ELM参数随机赋值造成的鲁棒性不足的问题。实验结果表明,该模型提高了LTE网络上行干扰分析的效率,并具有较强的鲁棒性。
Aiming at the problems of noise sensitivity and long training time in the uplink interference classification model of LTE(long Term Evolution)network,this article establishes a LTE network uplink interference analysis model combining stacked denoising autoencoder and extreme learning machine,using uplink interference raw data to unsupervised pre-training of SDA(Stacked denoising autoencoder) to extract high-level abstract features,and to provide initial parameters for the ELM(Extreme learning machine)classifier. The model takes advantage of ELM’s fast convergence and SDA’s noise suppression,and at the same time overcomes the problem of insufficient robustness caused by random assignment of ELM parameters. Experimental results show that this model improves the efficiency of LTE network uplink interference analysis,and at the same time has strong robustness.
作者
许鸿奎
姜彤彤
李鑫
姜斌祥
王永雷
XU Hong-kui;JIANG Tong-tong;LI Xin;JIANG Bin-xiang;WANG Yong-lei(School of Information and Electrical Engineering,Shandong Jianzhu University,Jinan 250101,China;Shandong Key Laboratory of Intelligent Building Technology,Shandong Jianzhu University,Jinan 250101,China;Juvenile Crime&Justice Research Center,China University of Political Science&Law,Beijing 100088,china;AI Research Institute,Hunan ENHT Technology Co.,Ltd.,Qingdao 266000,China)
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2022年第1期195-203,共9页
Journal of Jilin University:Engineering and Technology Edition
基金
国家重点研发计划项目(2017YFC0803604)
山东省重大科技创新工程项目(2019JZZY010120).
关键词
LTE网络上行干扰
降噪自编码器
极限学习机
特征提取
long term evolution network uplink interference
denoising autoencoder
extreme learning machine
feature extraction
