摘要
针对移动通信系统中的无线网络,当前的拥塞控制方法在部署数据传输和处理节点时,部署方式多为单目标模式,即每个节点只关注自己的目标任务,导致控制效率较低、数据丢包率偏高。为此,文章基于负载预测结果,设计了新的无线移动网络数据拥塞控制方法。首先,根据当前的拥塞控制需求,采用多阶的方式部署拥塞控制节点。其次,根据网络负载预测结果构建拥塞控制模型。最后,采用深度Q-网络(Deep Q-Leaning Network,DQN)辅助处理的方式控制网络数据拥塞。测试结果表明:相比于传统控制方法,应用文章方法后,数据吞吐量更高,数据丢包率被控制在1.5%以下,说明文章方法具有实际应用价值。
For wireless networks in mobile communication systems,current congestion control methods often deploy data transmission and processing nodes in a single target mode,where each node only focuses on its own target task,resulting in low control efficiency and high data packet loss rate.Therefore,this study designs a new data congestion control method for wireless mobile networks based on load prediction results.Firstly,based on the current congestion control requirements,a multi-level approach is adopted to deploy congestion control nodes.Secondly,construct a congestion control model based on the network load prediction results.Finally,Deep Q-Leaning Network(DQN)assisted processing is adopted to control network data congestion.The test results show that compared to traditional control methods,the method proposed in this paper has higher data throughput and a data packet loss rate controlled below 1.5%,indicating that the method proposed in this paper has practical application value.
作者
张苑
ZHANG Yuan(School of Information Engineering,Baoding University,Baoding Hebei 071000,China)
出处
《信息与电脑》
2024年第5期175-177,共3页
Information & Computer
关键词
负载预测
无线移动网络
网络数据
拥塞控制
控制方法
数据整合
load forecasting
wireless mobile network
network data
congestion control
control method
data integration
作者简介
张苑(2001—),男,河北张家口人,本科在读。研究方向:计算机科学与技术。
