巢湖学院学报 ›› 2020, Vol. 22 ›› Issue (6): 86-90.doi: 10.12152/j.issn.1672-2868.2020.06.012

• 信息科学 • 上一篇    下一篇

基于卷积神经网络的轻量化通信网络动态路由算法设计

杨静:安徽三联学院 基础实验教学中心   

  1. 安徽三联学院 基础实验教学中心,安徽 合肥 230601
  • 收稿日期:2020-07-10 出版日期:2020-11-25 发布日期:2021-02-02
  • 作者简介:杨静(1982—),女,安徽合肥人,安徽三联学院基础实验教学中心讲师,主要从事计算机网络、计算机视觉研究。
  • 基金资助:
    安徽省省级质量工程项目(项目编号:2018MOOC083);安徽三联学院校级科研重点项目(项目编号:KJZD2020007)

Design of Dynamic Routing Algorithm for Lightweight Communication Network Based on Convolutional Neural Network

YANG Jing:Basic Experimental Teaching Center, Anhui Sanlian University   

  1. Basic Experimental Teaching Center, Anhui Sanlian University, Hefei Anhui 230601
  • Received:2020-07-10 Online:2020-11-25 Published:2021-02-02

摘要: 为提高轻量化通信网络动态传输能力,提出基于卷积神经网络的轻量化通信网络动态路由算法。在建立轻量化通信网络动态节点部署模型的基础上,通过网络节点轮换调度方法建立网络动态路由探测模型。然后结合分布式转换协议重组轻量化通信网络动态路由特征空间结构,在建立节点多重覆盖调度模型后,结合传感信息融合跟踪方法检测网络动态路由特征。基于此,建立路由转发和自适应控制协议,采用节点遍历方法进行轻量化通信网络的信道均衡控制,在最优节点通信覆盖约束下,实现轻量化通信网络动态路由调度,再采用卷积神经网络进行轻量化通信网络动态路由控制的收敛性调节。仿真结果表明,采用该方法后,信道干扰得到有效抑制,且网络吞吐量得以提高,证明轻量化通信网络动态路由的节点转发能力较好。

关键词: 卷积神经网络, 轻量化通信网络, 动态路由, 控制协议

Abstract: In order to improve the dynamic transmission capability of lightweight communication networks, a dynamic routing algorithm for lightweight communication networks based on convolutional neural networks is proposed. Based on the establishment of a lightweight communication network dynamic node deployment model, a network dynamic routing detection model is established through the network node rotation scheduling method. Then the distributed conversion protocol is used to reorganize the lightweight communication network dynamic routing feature space structure. After establishing the node multiple coverage scheduling model, the sensor information fusion tracking method is used to detect the network dynamic routing feature. Based on this, the routing forwarding and adaptive control protocol are established, and the node traversal method is used to control the channel balance of the lightweight communication network. Under the constraint of optimal node communication coverage, the lightweight communication network dynamic routing scheduling is realized, and then the convolutional neural network is used to carry out the convergence adjustment of the dynamic routing control of the lightweight communication network. The simulation results show that the channel interference can be effectively suppressed and the network throughput can be improved after adopting this method, which proves that the node forwarding ability of dynamic routing in the lightweight communication network is better.

Key words: convolutional neural network, lightweight communication network, dynamic routing, control protocol

中图分类号: 

  • TP393