首页> 外文期刊>Optical fiber technology >A real-time fiber mode demodulation method enhanced by convolution neural network
【24h】

A real-time fiber mode demodulation method enhanced by convolution neural network

机译:卷积神经网络增强的实时光纤模式解调方法

获取原文
获取原文并翻译 | 示例
           

摘要

Inspired by the outstanding performance of artificial intelligence for various tasks, in this work we explore a new method for the fiber modal demodulation based on convolutional neural networks (CNN). The neural network is trained by using a self-made data set and is utilized to obtain the rough modal information. Based on this initial information, the interior point (IP) algorithm is afterward used to explore the precise optimal value. Simulation results indicate that this approach has stable performance and excellent accuracy. In addition, the method also has a satisfactory real-time characteristic.
机译:受人工智能在各种任务中的出色表现的启发,在这项工作中,我们探索了一种基于卷积神经网络(CNN)的光纤模态解调的新方法。通过使用自制数据集来训练神经网络,并利用该神经网络来获得粗略的模态信息。基于此初始信息,内部点(IP)算法随后将用于探索精确的最佳值。仿真结果表明,该方法性能稳定,精度高。另外,该方法还具有令人满意的实时特性。

著录项

  • 来源
    《Optical fiber technology》 |2019年第7期|139-144|共6页
  • 作者单位

    Northeastern Univ, Coll Informat Sci & Engn, Shenyang 110819, Liaoning, Peoples R China;

    Northeastern Univ, Coll Informat Sci & Engn, Shenyang 110819, Liaoning, Peoples R China;

    Northeastern Univ, Coll Informat Sci & Engn, Shenyang 110819, Liaoning, Peoples R China|Northeastern Univ, State Key Lab Synthet Automat Proc Ind, Shenyang 110819, Liaoning, Peoples R China;

    Northeastern Univ, Coll Informat Sci & Engn, Shenyang 110819, Liaoning, Peoples R China;

  • 收录信息 美国《科学引文索引》(SCI);美国《工程索引》(EI);
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    Optical fiber; Modal demodulation; Convolutional neural networks;

    机译:光纤;模态解调;卷积神经网络;

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号