.png){kind=icon}
返回
Efficient combination policies for diffusion adaptive networks
.png){kind=icon}
PRE
AI
DOI：10.1007/s12083-019-00726-2.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Diffusion adaptive networks (DANs) have many applications such as signal processing, mobile wireless sensor network and the internet of things (IoT). Unlike the classical centralized networks, a DAN uses the information exchange among local neighbors to solve global problems. The performance of the DAN highly depends on the combination matrix policies, which raises the issue of the optimal selection of the combination matrix. However, traditional combination policies focus on either the steady-state error or the convergence speed. Inspired by the solution of minimizing the mean square deviation (MSD) of the DAN, this paper proposes two efficient adaptive combination policies: 1) relative-instantaneous-error combination policy and 2) relative-deviation combination policy. These two policies are related to the inverse of noise by different metrics. Computer simulations verify that the proposed combination policies outperform the existing combination rules in either steady-state error or convergence rate in various noise environments. Finally, we apply the two combined rules to the collaborative target-tracking problem and achieve expected results.
Keyword:
Combination policy
Collaborative target tracking
Diffusion adaptive network
Signal processing
Internet of things (IoT)
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
2.6
论文数:
2.2K
被引数:
2.9K
机构
暂无机构信息
引用论文
.png){kind=icon}
分享收藏
查看更多内容