.png){kind=icon}
返回
A Link Quality Estimation Method for Wireless Sensor Networks Based on Deep Forest
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2020.3047648.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
In wireless sensor networks, sensor nodes, the miniature embedded devices, have limitation of energy, storage, computing, and etc. One of the tasks of the nodes is to use their limited resources to complete work efficiently. Choosing high quality link communication can effectively save energy. In this paper, we propose a link quality estimation model that is based on deep forest. To avoid a noise sample becoming a center point in the clustering, we use an improved K-medoids algorithm based on step increasing and optimizing medoids (INCK) when dividing the link quality grades. During the sample preprocessing stage, the Pauta criterion is used to delete the noise link samples, and we fill the mean value of each grade into the missing values. The feature extraction performance of deep forest is improved by combining the stratified sampling to change the unbalance distribution of link quality samples. And then the Stratified Sampling Cascade Forest link quality estimation (SCForest-LQE) is constructed by combining stratified sampling with cascade forest. The experiments are conducted in three real application scenarios. Compared with the existing six link quality estimation models, SCForest-LQE has better estimation performance and stability.
Keyword:
Wireless sensor networks
Clustering algorithms
Wireless sensor networks
link quality estimation
deep forest
stratified sampling
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:
3.6
论文数:
9.8W
被引数:
29.4W
机构
引用论文
LSTM based link quality confidence interval boundary prediction for wireless communication in smart grid.png)
COMPUTING
COMPUTINGIF2.8
PREAI
分享收藏
查看更多内容