基于序列判定的多维网络双向链路稳定性预测
Multidimensional Network Bidirectional Link Stability Prediction Based on Sequence Decision
王少英 1王福忠2
作者信息
- 1. 黄河交通学院智能工程学院,河南 焦作 454950
- 2. 河南理工大学电气学院,河南 焦作 454000
- 折叠
摘要
为了确保多维网络双向链路的稳定运行,提出一种多维网络双向链路稳定性预测算法.根据多维网络维数定义和小世界网络特性分析结果,获取多维双向网络的主要特征,通过提取的特征建立多维网络模型,采用稀疏编码网络方法对模型降维处理.以结构相似性为基础,计算降维处理后多维网络双向链路中不同节点之间的相似度,通过相似度排序结果获取一个连边序列,以连边序列为判定依据实现多维网络双向链路稳定性预测.实验结果表明,采用所提算法可以获取高精度、高效率的预测结果,有利于提升网络链路的稳定性.
Abstract
In order to ensure the stable operation of bidirectional links in multi-dimensional networks,an algo-rithm for predicting the stability of bidirectional links in multi-dimensional networks was proposed.Based on the defi-nition of multi-dimensional network dimensions and analysis results of small-world network characteristics,the main features of multi-dimensional bidirectional networks were obtained.Moreover,a multi-dimensional network model was built by extracting the features.Furthermore,the sparse encoding network method was adopted to reduce the dimen-sionality of the model.After that,the similarity between nodes in the reduced-dimensional bidirectional link was cal-culated by the structural similarity.Meanwhile,a link sequence was obtained through the similarity ranking result.Fi-nally,the sequence was used as a basis for determining the stability of bidirectional links in the multi-dimensional network.Experimental results show that the proposed algorithm can obtain high-precision and efficient prediction re-sults,which is conducive to improving the stability of network links.
关键词
多维网络/双向链路/稳定性预测/稀疏编码网络方法/相似度Key words
Multi-dimensional network/Bidirectional link/Stability prediction/Sparse coding network method/Similarity引用本文复制引用
基金项目
河南省科技攻关项目(222102320218)
河南省科技攻关项目(212102210146)
出版年
2024
NETL
NSTL
维普
万方数据