基于节点伪近邻的电路板近邻网络排序算法

Circuit board neighboring network sorting algorithm based on node pseudo-adjacency

在运用近邻网络排序集生成边界扫描测试向量方法中,多以网络局部或全局信息进行节点近邻关系排序,导致伪近邻点的识别排序能力较差。该文结合LeaderRank算法引入节点伪近邻作为局部重要性指标,首先利用LeaderRank求得网络节点的全局重要度,然后基于相关邻居关系提出节点伪近邻比计算方法,最后综合LeaderRank的全局重要度值与节点伪近邻性求得总体重要度,从而获得近邻网络重要度排序。采用所提方法和以往近邻排序算法对实际电路板网络模型进行近邻关系排序,对排序结果进行比较,并用SIR传染病模型进行仿真分析。实验结果表明,所提方法能够弥补以往排序算法的不足,从而获得更为精确的排序结果。

In the method of generating the boundary scan testing vector with the neighboring network sorting sets,the node adjacent relationship is sorted mostly according to the local or global network information,which results in poor recognition and sorting abilities of the pseudo⁃adjacent nodes.The pseudo⁃adjacency node is introduced as the local importance index by means of the LeaderRank algorithm.The global importance of the network node is obtained with the LeaderRank,and then the calculation method of the node pseudo⁃adjacency ratio is proposed based on the related neighboring relationship.The overall importance of the network node is acquired by synthesizing the global importance value of LeaderRank and the pseudo⁃adjacency property of nodes,so as to obtain the importance sorting of the neighboring networks.The neighboring relationship of the real circuit board network models is sorted by means of the method proposed in this paper and the previous neighboring sorting algorithm.The sorting results are compared,and the SIR infectious disease model is used for the simulation analysis.The experimental results show this method can make up for the shortcomings of the previous sorting algorithms,so as to obtain the more accurate sorting results.