Reference node selection method based on the combination of spectral clustering and random selection
A reference node and random selection technology, applied in the field of information security, can solve problems such as the deviation of the predicted node delay from the actual value and prediction failure, so as to avoid the failure of prediction delay and reduce the effect of error
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0037] Such as figure 1 As shown, a reference node selection method based on the combination of spectral clustering and random selection in the present invention includes the following steps: Step S101: Perform spectral clustering on network nodes.
[0038] Step S102: Determine whether the network node is an outlier, if yes, remove the outlier; if not, directly proceed to the next step.
[0039] Step S103: Randomly select a reference node.
[0040] Step S104: Optimizing the selected reference nodes.
Embodiment 2
[0042] Such as figure 2 As shown, another reference node selection method based on the combination of spectral clustering and random selection of the present invention includes the following steps:
[0043] Step S201: performing spectral clustering on network nodes, including:
[0044] Step S2011: Construct all points in the network node delay data set into a weighted undirected graph. All vertices in the graph are all network nodes in the data set, and the weight is the similarity between two node delay vectors. In order to calculate The similarity of the delay vector, the Gaussian kernel function is introduced to calculate the similarity of the delay vector of two nodes, the weight calculation formula is shown in formula (1):
[0045]
[0046] Among them, F i and F j is the time-delay feature vector of the two nodes, and the correlation between the two can be calculated by using the above formula. The greater the weight of the edge, the more similar the time-delay vec...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


