Light Gradient Boosting Machine-Based Link Quality Prediction for Wireless Sensor Networks

Algorithm 1

Grading link quality based on AHCMG algorithm.

Input:  D (link quality data set), (the number of clustering), (the number of link quality grades).

Output: Link quality grades.

1: Take each sample in the link quality sample set as a cluster;

2: Calculate Euclidean distance for cluster merging by using;

3: When the number of clusters is reduced to , the agglomerative hierarchical clustering is changed to manual division;

4: According to the sample distribution among clusters, the distance between clusters, the link quality grade is manually graded into categories;

5: Rank link quality grade according to the average PRR of each cluster;

6: return: Link quality grades.