An Improved K-Means Radar Signal Pre-selection Method

The K-Means algorithm has a wide range of applications in radar signal pre-selection. Its traditional algorithm is highly dependent on the number of clusters and the initial setting of cluster centers, and is sensitive to noise and outliers. This article proposes an improved K-Means radar signal pre-selection method by introducing the Euclidean distance [1] operator in the clustering space and the category estimation method based on criterion functions. Simulation analysis shows that this method effectively reduces the sensitivity of traditional K-Means algorithm to noise and outliers, and improves the accuracy of radar signal pre-sorting.