Performance Analysis of Radar Scan Pattern Classification using Visibility Graph

Accurate radar scan pattern classification is essential for identifying and responding to threats in electronic warfare systems. In this paper, we present an approach to enhance classification accuracy by leveraging visibility graphs to extract critical features from radar signals and applying the GoogLeNet deep neural network. By transforming radar scan patterns into visibility graphs, the method captures both global and local characteristics that are crucial for distinguishing between similar patterns. Simulation results demonstrate that the proposed approach achieves over $\mathbf{9 2 \%}$ in terms of accuracy, precision, and recall, highlighting its effectiveness in radar scan pattern classification.