Radio Signal Classification Method based on Adaptive Similarity Entropy Estimation

Radio Signal Classification Method based on Adaptive Similarity Entropy Estimation

연구 분야: Infrastructure

학회: 2023 9th International Conference on Big Data and Information Analytics (BigDIA)

With the rapid development of wireless communications, the identification and classification of radio signals has become particularly critical. However, traditional classification methods based on statistical features often encounter challenges in non-stationary or dynamic signal environments. To solve this problem, this study proposes a radio signal feature extraction and classification method based on adaptive similarity entropy estimation. First, we segment the continuous radio signal data using a sliding time window technique, and then calculate similarity entropy as a dynamic feature for each window. In order to adapt to the non-stationarity of the signal, an adaptive strategy is designed to dynamically estimate the parameters of similarity entropy. Finally, support vector machine (SVM) is combined for feature classification. This method can not only effectively capture the nonlinear and non-stationary characteristics of radio signals, but also has higher robustness and accuracy than traditional methods. This provides a new and effective tool for classifying radio signals in complex electromagnetic environments.