원문정보
초록
영어
The existing recognition methods of radiation source signal are extremely sensitive to the change of signal-to-noise ratio (SNR). In order to solve this problem, in this paper, it proposes a radar signal recognition method based on fractal box dimension and neural network under the condition of unstable SNR. Firstly, it extracts box counting dimension features of four different radiation source signals, and then uses the characteristic values of stable box counting dimension as the inputs of the neural network. Finally, it can recognize the four different kinds of radiation source signals. Simulation results show that, taking the box counting dimension characteristics as the input of the neural network to train, test and classify, it has good recognition rate in a certain changed range of SNR, and it has more widely application in the application environment.
목차
1. Introduction
2. Basic Theoretical Knowledge
2.1. Theory of Fractal Dimension
2.2. Basic Theory of Neural Network
3. System Identification Block Diagram based on Fractal and Neural Network
4. Simulation Results and Analysis
5. Conclusion
Acknowledgements
References