Constant False Alarm Rate Detection Method Based on Convolutional Neural Networks

This paper proposes an improved variability index constant false alarm rate detector utilizing convolutional neural networks. By combining statistical measures and neural networks to jointly determine different environment and improve strategies for various environments, the proposed method enhances the detection performance of variability index constant false alarm rate detector in multi-target environments. Simulation results demonstrate that the proposed method achieves a higher probability of accurate environmental classification, while maintaining robust detection performance in various scenarios, significantly enhancing the algorithm's adaptability in complex environments.