基于轻量级深度网络的信号检测与样式识别OA

Signal detection and recognition are key technologies in wireless communication and spectrum sensing.Recently,deep learning-based approaches for signal detection and recognition have gained significant research attention.However,deep learning models typically involve a vast number of parameters and high computational complexity,making their deployment dependent on high-performance computing platforms.To enable the practical application of deep learning-based signal detection and pattern recognition on low-cost edge devices,this study focuses on lightweight techniques for deep learning models.A model sparsity method based on component importance measurement is proposed,along with a lightweight model parameter optimization strategy using imitation learning.These techniques significantly reduce the model size while maintaining the accuracy of signal detection and recognition.