Research on Atmospheric Duct Interference Model based on CNN-GRU

Atmospheric Duct Interference (ADI) is a natural phenomenon that significantly hampers long-distance wireless communications, necessitating robust prediction and mitigation techniques to maintain network quality. In contrast, machine learning has offered various predictive strategies. To solve problems such as signal degradation and loss by ADI in wireless communication systems, a novel deep learning mode for the Atmospheric Duct Interference Model is proposed based on CNN-GRU, which integrates Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU) to detect, predict, and mitigate ADI, enhance communication reliability and performance. Its application approach includes ADI Data Collection and Preprocessing and constructing Model Architecture based on CNN-GRU. The experimental results show that the deep learning-based model, employing Convolutional Neural Networks (CNNs) and Gated Recurrent Units (GRUs), significantly improves the prediction and mitigation of Atmospheric Duct Interference (ADI). This approach enhances the reliability and efficiency of wireless communication systems by dynamically adapting to changing atmospheric conditions and robustly detecting and characterising ADI events in real-time.