Research on Communication Signals Optimization based on Deep Learning

At present, communication signals has become an essential component in privacy information transmission and social media establishment. The quality and noises of communication signals has crucially influenced on information transmission. Therefore, the optimization for communication signal is an essential method to enhance transmission ability and improve the communication channels. However, existing optimization mechanism are concentrated on digital signals processing techniques and wavelet transform, which are primarily based on the Fourier transform theory and leak the basic function for optimization process. In this paper, we utilize the advanced deep learning method to train an automatic optimization model to eliminate the noises in communication signals. The model is consisted by a pro-processing module, which is a residual network to simulate as filter and full-connected layer to reduce output dimensions. Subsequently, we add the white Gaussian noise into communication signals 4ASK and 4PSK to evaluate the proposed with existing optimization methods. From our extensive experimental simulation results, we can conclude that our model can achieve approximately 90% optimization accuracy in the different signal-to-noise ratios with reasonable computation costs.