Noise Reduction Algorithm for Communication based on Improved Generative Adversarial Network

There are many researches on noise suppression of traditional noise suppression algorithms, and the suppression effect has reached the bottleneck. Now the widely studied artificial intelligence algorithm provides a new idea for very low frequency (VLF) communication noise reduction in ocean navigation. The generative adversarial network is introduced into the noise interference suppression algorithm.Based on the improved generative adversarial network model, the VLF noise interference suppression algorithm is studied. The generation model with strong learning ability is used to improve the accuracy of expected signal frequency estimation in the spectrum of noisy signal, and further improve the signal-to-noise (SNR) ratio in the signal bandwidth. The experimental results show that, by training the generative adversarial network, the received signals in the direction of - 45° to 45° of the receiving antenna show good noise reduction effect, and the accuracy rate can reach above 90%.