Due to the influence of the flight environment, the frequency spectrum of the signal is prone to distortion, which destroys the original modulation structure of the signal. It is difficult to accurately distinguish the characteristics of signals and noise at different frequencies and time scales during the denoising process. Even with denoising processing, it is difficult to fully restore signal quality. Therefore, this study proposes a denoising method for communication signals of aircraft ultra short wave radio based on wavelet packet decomposition. On the basis of constructing a signal recognition network structure, wavelet packet decomposition technology is used to perform modal decomposition on the signal and complete signal denoising processing. Wavelet packet decomposition can subdivide both the high-frequency and low-frequency parts of a signal, providing richer frequency sub bands and more accurately capturing the characteristics of the signal at different frequencies and time scales, thereby distinguishing the effective and noise components in the signal in a more detailed frequency range and optimizing the denoising effect. The experimental results show that after applying this method for denoising, the amplitude of the noise signal is relatively low, indicating that the denoising effect of this method is good.