The sea target detection ability of HF ground wave radar is closely related to the suppression effect of ionospheric clutter, and the complexity and diversity of ionospheric clutter bring difficulties to the suppression. In order to suppress ionospheric clutter, the chaotic characteristics of ionospheric clutter are analyzed. On this basis, a suppression method based on improved particle swarm optimization to optimize wavelet neural network is proposed, which solves the shortcomings of particle swarm optimization that it is easy to premature and fall into local optimization. An adaptive probability mutation strategy is proposed to enrich the population diversity, the particle swarm optimization can jump out of the current optimal and find the global optimal in the whole iterative process. The experimental results show that the wavelet neural network optimized based on the improved particle swarm optimization algorithm can basically predict the value of ionospheric clutter. After the suppression of ionospheric clutter, the signal-to-noise ratio is effectively improved, and the research on the suppression of ionospheric clutter is of great significance.