Clutter is the radar noise signal that is reflected from the elements surrounding the targets. Since clutter degrades the radar system's range and doppler frequency detection capability, clutter suppression is a critical signal-processing algorithm that can improve the performance of a radar system. This paper proposes a ground-clutter suppression method using denoising encoder-decoder deep learning network with dual encoding channels, residual connections, and skip connections. Radar signal dataset pipeline was generated using MATLAB in order to train the network. In this paper, deep learning-based clutter suppression method that can be applied in various operating conditions, is discussed.