Abstract: The detection of the global three-dimensional atmospheric wind field plays a crucial role in improving the accuracy of numerical weather prediction, enhancing the ability of meteorological disaster early warning, and ensuring the safety of aerospace activities. At present, the main methods for detecting the atmospheric wind field include ground-based wind profile radars, spaceborne Doppler lidars, and millimeter-wave cloud radars, etc. Ground-based wind profile radars are difficult to deploy in oceanic and remote terrestrial areas, making it impossible to achieve global networked observations. Spaceborne Doppler Lidars can accurately detect the atmospheric wind field in clear-sky regions, but they are unable to obtain data on the wind field within clouds. Compared with lasers, millimeter waves have better penetration capabilities and possess unique advantages in the detection of the wind field within clouds. However, existing spaceborne millimeter-wave cloud radars cannot provide information on the horizontal wind field. Spaceborne millimeter-wave cloud radars with a conical scanning system can achieve the detection of the three-dimensional atmospheric wind field within clouds. In this paper, a signal simulation system for the atmospheric wind field detection within clouds by a spaceborne W-band Doppler radar was established, which provided a theoretical basis and technical reference for in-cloud wind field detection. The required elements of the signal simulation system include atmospheric profile data, reflectivity factor, attenuation coefficient, radar system parameters, spaceborne observation geometry, echo signal amplitude, echo signal frequency, echo signal phase, echo signals, and radial velocity. The system processes actual cloud profile data and simulates radar echo signals to estimate the radial wind speed. The effects of Signal-to-Noise Ratio (SNR) and pulse-pair cumulative number on the estimation accuracy of radial wind speed were systematically analyzed. The results show that for a spaceborne W-band Doppler radar, when the polarization pulse interval is not greater than 20 μs, the polarization diversity pulse-pair technique can achieve a wind speed detection range of 0 to 40 m·s^–1, effectively obtaining high wind speed products within clouds. The estimation accuracy of radial wind velocity is positively correlated with the increase in SNR and the number of pulse-pair accumulations. Under the index conditions of the radar system described in this paper, when the SNR is 0 dB and the number of pulse-pair accumulations is 64, the estimation accuracy of the radial wind speed is 1.34 m·s^–1, which can meet the wind measurement accuracy requirement of 2 m·s^–1 for numerical weather prediction.