The moving target signal is usually masked in the strong and broadened clutter in a space borne early warning radar system, indicating the necessity of effective clutter suppression processing. To solve the problem of high computational complexity of the space-time adaptive processing (STAP) algorithm, the reduced-dimension STAP algorithm is proposed, which conventionally utilizes the discrete Fourier transform (DFT) to perform the Doppler localization processing. However, the non-synchronous sampling of clutter is unavoidable, leading to the clutter spectral leakage, degrading the clutter suppression performance. In this paper, an extended factored approach with weight window is analyzed, which applies the weight window to suppress the spectral leakage effect. The clutter suppression performances after using different weight windows are analyzed according to the simulated space borne radar data. As a conclusion, the extended factored approach with weight window shows a better clutter suppression performance, and the application of Hann window exhibits the best performance. Finally, the theoretical analyses are verified by the processing results of real-measured airborne radar data.