Abstract: Ground Penetrating Radar (GPR) echo signals are susceptible to various clutter interferences in complex environments during actual detection, which seriously affects the interpretation and identification of radar data. Conventional single clutter suppression methods are difficult to effectively suppress direct wave, inclined linear waves and background noise simultaneously. To address this issue, this paper proposes a GPR clutter suppression method based on F-K Gaussian polar angle filtering and multiple singular spectrum analysis. Firstly, a peak-matched mean subtraction filter is designed to filter out the direct wave with the strongest energy in the radar profile. Subsequently, in view of the distribution differences between target signals and inclined linear waves in the frequency-wavenumber (F-K) domain, a Gaussian polar angle filter is adopted to construct an angular filtering window in the F-K domain, so as to accurately suppress linear waves with different inclination angles. Finally, multiple singular spectrum analysis is adopted to reconstruct radar data containing different frequencies into Hankel matrices for singular value decomposition. K-means clustering is introduced to realize adaptive classification and screening of singular values and separate background noise, which further optimizes the overall denoising effect of radar profiles. Results of numerical simulations and actual measurements demonstrate that the proposed method can effectively suppress the interference caused by direct wave, inclined linear waves and background noise, improve the signal to clutter ratio of radar data, and provide reliable data support for the accurate interpretation and identification of underground targets.