For short-period dense array observations, the signal-to-noise ratio of the ambient noise cross-correlation function will seriously affect the picking accuracy of the dispersion curve and the quality and resolution of subsequent tomography. The existing theory and practice show that the non-uniformity of the spatial and temporal distribution of noise sources will affect the signal-to-noise ratio of the subsequent noise cross-correlation function to a certain extent, and in some cases, it is difficult to eliminate this effect through conventional data processing techniques. Therefore, this paper focuses on the design of the observation system. Based on the observation data of the short-period dense array, the dominant azimuth of the high-frequency noise source in Guangdong-Foshan area is analyzed. At the same time, the relationship between the signal-to-noise ratio of the noise cross-correlation function and the superposition duration in the research region is further analyzed. The results show that there is a strong noise source in the south of the research region, which may be related to the ocean activities in the South China Sea, and the ideal noise cross-correlation function can be superimposed through two days of observation. The results will provide an important basis for the optimization of the observation system for the subsequent short-period dense array detection work in the above areas or adjacent areas.