<p>Rainfall is one of the most important elements of meteorology, which varies in time and place. This study introduces a novel framework for optimizing the location and density of rain gauge stations by integrating the entropy of information transfer with the Cuckoo Optimization Algorithm (COA). Using rainfall data from April (1993–2013) across 38 stations in the Gavkhouni Basin, Iran, two scenarios were evaluated. The first scenario was obtained by estimating the border and joint entropy of each station, and the amount of transferable information of the desired stations. The stations whose amount of transferable information was classified as weak or very weak were identified, and, based on the zoning of transferable information in the basin, the locations for installing additional stations were determined. In this scenario, five stations should be added to the existing stations. In the second scenario, assuming a blank network, the COA was used to generate the synthetic rainfall data, leading to repositioning 13 stations to maximize the transferable entropy. The results indicate an 8% increase in overall information transfer, demonstrating the efficiency of this hybrid approach in improving the network design and hydrological data quality.</p>