In this paper, we present a centralized method for real-time rainfall estimation using carrier-to-noise power ratio ( C/N ) measurements from broadband satellite communication networks. The C/N data of both forward link and return link are collected by the gateway station from the user terminals in the broadband satellite communication network and stored in a database. The C/N for such Ka-band scenarios is impaired mainly by the rainfall. Using signal processing and machine learning techniques, we develop an algorithm for real-time rainfall estimation. Extracting relevant features from C/N , we use artificial neural network in order to distinguish the rain events from dry events. We then determine the signal attenuation corresponding to the rain events and examine an empirical relationship between rainfall rate and signal attenuation. Experimental results are promising and prove the high potential of satellite communication links for real environment monitoring, particularly rainfall estimation.