—With the rapid expansion of satellite communication, an increasing number of unattended ground terminals are spreading out to serve local customers. And there is a great demand to equip the terminals with fault identification functionality so that the remote satellite operators can be notified and make immediate response in the event of any local service impairment. In this paper, we correlate signal behaviors of signal-to-noise ratio (SNR) measured at the ground terminal with different type of possible faults, and propose a terminal fault identification (TFI) system that utilizes wavelet technique to filter the SNR measurements and extract from multiple time-scales the signal patterns which are then to be matched with pre-defined fault signatures. The effectiveness of the approach is verified through the analysis over real-world data collected.