Abstract: For the purpose of reducing the false alarms during the ADS-B message preamble detection, a double-threshold preamble detection method was proposed. The method analyzed the sliding relevant feature between the local preamble and the received signal, then given the Constant False Alarm Rate (CFAR) threshold and the half-amplitude threshold resrespectively for the focus signal and the background signal. By combining double threshold, the method was effective for low SNR environment and could significantly suppress false alarms before cyclic redundancy check. The simulation results showed that the double threshold detection method had similar detection performance to CFAR method in low SNR range and lower mean sample false alarm rate in high SNR range. The result of the present work implied that the double threshold detection method was an ADS-B message preamble detection method with false alarms suppression ability, which is applicable to a high dynamic SNR range.