Ambiguity function based on specific radiation sources of ADS-B signals identification

As Automatic Dependent Surveillance-Broadcast (ADS-B) devices are widely used, verification of the authenticity of ADS-B signals becomes increasingly important. False signals can disrupt normal aircraft navigation and seriously threaten airspace safety. This paper proposes an ADS-B signal recognition method based on the ambiguity function (AF). It utilizes two-dimensional convolutional images to represent radio frequency fingerprint (RFF) characteristics in both time and frequency domains by convolving the actual received signal with an ideal reconstructed signal. A convolutional neural network is used to extract the fingerprint information of the signal and verify the identity of the radiation source, achieving the purpose of determining the authenticity of the signal. The experiments analyzed the impact of the number of aircraft categories and signal-to-noise ratio on recognition results, confirming the effectiveness of ADS-B recognition based on the AF. This study demonstrates the feasibility of RFF in the aerospace industry and can be generalized for use in Internet of Things (IoT) devices.