In order to deal with the interference caused by false automatic dependent surveillance-broadcast (ADS-B) signals on flight trajectory information, a detection system for ADS-B false targets was designed based on Kalman filtering for the prediction of flight trajectories. The message decoding is based on an ADS-B demodulation system using a software-defined radio platform, with the decoding verification part completed on the Qt end and dynamically displayed using Gaode map. An ADS-B fake message transmission system was created, and the track prediction part was completed based on Kalman filtering. The jump rate detection part was designed based on the positional dispersion of the ADS-B predicted data and the root mean square error. According to experimental tests, for the given fake messages, 90.4% of the jumps were successfully detected. Therefore, this system has a certain capability to detect ADS-B false targets.