GLRT-Based Techniques for the Reception of Satellite ADS-B Messages

Satellite-based automatic dependent surveillance-broadcast (ADS-B) enables global-scale aircraft tracking, but introduces significant challenges in packet detection due to the severe signal attenuation experienced along the propagation path. This paper investigates detection strategies for satellite ADS-B receivers based on the generalized likelihood ratio test (GLRT), providing a statistically sound alternative to existing heuristic solutions. To enhance robustness against residual frequency offsets, we exploit the modulus of the received samples, which is invariant to phase distortions. Additionally, we extend the observation window beyond the packet preamble to improve detection performance. To reduce the computational cost of an exact GLRT implementation, we explore several alternatives, including an asymptotic approximation of the Rician distribution and a Gaussian model for the observation data. Simulation results show that the proposed schemes achieve reliable detection even at low signal-to-noise ratios, and outperform existing methods by leveraging both the preamble and part of the data payload.