| Abstract [eng] |
In this Master’s thesis, methods for detecting ADS-B packets transmitted by aircraft are analyzed. The objective of the work is to examine various methods for detecting the ADS-B signal preamble, test them through computer simulations to identify the most effective solution, and validate the developed algorithm using real hardware, demonstrating its functionality under real-world conditions. In the review section, various aircraft navigation systems are examined, ranging from satellite-based GNSS signals to stationary ILS stations. Their functional characteristics and technical capabilities are discussed. The ADS-B system is selected as the primary focus of the work, and a detailed analysis of its signal characteristics and data packet structure is conducted. Subsequently, different methods for detecting ADS-B signals are analyzed. Based on scientific literature, two primary approaches are identified and reviewed: the application of digital indicators and correlation methods. Various digital indicator algorithms are examined, analyzing the phase, amplitude, and time-domain pulse characteristics of the received signal. In the analysis of the correlation method, the specifics of correlation computation are explored, considering two key parameters: the shape and length of the correlation mask and the detection threshold value with possibility of applying adaptive methods for its determination. A brief overview is also provided of the hybrid CFAR method, the application of AI, and the impact of hardware solutions on signal acquisition. In the methodological section, the focus shifts to testing correlation computation using computer simulations. The goal is to replicate the methods discussed in scientific papers by applying different correlation mask shapes. To achieve this, sample ADS-B signals are synthesized, and their correlation is computed. By generating a large number of signals, the dependency between signal energy and preamble correlation amplitude is determined, and an adaptive correlation threshold criterion is derived for preamble detection. During implementation, performance metrics are collected for various signal-to-noise ratios. Finally, the correlation algorithm is implemented in a real-time signal acquisition system based on an FPGA matrix. The developed system is tested with sample ADS-B signals transmitted from an SDR transmitter. Conclusions - the analysis of ADS-B signal reception methods revealed two primary contenders, of which the correlation algorithm with six different correlation masks was further investigated and tested. A variable correlation threshold method was successfully applied to signals of varying amplitudes, demonstrating that a static threshold value would fail to detect the signal. The feasibility of synthesizing and broadcasting ADS-B signals into the environment using an SDR transmitter and the GNU Radio framework was also tested. Functionality was validated with an ADS-B signal reception system supported by an FPGA matrix. The implemented correlation and adaptive threshold computations demonstrated that the algorithm can be successfully realized in practice and aligns with the results of theoretical simulations. |
