Novel Confidence Level Labelling Methods for ADS-B Signals with Coherent Demodulation Scheme

In the satellite-based application, low signal-to-noise ratio (SNR) is a considerable obstacle for automatic dependent surveillance-broadcast (ADS-B), leading to some erroneous bit declarations. In this case, the confidence level of each bit should be declared legitimately to make full use of error correction. In this study, the authors focus on methods to optimise the confidence declaration for ADS-B signals with coherent demodulation scheme. A novel optimal-threshold labelling method is firstly proposed for known and unknown amplitudes. They also present a dynamic-threshold labelling method, in which a specific threshold is unnecessary. Detailed theoretical analyses are carried out for the proposed methods and are verified by simulations. Finally, they make a comparison between different confidence level labelling methods in terms of reception performance and complexity. A conclusion is set forth that the methods have a high probability of correct reception at low SNRs.