Research Article
ADS-B Signal Separation Via Complex Neural Network
@INPROCEEDINGS{10.1007/978-3-030-89814-4_49, author={Yue Yang and Haipeng Zhang and Haoran Zha and Ruiliang Song}, title={ADS-B Signal Separation Via Complex Neural Network}, proceedings={Mobile Multimedia Communications. 14th EAI International Conference, Mobimedia 2021, Virtual Event, July 23-25, 2021, Proceedings}, proceedings_a={MOBIMEDIA}, year={2021}, month={11}, keywords={Blind source separation ADS-B Complex neural network Hilbert transform}, doi={10.1007/978-3-030-89814-4_49} }- Yue Yang
Haipeng Zhang
Haoran Zha
Ruiliang Song
Year: 2021
ADS-B Signal Separation Via Complex Neural Network
MOBIMEDIA
Springer
DOI: 10.1007/978-3-030-89814-4_49
Abstract
In the sphere of air surveillance, the Automatic Dependent Surveillance-Broadcast is a valuable method. However, the ADS-B system suffers from a considerable overlap issue, and has a significant influence on signal decoding, resulting in incorrect decoding or even data loss. A complicated neural network-based separation approach for ADS-B overlap signals is presented in this paper. Taking the two-signal overlap as the research object, the simulation data set is generated. After the Hilbert transform of the overlap ADS-B signal, it is input into the complex neural network, and finally the predicted waveform of the source signal is output to realize ADS-B signal separation. Experiments have shown that the technique is more efficient and has a lower error rate than previously proposed algorithms.
