EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Recursive Radar Signal Deinterleaving by Semantic Segmentation Using Recurrent Neural Networks

Louis Lesieur, Jean‐Marc Le Caillec, Abdelmalek Toumi, Ali Khenchaf, Vincent Guardia

IET Radar, Sonar & Navigation · 2026-01 · 20 (1)

DOI: 10.1049/rsn2.70191 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IET Radar, Sonar & Navigation, online erschienen am 08.07.2026. Verfasst von Louis Lesieur und 4 weiteren Autor:innen. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Association Nationale de la Recherche et de la Technologie.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We propose two methods for radar signal deinterleaving. They are recursive deinterleaving methods, consisting of extracting the waveforms one after another. The problem mostly comes down to pulse selection: One must find pulses corresponding to a waveform. The R‐U segmentation method uses a neural network (NN) to segment the pulse vector in an observed scene. The R‐W hybrid method aims to reproduce the behaviour of a human operator in the task of labelling pulses. It is a combination of two models, the first proposing a set of pulse selections by a cognitive approach, and the second being a NN estimating the relevance of each of these selections to choose the best one. Although they have a different role and do not belong to the same method, the two NNs have a similar architecture. In addition to these two methods, we propose lobe detection algorithms to include the pulse amplitude within the feature set that describes an observed radar waveform along its frequency, pulse width and direction of arrival. To train NNs, we designed a data augmentation process with waveforms derived from real‐world recordings. We also defined metrics for the evaluation of emitter identification and pulse deinterleaving.

Förderung

  • Association Nationale de la Recherche et de la Technologie

Zitieren

APA

Lesieur, L., Caillec, J. L., Toumi, A., Khenchaf, A., & Guardia, V. (2026). Recursive Radar Signal Deinterleaving by Semantic Segmentation Using Recurrent Neural Networks. IET Radar, Sonar & Navigation, 20(1). https://doi.org/10.1049/rsn2.70191

Vancouver

Lesieur L, Caillec JL, Toumi A, Khenchaf A, Guardia V. Recursive Radar Signal Deinterleaving by Semantic Segmentation Using Recurrent Neural Networks. IET Radar, Sonar & Navigation. 2026;20(1). doi:10.1049/rsn2.70191

BibTeX
@article{Lesieur2026,
  author = {Lesieur, Louis and Caillec, Jean‐Marc Le and Toumi, Abdelmalek and Khenchaf, Ali and Guardia, Vincent},
  title = {Recursive Radar Signal Deinterleaving by Semantic Segmentation Using Recurrent Neural Networks},
  journal = {IET Radar, Sonar & Navigation},
  year = {2026},
  volume = {20},
  number = {1},
  doi = {10.1049/rsn2.70191}
}
Kostenlos starten →