EUVIMEDEUVIMED entdecken
Crossref

GSP-KalmanNet: Tracking Graph Signals via Neural-Aided Kalman Filtering

Itay Buchnik, Guy Sagi, Nimrod Leinwand, Yuval Loya, Nir Shlezinger, Tirza Routtenberg

IEEE Transactions on Signal Processing · 2024 · 72 · 3700-3716

DOI: 10.1109/tsp.2024.3435935 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Signal Processing, 2024. Verfasst von Itay Buchnik und 5 weiteren Autor:innen. Die Arbeit wurde bisher 37-mal zitiert (Crossref) und nennt 76 Literaturangaben. Gefördert von: Israeli Ministry of National Infrastructure, Energy, and Water Resources, Israel Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Dynamic systems of graph signals are encountered in various applications, including social networks, power grids, and transportation. While such systems can often be described as state space (SS) models, tracking graph signals via conventional tools based on the Kalman filter (KF) and its variants is typically challenging. This is due to the nonlinearity, high dimensionality, irregularity of the domain, and complex modeling associated with real-world dynamic systems of graph signals. In this work, we study the tracking of graph signals using a hybrid model-based/data-driven approach. We develop the GSP-KalmanNet, which tracks the hidden graphical states from the graphical measurements by jointly leveraging graph signal processing (GSP) tools and deep learning (DL) techniques. The derivations of the GSP-KalmanNet are based on extending the KF to exploit the inherent graph structure via designing a graph frequency domain filtering and replacing the Kalman gain (KG) with a graph filter that minimizes the prediction error. Thus, it considerably simplifies the computational complexity entailed in processing high-dimensional signals and increases the robustness to small topology changes. Then, we use data to learn the KG, namely, the graph filter, following the recently proposed KalmanNet framework, which copes with partial and approximated modeling, without forcing a specific model over the noise statistics. Restricting the KG to a graph filter in the proposed GSP-KalmanNet reduces learned parameters, thereby enhancing stability. Our empirical results demonstrate that the GSP-KalmanNet achieves enhanced accuracy and run time performance, and improved robustness to model misspecifications compared with both model-based and data-driven benchmarks.

Förderung

  • Israeli Ministry of National Infrastructure, Energy, and Water Resources
  • Israel Science Foundation

Zitieren

APA

Buchnik, I., Sagi, G., Leinwand, N., Loya, Y., Shlezinger, N., & Routtenberg, T. (2024). GSP-KalmanNet: Tracking Graph Signals via Neural-Aided Kalman Filtering. IEEE Transactions on Signal Processing, 72, 3700-3716. https://doi.org/10.1109/tsp.2024.3435935

Vancouver

Buchnik I, Sagi G, Leinwand N, Loya Y, Shlezinger N, Routtenberg T. GSP-KalmanNet: Tracking Graph Signals via Neural-Aided Kalman Filtering. IEEE Transactions on Signal Processing. 2024;72:3700-3716. doi:10.1109/tsp.2024.3435935

BibTeX
@article{Buchnik2024,
  author = {Buchnik, Itay and Sagi, Guy and Leinwand, Nimrod and Loya, Yuval and Shlezinger, Nir and Routtenberg, Tirza},
  title = {GSP-KalmanNet: Tracking Graph Signals via Neural-Aided Kalman Filtering},
  journal = {IEEE Transactions on Signal Processing},
  year = {2024},
  volume = {72},
  pages = {3700--3716},
  doi = {10.1109/tsp.2024.3435935}
}
Kostenlos starten →