Distributed maximum correntropy filtering for a class of multi-rate systems over sensor networks

•The distributed maximum correntropy filtering problem is investigated.•A class of multi-rate systems over sensor networks is considered.•The sampling periods are allowed to be different.•A novel correntropy-based filtering performance index is constructed.•The desired filter gain is calculated usin...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Information fusion Jg. 127; S. 103908
Hauptverfasser: Liu, Yang, Wang, Zidong, Hu, Jun, Chen, Yun, Dong, Hongli
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.03.2026
Schlagworte:
Distributed filtering
Fixed-point algorithm
Maximum correntropy criterion
Multi-rate systems
Non-Gaussian noises
Sensor networks
Maximum correntropy criterion
Sensor networks
Multi-rate systems
Non-Gaussian noises
Fixed-point algorithm
Distributed filtering
ISSN:1566-2535
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•The distributed maximum correntropy filtering problem is investigated.•A class of multi-rate systems over sensor networks is considered.•The sampling periods are allowed to be different.•A novel correntropy-based filtering performance index is constructed.•The desired filter gain is calculated using a fixed-point algorithm. In this paper, the distributed maximum correntropy filtering problem is investigated for a class of multi-rate systems over sensor networks. The system under consideration is monitored by a sensor network whose sensors are permitted to have different sampling periods. The maximum correntropy criterion is employed to handle the non-Gaussian noises effectively. Given the distributed nature and the asynchronous sampling of the sensor network, a novel filtering performance index based on the correntropy is constructed for each node in the sensor network. To maximize this correntropy-based index, the desired filter gain at each time step is calculated using a fixed-point algorithm. Sufficient conditions for ensuring the convergence of the fixed-point method are established. Finally, the efficacy of the proposed filtering scheme is demonstrated through a target tracking example.
ISSN:1566-2535
DOI:10.1016/j.inffus.2025.103908