.png){kind=icon}
.png){kind=icon}
Return
Widely linear complex valued adaptive graph filtering algorithm
.png){kind=icon}
PRE
AI
DOI：10.1016/j.dsp.2025.105144.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With the widespread application of graph signal processing in fields such as social networks and sensor networks, developing efficient graph filtering algorithms capable of handling multidimensional or complex signals has become a key research focus. In the context of centralized graph structures, this paper proposes a Widely Linear Complex Adaptive Graph Filtering Algorithm (WL-CAGF). This algorithm innovatively extends graph filtering to the complex domain (CAGF) and, based on a widely linear model, captures the non-circular characteristics of signals by considering both the input signal and its conjugate. This enables more effective processing of non- circular signals and enhances adaptability to complex and dynamic graph signals. Simulation results show that the proposed WL-CAGF can efficiently process input signals from sensor network graphs than the traditional CAGF algorithm, and especially performs well in processing non-circular signals, significantly improving signal denoising and feature extraction.
Keywords:
Graph signal processing
Graph filter
Adaptive filter
Widely linear