Joint Time-Vertex Fractional Fourier Transform

Tuna Alikaşifoğlu,Bünyamin Kartal,Eray Özgünay,Aykut Koç

2024-07-10

Abstract:Graph signal processing (GSP) facilitates the analysis of high-dimensional data on non-Euclidean domains by utilizing graph signals defined on graph vertices. In addition to static data, each vertex can provide continuous time-series signals, transforming graph signals into time-series signals on each vertex. The joint time-vertex Fourier transform (JFT) framework offers spectral analysis capabilities to analyze these joint time-vertex signals. Analogous to the fractional Fourier transform (FRT) extending the ordinary Fourier transform (FT), we introduce the joint time-vertex fractional Fourier transform (JFRT) as a generalization of JFT. The JFRT enables fractional analysis for joint time-vertex processing by extending Fourier analysis to fractional orders in both temporal and vertex domains. We theoretically demonstrate that JFRT generalizes JFT and maintains properties such as index additivity, reversibility, reduction to identity, and unitarity for specific graph topologies. Additionally, we derive Tikhonov regularization-based denoising in the JFRT domain, ensuring robust and well-behaved solutions. Comprehensive numerical experiments on synthetic and real-world datasets highlight the effectiveness of JFRT in denoising and clustering tasks that outperform state-of-the-art approaches.

Signal Processing,Social and Information Networks

What problem does this paper attempt to address?

The paper attempts to address the problem of how to perform more flexible and efficient analysis and processing of time-vertex joint signals in Graph Signal Processing (GSP). Specifically, the paper introduces the Joint Fractional Fourier Transform (JFRT), an extension of the traditional Joint Time-Vertex Fourier Transform (JFT), which enables fractional-order analysis in both the time and vertex domains. Through this extension, JFRT provides more degrees of freedom, allowing for signal processing in the intermediate time-frequency domain, which helps improve the performance of signal processing tasks such as denoising, signal reconstruction, and node classification.

The main contributions of the paper include:
1. **Introduction of JFRT**: A new transformation method, the Joint Fractional Fourier Transform (JFRT), is proposed for analyzing signals simultaneously in the fractional time domain and the fractional vertex domain.
2. **Tikhonov Regularization Denoising**: A denoising method based on Tikhonov regularization is derived in the JFRT domain, ensuring more robust and well-posed solutions when dealing with noisy data.
3. **Experimental Validation**: Extensive numerical experiments on synthetic and real datasets demonstrate the effectiveness of JFRT in filtering, denoising, and clustering tasks, proving its performance superior to existing state-of-the-art methods.

Overall, the paper aims to provide a new and more powerful tool for the field of graph signal processing by introducing JFRT to tackle complex time-vertex signal processing tasks.