Project

Robust Time-Varying Graph Signal Recovery

for Dynamic Physical Sensor Network Data

Eisuke Yamagata, Shunsuke Ono

Abstract

We propose a time-varying graph signal recovery method for estimating the true time-varying graph signal from observations that are corrupted by missing values, unknown position outliers, and some random noise. Most existing studies on time-varying graph signal recovery focus on online estimation and graph learning based on the assumption that the entire dynamic graphs are not available, and robust recovery formulations and algorithms that fully exploit the entire dynamically changing graph structure have not been well studied. On the other hand, there are many situations, especially when dealing with timevarying data obtained from physical sensor networks, where the appropriate dynamic graphs can be easily generated using simple algorithms such as k-nearest neighbor by exploiting information on time-varying spatial sensor locations. In this paper, we focus on such cases and formulate dynamic graph signal recovery as a constrained convex optimization problem that simultaneously estimates both time-varying graph signals and sparsely modeled outliers. In our formulation, we use two types of regularizations, time-varying graph Laplacian-based and temporal differencebased, and also separately modeled missing values with known positions and unknown outliers, to achieve robust estimation from highly degraded data. In addition, an algorithm is developed to efficiently solve the optimization problem based on a primaldual splitting method. Extensive experiments on simulated drone remote sensing data and real sea surface temperature distribution data demonstrate the advantages of the proposed method over existing methods.

Results

[6] P. Berger, G. Hannak, and G. Matz, “Graph signal recovery via primaldual algorithms for total variation minimization,” IEEE Journal of Selected Topics in Signal Processing, vol. 11, no. 6, pp. 842–855, 2017.

[13] K. Qiu, X. Mao, X. Shen, X. Wang, T. Li, and Y. Gu, “Time-varying graph signal reconstruction,” IEEE Journal of Selected Topics in Signal Processing, vol. 11, no. 6, pp. 870–883, 2017.

[43] S. Chen, A. Sandryhaila, J. Moura, and J. Kovacevic, “Signal recovery on graphs: Variation minimization,” IEEE Transactions on Signal Processing, vol. 63, no. 17, pp. 4609–4624, 2015.

Reference

E. Yamagata and S. Ono, "Robust Time-Varying Graph Signal Recovery for Dynamic Physical Sensor Network Data," arXiv: 2202.064432, 2023.

@misc{yamagata2023robust,

      title={Robust Time-Varying Graph Signal Recovery for Dynamic Physical Sensor Network Data}, 

      author={Eisuke Yamagata and Shunsuke Ono},

      year={2023},

      eprint={2202.06432},

      archivePrefix={arXiv},

      primaryClass={eess.SP}

}

Links