Abstract:Choosing an appropriate frequency definition and norm is critical in graph signal sampling and reconstruction. Most previous works define frequencies based on the spectral properties of the graph and use the same frequency definition and $\ell_2$-norm for optimization for all sampling sets. Our previous work demonstrated that using a sampling set-adaptive norm and frequency definition can address challenges in classical bandlimited approximation, particularly with model mismatches and irregularly distributed data. In this work, we propose a method for selecting sampling sets tailored to the sampling set adaptive GFT-based interpolation. When the graph models the inverse covariance of the data, we show that this adaptive GFT enables localizing the bandlimited model mismatch error to high frequencies, and the spectral folding property allows us to track this error in reconstruction. Based on this, we propose a sampling set selection algorithm to minimize the worst-case bandlimited model mismatch error. We consider partitioning the sensors in a sensor network sampling a continuous spatial process as an application. Our experiments show that sampling and reconstruction using sampling set adaptive GFT significantly outperform methods that used fixed GFTs and bandwidth-based criterion.

What problem does this paper attempt to address?

This paper attempts to solve the problem of choosing appropriate frequency definitions and norms in graph signal sampling and reconstruction. Specifically, traditional graph signal processing methods usually define frequencies based on the spectral characteristics of the graph and use the same frequency definition and $\ell_2$ norm to optimize all sampling sets. However, this method faces challenges when dealing with model mismatches in classical band - limited approximations and irregularly distributed data. This paper proposes a new method to improve these problems by choosing frequency definitions and norms adapted to specific sampling sets, especially when collecting data of spatially continuous processes in sensor networks. ### Main Contributions 1. **Adaptive Frequency Definition and Norm**: - A graph Fourier transform (GFT) method based on sampling - set adaptation is proposed, which can better handle model mismatches and irregularly distributed data. - By concentrating the band - limited model mismatch error in the high - frequency part and using the spectral folding property to track these errors, the robustness of reconstruction is improved. 2. **Sampling Set Selection Algorithm**: - A sampling set selection algorithm is designed to minimize the worst - case band - limited model mismatch error. - This algorithm is especially suitable for partitioning the nodes in a sensor network into multiple subsets, and each subset uses its own sampling - set - adaptive GFT for interpolation. 3. **Experimental Verification**: - It is experimentally verified that the proposed method is significantly superior to the fixed GFT and bandwidth - criterion methods when dealing with sensor network data of spatially continuous processes. - The experimental results show that the performance improvement of this method is particularly obvious when dealing with rapidly changing spatial signals. ### Technical Details - **Graph Fourier Transform (GFT)**: - An adaptive inner product based on the sampling set is defined, thus allowing different frequency representations. - Through the spectral folding property, the high - frequency energy is folded into the low - frequency part, making it easier to isolate and minimize the error. - **Sampling Set Selection**: - It is assumed that the graph Laplacian matrix is approximated to the inverse covariance matrix of the data, and the signal has a decaying power spectrum. - An adaptive GFT interpolator is used, which depends on the sampling set rather than a pre - fixed subspace. - By calculating the energy covariance of the high - frequency band, the sampling set selection problem is formulated as an E - optimal design problem. - **Application Case**: - The nodes in the sensor network are partitioned into multiple subsets, and each subset samples alternately within different time windows, and the unsampled sensor data is obtained by interpolation. - This method can reduce the active time of sensors and prolong the life of sensors. ### Conclusion The adaptive frequency definition and norm method proposed in this paper, as well as the corresponding sampling set selection algorithm, significantly improve the performance of graph signal sampling and reconstruction, especially in the case of dealing with irregularly distributed data and model mismatches. Future work can further extend this algorithm to support the distributed implementation of large - scale sensor networks.

Graph-Based Signal Sampling with Adaptive Subspace Reconstruction for Spatially-Irregular Sensor Data

Irregularity-Aware Bandlimited Approximation for Graph Signal Interpolation

Local-Set-Based Graph Signal Reconstruction

Greedy Local-Set Based Sampling and Reconstruction for Band-Limited Graph Signals

Learning to Sample: Data-Driven Sampling and Reconstruction of FRI Signals

Eigendecomposition-Free Sampling Set Selection for Graph Signals

Sampling and Reconstruction of Spatial Fields using Mobile Sensors

Adaptive Graph Signal Processing: Algorithms and Optimal Sampling Strategies

Low-complexity Graph Sampling with Noise and Signal Reconstruction via Neumann Series

Reconstruction of Bandlimited Graph Signals from Random Local Sampling

Reconstructing Graph Signals from Noisy Dynamical Samples

Iterative Reconstruction of Graph Signal in Low-Frequency Subspace.

Graph-based Reconstruction of Time-Varying Spatial Signals

Sampling of Graph Signals with Successive Aggregations Based on Graph Fractional Fourier Transform.

Fast Graph Sampling Set Selection Using Gershgorin Disc Alignment

Energy-Aware Sampling Set Selection for Signal Reconstruction in Internet of Things

Reconstruction-Cognizant Graph Sampling Using Gershgorin Disc Alignment

Adaptive Sampling for Signals Associated with the Special Affine Fourier Transform

Efficient Directed Graph Sampling via Gershgorin Disc Alignment

Sensor selection on graphs via data-driven node sub-sampling in network time series

Clustering-Aided Graph Signal Sampling and Reconstruction for Large-Scale Sensor Networks