Weighted Mean and Median graph Filters with Attenuation Factor for Sensor Network

TL;DR

This paper introduces a novel weighted attenuation k-hop graph and a node selecting graph to enhance graph filters, improving signal denoising performance in sensor networks.

Contribution

It proposes a new graph construction method and applies it to mean and median filters, achieving better denoising results than traditional filters.

Findings

Weighted attenuation k-hop graph effectively models spatial neighbors.

Proposed filters outperform traditional median filter in denoising tasks.

Node selecting graph improves median filter performance.

Abstract

This paper proposes a weighted attenuation k-hop graph, which depicts the spatial neighbor nodes with their hops from the central node. Based on this k-kop graph, we further propose a node selecting graph, which selects temporal neighbor nodes of multiple instances of the central node. With this node selecting graph, we propose a graph mean filter. In addition, we also apply the proposed node selecting graph to the median filter. Finally, the experimental results show that the proposed mean filter performs better than the original median filter in the signal denoising polluted by white noise and the median filter using node selecting graph also has better performance than the original median filter.