Graph Signal Adaptive Message Passing

TL;DR

This paper introduces GSAMP, an innovative message passing technique that adaptively processes time-varying graph signals for online prediction, missing data imputation, and noise removal using localized computations.

Contribution

GSAMP is a novel adaptive message passing method that performs localized updates for real-time processing of dynamic graph signals, unlike traditional global filtering approaches.

Findings

Effective in handling real-world, time-varying graph signals

Robust against Gaussian and impulsive noise

Improves prediction and data imputation accuracy

Abstract

This paper proposes Graph Signal Adaptive Message Passing (GSAMP), a novel message passing method that simultaneously conducts online prediction, missing data imputation, and noise removal on time-varying graph signals. Unlike conventional Graph Signal Processing methods that apply the same filter to the entire graph, the spatiotemporal updates of GSAMP employ a distinct approach that utilizes localized computations at each node. This update is based on an adaptive solution obtained from an optimization problem designed to minimize the discrepancy between observed and estimated values. GSAMP effectively processes real-world, time-varying graph signals under Gaussian and impulsive noise conditions.