Translation and Stationarity for Graph Signals

Benjamin Girault 1, 2 Paulo Gonçalves 1, 2 Eric Fleury 1, 2
1 DANTE - Dynamic Networks : Temporal and Structural Capture Approach
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme, IXXI - Institut Rhône-Alpin des systèmes complexes
Abstract : We propose a novel time shift like operator called the graph translation. We enforce that it is isometric and define it such that it shares with the time shift operator its key properties. Using the graph translation operator and its isometry, we propose a tractable definition of stationary graph signals and characterise stationary graph signals using their spectral properties. We illustrate this characterisation on synthetic stationary random graph signals. Finally we propose a method to test the stationarity of a real dataset.
Document type :
Reports
Complete list of metadatas

Cited literature [13 references]  Display  Hide  Download

https://hal.inria.fr/hal-01144991
Contributor : Benjamin Girault <>
Submitted on : Thursday, April 23, 2015 - 2:51:13 PM
Last modification on : Thursday, February 7, 2019 - 3:44:29 PM
Long-term archiving on : Wednesday, April 19, 2017 - 3:42:11 AM

File

RR-8719.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01144991, version 1

Citation

Benjamin Girault, Paulo Gonçalves, Eric Fleury. Translation and Stationarity for Graph Signals. [Research Report] RR-8719, École Normale Supérieure de Lyon; Inria Rhône-Alpes; INRIA. 2015. ⟨hal-01144991⟩

Share

Metrics

Record views

425

Files downloads

668