Graph signal processing: Analyzing and filtering data over networks

Elvin Isufi

Nowadays, we are surrounded by massive data that reside on irregular structures. Examples include data generated by biological, financial, and sensor networks. Graphs indeed offer the ability to model the interactions between them. Current efforts in signal processing are being focused on providing analysis and processing tools for these data such that the underlying structure is taken into account. These include the development of a Fourier transform and filtering operations for graph data. The field that gathers these tools is called graph signal processing (GSP).

     This talk will be focused on the fundamentals of GSP spanning concepts like the signal variation over a graph, the graph Fourier transform, the graph filter, and its distributed implementation. The talk will be concluded with some illustrative examples and future work directions.

Additional information ...