Lab Home | Phone | Search
Center for Nonlinear Studies  Center for Nonlinear Studies
 Home 
 People 
 Current 
 Affiliates 
 Visitors 
 Students 
 Research 
 ICAM-LANL 
 Publications 
 Conferences 
 Workshops 
 Sponsorship 
 Talks 
 Colloquia 
 Colloquia Archive 
 Seminars 
 Postdoc Seminars Archive 
 Quantum Lunch 
 Quantum Lunch Archive 
 CMS Colloquia 
 Q-Mat Seminars 
 Q-Mat Seminars Archive 
 P/T Colloquia 
 Archive 
 Kac Lectures 
 Kac Fellows 
 Dist. Quant. Lecture 
 Ulam Scholar 
 Colloquia 
 
 Jobs 
 Postdocs 
 CNLS Fellowship Application 
 Students 
 Student Program 
 Visitors 
 Description 
 Past Visitors 
 Services 
 General 
 
 History of CNLS 
 
 Maps, Directions 
 CNLS Office 
 T-Division 
 LANL 
 
Tuesday, January 24, 2017
1:00 PM - 2:00 PM
CNLS Conference Room (TA-3, Bldg 1690)

Colloquium

Cascading Overload Failures in Spatially-Embedded and Random Networks with Distributed Flows

Gyorgy Korniss
Rensselaer Polytechnic Institute

In complex information or infrastructure networks, even small localized disruptions can give rise to congestion, large-scale correlated failures, or cascades, -- a critical vulnerability of such networks. Here, we study cascades of overload failures for distributed flows in spatial and non-spatial random graphs, and empirical networks. Our recent results on load-based cascading failures in spatially-embedded random networks (applicable to power grids) underline both the conceptual and computational challenges and difficulties identifying critical nodes, lines, and regions to mitigate cascades. We observed that cascading failures are non-self-averaging in spatial graphs, hence predictability is poor and conventional mitigation strategies are largely ineffective. Among our main findings is that protecting all nodes (or edges) by the same additional capacity (tolerance) may actually lead to larger global failures. I.e., indiscriminately investing resources in the protection of nodes or links can actually make the network more vulnerable against cascading failures (“paying more can result in less”, in terms of robustness).

Host: Angel E. Garcia