Eur. Phys. J. B 49, 513-522 (2006)
https://doi.org/10.1140/epjb/e2006-00085-1

Topological patterns in street networks of self-organized urban settlements

¹ Centre de Recherches sur la Cognition Animale, CNRS UMR 5169, Université Paul Sabatier, 118 route de Narbonne, 31062 Toulouse Cedex 4, France
² Laboratoire NXI GESTATIO, Département de Design, Université du Québec à Montréal, Case Postale 8888, Succursale Centre-Ville, Montréal Québec, H3C 3P8, Canada
³ ICREA-Complex Systems Lab,Universitat Pompeu Fabra, Dr Aiguader 80, 08003 Barcelona, Spain
⁴ Laboratoire d'Informatique Nantes-Atlantique, Université de Nantes, BP 92208, 2 rue de la Houssinière, 44322 Nantes Cedex 03, France

Corresponding author: ^a buhl@cict.fr

Received: 7 September 2005
Revised: 5 January 2006
Published online: 31 March 2006

Abstract

Many urban settlements result from a spatially distributed, decentralized building process. Here we analyze the topological patterns of organization of a large collection of such settlements using the approach of complex networks. The global efficiency (based on the inverse of shortest-path lengths), robustness to disconnections and cost (in terms of length) of these graphs is studied and their possible origins analyzed. A wide range of patterns is found, from tree-like settlements (highly vulnerable to random failures) to meshed urban patterns. The latter are shown to be more robust and efficient.