References & Citations
Condensed Matter > Disordered Systems and Neural Networks
Title: Loops of any size and Hamilton cycles in random scale-free networks
(Submitted on 23 Feb 2005 (v1), last revised 28 Feb 2005 (this version, v2))
Abstract: Loops are subgraphs responsible for the multiplicity of paths going from one to another generic node in a given network. In this paper we present an analytic approach for the evaluation of the average number of loops in random scale-free networks valid at fixed number of nodes N and for any length L of the loops. We bring evidence that the most frequent loop size in a scale-free network of N nodes is of the order of N like in random regular graphs while small loops are more frequent when the second moment of the degree distribution diverges. In particular, we find that finite loops of sizes larger than a critical one almost surely pass from any node, thus casting some doubts on the validity of the random tree approximation for the solution of lattice models on these graphs. Moreover we show that Hamiltonian cycles are rare in random scale-free networks and may fail to appear if the power-law exponent of the degree distribution is close to 2 even for minimal connectivity grater than 3.
Submission history
From: Ginestra Bianconi [view email][v1] Wed, 23 Feb 2005 12:48:19 GMT (52kb)
[v2] Mon, 28 Feb 2005 13:58:34 GMT (49kb)
Link back to: arXiv, form interface, contact.