TY - JOUR
T1 - Multiple discontinuous percolation transitions on scale-free networks
AU - Chen, Wei
AU - Zheng, Zhiming
AU - Jiang, Xin
AU - D'Souza, Raissa M.
N1 - Publisher Copyright:
© 2015 IOP Publishing Ltd and SISSA Medialab srl.
PY - 2015/4/29
Y1 - 2015/4/29
N2 - Percolation transitions in networks, describing the formation of a macroscopic component, are typically considered to be robust continuous transitions in random percolation. Yet, a class of models with various rules of connecting edges were recently devised which can lead to discontinuous transitions at percolation threshold. Here we study the Bohman-Frieze-Wormald process on scale-free networks constructed via a modified configuration model. We show via numerical simulation that multiple discontinuous transitions appear in the thermodynamic limit for the degree distribution exponent λ ∈ [2, λc) with λc ∈ (2.3, 2.4). For λ ∈ (λc, 5] this model undergoes a unique discontinuous transition in the thermodynamic limit, but for any finite system a second discontinuous transition occasionally appears at some point above percolation threshold due to the aggregation of two existing giant components. For all values of the exponent λ ∈ [2, 5] we observe a pronounced right-hump in the evolution of component size distribution providing further evidence that the percolation transition is discontinuous at percolation threshold.
AB - Percolation transitions in networks, describing the formation of a macroscopic component, are typically considered to be robust continuous transitions in random percolation. Yet, a class of models with various rules of connecting edges were recently devised which can lead to discontinuous transitions at percolation threshold. Here we study the Bohman-Frieze-Wormald process on scale-free networks constructed via a modified configuration model. We show via numerical simulation that multiple discontinuous transitions appear in the thermodynamic limit for the degree distribution exponent λ ∈ [2, λc) with λc ∈ (2.3, 2.4). For λ ∈ (λc, 5] this model undergoes a unique discontinuous transition in the thermodynamic limit, but for any finite system a second discontinuous transition occasionally appears at some point above percolation threshold due to the aggregation of two existing giant components. For all values of the exponent λ ∈ [2, 5] we observe a pronounced right-hump in the evolution of component size distribution providing further evidence that the percolation transition is discontinuous at percolation threshold.
KW - networks
KW - percolation problems (theory)
KW - random graphs
UR - https://www.scopus.com/pages/publications/84947205265
U2 - 10.1088/1742-5468/2015/04/P04011
DO - 10.1088/1742-5468/2015/04/P04011
M3 - 文章
AN - SCOPUS:84947205265
SN - 1742-5468
VL - 2015
JO - Journal of Statistical Mechanics: Theory and Experiment
JF - Journal of Statistical Mechanics: Theory and Experiment
IS - 4
M1 - P04011
ER -