Font Size: a A A

Research On Node Importance And Cascading Failures Oriented On Invulnerability Of Networks

Posted on:2019-01-31Degree:MasterType:Thesis
Country:ChinaCandidate:Z H ZhangFull Text:PDF
GTID:2428330566495979Subject:Information security
Abstract/Summary:PDF Full Text Request
The rapid development of networks brings convenience to people's life.However,it also arouses security problems and catastrophic incidents,thus causing large-scale cascading failures.These incidents seriously affect people's life and threaten the stability of the society.Based on these,details of this thesis are as follows.(1)An improved weighted Leader Rank is proposed,which aims to identify influential spreaders in social networks.The effects of clustering and degree are considered to characterize the weights of links.The effectiveness of the proposed algorithm can be verified by conducting SIR model simulations on real online social networks.(2)A novel load capacity model with a tunable proportion of load redistribution is proposed.Clustering and degree are considered to redistribute the loads.The proposed model is applied on artificial networks as well as two real networks.Simulation results show that the critical threshold from collapse to intact states is affected by the tunable parameter,and the scale of cascading failures can be rapidly controlled by adjusting the tunable parameter.(3)Based on the proposed load capacity model,how topology structure affects the invulnerability of networks is studied in this section.Degree distribution and average degree are major factors that influence invulnerability of networks when networks are faced with intentional attacks.Highly clustered structure cannot guarantee the invulnerability,and plays a negative role.Under the random attacks,scale-free networks and random graphs get more robust with the increase of average degree.Especially for scale-free networks,the changes of critical threshold present power exponential decline.The changes of critical threshold for small-world networks present normal distribution.Overall,the changes of invulnerability for small-world networks are quite stable,except for some situations corresponding to several parameters.Oriented to invulnerability of networks,identifying influential nodes,modeling cascading failures and analyzing properties affecting the invulnerability are major research.
Keywords/Search Tags:influential nodes, LeaderRank, clustering coefficient, cascading failures, invulnerability, transition phase
PDF Full Text Request
Related items