Zhen-Hao Zhang’s research while affiliated with China People's Public Security University and other places

Cascading failures in real systems are common phenomena. Considering the risk awareness during cascade failure and the lack of information of initial loads, a cascading failure model based on local real-time information is proposed in this paper, where the load-redistribution is based on node’s degree. The impact of the load-redistribution strategy on network robustness is analyzed. Our analysis indicates that when the initial load-distribution and the load-redistribution strategy based on the degree in our paper are the same, the network shows better robustness against node failure. What makes more sense is that, without knowing the initial load-distribution, keeping the transfer rate of load-redistribution linearly proportional to the node’s degree is an effective way to improve the network robustness.

How clustering affects network robustness against epidemic propagation is investigated in this paper. The epidemic threshold, the fraction of infected nodes at steady state and epidemic velocity are adopted as the network robustness index. With the help of the networks generated by the 1K null model algorithm (with identical degree distribution), we use three network propagation models (SIS, SIR, and SI) to investigate the influence of clustering against epidemic propagation. The results of simulation show that the clustering of heterogeneous networks has little influence on the network robustness. In homogeneous networks, there is limited increase in epidemic threshold by increasing clustering. However, the fraction of infected nodes at steady state and epidemic velocity evidently decrease with the increase of clustering. By virtue of the generated null models, we further study the relationship between clustering and global efficiency. We find that the global efficiency of networks decreases monotonically with the increase of clustering. This result suggests that we can decrease the epidemic velocity by increasing network clustering.

Defence against cascading failures is of great theoretical and practical significance. A novel load capacity model with a tunable proportion is proposed. We take degree and clustering coefficient into account to redistribute the loads of broken nodes. The redistribution is local, where the loads of broken nodes are allocated to their nearest neighbours. Our model has been applied on artificial networks as well as two real networks. Simulation results show that networks get more vulnerable and sensitive to intentional attacks along with the decrease of average degree. In addition, the critical threshold from collapse to intact states is affected by the tunable parameter. We can adjust the tunable parameter to get the optimal critical threshold and make the systems more robust against cascading failures.