Cascading failures are common in most of the networks, which can result in a congestion regime with degradation in the network performance. In order to investigate how a small shock can trigger avalanches mechanisms affecting a considerable fraction of the network, lots of failure models have been constructed. In this paper, Based on our AHP network cascading model, we have estimated how the efficiency will be affected when coefficients of K, S, T changed, we find the fact that the network efficiency of BA network is determined by its attacked types, and the efficiency is largely influenced by the attacked types of K and S, and under the same number of failure node, the efficiency under attacks of types T and I are relatively higher than that of the efficiency under attacks of types S and K , and the importance I is largely determined by the proportion of T, when the node failure number is equal, the higher proportion of T, the higher efficiency it is under I type attacks. We also tabled some proposals for reducing the damage that the networks suffered from the cascading failures.
Copyrights © 2013