t-Diagnosability and Conditional Diagnosability of BCube Networks

With the expansion of data center networks scale and the increase of number of nodes, how to quickly locate failure nodes in the network is necessary in order to avoid network anomalies, so fault diagnosis is significant to measure the reliability of data center network. BCube is proposed as a new server-centric data center network structure, where servers with multiple network ports connect to multiple layers of commodity mini-switches, which makes Bcube have many excellent attributes. In this paper, we studied the t-diagnosability and conditional diagnosability of BCube network structure, and we proved that the t-diagnosability for BCube_k, n under the Preparata, Metze, and Chien's (PMC) model is (n-1)(k+1)-1 for n ≥ 2, k>0. And the conditional diagnosability is larger than ⌊((n-1)(5k+1)-4)/2⌋ for n ≥ 2, k>0.