Reliability Analysis Toward a Family of Interconnection Networks and Data Center Networks

• Published in: IEEE transactions on reliability pp. 1 - 14
• Main Authors: Wang, Yihong, Fan, Weibei, Cheng, Baolei, Wang, Yan, Fan, Jianxi
• Format: Journal Article
• Language: English
• Published: IEEE 2025
• Subjects: <inline-formula xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <tex-math notation="LaTeX"> h</tex-math> </inline-formula>-extra connectivity; <inline-formula xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <tex-math notation="LaTeX"> h</tex-math> </inline-formula>-extra diagnosability; <inline-formula xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <tex-math notation="LaTeX"> t/m</tex-math> </inline-formula>-diagnosability; Data centers; Fans; Fault diagnosis; Fault tolerance; Fault tolerant systems; fault-tolerant routing (F-TR); Hardware; Optical switches; Reliability; Routing; Topology
• ISSN: 0018-9529; 1558-1721
• DOI: 10.1109/TR.2025.3572117

With the increasing network scale, hardware faults are inevitable. Therefore, research on network reliability under the condition of hardware failure is an important subject. In this article, we investigate <inline-formula><tex-math notation="LaTeX">h</tex-math></inline-formula>-extra connectivity, <inline-formula><tex-math notation="LaTeX">h</tex-math></inline-formula>-extra diagnosability under the PMC and MM<inline-formula><tex-math notation="LaTeX">^{*}</tex-math></inline-formula> models, and <inline-formula><tex-math notation="LaTeX">t/m</tex-math></inline-formula>-diagnosability under the PMC model of recursive networks based on complete graphs (RNCGs) that include not only interconnection network Dragonfly but also data center networks-DCell, generalized DCell, and other unknown networks. In addition, we propose the fault-tolerant routing (F-TR) algorithm, FTPath, which constructs a F-TR in the largest component when the number of faulty nodes is less than the <inline-formula><tex-math notation="LaTeX">h</tex-math></inline-formula>-extra connectivity. Moreover, we evaluate the performance of RNCGs. First, we compare the fault-tolerant performance of RNCGs through experiments. The results show that the average path length constructed by the FTPath algorithm is close to that of the breadth-first search algorithm, and shorter than that of the depth-first search algorithm. Second, we evaluate the performance of this network for different parameters: <inline-formula><tex-math notation="LaTeX">h</tex-math></inline-formula>-extra connectivity, diagnosability under different strategies. The results show that the network has good fault tolerance and fault diagnosis capabilities.