Effectiveness investigation of a Fault-Tolerant Ethernet scheme for large-scale star network

In this paper, we analyze and investigate the effectiveness of a novel fault-tolerant scheme for a large-scale star network called Scalable Autonomous Fault-Tolerant Ethernet (SAFE) which overcomes the inherent scalability issue in the conventional software approach based on heartbeat mechanism by dividing a whole network into several subnets and limiting the number of nodes in each subnet. SAFE generates management data traffic to handle fault tolerance in both single-subnet and inter-subnet communication. We develop a simulation which performs the basic operation of SAFE scheme and investigate the influence of the management data traffic to the normal data traffic via several experiments. The results validates the effectiveness of SAFE scheme.