A layered approach to fault-tolerance and timeliness issues in SAFENET

We investigate the fault-tolerance issue of networking and propose a layered solution to it. By evaluating the performance of fault-handling schemes at each layer of a protocol stack, we can find a set of fault-handling schemes whose combined performance will guarantee the required end-to-end delay bound for real-time communication services while maintaining a certain level of fault-tolerance. The Survivable Adaptable Fiber Embedded Network (SAFENET) is used as the main vehicle of this study.<
>