An automatic failure isolation and reconfiguration methodology for fiber distributed data interface (FDDI)

Abstract

An automatic failure isolation and reconfiguration methodology for the fiber distributed data interface (FDDI) is described. The authors categorize physical-related errors into several types such as link failure at the primary ring, link failure at the secondary ring, cable failure, and subsequent failures after ring reconfiguration. A reconfiguration algorithm was designed to locate and bypass both single and multiple link failures as well as cable failures. By introducing several flags and management frames, the network is capable of locating the errors and can carry out different ring reconfiguration strategies based on the type of errors. Each node stores the knowledge of the ring status by using these flags. This technique is capable of providing the maximum possible connectivity among the nodes and avoiding any unnecessary extension of transmission length. The new reconfiguration methodology is explained by using pseudocode.<>