Reliability is ascertained in an efficient agile network through structural simplicity, storage of route-set description at each edge node, rapid reporting of a failure to edge nodes that may be affected, and rapid rerouting of affected connections. Structural simplicity results in routes of a small number of hops each, thus enabling rapid rerouting. A proposed wide-coverage high-capacity network that scales to multiples of petabits per second and can encompass millions of edge nodes employs both fast-switching time-shared optical nodes and channel-switching optical core connectors. The core capacity is adaptively shared. Thus, failure of a proportion of core connectors may reduce network connectivity but would not result in a major service discontinuity. Several other simple structures can be devised to enable rapid reporting and rerouting, thus realizing economical highly-reliable networks.
展开▼