Abstract: Efficient methods for rerouting failed channels on to spare channels in a multichannel transmission system are described. One method uses a minimal amount of rerouting to reroute all the channels if one or two line cards fail. Another set of methods reroute up to two failed channels on to spare channels in the event of channel failures. With this method only the failed channels are rerouted. However the failed channels may have to be routed through a slightly larger, but minimal, number of additional hops than with the first method.

Abstract: The invention provides a method for improving network routing calculations for routing data packets or connections between two points in a network through many incompatible sub-networks. Each node in the network comprises a plurality of agents, written in migrating executable code, which represent each sub-network or domain in the overall network. Hence, an originating node wishing to send data over the network need not concern itself with the topology of the links or communicate directly with other sub-networks. Instead, the originating node can calculate a path through the network locally, simply by querying its agents. Hence, this paradigm is based on an object-oriented design, in which the network controller (NC) at one node does not attempt to interpret topological data of a different node or sub-network. Instead, a network controller at the originating node activates the agents which represent the other sub-networks, and queries the agents (a black-box approach).

Abstract: A fault detecting apparatus and method for a network node of an optical transmission system receives a wavelength division multiplexed (WDM) optical signal which includes a group of optical channels. The network node includes an optical demultiplexer, optical switch, optical multiplexer, and an optical channel add/drop device. Supervisory optical signals having a different wavelength than the wavelengths of the group of optical channels are generated and spliced with each optical channel by the optical demultiplexer. The supervisory channel optical power is tapped off at various points in the network node, such as the output optical channels of the demultiplexer, dropped optical channel outputs, output optical channels from the optical switch, and the output WDM optical signal of the multiplexer. The supervisory channels are monitored by a wavemeter for bypass lightpaths and dropped lightpaths.

Abstract: A wiring scheme for providing loop-back capability for individual failed channels around a ring network composed of nodes with limited space-switching. With this invention each node has a switch for interconnecting channels between incident links of the node. Each secure channel i has a backup channel B(i). In the normal mode of operation, the switch is configured so that the ports for channel B(i) of the incident links are connected to each other. Upon failure of channel i in one of the incident links of the node, the port for channel i of the failed link is connected to the port for channel B(i) on the other incident link of the node. With this invention when there is a failure of channel i on one of the links, the connection using this failed channel can be restored by switch configurations at only two nodes of the ring since the backup channels at the other nodes are already connected to each other.

Abstract: Efficient methods for rerouting failed channels on to spare channels in a multichannel transmission system are described. One method uses a minimal amount of rerouting to reroute all the channels if one or two line cards fail.Another set of methods reroute up to two failed channels on to spare channels in the event of channel failures. With this method only the failed channels are rerouted. However the failed channels may have to be routed through a slightly larger, but minimal, number of additional hops than with the first method.