Abstract: A method and system of routing variable-length packet data across a wave division multiplex (WDM) communications network having a plurality of data communications channels comprises inverse-multiplexing each data packet into a respective frame. Each frame includes a label block containing label information of the frame, and two or more respective payload blocks having a predetermined length. The label block contains encoded routing information, a start time, and, possibly an end time, and is transmitted over a label channel of the communications network. The start time preferably indicates a delay between launching a first bit of the label clock and the first bit of the payload blocks. The end time may be a bit count indicative of the location of the last bit of the data packet within the frame. The data packet is divided into a plurality of data segments, each of which is transported across the network within a respective payload block.

Abstract: In a method of providing protection switching in a meshed communications network, a protection path between two nodes on the network is only established, on demand, as it is required, after a fault along a working path between the two nodes has been detected. After the protection path is established, carriage of traffic from the working path is switched to the protection path. Once the working path is restored, traffic is switched back from the protection path to the working path, and the protection path is torn down. Switching between working and protection channels may be effected using a path establishment protocol or using protection switching mechanisms of existing network protocols, such as SONET or ATM APS. Software and network nodes embodying the method are also disclosed.

Abstract: In a mesh communications network, in which working and protection paths may be established, channels used to carry protection traffic between nodes are shared across multiple protection paths. Channels need only be shared if sharing does not adversely impact network usage. If working and protection paths become susceptible to single points of failure, channels need not be shared.

Abstract: In a method of label selection for end-to-end transport of label switched traffic through a communications network between a source node and a destination node, a request message is launched toward the destination node from the source node. The request message includes a label list having one or more label identifiers indicative of respective corresponding labels available for use by the source node. The label list is revised, at a cross-connect service each successive hop between the source node and the destination node, based on labels available for use by each respective hop, to produce a reduced label list. The reduced label list includes label identifiers indicative of respective corresponding labels available for end-to-end transport of label switched traffic between the source node and the destination node.

Abstract: Resources of a shared physical network element of a communications network are dynamically allocated between connection-oriented traffic and connectionless traffic. For each shared physical network element of the network, a resource requirement of the connection-oriented traffic is determined; and a respective traffic metric to be used for routing connectionless traffic is dynamically adjusted based on the determined resource requirement of the connection-oriented traffic. As a result, resources of the shared physical network element can be efficiently utilized, and congestion of connectionless traffic being routed through the shared physical network element is avoided.

Abstract: In a switching protection system, a protocol is followed by a first connection controller to help facilitate the resynchronization of the network after the failure has occurred. This protocol contains the messaging system of the ability to query and store the exact sequence of logical ports and time slots that make up the data path, as a resynchronization table coupled to the first controller. The head/ingress end of the path can receive this information during set-up and can store it both locally and in shadow locations. The system also contains the ability to distinguish between signaling and data channel failures, as signaling channel failures to do not clear cross connect data but simply leave it “ownerless”. The system also contains the ability to populate the exact sequence of logical ports and time slots accumulated in the definition of the network, as the set-up message initiated by the head controller to re-claim “ownerless” cross connections contained on respective network elements.

Abstract: A process for selecting the best possible route for an alternate communications path in connection-oriented networks takes into account the reserved network resources and the set of links and/or nodes used by a current communications path. Whenever a request for route selection includes the resources reserved by the current communications path, the alternate communications path is computed considering the network resources reserved by the current communications path as available, in addition to the available network resources existing in the topology database. In another embodiment of the invention, the topology database is initially modified such that the network resources reservations used by the current communications path are removed from the topology database. The alternate communications path is then computed. Finally, the topology database is restored to reflect again the network resources reservations used by the current communications path.

Abstract: A method and system of routing variable-length packet data across a wave division multiplex (WDM) communications network having a plurality of data communications channels comprises inverse-multiplexing each data packet into a respective frame. Each frame includes a label block containing label information of the frame, and two or more respective payload blocks having a predetermined length. The label block contains encoded routing information, a start time, and, possibly an end time, and is transmitted over a label channel of the communications network. The start time preferably indicates a delay between launching a first bit of the label clock and the first bit of the payload blocks. The end time may be a bit count indicative of the location of the last bit of the data packet within the frame. The data packet is divided into a plurality of data segments, each of which is transported across the network within a respective payload block.

Abstract: User traffic of one type must traverse many diverse networks, each operating by different paradigms and requiring different protocol stacks, before it reaches the destination. MPLS (Multiprotocol Label Switching) is being developed to solve some of the problems which involve the existence of different protocols along the traffic path. There is currently no accepted way of transporting ATM cells over an MPLS domain. New ways of packetizing ATM cells into labeled MPLS packets for transporting over an MPLS domain are described.

Abstract: In a method of label selection for end-to-end transport of label switched traffic through a communications network between a source node and a destination node, a request message is launched toward the destination node from the source node. The request message includes a label list having one or more label identifiers indicative of respective corresponding labels available for use by the source node. The label list is revised, at a cross-connect service each successive hop between the source node and the destination node, based on labels available for use by each respective hop, to produce a reduced label list. The reduced label list includes label identifiers indicative of respective corresponding labels available for end-to-end transport of label switched traffic between the source node and the destination node.

Abstract: In a method of providing protection switching in a meshed communications network, a protection path between two nodes on the network is only established, on demand, as it is required, after a fault along a working path between the two nodes has been detected. After the protection path is established, carriage of traffic from the working path is switched to the protection path. Once the working path is restored, traffic is switched back from the protection path to the working path, and the protection path is torn down. Switching between working and protection channels may be effected using a path establishment protocol or using protection switching mechanisms of existing network protocols, such as SONET or ATM APS. Software and network nodes embodying the method are also disclosed.

Abstract: In a mesh communications network, in which working and protection paths may be established, channels used to carry protection traffic between nodes are shared across multiple protection paths. Channels need only be shared if sharing does not adversely impact network usage. If working and protection paths become susceptible to single points of failure, channels need not be shared.

Abstract: In a switching protection system, a protocol is followed by a first connection controller to help facilitate the resynchronization of the network after the failure has occurred. This protocol contains the messaging system of the ability to query and store the exact sequence of logical ports and time slots that make up the data path, as a resynchronization table coupled to the first controller. The head/ingress end of the path can receive this information during set-up and can store it both locally and in shadow locations. The system also contains the ability to distinguish between signaling and data channel failures, as signaling channel failures to do not clear cross connect data but simply leave it “ownerless”.

Abstract: An optical medium containing a transport signal that has a plurality of sub-signals, each sub-signal being characterized by a known wavelength of light and a known bandwidth, the bandwidth of the sub-signals being non-uniform. Optionally, the spacing between the sub-signals (spacing in the optical frequency domain) is non-uniform as well. The spacing between two adjacent sub-signals can be varied on the basis of the bandwidth of the sub-signals.

Abstract: In a methodology for route selection for path balancing in a connection-oriented packet switching network, a comparison is made of the measures of utilizations of a network resource by at least two links of a link group in the network, and at least one link is identified as a candidate for carrying traffic path based on the comparison. Network traffic path on the link is then moved to the candidate link to reduce variation in the utilization of the network resource by each of the links.

Abstract: An alternate path can be established in a connection-oriented packet switching network by correlating the bandwidth reservation of this alternate path with the bandwidth reservation of an active path on trunks through which both paths pass. The alternate path that is established can have one or more of a number of different attributes that vary from the attributes of the active path. The alternate path is established in a connection-oriented network with bandwidth correlation in three stages: a keying stage, a correlation stage and, a transfer stage. An alternate path may be established without a double reservation of bandwidth, thus making efficient use of limited network resources.