Network Protection

Pre-cross-connected trails (PXT) are a type of network protection technique similar to p-cycles. Rather than setting up complete circles like p-cycles, the PXT technique establishes open pre-cross-connected trails as protection path. Then depending on the location where a failure occurs, switch-over actions are taken at the two end nodes of a protected working path.

Fig. 1 Pre_cross_connected_trails (PXT)

Fig. 1 shows an example of PXT protection. A pre-cross-connected trail (1-2-4-6-7-8) is pre-established in the network; all the switch states are pre-cross-connected in all the intermediate nodes. This trail offers protection for the working paths W1 and W2. For example, if working path W1 gets failed, we may perform switching-over actions at the two end nodes 2 and 8 to fast recover the failure without needing to set up the cross-connect state on nodes 4, 6, and 7 since they have been pre-cross-connected. Similarly, a failure on working path W2 can be recovered by the switching-over actions at nodes 1 and 6. Different from the traditional span-protecting p-cycles, which are a span-based protection technique, PXT can be classified as a path-based protection technique. As shown in Fig. 1, the trail is essentially protecting two working paths. Of course, by nature PXT can also protect spans if a working path consists of only a single link.

Fig. 1 Example of ring cover

Ring cover is a type of mesh-based protection technique to apply the traditional ring protection techniques such as unidirectional path switched ring (UPSR) and bidirectional line switched ring (BLSR) to a mesh-based network. For a simple network, it may be possible to employ a single ring to cover all the links of the network. However, for a large network, more than one ring is often required. Under the ring cover technique, we need to ensure that each link in the network should be covered or traversed by at least one of the rings. This is due to the facts that many nodes in a mesh network may have nodal degrees higher than two and also the number of nodes on a single ring is often limited. For example, under the traditional SDH/SONET technique, the maximal number of nodes on a ring is 16. Figure 1 shows an example of ring over for a mesh network, in which three rings are placed, some links are traversed by a single ring, and other links are traversed by two rings. The design of a ring-covered network is essentially an optimization problem to use a minimal number of rings and protection capacity to cover all the links and fulfill protection. Though the ring cover technique can support protection for a mesh network, it is still a ring-based technique. Thus, its spare capacity redundancy is still high, similar to the traditional single-ring-based techniques, and worse than other mesh-based protection techniques such as span-restorable network, span-protecting p-cycles, path restoration, and so on.