Network Protection

Fig. 1 Path segment restoration

Path segment restoration is also called sub-path restoration. It is similar to the path restoration technique, which is however more generic than the latter. Path restoration carries out restoration always at the two end nodes of a working paths, while path segment restoration can perform restoration between any node pair on a working path as long as one node is on the left-hand side and the other node is on the right-hand side of a network failure. Fig 1 shows an example of path segment restoration. Under this technique, for a link failure (3-4), we can use many protection paths as shown by dotted lines to recover the failure. Moreover, the dotted line actually represents a protection path group (more than one path) existing between the node pair. Thus, it can be seen that the path segment restoration technique is very flexible in its restoration. Since it is any node pair, the two end nodes of the working path is just one of these node pairs. Thus, path restoration can be seen as a special case of the path segment restoration. As a more generic technique, the path segment restoration technique shows advantages of better spare capacity efficiency and faster restoration speed than the path restoration technique. Of course, this is the at the cost of more complicated network control and operation as rather than a single end-to-end protection path, we have to handle all the eligible restoration paths corresponding to any working path segments. For example, in Fig. 1 there are a total of 3×2=6 protection path groups that recover the failure of link (3-4). The segment-protecting p-cycle technique is essentially also a type of path segment protection technique which, compared to path segment restoration, is more restrictive to require deploying preconfigured p-cycles for working segment recovery.

For network protection, two aspects are important: (a) spare capacity pre-planned for restoration and (b) restoration speed for failure recovery. For the latter, the current protection or restoration techniques can be classified into two groups. One is pre-planned techniques and the other is pre-configured techniques. A pre-configured technique inherent embeds spare capacity pre-plan. The pre-planned protection technique include span restorable networks, shared backup path protection (SBPP), and path restoration (PR), while the pre-configured techniques include ring techniques, span-protecting p-cycles, segment-protecting p-cycles, and pre-cross-connected trails (PXT).

Fig. 1 Difference between preplan and preconfiguration

The key difference between pre-planned and pre-configured techniques is that the pre-configured techniques can achieve faster restoration speeds than the pre-planned techniques. In addition to pre-planning sufficient spare capacity for restoration upon a failure, the pre-configured techniques actually have set up protection path and cross-connected intermediate node switch states even before a failure. The benefit of this is that when a failure occurs, only the two end nodes of a working path or span need to perform switching-over operation to recover the failure. In contrast, under the pre-planned schemes, only sufficient protection capacity is reserved, but their connectivity on the intermediate nodes is not pre-cross-connected. When a failure occurs, a signaling session is required to activate the protection path and set up switch states on the traversed intermediate nodes. Thus, in general the pre-planned techniques require longer restoration time. In the aspect of spare capacity efficiency, these two types of techniques are essentially very close, although the pre-planned techniques can sometimes show a little bit better spare capacity efficiency. Fig. 1 shows an example for the difference between the concepts of pre-plan and pre-configuration. There is a preplanned protection path (dotted line), for which 3 units of protection capacity are reserved on all the three traversed links. Between the first and second links, the cross-connectivity between reserved protection capacity units on the two links are not preconfigured, which need to be activated through a later signaling process. In contrast, between the second and third links, the cross connectivity between protection capacity units are preconfigured, which do not need to be reconfigured in the future.