Span restorable networks
Fig. 1(a) Span restorable networks
Span restoration is type of mesh-based restoration technique. Upon a link failure, the technique performs restoration at the end two nodes of a failed link, which loopbacks all the affected service flows locally onto a route that bypass the failed link. Fig. 1(a) shows an example of span restoration. A fiber cut of link (1-2) affects 5 units of working capacity traversing the link. To recover these 5 units of working capacity, three restoration routes are used to restore 2, 2 and 1 units of capacity, respectively. These restoration routes include 1-3-2 for 2 units, 1-3-4-2 for 2 units, and 1-3-4-5-2 for 1 unit. Because the failure recovery is carried out at the two end nodes of a failed link, span restoration generally has a fast restoration speed due to fast failure detection.
Fig. 1(b) Spare capacity sharing in span restorable networks
Meanwhile, span restoration also shows the advantage of efficient spare capacity utilization due to its spare capacity sharing feature between protection paths. Under single-failure protection, each time there is only a single network failure. Thus, the protection capacity used for one network failure can also be used for other network failures since they would not occur at the same time. Regarding spare capacity sharing, Fig. 1(b) shows a simple example. The spare capacity reserved for the failure recovery of link failures (1-2) and (4-5) can be shared on link (2-4), where only 2 units rather than 4 units of spare capacity are required to fully recover the two independent link failures.
