Information Technology Reference
In-Depth Information
•
Recovery time
: the time from the moment of the failure until the traffic is up and
running again. The recovery time is dependent mainly on the protection scheme used.
The failure notification, the backup path (or backup segment), and the sum of the
lengths of the working path (or active segment) are the main factors in determining
the recovery time. Note that the recovery time has other components that are
independent of protection schemes that are used.
•
Configuration:
it refers how the protection existences are inaugurated to restore
working traffic upon failure.
•
Link-disjoint
: two paths are called link-disjoint if they do not pass through any
(bidirectional) WDM links in common.
•
Path pair
: it is defined as two paths from a source to each destination node in such a
way that paths are link disjoint from each other.
•
Spanning path
: it is defined as a path from a leaf node to any other leaf node in a
multicast tree.
•
Intersection node
: if there are two nodes at which a path pair intersects with other
exiting path pairs, these two nodes are called intersection nodes.
•
Straddling path
: it is defined as a path between two intersection nodes.
•
Overlapping path
: the sub-path on the path pair from a source to intersection node.
•
Blocking
: if a lightpath for a connection request cannot be found, blocking occurs.
•
Ligthpath:
is an optical channel that spans multiple fiber links to provide a
connection between two network nodes.
•
Multicast Capable (MC):
An MC node is a node which has light splitting switch and
can forward an incoming message to multiple output channels.
•
Multicast Incapable (MI)
: a node that cannot split the light.
•
Light-tree
: it is a point-to-multipoint extension of a lightpath where its branching
nodes are equipped with multicast-capable optical switches [9].
•
Idle-backup edges
: the backup edges which do not carry traffic and are reserved to be
used when a failure occurs [10].
Protection against single link failure can be divided into two main groups: link-based
protection and path-based protection. Some of their subgroups include Link Protection, p-
cycle protection (which also sometimes classified under link protection class), Path
Protection, Segment Protection, Partial Path Protection, and Virtual Link Protection. In Link
Protection, all connections around a failed link are rerouted. The
p
-cycle is a virtual cycle
configured in the spare capacity of a network to protect the working capacity in all the links
that have their endpoints on the cycle. Path Protection is a technique in which network
resources (bandwidth, buffer, and process capacity) are saved for a unique path protection
regarding to a main path. In Segment Protection, a working path is divided into multiple
active segments and each working path is protected with a backup segment. In Partial Path
Protection, when a link of any path fails, the network can use various protection paths for
sending data. In Virtual Link Protection, the protection path is computed for each virtual link,
reserved in advance. Sub-Path Protection has two main ideas. First, a large network is divided
into several smaller areas by using the open shortest path first (OSPF) routing algorithm.
Then, two fiber-disjoint paths are computed for a given connection request.
