Multiplex communications – Fault recovery – Bypass an inoperative channel
Reexamination Certificate
1998-06-03
2001-11-27
Chin, Wellington (Department: 2664)
Multiplex communications
Fault recovery
Bypass an inoperative channel
C370S216000, C340S870030, C340S870030, C714S002000
Reexamination Certificate
active
06324162
ABSTRACT:
TECHNICAL FIELD
This invention relates to a technique for restoring service in a mesh network upon the failure of path.
BACKGROUND ART
Manufacturers of optical transport systems now offer an ever-increasing amount of available capacity on individual optical fibers. For this reason, achieving fast self-healing in a network from an accidental failure is of paramount importance. Presently, SONET/SDH ring networks comprise the most common type of self-healing network design. While SONET/SDH ring networks offer quick restoration and are easy to implement, such ring networks are not cost effective for many applications. For example, a highly connected network with most of the nodes each having at least three links requires more spare capacity when designed in a ring configuration than a comparable design with mesh configuration in which spare capacity is shared by the entire network more efficiently.
While mesh networks achieve more efficient use of spare capacity, restoration in a mesh network is more complex and slower than in a ring network. One type of mesh network restoration scheme is Link-Based Restoration (LBR) in which failed traffic is restored between the two nodes adjacent to the failure location regardless of the originating and terminating nodes of the failed signals. Another mesh restoration scheme is Path-Based Restoration (PBR) in which failed traffic is restored between the pairs of nodes where the failed signals originate and terminate regardless of the failure location. PBR requires less spare capacity to effect restoration as compared to LBR. On the other hand, PBR algorithms for establishing alternate paths when a failure occurs are more complex and therefore PBR may be slower, as compared to LBR.
Utilizing pre-computed routes can increase the PBR speed. One proposed restoration path computation approach employs node and link disjoint restoration paths. link- and node-disjoint restoration path is a path (i.e., a collection of individual channels in a series of links, each coupling a pair of nodes) that does not share any link or any intermediate node with the original path. However, accomplishing PBR using node-and link-disjoint restoration paths is inefficient in terms of maintaining sufficient spare capacity. On the other hand, it is necessary to determine which of the links on the signal path failed in order to implement a PBR method using unrestricted restoration paths. It is often problematic to make such a determination at the endpoint nodes. Further, computing a restoration path that is distinct from other restoration paths when multiple paths fail simultaneously is often not fast.
Thus, there is a need for a restoration technique that is not restricted to link-and node-disjoint paths, thus making the most efficient use of spare capacity. Additionally, there is a need for a mesh network restoration technique that affords restoration speeds comparable to those achieved in SONET/SDH rings. Further, there is a need for a mesh network restoration technique that is self-healing and self-managing.
BRIEF SUMMARY OF THE INVENTION
Briefly, the invention comprises a method for restoring service in a mesh network formed of a plurality of nodes, at least two of which serve as end points for originating and/or terminating traffic. A plurality of links, each having one or more working channels and one or more restoration channels, selectively couple pairs of the nodes. The collection of connected channels in a series of links carrying traffic between a pair of end-point nodes comprises a path. In accordance with the invention, at least one of a pair of connected nodes monitor the channels in connected links to detect a link in failure, that is, a link having at least one failed channel carrying traffic. Upon detecting a link in failure, one node determines whether the link in failure includes at least one available channel (typically, a restoration channel) to carry the traffic unable to pass on the failed channel. If so, the one node signals the node at the other end of the link in failure to route traffic on the available channel in the link, thus achieving “localized” restoration in a timely fashion.
Unfortunately, not all links may possess an available channel. (Indeed, a multiple-channel link, when severed, will suffer a failure of all its channels.) In the event that a link in failure lacks available restoration capacity (i.e., one or more available channels), then the nodes, which detected the failure, alert the path end-point nodes to restore the failed paths. Each end-point node accesses a database storing a plurality of pre-computed paths corresponding to the failed path. Each pre-computed path corresponds to a pre-selected link and identifies, for an associated pair of end-point nodes coupled by a path having a link in failure, an alternate path that bypasses the link in failure. After accessing the database, each end-point node selects the alternate path corresponding to the link in failure. The end-point nodes then route traffic in accordance with the pre-computed alternate path information to restore the network.
REFERENCES:
patent: 5182744 (1993-01-01), Askew et al.
patent: 5235599 (1993-08-01), Nishimura et al.
patent: 5495471 (1996-02-01), Chow et al.
patent: 5852600 (1998-12-01), Russ
AT&T Corp.
Chin Wellington
Levy Robert B.
Tran Maikhanh
LandOfFree
Path-based restoration mesh networks does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Path-based restoration mesh networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Path-based restoration mesh networks will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2598252