Optical communications – Multiplex – Optical local area network
Reexamination Certificate
1999-07-07
2003-11-04
Swarthout, Brent A. (Department: 2636)
Optical communications
Multiplex
Optical local area network
C398S050000
Reexamination Certificate
active
06643464
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to optical transmission networks, to nodes for use in such networks, to link selectors for use in such nodes, to controllers for link selectors, to protection path sharing arrangements for optical transmission links, to methods of transmitting data over such networks, to software for managing such networks, and to software for determining configuration of secondary paths in such networks.
BACKGROUND TO THE INVENTION
Optical transmission systems are often constructed with a fault recovery mechanism so that if there is a complete loss of transmission capability, e.g. from a cut in the fibre, or a failure in the transmission path for any other reason, the traffic can be reallocated to other physically diverse routes. Known fault recovery mechanisms can be classed in two categories, firstly mesh restoration schemes, and secondly linear or ring protection methods. The term “protection” implies a very fast recovery. The term “restoration” implies a slower recovery with correspondingly more disruption to traffic.
Mesh restoration method speeds may be in the order of one to fifteen minutes, but are relatively efficient in terms of the amount of bandwidth set aside for recovery. Linear or ring protection methods may complete their operation in the order of ten to sixty milliseconds, but are less bandwidth efficient, typically requiring fifty to seventy percent of network capacity to be set aside for protection.
Restoration may take place at various layers of the well-known OSI model. There are advantages to conducting the restoration at the lowest feasible level, to minimise congestion that can be caused by delays and retransmissions of traffic. Restoration at the optical level, by switching a signal to another optical path, can be done either by optical switching, or by converting into the electrical domain, and switching in the electrical domain, if necessary, using demultiplexing if the bitrate of the optical signal is too high for electrical switching.
Mesh restoration operates by identifying several different alternative paths through the mesh to the destination node according to availability of bandwidth on these paths. The traffic is divided between these alternative paths, and recombined at the destination node, at the far side of the fault. It is bandwidth efficient because the amount of bandwidth on each link set aside for protection purposes need only be a fraction of the bandwidth of each working transmission path. However, it is usually slow to operate because for a given fault in the mesh, the alternative routes need to be determined, often by a central controller, and it may take time for the numerous nodes in the different paths to be configured. Then the traffic can be divided appropriately between the various protection paths to the destination node.
The software for controlling such restoration may be complex, and may require manual intervention, for a large mesh network.
Linear or ring protection methods involve providing a preconfigured (and often dedicated, though it can be shared) protection path for each link. Since the working path will normally take the shortest route between nodes, the protection path will normally involve more links than the working path. Accordingly, over the entire network, more protection path bandwidth needs to be provided than working bandwidth, and so working bandwidth may be as low as twenty to fifty percent of total bandwidth. This is expensive but it enables the protection path to be switched with a minimum of processing. Ideally, as soon as the fault is detected, a signal is sent to a switch at each end of the protection path to bring the protection path into operation. In this case there are only two switches to operate, there is no path determination, or splitting up and recombination of the signal. The process needs no central control, and can therefore happen very quickly, in the order of ten to sixty milliseconds.
Many variations have been proposed, within each of these two categories, attempting to achieve a physical layer fault recovery scheme that is both fast, and offers efficient use of bandwidth. In the category of protection schemes, one known option is to provide a pre-configured, dedicated protection path for each working path. A simple example is shown in FIG.
1
.
Another example is known from U.S. Pat. No. 5,159,595, (Nortel Networks reference RR1110) showing a bidirectional ring, and a capability at nodes next to a fault, to couple a working path onto the path going in the other direction around the ring, to create a folded loop. Another example making use of an optical switch to enable a protection ring to bypass a node is shown in WO97/09803 (Nortel Networks reference RR2473).
Another proposal for a protection scheme is shown in patent publication WO9923773 (Nortel Networks reference RR2258) involving an optical network made up of working paths in the form of SONET rings. On a link where protection paths for these SONET rings overlap, instead of providing two fibers for the two protection paths, a single fiber is shared by the two protection paths. An optical selector needs to be provided at the nodes at each end of the link. At these nodes, the protection paths pass through ADMs (Add Drop Multiplexers), where the protection path would be converted to the electrical domain to enable the working path to be switched onto the protection path. Although offering a bandwidth saving, the disadvantage of the optical degradation introduced by the optical selector, or the cost of additional equipment to overcome this, have tended to outweigh the benefit. Accordingly, this proposal has not been adopted on a commercial scale. More recently, sharing of protection bandwidth has been achieved in a different way, by optical shared protection rings (OSPR). This technique has been adopted for commercial use.
FIG. 2
shows an OSPR. The protection ring extends over several links in the network. On each link, the single ring can effectively replace several dedicated protection paths. Each of the protection paths still follows a preconfigured route around part of the ring. Accordingly, there is no need to determine a route and divide up the traffic, and so the speed of recovery is relatively good. As this protection ring can be shared between multiple protection paths, there is an improvement in bandwidth efficiency over using dedicated paths, to a ratio in the illustrated example of 1 protection path to 1 working path, i.e. 50% bandwidth efficient.
An OSPR scheme has been published in “Technical Digest of the Optical Fiber Communication Conference and the International Conference on Integrated Optics and Optical Fiber Communication, San Diego, Calif., Feb. 21-26, 1999, paper entitled Availability analysis of optical shared protection rings for long haul networks, by Philippe Neusy and Richard Habel, pages 176/TuL5-1 to 179/Tul5-4”.
OSPR has been regarded as an optimum solution that can be tailored to suit the networks of working paths arranged in either rings or meshes. Efforts to achieve further bandwidth efficiency improvements have therefore focussed on how to arrange the protection rings to be shared between more protection paths on the rings.
References to optical networks in this document are not intended to be limited to all-optical, but are intended to include for example networks making use of optical transmission and electrical domain switching.
References to nodes in the network can encompass add/drop points where traffic is added or dropped from the network. They can also encompass re-routing junctions where no traffic is added or dropped from the network. References to links between nodes are intended to encompass lines between add/drop points, or re-routing junctions, which may include a proportion of a span, or one or more spans. Spans are optical paths between electrical regeneration or optical amplification points.
References to working paths are intended to encompass end to end paths set up over multiple networks, or over multiple links in a network, or over parts of r
Baker Nigel
Roorda Peter
Sparks Adrian P
Barnes & Thornburg
Nortel Networks Limited
Swarthout Brent A.
LandOfFree
Constrained optical mesh protection for transmission systems does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Constrained optical mesh protection for transmission systems, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Constrained optical mesh protection for transmission systems will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3129177