Optical: systems and elements – Deflection using a moving element – Using a periodically moving element
Reexamination Certificate
1999-06-01
2002-11-05
Pascal, Leslie (Department: 2633)
Optical: systems and elements
Deflection using a moving element
Using a periodically moving element
C359S199200
Reexamination Certificate
active
06476945
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention is directed to wavelength division multiplexed (WDM) networks, and more particularly to a method of allocating the wavelengths in a dual ring.
2. Background Art
Current optical networks generally have a linear or a ring configuration. The ring configuration is considered a cost-effective network architecture allowing bandwidth sharing and improved survivability in the event of span failure. Generally, a ring is formed with add/drop multiplexers (ADMS) which insert/extract traffic into/from a working and a protection fiber/time-slot. But self-healing rings have a fundamental limitation. Namely, if one span only needs a bandwidth upgrade, all nodes have to be replaced to support higher rates. A solution to increase the capacity of the ring is to use a plurality of channels on the same fiber, the channels being routed separately according to their wavelength, a technique termed wavelength division multiplexing (WDM).
A key feature of WDM and dense WDM systems is that the discrete wavelengths form an orthogonal set of carriers which can be separated, routed, and switched without interfering with each other, as long as the total light intensity is kept sufficiently low. By using WDM, the capacity of a ring can be increased in an efficient and cost effective way with minimal changes to the nodes hardware or to the automatic switching protocol (ASP).
A WDM ring network comprises a plurality of add/drop nodes connected in a ring along two unidirectional fibers, one for each direction around the ring. Each add/drop node communicates with an associated edge node(s) using a number of local wavelengths (channels). An edge node is also referred to herein as a user node, or a local user. Edge nodes can be electronic or optical nodes, this is not relevant to the present invention.
A WDM node also comprises transmitters and receivers for each wavelength, and a wavelength add/drop multiplexer (ADM). The ADM multiplexes the local wavelengths received from the associated electronic edge node with the pass through wavelengths, before launching the multiplexed signal over the ring in the direction of interest. The ADM also demultiplexes the traffic received from the ring and directs it to the associated edge node and to the downstream nodes, respectively. Thus, a wavelength may be filtered at a node to drop traffic from the ring to the local user, and new traffic may be added on this wavelength from a local user into the ring. Alternatively, a wavelength can pass through an intermediate node so that it will be terminated at a later node of the ring.
A unidirectional channel is created from a source node to a destination node by injecting a wavelength at the source node and dropping it the destination node. This wavelength cannot be added or dropped at any other nodes between the source and destination nodes, in this unidirectional fiber. A node which is not the source or destination for a certain wavelength is called herein an intermediate node.
Multi-wavelength dual rings can be used to create a high capacity all-optical core network interconnecting several edge nodes. One example of a meshed architecture is a set of parallel dual optical rings interconnecting a number of edge nodes, with each edge node accessing any of the rings. Within a ring, wavelengths are used to create channels between optical node pairs, as described above. The capacity of the ring may be increased by scaling up the number of nodes per ring. This architecture could be attractive for metropolitan networks, since the cost of fibers from the electronic edge nodes, which is proportional to distance, would be small.
Any ring configurations for mesh interconnectivity will benefit, in terms of cost savings, from using the minimum number of wavelengths around the ring, while connecting all nodes. Even though it is not a trivial task, solutions can be derived manually for small numbers of nodes.
For example, U.S. Pat. No. 5,751,454 (MacDonald et al., issued on May 12, 1998, and assigned to Northern Telecom Limited) discloses a wavelength allocation method for full-mesh networks with a small number of nodes. The method provides direct node to node routes, and complete transparent interconnections with extra capacity for heavy used routes on a portion of the ring. For networks with large number of nodes, this patent proposes under-connected networks with a number of accelerated, direct routes between some of the nodes.
It is evident that full connectivity and wavelength allocation become extremely difficult problems to solve for rings with high number of nodes. A method that gives the minimum number of wavelengths and the way that these wavelengths may be allocated between all pairs of nodes of a WDM ring network, and that applies to any number of nodes, is therefore highly desirable.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method of wavelength allocation that allows creating a fully meshed network at the optical layer when the nodes are connected, at the physical layer, in a ring.
In one aspect of the invention there is provided in a D/WDM ring with n add/drop nodes connected over a forward and a reverse fiber, a method of allocating a wavelength between each pair of nodes for obtaining a fully meshed network, comprising, determining an add/drop requirement N
a
for all nodes of the ring, a fiber requirement N
f
for each span of the ring and the minimum number N of wavelengths/span, preparing a hop table with all nodes nj and all wavelengths &lgr;
i
for the ring and selecting an origin for the hop table by defining a node of origin and a first wavelength, if N
a
≦N
f
, determining an initial hop vector comprising a set of n initial forward hop values, generating all hops for all nodes and all wavelengths using the initial hop values and recording all the hop values in the hop table, and equipping each node with wavelength-specific receivers and transmitters according to a source-destination table prepared from the hop table. The wavelengths requirement is N
f
.
Advantageously, the method of the invention uses a minimum number of wavelengths and a same wavelength set on each fiber, with the constraint that a wavelength is used at most once on any resource (add/drop node, fiber).
This method can be readily incorporated in an engineering tool, which significantly simplifies design of the network configuration.
The benefits of this method are thus a reduction of the cost of optical equipment, and a simplification of the network configuration.
REFERENCES:
patent: 5751454 (1998-05-01), MacDonald et al.
patent: 5774244 (1998-06-01), Tandon et al.
patent: 5999288 (1999-12-01), Ellinas et al.
patent: 6339488 (2002-01-01), Beshai et al.
patent: 0 716 521 (1996-06-01), None
patent: WO 99 39470 (1999-08-01), None
Elrefaie A: “Multiwavelength survivable ring network architectures” Proceedings of the International Conference on Communications (ICC), Geneva, May 23-26, 1993.
Obara, H: “Multifiber wavelength-division multiplexed ring network architecture for Tera-bit/s throughput” Communications, 1998. ICC98. Conference Record. 1998 IEEE International Conference on Atlanta, GA USA Jun. 7-11, 1998.
Nortel Networks Limited
Pascal Leslie
Singh Dalzid
LandOfFree
Wavelength allocation in a ring does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Wavelength allocation in a ring, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Wavelength allocation in a ring will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2939220