WO2011085823A1

WO2011085823A1 - Routing through network having optical and electrical layers

Info

Publication number: WO2011085823A1
Application number: PCT/EP2010/051287
Authority: WO
Inventors: Paola Iovanna; Giulio Bottari; Gianpaolo Oriolo
Original assignee: Telefonaktiebolaget L M Ericsson (Publ)
Priority date: 2010-01-12
Filing date: 2010-02-03
Publication date: 2011-07-21
Also published as: EP2524457A1

Abstract

Routing traffic across a network having an optical layer (40) and an electrical layer (30), involves using a common topology representing the optical and electrical layers of the network. The common topology is such that an individual port of a respective optical device in the optical layer (40) is represented as corresponding virtual nodes (145), coupled by virtual links (155), in the common topology together with a representation of nodes and links of the electrical layer. The routing path in terms of the common topology is mapped back into a routing path through the actual optical devices (50) of the optical layer. By moving the handling of colour (including wavelength assignment, wavelength continuity and colour selection where it is tunable/selectable) and direction constraints to the preliminary step of creating the common topology, the routing process can be simplified.

Description

ROUTING THROUGH NETWORK HAVING OPTICAL AND

ELECTRICAL LAYERS

Technical Field: This invention relates to methods of routing traffic across a network, to methods of operating a node of a network, to a network management system for a network, and to corresponding computer programs for such methods and apparatus.

Background: The integration of packet and optical technologies (both in case of multilayer nodes and networks) requires routing solutions that allow handling such heterogeneous layers in very efficient way. Two main groups of routing solutions have been defined:

• the first group involves multi-layer solutions where the optical layer is composed by nodes with full wavelength conversion capability;

the second group provides solutions that deal with all optical networks but without integration with packet layer;

This last group includes both off-line provisioning, and with on-line routing of new traffic in a live network.

Multi-layer routing is a very complex task because the different layers tend to use heterogeneous technologies which have different constraints. Moreover this task is more complex when one of the two layers is all-optical. One of the critical problems to address routing (R) in case of all optical networks is due to the fact that such networks require considering a set of issues such as wavelength assignment (WA) and physical impairment validation (IV).

Moreover a set of nodes limitations, like color and direction constraints, must be considered to describe all optical nodes. The information required to describe an optical node, can be contained in several matrices representing the connectivity of a node, the switching connectivity matrix, the wavelength conversion matrix, and so on. In conventional multi-layer routing, the following issues arise:

packet layer routing does not deal with WA and IV;

all optical nodes have peculiar features (direction, color, transponder type, available wavelengths, etc..) that are different with respect packet nodes (in general not optical nodes);

solutions must be scalable and efficient in terms of resource optimization. Summary of the Invention:

An object of the invention is to provide improved apparatus or methods. According to a first aspect, the invention provides:

A method for routing traffic across a network, the network having an optical layer and an electrical layer, by determining a routing path for the traffic across the network using a common topology representing the optical and electrical layers of the network. The common topology is such that an individual port of a respective optical device in the optical layer is represented as a corresponding virtual node or nodes, coupled by one or more virtual links, in the common topology together with a representation of nodes and links of the electrical layer. The determined routing path through the virtual nodes and links of the common topology is mapped back into a routing path through the actual optical devices of the optical layer.

This can help enable more efficient routing by moving the handling of colour (including wavelength assignment, wavelength continuity and colour selection where it is tunable/selectable) and direction constraints to the preliminary step of creating the common topology, which simplifies the routing process. The routing is effectively simplified because the topology is homogenised, despite the topology also being expanded by the use of virtual nodes and links.

Any additional features can be added to those discussed above, and some are described in more detail below.

Another aspect of the invention can involve a corresponding method of using such a network by requesting a routing and sending the new traffic according to a path determined by the above methods for routing.

Another aspect provides a program on a computer readable medium and having instructions executable by a processor to cause the processor to carry out the steps of the routing method.

Another aspect provides a network management system having a processor and the program for carrying out a method of routing.

Another aspect provides a node for a network having an optical layer and an electrical layer, the node having a program as set out above, as part of a distributed routing system. Any of the additional features can be combined together and combined with any of the aspects. Other advantages will be apparent to those skilled in the art, especially over other prior art. Numerous variations and modifications can be made without departing from the claims of the present invention. Therefore, it should be clearly understood that the form of the present invention is illustrative only and is not intended to limit the scope of the present invention.

Brief Description of the Drawings:

How the present invention may be put into effect will now be described by way of example with reference to the appended drawings, in which:

Fig. 1 shows a schematic view of a network, showing an electrical layer and an optical layer.

Fig. 2 shows steps according to an embodiment,

Fig. 3 shows a schematic view of nodes of an optical layer, with corresponding views of some nodes represented as virtual nodes and links,

Fig. 4 to 7 show views of different nodes represented as virtual nodes and links,

Figs 8 and 9 show a part of a network and a corresponding common topology for that part of the network,

Fig. 10 shows a view of a another network having multi layer node and having optical layers in different domains,

Fig 11 shows steps according to another embodiment,

Fig 12 shows steps according to another embodiment, and

Fig 13 shows a schematic view of a hybrid node in the form of a ROADM for use in a network.

Detailed Description:

The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes.

Definitions

Where the term "comprising" is used in the present description and claims, it does not exclude other elements or steps. Where an indefinite or definite article is used when referring to a singular noun e.g. "a" or "an", "the", this includes a plural of that noun unless something else is specifically stated.

The term "comprising", used in the claims, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps.

Elements or parts of the described nodes or networks may comprise logic encoded in media for performing any kind of information processing. Logic may comprise software encoded in a disk or other computer-readable medium and/or instructions encoded in an application specific integrated circuit (ASIC), field programmable gate array (FPGA), or other processor or hardware.

References to nodes can encompass any kind of node, including for example amplifying or filtering or switching nodes, not limited to the types described, not limited to any level of integration, or size or bandwidth or bit rate and so on.

References to optical nodes can encompass transparent nodes (such as amplifiers and all-optical wavelength crossconnects), nodes with OEO lightpath termination capabilities (like xOADMs) and translucent nodes (optical nodes with partial regeneration/wavelength conversion capabilities).

References to software can encompass any type of programs in any language executable directly or indirectly by a processor.

References to hardware, processor or circuitry can encompass any kind of logic or analog circuitry, integrated to any degree, and not limited to general purpose processors, digital signal processors, ASICs, FPGAs, discrete components or logic and so on.

Some Abbreviations:

IETF: Internet Engineering Task Force

ITU: International Telecommunication Union

OIF: Optical Interworking Forum

PCE: Path Computation Element

TE: Traffic Engineering

WSON: Wavelength Switched Optical Network

Introduction

By way of introduction to the embodiments, some issues with conventional methods will be explained. Path computation in WSON can be very complex because it can involve: Routing to find the lightpath connecting a pair of source and destination nodes at low cost, Wavelength Assignment to ensure the e2e (end to end) wavelength continuity, and Impairment Validation, to assess the quality of the wavelength to guarantee its correct reception despite the optical impairments which affect the analogue signal propagation. All this is different to Packet Routing. Node modeling in WSON can be useful to provide sufficient node information to support the complex path computation of WSON, such information including for example:

Direction constraints: a wavelength originated from a local port can reach any/one direction (directionless/bound);

Color constraints: ability to drop or add any/one wavelength at a port (colorless/bound);

Node limitations: number of channels, number of ways, bandwidth access;

Node capabilities: interface types, switching capabilities, wavelength conversion, signal regeneration.

Some embodiments address these issues by building up a common topology in the form of a virtual network topology where packet and optical layers can be represented as homogeneous links and nodes with different attributes in order to allow routing to run concurrently across the two layers in a more efficient way. The basic concept can be summarized as follows:

a) Transformation of the physical network topology into a common topology which represents the optical nodes (with their main constraints like direction, color, number of ways) and packet nodes in a homogeneous way. The considered node constraints can be static features which are dependent on the hardware of the optical node, or dynamic features selectable or controllable by a control plane. b) Definition of a method that run on the common topology where a subset of operations are performed as a preliminary off line step, independently of the traffic demands (e.g. impairment validation and partial wavelength assignment), and another subset of operations which are performed in real-time (path setting and wavelength assignment). This can help enable path computation to be simplified to have reasonable computing time and still be usable in dynamic scenarios as well.

Additional features of some embodiments

One or more of the virtual nodes and links can have attributes to represent at least possible wavelengths, capacity and optical impairment, and the routing being determined at least according to the attributes, and according to other nodes and links in the common topology. It can be easier to compute if these constraints are attributes. The network can have at least one multilayer node in both the electrical layer and the optical layer, and the multilayer node is represented in the common topology by a corresponding set of virtual nodes and links. Such multilayer nodes otherwise add particular complexity to the routing process.

The network can be partitioned into more than one domain, the common topology being determined for one of the domains. This enables the partition to be preserved and respected whatever the reason for the partitioning, such as administrative control by different operators, or preservation of confidentiality such as for an intranet or for islands of different vendor's proprietary equipment or different network management systems.

The network can have an optical device having a selectable operating wavelength, and the common topology can have a corresponding representation having virtual links and nodes corresponding to each of the different operating wavelengths. This can help enable wavelength assignment to be simplified.

The determining of the routing can have the step of creating a lightpaths matrix for light paths in the common layer. This can help enable the routing process to be calculated efficiently, especially in checking physical feasibility by accumulating optical impairments and other physical factors such as optical amplification along the path.

The method can have the step of assigning an optical source node and an optical destination node in the common topology to a respective traffic item of the traffic being routed. This can help simplify the overall routing task.

The routing step can involve determining relative cost for possible paths through the common topology for a particular item of the traffic being routed, and selecting a physically feasible path according to the cost. The common topology can make this step simpler or more efficient than trying to compare routes in multiple layers for example. The determining of the routing can involve any of: determining a path along the nodes and virtual nodes of the common topology, determining where items of the traffic being routed can be aggregated, and determining recovery paths for the items of the traffic being routed. Any or all of these steps can become simpler or more efficient by having a common topology, rather than having to carry out calculations with multiple layers. The method can have the subsequent step of configuring the optical device according to the determined routing path. This can enable the routing to be used in the real network. The method can have the preliminary step of generating the common topology.

The method can have the step of updating the common topology with attributes indicating network occupation, retrieved from the network. The routing can then be determined for a new traffic demand without using parts indicated by the attributes as being occupied.

The description of embodiments can be divided into steps involved in creating the common topology, and steps involved in using the common topology to determine routing paths.

The first of these sets of steps can work on the network topology and perform a network transformation in order to provide a common topology for example in the form of a graph where both packet and optical nodes/links are represented homogeneously. This allows performing routing on the two layers concurrently. Such operations are performed just on network topology and node/links constraints independently of the traffic demand features. Such set of steps can be performed off-line in which case it has no real-time constraints. The output is a common topology in the form of a virtualized network composed by links and nodes with attributes where the packet nodes are described according to the real network topology while optical nodes are described as suitable lightpaths list. Moreover in this set of steps the impairments validation is performed as well, for each of the lightpaths in the list.

The second set of steps can run on the common topology having virtual links and nodes. This can be a single topology where both packet and all optical nodes are represented as homogeneous links and nodes with different attributes. In this set of steps the multilayer path is computed including wavelength assignment on validated links.

The steps for the two sets are reported in the following:

Set of operations independently of the traffic demands. Such operations can be performed off-line with respect the time traffic requires to be served:

1) Topology transformation by network virtualization of the all-optical nodes. This operation can be performed just once and it is based on static information of the nodes/links.

2) Pre-computing of available lightpaths according to some optimization criterion (e.g. min-hop) and measuring QoT in order to have a list of validated lightpaths. This operation can be performed either just once (i.e. a complete computing of all lightpaths in advance with respect to network running) or more times while network is running. In this last case, just a subset of possible paths is computed, when there are not available resources, another set of lightpaths is computed.

3) Assigning the possible wavelengths to be used and assignment of a cost to such links according to the available wavelengths

Set of operations independent of the traffic demands that can be performed on-line:

1) Computation of the path crossing packet links and lightpaths by minimization of costs and assignment of wavelength where more choices are possible.

2) Updating the available wavelengths on the paths.

Fig 1 , network view

The network 20 can be composed of an electrical layer 30, in the form of a packet layer or TDM layer or any other type of electrical layer, and an optical layer 40. The electrical layer is composed of nodes 60 coupled by links, which are able to process and carry items of traffic such as packets. Each link is defined in terms of administrative cost and capacity (bandwidth).

Along these links the traffic demands (also named commodities) can be aggregated to have bigger capacities so that several signals can be routed along the same link.

The optical layer has optical nodes 45 comprising optical devices 50, some of these optical devices may have switching functions, some may couple the optical nodes to the electrical layer, others may be tunable to use one or more selectable wavelengths for example. The nodes may be configured by a network management system NMS 10, being centralized or distributed, and there can be an off line planning tool 5 in the form of a program run on a personal computer PC 15 for example. The routing methods described can either be carried out by the off line planning tool, or by the NMS. If the NMS is distributed, it can be incorporated in one or more of the nodes, so that a processor in the node carries out the routing methods described, at least for traffic entering the network at that node.

Fig 2, view of steps according to an embodiment

Some main parts of the solution are as follows. In the first part the original network topology is transformed into the virtual one G(N,A), representing the features described in the previous paragraph. This transformation especially concerns the optical nodes and links, in order to represent their characteristics, while the electric elements are not modified and for this reason they can be seen as nodes and links without any attributes.

At the end of the first part the virtual network is obtained and this is the input of the second part, the routing solutions, described in more detail below. Finally the third part, is the opposite of the first part, mapping the virtual network back to the original one, and the commodities are now routed and the wavelengths assigned along the actual optical paths, and the optical and electrical devices can be configured accordingly by a control plane, and the traffic flow can start.

As shown in figure 2, a common topology is generated at step 100. At step 1 10 electrical nodes and links are added to the common topology, without needing amendment or transformation. The common topology can be stored in store 160 as shown. At step 120, each port of a first optical device is represented as a virtual node coupled by virtual links in the common topology. At step 130, a next port of the optical device is similarly processed, to represent it in the common topology as a virtual node coupled by virtual links. At step 140, a next optical device is processed.

At step 150, one or more items of traffic have routing paths determined using the common topology. At step 155, the routing paths determined in terms of the common topology are mapped back to the actual devices in the optical layer.

Fig 3, Network transformation into virtual graph, for optical nodes

Fig. 3 shows a schematic view of nodes of an optical layer, with corresponding views of some nodes represented in an expanded form with ports represented as virtual nodes and links. Network transformation principally impacts all-optical nodes in order to represent specific features of such nodes in homogeneous way with respect to the packet nodes. The optical layer is composed of nodes 45 and links, which are able to process and carry optical signals. Each link is defined in terms of administrative cost and the number Y of wavelengths available on it, currently typical values for Y are 40 or 80. Each wavelength is characterized by a bandwidth (e.g. 1, 2.5 Gbit). The optical nodes can be divided into two sets, depending on their connections in the network.

In the first set there are all the optical nodes 45 connected to only other optical nodes (OXC, Optical Cross Connects). The nodes of this kind are named from now on "simple nodes". In the second set there are the "hybrid nodes" 70 (OADM, Optical Add-Drop Multiplexer); these are the nodes along the borders between the two layers, connected both to optical nodes and electric nodes 60.

The hybrid nodes allow the passage of the commodities from an optical layer to the electrical layer, thanks to a set of transponders and muxponders. The electric signal, coming from an electric node (router), arrives to an optical hybrid node and here it is transformed and routed along a wavelength (in the same way a wavelength is transformed in electric signal and dropped from the optical domain).

If more than one electric signal arrives at the hybrid node, the arriving signals can be eventually combined and transformed in a single optical signal but then, along all the optical domain, they cannot be combined or split any more. All the optical signals are considered unsplittable.

All the optical nodes (simple and hybrid nodes) have a number of optical ports (line ports), by which the commodities cross the node. These ports, if belonging to the same node, are fully connected and a node is composed of a number of ports equal to its optical degree (only connections between two optical nodes are considered). Figure 3 shows as an example how one optical node has its port structure represented in the common topology by virtual nodes 145 and virtual links 155.

A simple optical node can be designed as a set of nodes, corresponding to the optical ports (the line ports). These ports are always fully connected, if belonging to the same node, and their number is equal to the node's optical degree (only the optical links are considered). Each optical link represents a set of Y wavelengths (typically 80).

Each electric port accepts one or more electric signals (it depends if it is a transponder or a muxponder) and transforms them in a single wavelength (^Λ ί).

In the hybrid nodes, in addition to the optical ports, there are specific A/D sections (Add/Drop section) to allow the adding and dropping of the signal between the two domains. These sections are composed by a number X of transponders and muxponders (named electric ports) by which the electric signal is transformed into an optical signal of a given wavelength.

If an electrical port represents a coupling to a transponder, it accepts only one single electric signal which can be output as a wavelength; if an electric port represents a coupling to a muxponder it can accept and combines different electric signals and transforms them in one single wavelength. Figure 3 shows as an example how one hybrid optical node has a number of transponders and a muxponder, represented in the common topology as multiple electrical ports 62 of the electrical node 60, coupled to virtual nodes 170 in a one to one manner for the transponders, and coupled to virtual node 174 representing the muxponder having two electrical paths and a single optical path. Virtual links 155 represent paths having a fixed or selectable wavelength. These are then multiplexed by virtual node 175 onto a single path having many wavelengths which can be listed as occupied or free.

If an electric node is connected to the optical domain it is linked to a hybrid node and, more specifically, to an electric port of that node by a bidirectional electric link. Each A/D section is always associated to an optical port, in which it inserts the signal received in an add port or from which it removes the signal using a drop port. All the A/D ports are connected to this optical port by a bidirectional hybrid links set.

This link is called hybrid because the electric signal is transformed in wavelength (add) along it and vice versa (drop). On the other hand each optical port is associated to at most one A/D section and in this case it is connected to all the electric ports belonging to this A D section, by a bidirectional hybrid links set.

Each A/D section has a number X of electric ports 62 so X wavelengths are available on it. The number X can be different for every A/D section, but it is always less than (or at most equal to) the total of wavelengths Y. Each hybrid link, connecting an electric port to the correspondent optical port, represents a single wavelength.

Figs 4-6, examples having directionality and colour

Each couple of twin add/drop ports, belonging to an A/D section, can be defined in terms of directionality and colour tunability: the couple of port can be defined as directionless/directionbound and colorless/colored as will be explained now in more detail.

As for direction:

If an electric port is directionless the commodity added using this port, coming from the electric domain, can be routed to any optical port of that node. In the same way a commodity dropped from a directionless port can come from any optical port of the node it.

If an electric port is directionbound the commodity added using this port, coming from the electric domain, is routed to the optical port associated to the relevant A/D section. In the same way a commodity dropped from a directionless port comes from the optical port associated to the relevant A/D section.

If all the ports of an A/D section have directionless capability, the section can be defined as a directionless section (as a consequence, the relevant way has directionless capability). If all the ports of an A D section have not directionless capability, the section can be defined as a directionbound section (as a consequence, the relevant way is directionbound). In a realistic scenario, a node has just one directionless way while the other ways are directionbound.

As for tunability:

If an electric port is colorless, the commodity added using this port, can be tuned on each of the Y wavelengths. The drop port can be tuned to receive each of the Y wavelengths (the same of the relevant add section).

If an electric port is colored, the commodity added using this port, shall use one predefined wavelength among the Y wavelengths. The drop port is able to receive one predefined wavelength among the Y wavelengths (the same of the relevant add section).

In general, an A/D section can have a mix of colorless and colored ports but configurations in which all the sections ports are colorless or all the section ports are colored are also possible. However, the section must always have the same list of colours in the Add and in the Drop side of the section (biderectionality is required) and all the colours belonging to the same section (Add or Drop) must be different from each other.

As previously described, each electric port can be directionless/directionbound and colorless/colored. Each hybrid node can have both directionless and directionbound ways and each of these can have a mix of colorless and colored electric ports. Figures 4 to 6 show some possible combinations.

The example of figure 3 shows a directionless and colored example. Figure 4 shows an example which is direction bound and colored. As shown in Figure 4 the optical port, connected to a directionbound section, is divided into two nodes 180 coupled by direction bound virtual links 165. This is useful to represent the directionbound feature, as better described with respect to Figure 7.

The example of figure 5 is direction bound and has both coloured and colourless paths. The example of figure 6 is directionless and has a mix of coloured and colourless paths. If an electric port is colored a specific colour is pre-assigned to it, if it is colorless then any colour (different from the colours belonging to the same A/D section) can be associated to it and for this reason it is represented as a set of colored electric ports 172 and corresponding virtual links. However that representation is just to explain the colorless concept, but in the virtual model it can be represented as a single node and have a list of colours associated to it.

Figure 7 Hybrid node with two optical ports and one A/D directionbound section

Figure 7 shows a hybrid node with two optical ports 270 and 180 respectively, and one A/D direction-bound section: when the electric signal arrives at an electric port of that section it has to go to the corresponding optical port 180 and then it is forced, by the direction-bound feature, to exit from the node to device 290 in neighbouring node 280, following the arrows shown in the figure. It cant reach the device 310 in the other neighbouring node 300.

In the same way a wavelength dropped from a direction-bound electric port cannot come from optical ports (belonging to the same hybrid node) different from the corresponding one. In this Figure there are different kinds of links: the broken lines are the electric links and the continuous ones are the optical and hybrid links (it depends if they connect two optical ports or an electric port to an optical one).

A list of colours is associated to each node and link of the network to represent the different wavelength available on the same link. More specifically to each electric port and hybrid link is associated one colour or a list of colours (it depends on the colored/colorless feature) and a list of Y colours is associated to each optical node and optical link.

Figs 8,9, common topology

Figures 8 and 9 show another example of a real network topology and its corresponding common topology respectively. The real network has an optical layer having simple optical nodes 45, and hybrid nodes 70, coupled to electrical nodes 60. In figure 9, each of the hybrid nodes is represented by virtual nodes 172 representing transponders, coupled by virtual links to a virtual node 175 to represent the multiplexing of optical wavelengths. The optical devices of simple optical nodes are represented as in figure 3 by virtual nodes 145 and virtual links for each of the optical ports. This enables a homogeneous single layer topology which facilitates the routing process.

Figs 10, 11, further embodiments Fig 10 shows another view of a network according to an embodiment. This is similar to that of figure 1 , but also shows a control plane 12 distributed across some or all of the nodes and coupled to the NMS. It also shows a second optical layer 42, which may be part of a separate domain from the first optical layer 40, to enable separate management of domains for any reason, as discussed above. The second optical layer may be arranged as a layer below the corresponding electrical layer, to implement some or all of the links between the nodes of the electrical layer, a so-called vertical layering. An alternative is a horizontal layering where the optical layer can be seen as effectively at the same level as the electrical layer, rather than being used to implement links between the nodes of the electrical layer. Also shown is a multi layer node 70 effectively part of the electrical layer and part of the second optical layer. This may have electrical switching and optical add drop capabilities, and therefore be a particular type of the hybrid node discussed above.

Fig 1 1 shows steps according to another embodiment. In figure 1 1 , some similar steps are shown to those of figure 2. At step 90, network information is acquired, such as topology of the electrical and optical layers. At step 102, the common topology is generated as described above, by transformation of information about optical nodes and their ports, into virtual nodes and links that are homogenous with the nodes of the electrical layer. So far, these steps could be carried out off line by a planning tool, or by an NMS . At step 104, network occupation attributes are updated according to information retrieved from the network about the current occupancy. At step 105 a new traffic demand is received. A routing for this new traffic demand is determined using the free resources in the common topology, that is those resources not indicated as being occupied, as shown by step 135. At step 147, the routing in terms of virtual nodes and links is converted to a routing in terms of actual nodes and links. The new routing for the new traffic demand is sent out to the actual nodes using the NMS and control plane, at step 155.

Fig 12, Routing solution running on virtual network

As has been described, the common topology can be created by transforming each optical node in a real topology of the optical layer, into a virtual graph representing direction, wavelength availability and interface type, together with electrical nodes and links. A suitable routing algorithm can be run on the common topology to serve the traffic matrix. The algorithm can include the wavelength assignment and the feasibility assessement (QoT) inside the optical layer (WSON area or domain). Routed paths are then mapped from the common topology to the real topology.

This can involve for example creating a Lightpaths Matrix, by finding a lightpath , with the minimum number of hops and satisfying the QoT, for each possible pair of electric ports at the interface of the optical domain of the network. In the matrix, attributes can be assigned to each lightpath for example indicating a colour (or list of colours), a capacity and a cost.

Various traffic demand items, referred to as commodities, can be sorted according to the parameters and the objective function, for example:

From lower to higher bandwidth value to maximize the number of traffic demands, or

From higher to lower bandwidth to maximize the accommodated bandwidth The path computations can then be made on the virtual topology for each item in the order specified. In the WSON domain it is useful to provide the routing engine with a robust but fast way to assess the feasibility of the lightpaths under computation. The proposed algorithm can include the feasibility verification based on the estimation of a cumulative parameter called Quality of Transmission (QoT). Other ways can be conceived. The proposed algorithm is able to check the QoT while routing a path inside an optical domain.

Fig 12, Routing Problem formulation:

For sake of clarity the routing problem can be formulated as being to find the set of paths along which the commodities can be routed; and the colour or the list of colors to assign to each commodity along the optical path, where there is a choice, so that desired criteria are met to suit the application, such as the total cost is minimized and the colour or the list of colors assigned to each commodity is the same along all its optical path in the same optical domain, given certain inputs to the define the problem as follows, some of which are summarized in step 510 of fig 12:

a vector (^a ' for j =l, ...,k, where ^{fc R} is the capacity required by the generic commodity j to be routed from the source node ¾ to the destination node ( *i ³* J. More specifically the couple of nodes ( i ^" ~ϊ* can represent the generic optical/electric node or a specific (colorless/colored) electric port of the optical node. An oriented graph G(N,A) representing the virtual network, where N is the nodes set and A the links set.

N = ^ϋ ^ ) is the set of electric and optical nodes.

Each optical simple node is represented as a set of optical ports ■> and each optical hybrid node as a set of optical and electric ports iP^ ^υ

an optical node is simple (OXC) if it is connected only to optical nodes;

an optical node is hybrid (OADM) if it is connected to optical and electric nodes; each {simple and hybrid) node v has ) optical ports, where Λ^ν) is equal to the node's optical degree. The optical ports are fully connected if belonging to the same node.

each hybrid node v has k_el>^'J A/D section, where ^KeK )— ^Ks {^¾);

each A/D section j is composed by electric ports <°^" ;

to each A/D section is associated an optical port and all its electric ports are connected to this optical port (by bidirectional hybrid links);

to each optical port is associated at most an A/D section and in that case it is connected to all the electric ports of this section (by bidirectional hybrid links); each electric port can be directionless {dl) or directionbound {db);

if an electric port is dl the commodity added using this port, coming from the electric domain, can be routed to any optical port of that node itself. In the same way a commodity dropped from a directionless port can come from any optical port of the node itself;

if an electric port is db the commodity added using this port, coming from the electric domain, is routed to the optical port associated to the relevant A/D section. In the same way a commodity dropped from a directionless port comes from the optical port associated to the relevant A/D section;

an A/D section can be defined dl if all its electric ports have dl capability;

an A/D section can be defined db if all its electric ports have db capability;

A/D sections with a mix of dl and db electric ports do not exist;

each hybrid node can have both dl and db A/D sections;

to each optical port ¾ is assigned a list Ι ^³») of Y colours;

each A/D section j has ^ '^) electric ports, where ^ iO— γ ^ i; each electric port can be colorless (cl) or colored (cd);

if an electric port is cl, the commodity added using this port, can be tuned on each of the Y wavelengths. The drop port can be tuned to receive each of the Y wavelengths (the same applies to the relevant add section);

if an electric port is cd, the commodity added using this port, shall use one predefined wavelength among the Y wavelengths. The drop port is able to receive one predefined wavelength among the Y wavelengths (the same of the relevant add section);

an A/D section j can have only cl ports or only cd ports or a mix of them

if an electric port is cl a list of ^{~ A} ^Ui) colours is associated to it and each one represents a possible colour;

if an electric port is cd only one colour is associated to it;

a db electric port can be cl or cd;

a dl electric port can be cl or cd;

each A/D section must always have the same list of colours in the Add and in the Drop side (bidirectionality is required) and all the colours belonging to the same section (Add or Drop) must be different from each other;

each electric port has a capacity (^U"P ^ ) equal to the capacity of the wavelength connected to it.

A = (^rf -^β ) is the set of electric, hybrid and optical links.

All the links are oriented and each one has a cost (^Ci ^^→ ^ ) and a capacity " ). The cost vector can be defined as shown at step 520. The cost vector can be defined as a set of parameters that will affect the routing progress. A simpler version of the cost vector contains just an administrative cost. Additional costs can considered, for example: a cost which increases with bandwidth occupation (i.e. for Traffic Engineering purposes), an extra cost to discourage the routing across critical links, the delay introduced by the link, cost related to physical considerations which the operator would like to consider outside the QoT.

Each electric link connects an electric node to one other electric node or to one electric port; each hybrid link connects an electric port to the correspondent optical port; each optical link connects two optical ports; to each hybrid link is associated one colour/a list L of colours (it depends on the cd/cl electric port); to each optical link ¾ is associated a list L(^a*) of Y colours (typically 80).

A cost vector is defined at step 520, to take into account what is to be optimized to suit the application. Three main steps in the determination of routing paths can be considered: lightpath inizialization at step 530, Source and Destination Assignment at step 540, and Routing and Updating at steps 550 to 570. Each will be described in more detail, by way of example, other ways can be conceived.

Step 530 of fig 12, Lightpaths initialization:

For each separate optical domain z: Create the 'Lightpaths Matrix'as shown by step 530; This matrix is composed of a number of rows equal to the number of lightpaths ^i.: _i ^ =^.p_{k i} ·_{η z an(}j g_{ye co}i_umns (respectively representing the source i, the destination k, the cost c^ "*), the capacity u^«¾) and the colour L(^i;^*)/list of colors of each lightpath

Sort the electric ports pair 0-' ^), such that CA O don't belong to the same Add/Drop section, in a list ^ ^' (put the colorless ports pairs at the end of the list):

For each electric orts pair ^{;, '} "v in & ^' :

Find the oriented lightpaths ^i&' ^fei with the minimum number of hops, such that the QoT is satisfied and with a colour list L(ⁱ^'*)=L(ⁱ Ϊ⁵*:) ^ ;

colours availability: add a node i to ^lP only if L(^i'^=^"¾) ^{¾ 5} L( ) ^ ®, in the same way

Assign to ^L^^ik and ^{i ii5¾£} a colour ^ , arbitrary choosing from their colours lists and update L(^f¾) = L(^lP*t) =

Delete ^Λ from the colours list of every link and node (different from i and k) belonging to ¾*ifc and

Assign capacities to the lightpaths: u

Assign a cost to the lightpaths fc), c^: where c has a value which is a function of the costs assigned to the links composing the lightpath and other parameters, as defined at step 520;

Update the Lightpaths Matrix;

Step 540, of fig 12, Source and Destination Assignment:

For each each commodity j:

If ^Λ?! is an hybrid optical node:

Assign to it an electric port belonging to the hybrid optical node, with a spare capacity at least equal to ^ui and such that L(^s ) ^Ω L(¾^')≠ ^ (only if ^s} and ¾ belong to the same optical domain), choosing according to fixed parameters (direction features, number of available colours on it, cost, other commodities with the source/destination on it)

Steps 550 to 570 of fig 12, Routing and Updating:

For each commodity j, as shown in step 550:

Find the minimum cost path from ^'ϊ- to ₉ with a capacity at least equal to where

(add an electric port

P^ to ^ only if ^ ^€

As shown b step 570, update the capacities along the electric links belonging to ;

Update L(i) = L(k) ^A'

Delete from the Lightpaths Matrix all the lightpaths ^ijJ3iS, such that one of these condition is satisfied:

i = x and k Φ y

i≠ x an d k = y

i≠ and k = x

y and k≠ x

Repeat for a next value of j until as shown in step 560, end when j =k. Fig 13, view of a hybrid node in the form of a ROADM

Figure 13 shows an example of a ROADM which could be used as a hybrid node, or could be part of a multi layer node, if combined with some electrical switching of the added or dropped electrical signals. It shows four similar modules labelled north, south, east and west, each of which have similar components, so only south, 680, will be described further. This shows an optical power splitter 650 arranged to receive an incoming single wavelength signal and broadcast this over four or more optical outputs in the form of waveguides generally labelled 600 to the other three or more modules, and to one local drop path, local to that module. This drop path leads to an array wave guide 660 which has an optical wavelength demux or separation function, for separating different wavelengths onto separate physical paths to receivers Rx 670. These receivers output electrical signals which can be fed to further electrical circuits for TDM demux or electrical switching for example, or straight to local destinations such as local networks.

The module also has a wavelength selective switch WSS 640, for selecting a wavelength to be sent out on the outgoing path from the south module. This WSS receives wavelengths from other modules East, North, and West along internal waveguides labelled generally as 600, and one or more wavelengths for adding at that module. The added wavelength is selected by AWG 620 which combines different physical paths from separate transmitters 610 for each wavelength, onto a single input of the WSS. Any one of the transmitters can be activated, which determines which wavelength is being added. An electrical signal to be added can be fed to the appropriate transmitter for the desired wavelength. To be able to send out WDM signals, the WSS could be made as a WDM multiplexer, or a WDM multiplexer could be provided downstream of the WSS. In this case, the AWG could feed the WDM multiplexer directly, bypassing the WSS.

As described, embodiments can enable: more efficient multi-layer routing, resource optimization that allows capex reduction, a scalable solution, and a method of optical nodes modeling that allows applying several dynamic routing approaches in a multilayer scenario. Other variations and embodiments can be envisaged within the claims.

Claims

Claims:

1. A method for routing traffic across a network, the network comprising an optical layer and an electrical layer, the method comprising the steps of:

determining a routing path for the traffic across the network using a common topology, the common topology representing the optical and electrical layers of the network, such that an individual port of a respective optical device in the optical layer is represented as a corresponding virtual node or nodes, coupled by one or more virtual links, in the common topology together with a representation of nodes and links of the electrical layer, and mapping the determined routing path through the virtual nodes and links of the common topology back into a routing path through the actual optical devices of the optical layer.

2. The method of claim 1, one or more of the virtual nodes and links having attributes to represent at least possible wavelengths, capacity and optical impairment, and the routing being determined at least according to the attributes, and according to other nodes and links in the common topology.

3. The method of claim 1 or 2, the network comprising at least one multilayer node in both the electrical layer and the optical layer, and the multilayer node is represented in the common topology by a corresponding set of virtual nodes and links.

4. The method of any preceding claim, the optical device having a selectable operating wavelength, and the common topology having a corresponding representation comprising virtual links and nodes corresponding to each of the different operating wavelengths.

5. The method of any preceding claim, the determining of the routing comprising the step of creating a light paths matrix for light paths in the common topology.

6. The method of claim 5 the routing step involving determining relative costs for possible paths through the common topology for a particular item of the traffic being routed, and selecting a physically feasible path according to the cost.

7. The method of any preceding claim, the determining of the routing involving any of: determining a path along the nodes and virtual nodes of the common topology, determining where items of the traffic being routed can be aggregated, and determining recovery paths for the items of the traffic being routed.

8. The method of any preceding claim and comprising the step of configuring the optical device according to the determined routing path.

9. The method of any preceding claim and comprising the step of updating the common topology with attributes indicating network occupation, retrieved from the network, and the routing being determined for a new traffic demand without using parts indicated by the attributes as being occupied.

10. A method of using a network comprising an optical layer and an electrical layer, the method comprising the steps of requesting a routing for a new traffic item, and sending the new traffic through the network along the routing path determined using the method of routing of any preceding claim.

11. A program on a computer readable medium and comprising instructions executable by a processor to cause the processor to carry out the steps of any preceding claim.

12. A network management system for a network having an electrical layer and an optical layer, the system comprising a processor and a program store having stored instructions executable by the processor to cause the processor to determine routing path for traffic across the network using a common topology, the common topology representing the optical and electrical layers of the network, such that an individual port of a respective optical device in the optical layer is represented as a corresponding virtual node or nodes, coupled by one or more virtual links, in the common topology together with a representation of nodes and links of the electrical layer, and the stored instructions also being executable to cause the processor to map the determined routing path through the virtual nodes and links of the common topology back into a routing path through the actual optical devices of the optical layer.

13. The system of claim 12, one or more of the virtual nodes and links having attributes to represent at least possible wavelengths, capacity and optical impairment, and the routing being determined at least according to the attributes, and according to other nodes and links in the common topology.

14. A node for a network comprising an optical layer and an electrical layer, the node comprising a processor arranged to route traffic across a network, the processor being arranged to determine a routing path for the traffic across the network using a common topology, the common topology representing the optical and electrical layers of the network, such that an individual port of a respective optical device in the optical layer is represented as a corresponding virtual node or nodes, coupled by one or more virtual links, in the common topology together with a representation of nodes and links of the electrical layer, and the processor being arranged to map the determined routing path through the virtual nodes and links of the common topology back into a routing path through the actual optical devices of the optical layer.

15. The node of claim 14, one or more of the virtual nodes and links having attributes to represent at least possible wavelengths, capacity and optical impairment, and the processor being arranged to determine the routing at least according to the attributes, and according to other nodes and links in the common topology.

16. The node of claim 14 or 15, the network comprising at least one multilayer node in both the electrical layer and the optical layer, and the multilayer node is represented in the common topology by a corresponding set of virtual nodes and links.

17. The node of any of claims 14, 15 or 16, the optical device having a selectable operating wavelength, and the common topology having a corresponding representation comprising virtual links and nodes corresponding to each of the different operating wavelengths.