JP2006527543A - Optical network topology database and optical network operations - Google Patents

Abstract

Optical network topology database and optical network operations. According to one embodiment of the invention, a number of wavelength division multiplexed access nodes use distributed search to build a network topology database based on one or more sets of connectivity constraints. According to another embodiment of the present invention, a set of one or more connectivity constraints including quality of service (QoS) based criteria is applied to the physical network topology of a wavelength division multiplexed optical network. Divide the optical network into different service levels. Furthermore, a service level topology is determined for each of the service levels. In accordance with another embodiment of the present invention, a number of wavelength division multiplexed access nodes in an optical network employ a source-based scheme to establish a communication path. Each of these access nodes stores a set of one or more network topology databases based on a set of one or more connectivity constraints. According to another embodiment of the present invention, a wavelength division multiplexing optical network includes a number of nodes, each node comprising an optical cross-connect, each node in which from that node to another node. Stores a database that represents connectivity without conversion. In addition, each of these nodes uses a messaging scheme to propagate notifications of changes in the optical network to other nodes and maintain its database. Each node's messaging scheme only sends messages to selected nodes among other nodes based at least in part on connectivity without conversion, thereby minimizing the number of communications between nodes. suppress.

Description

(Cross-reference to related applications)
This application is a continuation of application Ser. No. 10 / 455,933 filed Jun. 6, 2003, which is incorporated herein by reference. This application is further incorporated by reference in application Ser. No. 10/455, filed Jul. 23, 2003, which is a continuation of application No. 10 / 455,933, filed Jun. 6, 2003, which is incorporated herein by reference. It is also a continuation application of 626,055. This application is further incorporated by reference in application Ser. No. 10/455, filed Jul. 23, 2003, which is a continuation of application No. 10 / 455,933, filed Jun. 6, 2003, which is incorporated herein by reference. It is also a continuation application of 626,363. This application is further incorporated by reference into “Selective Distribution Messaging” filed on June 3, 2004, which is a continuation of application No. 10 / 455,933, filed June 6, 2003, which is incorporated herein by reference. It is also a continuation of the application _________ entitled “Schem for an Optical Network”

  Embodiments of the present invention relate to the field of networking, and more specifically to optical networks.

  Generalized Multiprotocol Label Switching (GMPLS) [RFC3471] is a multi-protocol label switching (MPLS) architecture [RFC3031] that is time-shared (eg, synchronous optical network and synchronous digital hierarchy, ie SONET / SDH). It is expanded to incorporate functions of switching, wavelength (optical lambda) switching, and spatial switching (eg, from incoming port or fiber to transmitting port or fiber).

  GMPLS is an extension of MPLS that captures network devices where the forwarding plane does not recognize packets or cell boundaries, and therefore cannot forward data based on information carried in the packet or cell header. In particular, such network devices include devices whose forwarding decisions are based on time slots (TDM), wavelengths (lambda), or physical (fiber) ports. GMPLS uses unidirectional label switch paths (LSPs) and bidirectional LSPs (for bidirectional LSPs, the term “initiator” is used to refer to the node that initiates the establishment of the LSP, and the node that is the target of the LSP. The term “terminator” is used to refer to: Note that there is only one “initiator” and one “terminator” for bidirectional LSP), and also called wavelength band switching. Support a special case of switching (wavelength band represents a set of continuous wavelengths that can be switched together to a new wavelength band. A wavelength band label is defined to support this special case. Waveband switching is, of course, another level of label hierarchy. As far as the .MPLS protocol to enter, little difference is not in the wavelength band label and wavelength label).

  There are several new forms of “labels” to handle the extension of MPLS into the optical and temporal domains. These new forms of labels are collectively referred to as “generalized labels”. The generalized label contains enough information for the receiving node to make the cross-connect programmable regardless of the type of cross-connect so that the path ingress segments are properly concatenated. Generalized labels allow not only labels that travel in-band with associated data packets, but also representations of labels that identify time slot division multiplexing positions, wavelength division multiplexing positions, or space division multiplexing positions By expanding the conventional label. For example, a generalized label can be (a) a single fiber in a bundle, (b) a single wavelength band in a fiber, (c) a single wavelength in a wavelength band (or fiber), or (d) a wavelength (or A label representing a set of time slots in the fiber) can be carried. It can also carry a general MPLS label, a frame relay label, or a label representing an ATM label (VCI / VPI).

  Therefore, GMPLS forms a path called a label switch path (LSP) through the network. These paths can be connection-type or connectionless-type. For example, Resource Reservation Protocol (RSVP) is often used to deploy connection-oriented LSPs, while Label Management Protocol (LMP) such as Label Distribution Protocol (LDP) is provisioning for connectionless LSPs. Often used for.

Optical network An optical network is a collection of optical network devices interconnected by links comprised of optical fibers. Thus, an optical network is a network whose physical layer technology is fiber optic cable. The cable trunk is interconnected with an optical cross-connect (OXC), and signals are inserted and removed by an optical add / drop multiplexer (OADM). An optical network device that can use traffic entering and / or exiting an optical network is called an access node, in contrast, an optical network device that cannot do so is called a pass-through node (an optical network is a pass-through node). -You do not have to have a node). Each optical link interconnects two optical network devices and typically includes optical fibers that carry traffic in both directions. There can be multiple optical links between two optical network devices.

  A given fiber can simultaneously carry multiple communication channels using a technique called wavelength division multiplexing (WDM), which is a form of frequency division multiplexing (FDM). When implementing WDM, each of a plurality of carrier wavelengths (or equivalently, frequency or color) is used to implement a communication channel. Thus, a single fiber looks like multiple virtual fibers, with each virtual fiber carrying a different data stream. Each of these data streams may be a single data stream or a time division multiplexed (TDM) data stream. Each of the wavelengths used for these channels is often called lambda.

  The light path is a one-way path in the optical network where lambda does not change. When a light path is given, the optical nodes that are the start and end points of the path are called the source node and the destination node, respectively, and are on the light path between the source and destination nodes. Multiple nodes (if any) are called intermediate nodes. An optical circuit is a bi-directional end-to-end path (between access nodes entering and leaving the optical network for traffic carried by the optical circuit) through the optical network. Each of the two directions of the optical circuit is composed of one or a plurality of light paths. In particular, if a single wavelength is used in a given direction of an optical circuit end-to-end path, a single end-to-end light path is provisioned for that direction (that The source node and destination node of the light path are the access nodes of the optical network and are the same as the end nodes of the optical circuit). However, if a single wavelength is not used for a given direction, wavelength conversion is necessary and more than one light path is provisioned for that direction of the end-to-end path of the optical circuit. Thus, a light path includes a lambda and a path (a set of optical nodes through which traffic travels with the lambda).

  In other words, when using GMPLS over an optical network, the optical network can be considered circuit switched, in which case the LSP is a circuit. Each of these LSPs (one-way or two-way) has an end label whose generalized label (s) is the wavelength (s) of the light path (s) used. Form a two-end path. If wavelength conversion is not used for a given bi-directional LSP, there is a single end-to-end light path in each direction (hence a single wavelength and therefore a single generalized label).

  An optical network device is considered to include two planes: a data plane and a control plane. Data planes are components through which light passes (eg, switch fabrics or optical cross-connects, input and output ports, amplifiers, buffers, wavelength splitters or optical line terminators, adjustable amplifiers, etc.), add / drop components ( For example, a transponder bank or an optical add / drop multiplexer), as well as light monitoring components. The control plane includes components that control the data plane components. For example, the control plane is often composed of software running on a set of one or more microprocessors in an optical network device that controls the components of the data plane. Specifically, for example, software executed on one or a plurality of microprocessors determines that the switch fabric needs to be changed, and then instructs the data plane to execute the switching. It should also be noted that the control plane of the optical network device is in communication with the centralized network management server and / or multiple control planes of one or more other network devices.

  Various network topologies have been developed for optical network devices, including ring and mesh topologies. Similarly, various control and data planes have been developed for optical network devices. For example, in wavelength division multiplexing (WDM), it is necessary to change the data plane and the control plane. As another example, various approaches have been used to implement switch fabrics, including optical cross-connects such as MEMS, acousto-optics, thermo-optics, holographic, optical phased arrays.

  Usually required for operation of an optical network are A) construction and maintenance of a network database, and B) establishment of a light path. For example, the network database may be 1) information about neighboring optical nodes (eg, using link management protocol (LMP)) (eg, one or more links, one or more lambdas, lambda bandwidths) A link state database that tracks (such as width), and 2) information about the physical connectivity of nodes throughout the domain and / or network (eg, using OSPF-TE) (eg, nodes, links, lambdas, etc.) Contains the topology database to be tracked. To establish an LSP, operations that are typically performed offline are: 1) the shortest path / between source and destination using a shortest path first algorithm based on network database (s). Determining the wavelength 2) Assigning the path / wavelength (often referred to as path signaling. Effectively informs the involved optical network device how to configure the switch fabric, eg RSVP with GMPLS Or CR-LDP based signaling). Steps 1 and 2 can be reversed.

  There are generally three ways to operate an optical network: 1) centralized static provisioning, 2) source-based static provisioning, and 3) hybrid static provisioning. In centralized static provisioning, an independent centralized network management server maintains a network topology database and communicates with each of the optical network devices in the network. In response to some predefined requests for the optical circuit, the network management server finds the shortest path / wavelength. Thereafter, the network management server assigns the path / wavelength and causes the switch fabric to be configured.

  In source-based static provisioning, each access node in the network performs the task of building / maintaining a network topology database. In response to some predefined request of optical circuitry received by the access node, the node 1) buffers traffic as needed, 2) finds the shortest path / wavelength, 3) the path / wavelength Have the assignment performed and configure the switch fabric.

  In hybrid static provisioning, each node of the network uses OSPF-TE to build a network topology database from which the network topology database is built and maintained in a centralized network management server. Is done. The network management server initiates a form of source-based provisioning. This allows the network administrator to maintain control of provisioning of each provisioned light path.

  One problem with existing optical networks is the network topology database used and how it is built and maintained. In particular, these monolithic physical topology databases (eg OSPF-TE) must store all data to give a physical view of the network (there are multiple lambdas per link, given Because different lambdas on the link give different bandwidth, connectivity at the lambda level as well as connectivity at the link level). These large network databases are relatively time consuming to analyze and require a relatively long time and a relatively large amount of node intercommunication to propagate changes. Furthermore, such a network topology database will be even larger if it is necessary to record QoS type information.

  Another problem with existing optical network devices is the static and offline nature of operations (eg, using centralized static provisioning, source-based static provisioning, hybrid static provisioning, etc.). More specifically, determining the expected needs of the end node's connection to the end node via the optical network, and the current state of the optical network (basically a snapshot of the network parameters) Based on the above, provisioning of an optical circuit is performed through an optical network through which traffic is transmitted. This optical circuit is static in the sense that it will not be changed on-the-fly based on current bandwidth requirements (demand changes), network current status, and the like. Instead, this optical circuit is only modified if it is reprovisioned (eg if a customer requests an upgrade to a wide or narrow bandwidth) and / or some form of protection switch is redundant Based on the scheme.

  Thus, a given optical circuit is established with the maximum bandwidth deemed necessary at any point in time, and this maximum bandwidth is provisioned for that purpose. That is, the designated customer is provisioned with a certain amount of bandwidth for all classes of traffic, whether the customer is using some or all of that bandwidth at any time, It is done regardless of whether it is not used. Bandwidth becomes useless due to changes in bandwidth requirements and / or network status.

  Therefore, the optical layer is operated to realize a point-to-point link that does not have intelligence functions or real-time decision making functions. In addition, the need to know demand in advance places challenges on service creation and service provisioning. For this reason, the use of resources in the optical layer becomes inefficient.

  In addition, work is being done to implement a quality of service (QoS) mechanism at the IP layer and / or using MPLS, both of which are via SONET, Does not distinguish between traffic types (does not have QoS mechanism). As a result, the regular optical control plane does not have the ability to divide traffic into different classes based on service level requirements (that is, it does not incorporate service level requirements for different types of traffic into the lightpath calculation. ).

  The optical network topology database and the optical network operation will be described. According to one embodiment of the invention, a number of wavelength division multiplexed access nodes use a distributed search-based scheme to create a network topology database based on a set of one or more connectivity constraints. To construct.

  According to another embodiment of the invention, a set of one or more connectivity constraints including quality of service (QoS) based criteria is applied to the physical network topology of a wavelength division multiplexed optical network. Divide the optical network into different service levels. In addition, a service level topology is determined for each of the service levels.

  According to another embodiment of the present invention, a number of wavelength division multiplexed access nodes in an optical network employ a source-based scheme to establish a communication path. Each of these access nodes stores a set of one or more network topology databases based on a set of one or more connectivity constraints.

  In accordance with another embodiment of the invention, a wavelength division multiplexed optical network includes a number of nodes, each node comprising an optical cross-connect, each node in which from that node to another node. Stores a database that represents connectivity without conversion. In addition, each of these nodes uses a messaging scheme to propagate notifications of changes in the optical network to other nodes and maintain its database. Each node's messaging scheme only sends messages to selected nodes among other nodes based at least in part on connectivity without conversion, thereby minimizing the number of communications between nodes. suppress.

  The invention may best be understood by referring to the following description and the accompanying drawings, which are used to illustrate some embodiments of the invention.

  In the following description, numerous specific details are set forth (eg, logical resource partitioning / sharing / overlapping implementation, system component types and relationships, and logical partitioning / integration selection). However, it will be understood that some embodiments of the invention may be practiced without these specific details. In other instances, well known circuits, software instruction sequences, structures, techniques have not been shown in detail in order not to obscure the understanding of this description.

  References to “one embodiment”, “an embodiment”, “an exemplary embodiment”, etc. in the specification may include certain features, structures, or characteristics of the described embodiment. Indicates that not all embodiments necessarily include a particular function, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, where a particular function, structure, or characteristic is described with respect to one embodiment, such function, structure, or characteristic is achieved with respect to other embodiments, whether or not explicitly described. Is considered to be within the knowledge of those skilled in the art.

  In the following description and claims, the terms “coupled” and “connected” may be used with their derivatives. It will be understood that these terms are not intended as synonyms for each other. Rather, in certain embodiments, “connected” can be used to indicate that two or more elements are in direct physical or electrical contact. “Coupled” also means that two or more elements are in direct physical or electrical contact. However, “coupled” also means that two or more elements cooperate or interact even if they are not in direct contact.

Overview According to some embodiments of the present invention, a set of one or more connectivity constraints is imposed on the construction / maintenance of a network topology database. As a result of the set of connectivity constraints, such a network topology database is small compared to a network topology database that represents all physical connectivity in the network. In accordance with one aspect of the invention, the set of one or more connectivity constraints includes one or more QoS-based criteria, and thus different classes of traffic based on different QoS requirements. Can be effectively divided into QoS-based logical network views that can be used to provision different wavelengths for. According to another aspect of the present invention, the set of one or more connectivity constraints includes no conversion constraints, so that an optical circuit without conversion can be established. According to another aspect of the invention, a distributed search-based approach is used to build and maintain a network topology database based on a set of connectivity constraints. According to another aspect of the invention, the optical network uses a source-based scheme in which a network topology database based on a set of connectivity constraints is maintained in the access node. Because the network topology database is small (compared to the physical network topology database) and this source-based scheme is distributed, optical circuit provisioning is real-time (or on-the-fly, or demand) Does not need to be known in advance).

  Since each of the above aspects is independent, different embodiments can implement different aspects and / or combinations of the above aspects of the invention. For example, in some embodiments of the present invention, both QoS criteria and no-transform constraints are included in the connectivity constraints. The network topology database based on this set of connectivity constraints has 1) reduced size across the physical connectivity network topology database, and 2) different wavelengths based on the QoS of different classes of traffic. 3) An optical circuit without conversion can be established. Some of these embodiments use a distributed search-based approach to implement a source-based scheme and build / maintain a network topology database, while others in these embodiments are Different schemes and / or different database construction / maintenance techniques can be used.

Exemplary QoS Embodiments Different embodiments of the present invention support different QoS criteria. For example, QoS criteria include bandwidth, bit error rate, optical signal-to-noise ratio, peak noise level, rerouting priority, and the like. That is, the QoS criteria can include criteria that can distinguish between different wavelengths based on the quality of service. For a given wavelength on a given link, the QoS criteria value (wavelength parameter) may be determined based on the configuration (eg, the type of laser used) and / or by starting the light. it can.

  The QoS criterion is used to classify the wavelengths on the link into one of the supported set of service levels. In particular, for each QoS criterion, a service level parameter is provided for each service level. The wavelength parameter for a given wavelength on a given link is compared with the service level parameter and the wavelength is classified into one of a plurality of service levels.

Exemplary Network In some optical networks, different wavelengths of at least some nodes have different wavelength parameters. For example, a given optical network device has different groups of wavelengths implemented to operate at different bandwidths (eg, group A for OS-X, group B for OS-Y, group C for OS-Z). And service level parameters that differentiate based on bandwidth. As a result, optical networks not only have interconnections given at the physical link level (physical topology), but also at each service level (for each service level, for the network and each node). A given interconnection for a given interconnect to a service level topology) and a service level without each translation (a service level topology without translation for each node for each service level) Have a connection.

  FIG. 1 is a block diagram illustrating an exemplary optical network according to an embodiment of the present invention. The optical network of FIG. 1 includes five access nodes labeled N1, N2, N3, N4, N5. The ability to implement multiple lambdas on a single link is expressed in a simplified form by numbering the lambdas, and lambdas with the same number are the same wavelength. FIG. 1 shows numbered lambdas that can be used on each optical link of an exemplary optical network. The term “available” means that when used in conjunction with a lambda number, the node has the ability to generate such a wavelength and can be used using assigned or unassigned terms. Identify whether a specific wavelength is currently provisioned. In order to explain the physical connectivity shown in FIG. 1, “node number: node number” indicates that there is an optical link between the nodes, and “node number: node number = lambda number ( The number of wavelengths that can be used on the link. Using this format, FIG. 1 shows that N1: N2 = lambda 1, 2, 3, 4, N2: N4 = lambda 1, 3, 5, N4: N5: = lambda 1, 2, 3, 4, N1: N3 = Lambda 1, 2, 4, N3: N4 = Lambda 1, 2, 4

  It will be appreciated that the topology of FIG. 1 is an example and that any number of different topologies may be used with the present invention. Further, although FIG. 1 shows that different wavelengths can be used over different optical links, it will be appreciated that the same wavelength can be used on all optical links. Furthermore, although specific wavelengths have been identified as usable in FIG. 1, optical network devices can be implemented to generate a variety of different wavelengths with a laser, and the present invention provides one Or equally applicable to an optical network including a plurality of such optical network devices. However, for purposes of explanation, some embodiments of the invention will be described with reference to the wavelengths illustrated in FIG. Although the exemplary optical network of FIG. 1 is comprised of multiple access nodes, some embodiments of the present invention are equally applicable to optical networks that include pass-through nodes.

  FIG. 2 is a block diagram illustrating an example QoS-based logical network diagram of the example optical network of FIG. 1 according to some embodiments of the present invention. In the example of FIG. 2, the set of supported service levels includes service levels A, B, and C. Each wavelength on each link is classified into one of service levels A, B, and C by comparing the wavelength parameter of each wavelength on each link with the service level parameter. . To refer to a given service level, place S in the front part of the service level label (SA, SB, SC). The format for identifying the wavelengths on a given link classified for a given service level is the best given by example. In particular, SA (N1: N2) = Lambda 1, 2 has an optical link between N1 and N2, and the wavelength parameters of Lambda 1 and Lambda 2 on that link are those for service level A (Lambda 1 and 2 are called the set of service level channels on link N1: N2 for service level A). Using this form, FIG. 2 shows that service level A connectivity is SA (N1: N2) = Lambda 1, 2, SA (N2: N4) = Lambda 1, SA (N4: N5) = Lambda 1, 2, SA (N1: N3) = lambda 1, 2, SA (N3: N4) = lambda 1,2. Service level connectivity for service level B is: SB (N1: N2) = Lambda 3, SB (N2: N4) = Lambda 3, SB (N4: N5) = Lambda 3, SB (N1: N3) = X , SB (N3: N4) = X (where X indicates an empty set). Service level C connectivity is SC (N1: N2) = Lambda 4, SC (N2: N4) = Lambda 5, SC (N4: N5) = Lambda 4, SC (N1: N3) = Lambda 4, SC (N3: N4) = Lambda 4.

  Thus, FIG. 1 shows the physical link level connectivity, and FIG. 2 shows the respective service level (service level topology for the network) connectivity. In fact, this service level node connectivity divides the optical network into QoS-based logical network views as shown in the figure. Thus, for the first access node, there are one or more paths on the physical link to the second access node (physical topology). For a given one of these paths, each of one or more links that make up the path may have the same service level wavelength. For a given one of these paths, there may also be one or more of the same wavelength at the same service level on one or more links that make up the path.

  3A-C illustrate service level topologies without service level AC conversion for N1 of the optical network of FIG. 2 according to some embodiments of the present invention. In particular, FIGS. 3A-C illustrate an untransformed service level topology in the form of a tree with N1 as the root with branches representing nodes to nodes links through the network. As used below, the phrase “set of path service level channels” means the intersection of a set of link service level channels on the link of the path. For example, the set of path service level channels for service level A paths N1: N2: N4: N5 are the link service level channel sets SA (N1: N2), SA (N2: N4). It is a common part of SA (N4: N5).

  FIG. 3A illustrates a service level A untranslated service level topology for N1 of the optical network of FIG. 2 according to some embodiments of the present invention. In FIG. 3A, N1 has branches to N2 and N3. For the N2 branch, Lambda 1 and Lambda 2 are available at service level A. Therefore, in the path from N1 to N2, lambda 1 and lambda 2 constitute a set of path service level channels for service level A. Similarly, in the branch from N1 to N3, the set of path service level channels includes lambda 1 and 2. For paths between adjacent nodes, the link service level channel set (eg, SA (N1: N2) = lambda 1, 2) is the same as the path service level channel set. Please keep in mind.

  There is a branch from each of N2 and N3 to a different representation of N4. The branch from N2 to N4 represents the path N1: N2: N4. Since the link service level channel sets for N1: N2 and N2: N4 include lambda 1, 2 and lambda 2, respectively, the common part of these link service level channel sets is lambda 1 only. (Only N1: N2: N4 path without conversion uses lambda 1 in both N1: N2 and N2: N4). Thus, the set of path service level channels for paths N1: N2: N4 includes only lambda 1. In contrast, the branch from N3 to N4 represents the path N1: N3: N4. The intersection of the link service level channel set for N1: N3 and N3: N4 includes lambda 1 and 2, so the set of path service level channel for path N1: N3: N4 is lambda 1 And 2.

  The branch from N1 to N4 through N2 branches to 1) N3 with path service level channel set lambda 1 and 2) to N5 with path service level channel set lambda 1 To do. The branch from N1 to N4 through N3 branches to 1) N2 with path service channel set lambda 1 and 2) to N5 with path service level channel set lambda 1,2. To do. Even though the set of link service level channels for N4: N5 includes lambda 1,2, the set of path service level channels for N1: N2: N4: N5 includes only lambda 1 and no conversion Note that it remains. This is in contrast to the set of path service level channels for N1: N3: N4: N5, which includes both lambda 1 and 2, but it can both be used on each link of this path That's why.

  FIG. 3B illustrates a service level B untranslated service level topology for N1 of the optical network of FIG. 2 according to some embodiments of the present invention. The tree of FIG. 3B has no branch from N1 to N3 because there is no lambda suitable for service level B on link N1: N3. However, there is a branch from N1 to N2, and the set of path service level channels for the branch from N1 to N2 includes lambda 3. N2 has a branch to N4, which has it as a set lambda 3 for the path service level channel. Since there is no lambda suitable for service level B on the link from N4 to N3, there is no branch from N4 to N3. However, there is a branch to N5 and the set of path service level channels for N1: N3: N4: N5 includes lambda 3.

  FIG. 3C illustrates a service level C untranslated service level topology for N1 of the optical network of FIG. 2 according to some embodiments of the present invention. The tree of FIG. 3C branches from N1 to 1) N2 with path service level channel set lambda 4 and 2) N3 with path service level channel set lambda 4. There will be no branch from N2 because wavelength conversion will be required (N2: the link service level channel set for N4 is lambda 5, but the path service level for the path from N1 to N2 The set of channels includes lambda 4). There is a branch from N3 to N4, which has it as a set lambda 4 for the path service level channel. There is a branch from N4 to each of N2 and N5, for which both sets of path service level channels include lambda 4.

  It will be appreciated that the topology given for a node can be service level based and / or no conversion based (depending on the set of connectivity constraints used). Thus, the phrase “service level topology” for a node uses at least one QoS-based criterion, but does not exclude the use of a no-conversion criterion (unless otherwise noted herein). Similarly, the phrase “no transformation topology” for a node at least uses no transformation criteria, but does not exclude the use of QoS based criteria (unless otherwise noted herein). That is, one topology for a node is service level based does not indicate whether it is a non-transformed base, and one topology for a node is untransformed based. Does not indicate whether it is service level based, but one topology for one node is untransformed and QoS based indicates that it must be both.

  If the set of connectivity constraints includes QoS-based criteria, the service level topology for the network and the service level topology for each node (or no translation service level if no translation criteria are also used) Topology), and if the set of connectivity constraints includes no-conversion criteria, there is no-conversion topology for each mode (or no-conversion service-level topology if QoS-based criteria are also used) Will also understand. Different embodiments may store a network topology database representing one or more of those different topologies in different devices depending on the implementation and connectivity constraints used. For example, the centralized network management server may include a service level topology for the network, a service level topology for each node, one or more no-conversion topologies for each node, and / or no-conversion services for each node. A network topology database (s) representing the level topology can be stored. For example, each access node may have a service level topology for the network, a service level topology for the node, one or more untransformed topologies for the node, and / or an untransformed service level for the node. A network topology database (s) representing the topology may be stored. It will be appreciated that other configurations are within the scope of the present invention.

  FIG. 4 is a block diagram illustrating a hierarchy of terms according to some embodiments of the invention. The terms illustrated in FIG. 4 are used in connection with some embodiments of the invention described below. With respect to FIG. 4, the network is divided into sets of one or more service levels, each service level including a set of zero or more possible end-to-end paths, Each possible end-to-end path includes one or more links, and each link includes one or more available lambdas. A possible end-to-end path for a given service level is called a set of possible end-to-end service level paths (between access nodes with lambdas available at that service level. All paths that can be formed). The union of possible end-to-end service level paths for all service levels is called a set of end-to-end network paths. The links that make up a given path are called a set of path links, but a union of all possible end-to-end path links within a set of possible end-to-end service level paths Is called a set of service level links. A lambda on a service level possible end-to-end path link is called a link lambda, whereas a lambda union on a service level possible end-to-end path link is a path lambda. Called. The term service level link lambda is used to refer to a service level link and a link of the lambda above that suitable for that service level.

  The hierarchy illustrated in FIG. 4 forms a framework for a set of connectivity constraints that include one or more QoS-based criteria that divide the network into multiple service levels. If the set of one or more connectivity constraints further includes a no-conversion constraint, all link lambdas for a given service level possible end-to-end path link will be the same. In other words, the same lambda must be used on each link of the end-to-end path to achieve an end-to-end path without translation (the lambda must be the same on each link of the path) Must be appropriate for the service level). In contrast, if the set of connectivity constraints does not include the no-conversion constraint, the set of link lambdas may be different for different links in the service level possible end-to-end path .

Building and maintaining a network topology database with a set of connectivity constraints Various techniques have been described for building and maintaining a network topology database with a set of connectivity constraints. Is an aspect of the present invention that is independent of any aspect of the present invention, and thus the present invention is an exemplary approach for building and maintaining a network topology database with a set of connectivity constraints as described herein. It will be understood that the present invention is not limited to this.

Overview FIG. 5 is a flow diagram for building and maintaining a network topology database with a set of connectivity constraints according to some embodiments of the present invention. As will be described in detail below, it will be appreciated that different blocks of FIG. 5 can be implemented in a distributed and / or centralized manner.

  At block 505, the lambda for each link is tracked and control is passed to block 505. While some embodiments of the present invention use Link Management Protocol (LMP) to discover adjacent links between nodes, other embodiments of the present invention may use other approaches. (For example, a manual input method to each node, a manual input method to the centralized network management server, etc.). Further, although some embodiments of the present invention include a monitoring unit that measures wavelength parameters at one or more nodes of the network, alternative embodiments of the present invention may use other approaches. (E.g., periodic external test devices, manual input of wavelength parameters to each node, manual input of wavelength parameters to a centralized network management server, etc.).

  As shown in block 510, classification according to QoS criteria is maintained such that each link lambda determines the service level link lambda and control flow to block 515. In some embodiments, block 510 is performed by each node for adjacent links, although alternative embodiments of the invention use an alternative approach (eg, a centralized network management server see block 505). However, block 510 is executed in response to receiving the wavelength parameter information as described).

  At block 515, service level connectivity based on conversion criteria is maintained for each service level. Maintained service level connectivity includes lambdas and status that can be used depending on whether they are assigned or unassigned. In some embodiments of the present invention, service level connectivity is built in a distributed manner and maintained at the access node, although alternative embodiments of the present invention can use alternative approaches (eg, centralized network management). Run in the server). The conversion criteria represents the number of wavelength conversions that are acceptable for a given optical circuit. For example, if one of the connectivity constraints is a non-conversion connectivity constraint, the number of allowable wavelength conversions is zero.

  FIG. 6 is a flow diagram illustrating write path provisioning according to some embodiments of the invention. In different embodiments of the present invention, such provisioning may be implemented using source-based, centralized, hybrid, or other provisioning schemes.

  At block 605, request criteria are received. A demand criterion represents a demand for a communication path (eg, optical circuit, light path, end-to-end unidirectional path, etc.). Control from block 605 is passed to block 610. In some embodiments of the invention that use a source-based scheme, the demand criteria is received by an access node in the optical network. In other embodiments of the invention using a centralized network management server, such demand is received by the network management server directly from the requester and / or directly from an access node in the optical network that receives the demand criteria. Of course, in alternative embodiments of the invention, other schemes may be used and / or implemented in other ways.

  As shown in block 610, if a service level is not specified, it is determined and control is passed to block 615. For example, some demand requests may come from an entity that recognizes the service level provided by the optical network, while other entities that make requests may not. These latter entities may not contain any parameters or may contain parameters that can determine the service level.

  At block 615, it is determined whether there are any end-to-end paths available at the determined service level. If there is an available path, control is passed to block 620. Otherwise, control is passed to block 625.

  At block 620, the path and required lambda are selected (assigned) and assigned. The number of different wavelengths assigned depends on the wavelength conversion criteria (for example, if no conversion connectivity constraints are used, the same wavelength or wavelengths are used across each link in the selected path. ) Since this assignment affects the service level connectivity of block 515, block 515 is updated (eg, some action is taken in response to periodic checks of assignment, formation, etc.). In some embodiments of the invention that use a source-based scheme, the source node 1) selects a path and one or more lambdas, and 2) communicates with other nodes in the optical network to allocate. To execute block 620. In another embodiment of the invention where a centralized network management server is used, block 620 is performed by a network management server that performs selection and communicates assignments to nodes on the path. Of course, in alternate embodiments of the invention, other schemes may be implemented in other ways.

  At block 625, alternative actions are performed depending on how the optical network is managed. For example, use a path from a higher service level, multiplex two or more paths from a lower service level, assign a single path from a lower service level, refuse or allow to perform more wavelength conversions One or more can be optional.

  In addition to the need to allocate in response to demand criteria, other operations are performed as part of optical network management (eg, deallocation, adding new wavelengths, adding new links, wavelength failure / restoration) Link failure / restore, node failure / restore, etc.). To implement an optical network for the current view, one or more blocks in FIG. 5 are updated in response to those changes. For example, if a request is made to release a path assignment, the node (source) that initiated the assignment is instructed to release the assignment of that path, and in block 515 the service including the deallocated path is serviced. • Update service level connectivity for the level. Updates are made through blocks 505, 510, 515 as a result of the addition of wavelengths, links, or nodes (as well as the removal of wavelengths, links, or nodes that were not carrying active traffic). Loss of a wavelength, link, or node is a failure that some action is taken depending on the redundancy scheme implemented (different redundancy schemes may be used in different embodiments of the invention), or that wavelength from the network , Link, or node exclusion.

  Also, a request to change the demand criteria for a given provisioned service (eg, a request to modify the service level of a given provisioned service) is addressed by some embodiments of the present invention. Also note that. In particular, some such embodiments are successful in assigning a new path and, if necessary, migrating traffic from the old path to the newly assigned path and releasing the old path assignment. To respond to such a request. While the above uses of a variety of different approaches can be implemented in different embodiments, embodiments using a source-based scheme are described below by way of example and are not intended to be limiting.

  Of course, one or more portions of an embodiment of the present invention may be implemented using any combination of software, firmware, and / or hardware. Such software and / or firmware may be a magnetic disk, optical disk, random access memory, read only memory, flash memory device, electrical, optical, acoustic, or other form of propagated signal (eg, carrier wave, infrared Can be stored and communicated using machine-readable media such as signals, digital signals, etc. (internally and at other access nodes on the network).

Exemplary Distributed Search Techniques Some embodiments of the present invention are for building and maintaining a network topology database at a source node based on a set of connectivity constraints including QoS criteria and no transformation constraints. This will be described with reference to the distributed search-based method. However, in alternative embodiments, a distributed approach can be used, but a service level topology database cannot be built and maintained in the source node (eg, built and maintained in a centralized network management server). Will be understood). In addition, although a distributed search-based approach has been described, alternative embodiments (eg, a centralized approach) can be used in alternative embodiments. Similarly, alternative embodiments of the present invention do not include a no-conversion connectivity constraint, or if necessary (eg, where there is no end-to-end path without a requested service level conversion). You can relax it.

  FIG. 7 is a block diagram illustrating an exemplary access node according to some embodiments of the present invention. FIG. 7 shows a control plane 700 that is coupled to the data plane 701. The control plane 701 comprises a node database 702 coupled with a node module 735. Of course, the control plane 700 includes other items (eg, protocols).

  FIG. 7 shows that the node database 702 includes a service level connectivity database 705, a service level parameter database 710, a link state database 715, and a routing database 720. The link state database 715 includes a set of one or more link state structures 725, one for each link connected to that node. In some embodiments, the link is discovered through a link management protocol, but in alternative embodiments, other approaches can be used as described above. Each link state structure consists of neighboring nodes, ports used when neighboring nodes are connected (fiber links terminate at one port on the node), wavelengths available on that link (through ports) As well as record the parameters of each wavelength.

  Service level parameter database 710 stores the service level parameters already described herein. The service level connectivity database includes a set of one or more service level topology structures 730, one for each service level. Each of these service level topology structures stores a representation of the untransformed service level topology for that node (see, eg, FIGS. 3A-C). In addition, the service level topology structure for each service level tracks the assigned / unassigned status for each lambda in that topology. The status is not limited to being assigned or unassigned. For example, a “damaged” status can be assigned to a lambda that has failed due to a fiber being cut. In embodiments where only bi-directional paths can be assigned, the granularity of the assigned / unassigned status is merely a lambda level. However, in embodiments where one-way path assignment is possible, the granularity of assigned / unassigned status is the status of each direction for each lambda.

  The node module 735 includes a startup module 740, a connectivity request module 745, an allocation module 750, an allocation release module 755, and an add / delete module 760. The operation of these modules in some exemplary private embodiments is described with respect to FIGS. 9-10, 11, 12-14, 15-16, and 17-18, respectively.

Startup FIG. 8 is an exemplary data flow diagram of formation of a service level A service level topology for N1 of the optical network of FIG. 2 according to some embodiments of the present invention according to a distributed search-based technique. FIGS. 9-11 are flow diagrams of a distributed search-based approach for building a service level topology using a set of connectivity constraints including QoS criteria and no conversion criteria within an access node of an optical network. . By way of example, FIGS. 9-11 will be described with reference to the exemplary data flow diagram of FIG. The operations of this flowchart and other flowcharts are described with reference to exemplary embodiments of other charts. However, the operations of the flowchart can be performed by some embodiments of the invention other than those described with reference to these other figures, and will be described with reference to these other figures. It will be appreciated that some embodiments of the present invention may perform different operations than those described with reference to the flowchart.

  FIG. 9 is a flow diagram performed by each access node in joining an optical network according to some embodiments of the present invention. The flow diagram starts in response to provision of wavelength parameters and service level parameters (905). Referring to the database of FIG. 7, this is performed in response to populating the service level parameter database 710 and the link state database 715. In some embodiments, populating the service level parameter database is performed by the service provider through a network management interface.

  At block 910, the number of service levels is determined and control is passed to block 915. In some embodiments of the present invention, block 910 is performed by analyzing a service level parameter database.

  As shown in block 915, for each link to an adjacent node, the lambda on that link is classified by the service level parameter to form a set of link service level channels. Referring to FIG. 8, the set of link service level channels for service level A of each node shown in FIG. 8 is represented by a box next to that node. Control from block 915 is passed to block 920.

  At block 920, service level topology construction is initiated for each service level.

  FIG. 10 is a flow diagram illustrating service level topology enhancement over a single service level according to some embodiments of the invention. Accordingly, the flow of FIG. 10 is performed for each service level responsive to block 920.

  At block 1005, the service level topology structure is instantiated and populated by the qualifying neighbor (the neighbor whose source node has a non-empty set of link service channels at this service level). Control is passed to block 1010. Referring to FIG. 8, N1 instantiates a service level topology structure 730 in its service level connectivity database 705. The service level topology structure includes not only N1 in its route, but also branches to N2 and N3 respectively.

  As shown, at block 1010, one or more connectivity request messages are transmitted to the neighboring node (s) to be modified and control is passed to block 1015. In some embodiments of the invention, each connectivity request message includes a request ID, a source node ID, a forwarding node ID, a service level, a calculated set (along with a set of one or more paths) Path service level channel set). Not all of these fields are required for block 1010 (eg, the source node is the same as the forwarding node and the necessary information in the calculated set is already known by the neighboring node), but they are , Used when a search is performed through the network (see FIG. 11). In some embodiments of the present invention, each connectivity request message includes the above-described fields, but alternative embodiments can be implemented by other methods (eg, full version connectivity request message). Can be used for FIG. 11 and a reduced version of the connectivity request message can be used for block 1010, where such a reduced version simply includes only the request ID, source node ID, and service level. it can). Referring to FIG. 8, N1 transmits a connectivity request message to each of N2 and N3 (source node ID is N1 and service level is A).

  At block 1015, the service level topology structure is updated in response to receiving the connectivity response message. The node sending such a connectivity response message and the content of such connectivity response message will be described later with respect to FIG. For example, upon receiving a connectivity response message, the received data is added to the appropriate branch of the appropriate service level topology structure. After receiving the connectivity stop message, the path identified in the received message of the service level topology structure is complete for the service level identified in the received message. For the example optical network described herein, the service level topology structure for service level A represents something similar to that shown in FIG. 3A. For example, in some embodiments of the invention, a table is maintained with each available path and its corresponding set of path service level channels (eg, each entry in the table is available for use). One of the active paths and its corresponding set of path service level channels).

  FIG. 11 is a flow diagram illustrating operations performed by a node in response to a connectivity request message received via a link according to some embodiments of the invention. Referring to FIG. 8, node N2 receives a connectivity request message from node N1.

  At block 1110, it is determined whether a connectivity request message has already been processed. If so, control is passed to block 1115, at which point the flow is complete. Otherwise, control is passed to block 1120. The connectivity request message may have already been processed because it has been received from a different neighboring node. The determination as to whether the connectivity request message has already been processed can be performed in various ways. For example, in one embodiment, where the connectivity request message includes a request ID and a source node ID, this determination may include this request ID or source node ID of the current connectivity request message and that for the previous connectivity request message. This can be done by comparing with such a log.

  As shown in block 1125, the intersection of the received set of path service level channels for the path to this node and the set of link service level channels for each propagation port is determined. The Control from block 1125 is passed to block 1130. The phrase propagation port refers to a port other than 1) the port from which the connectivity request message was received, and 2) the port connected to the source node (ie, the source node is adjacent to this node). In some embodiments of the invention, the propagation port is determined by selecting a link from a forwarding state ID identified in the connectivity request message and a link state database not connected to the source node ID. Referring to FIG. 8, N2 receives the connectivity request on the link to N1, and since N1 is the source node, N2 selects the port used to connect the link to node N4. N2 then determines the intersection of the set of path service level channels for N1: N2 and the link service level set for N2: N4. This common set is the set of path service level channels for N1: N2: N4 and is included in the calculated set (the set of path service level channels is below N2 in FIG. (Shown as a dashed rectangle). In particular, N2 is the intersection of the path service level channel set (SA (N1: N2) = Lambda 1, 2) and the link service level set (SA (N2: N4) = Lambda 1): , Lambda 1 (this is also represented here using the form SA (N1: N2: N4) = lambda 1).

  Thus, the calculated set represents the intersection of the set of preceding link service level channels for the path that carried the connectivity request message. For block 1010, the calculated set is the same as the set of link service level channels for the link used to transmit the connectivity request message. However, when the connectivity request message is retransmitted to other nodes, the calculated set represents the set of intersections for the paths that have been taken and for each such path.

  At block 1130, it is determined whether there is a non-empty set of intersections. If so, control is passed to block 1140. If not, control is passed to block 1135.

  As indicated by block 1140, the connectivity response message is transmitted to the source node along with the common set (s) as a set of one or more path service level channels. Control is then passed to block 1145. In some embodiments of the invention, each connectivity response message includes a service level, a request ID, a response node ID, and a calculated set. For node N2, N2 sends to N1 a connectivity response message containing lambda 1 as the calculated set N1: N2: N4.

  At block 1145, the connectivity request message is transmitted on the propagation port along with the set of common part (s) as the set of path service level channel (s) and control is provided. Passed to block 1115. For node N2, N2 sends a connectivity request message to N4. For some embodiments of the invention in which the connectivity request message includes a request ID, source node ID, forwarding node ID, service level, calculated set, N2 has a request ID, N1 in their fields, respectively. , N2, A, calculated set N1: N2: N4, lambda 1 are written.

  At block 1135, the connectivity stop message is sent back to the source node and control is passed to block 1115. In some embodiments, such a connectivity stop message includes a source node ID.

  To complete the example of FIG. 8, in response to the connectivity request message from N1, N3 determines a set of intersections to N4 for service level A. N3 sends this set of common parts back to N1 in a connectivity response message and further sends it to N4 in a connectivity request message. Meanwhile, in response to N2's connectivity request message, N4 determines a set of intersections for N3 and N5 for service level A. N4 sends them back to N1 with a connectivity response message and sends them to N3 and N5 with a connectivity request message. In response to N4's connectivity request message, 1) N3 does not do anything because it has already seen this request ID before (the above connectivity request message from N1), 2) N5 stops connectivity Send the message back to N1. In response to N3's connectivity request message, N4 does nothing because it has already seen this request ID before (the above connectivity request message from N2).

Allocation FIGS. 12-14 are flow diagrams illustrating path allocation according to some embodiments of the invention. FIG. 12 is a flow diagram illustrating operations performed by an access node to assign a path according to some embodiments of the invention. As a result of the operation of FIG. 12, 1) update the routing database message (s) sent to the node along the selected path to be assigned, and 2) the assignment sent to several nodes. Update channel message (s). FIG. 13 is a flow diagram illustrating operations performed by an access node in response to an update routing database message according to some embodiments of the invention, and FIG. 14 is a diagram of some embodiments of the invention. Is a flow diagram illustrating operations performed by an access node in response to an update assignment channel message.

  Referring to FIG. 12, the access node (acting as the source node) receives the demand for the path and control is passed to block 1210. There are various mechanisms for receiving such demand for paths by the access node. For example, in some embodiments of the present invention, OIF-UNI and / or OIF-NNI interface protocols are used to communicate with nodes or domains that do not support GMPLS or MPLS, respectively.

  At block 1210, the service level and destination node for the demand are determined and control is passed to block 1220. Block 1210 may be performed in the same manner as block 605 of FIG.

  As indicated by block 1220, it is determined whether there are any paths available at that service level. If not, control is passed to block 1225. If so, control is passed to block 1230. Block 1220 can be implemented in various ways. With respect to the exemplary embodiment of FIG. 7, the service level topology structure is analyzed to determine if the destination node is reachable and if there are any unassigned lambdas available. In some embodiments, the service level topology structure is analyzed in response to demand, but in alternative embodiments of the present invention, various paths are analyzed and / or preselected faster (and preassigned). Generate a derived structure (see below for example). Block 1220 is similar to block 615 of FIG.

  As indicated at block 1225, an alternative action is taken. Block 1225 is similar to block 625, and the various alternatives described therein are equally applicable here.

  At block 1230, a path and channel are selected based on the selection criteria, and control is passed to block 1235. In some embodiments, path and channel selection includes selection of a node: channel: port sequence for that path (as in this case, if no conversion connectivity constraint is used, a single Note that the same channel is used). Different embodiments may use different selection criteria for path and channel selection. For example, some embodiments of the present invention utilize load balancing as described later in this document. It will also be appreciated that various path computation techniques can be used, including Dijkstra's algorithm.

  As indicated by block 1235, the routing database is updated and control is passed to block 1240. The routing database is updated to reflect incoming port connections identified by demand in the outgoing channel: port with block 1205 of the selected path. In some embodiments, well-known techniques are used to respond to routing database updates and modify the access node's data plane accordingly (of course the data plane first And / or alternative embodiments can be implemented to modify through different mechanisms). Whether the opposite direction path is also assigned depends on whether the implementation requires all paths to be bidirectional and / or whether a bidirectional path is required on demand.

  At block 1240, one or more update routing database messages are transmitted to the nodes on the selected path, and control is passed to block 1245. In some embodiments of the invention, each update routing database message includes channel and port information related to the receiving node of the message as well as the update ID.

  As indicated by block 1245, the selected service level topology structure is updated and control is passed to block 1250. In particular, the selected channel is marked as assigned in the set of all path service level channels that are downstream of the links in the selected path. That is, the selected channel is within the set of path service level channel (s) of the available path (s) that include one or more links of the selected path. Marked as assigned. To illustrate, assume that paths N1: N2: N4 are assigned with lambda 1 of FIG. Referring to FIG. 3A, lambda 1 has N1: N2, N1: N2: N4, N1: N2: N4: N3, N1: N2: N4: N5, N1: N3: N4: N2 path service level It needs to be marked as assigned from a set of channels because each contains one or more links on the selected path.

  At block 1250, an update assignment channel message is sent to the selected service level topology structure node. In some embodiments, each update assignment channel message includes an update ID, service level, path, assigned channel, destination set. The destination set represents the set of nodes to which the message will be sent. Although the node to which a message is sent can be determined in various ways, some embodiments of the present invention analyze the service level topology structure and analyze all of the nodes away from the source node (excluding duplicates). ).

  FIG. 13 is a flow diagram illustrating operations performed by an access node in response to an update routing database message according to some embodiments of the invention. At block 1310, the routing database is updated. The receiving node's routing database is updated to reflect the connection identified in the received message. In some embodiments, well-known techniques are used to respond to routing database updates and modify the access node's data plane accordingly (of course the data plane first And / or alternative mechanisms can be implemented to modify through different mechanisms, eg, signaling).

  Block 1310 is performed in the same manner as block 1235 of FIG.

  FIG. 14 is a flow diagram illustrating operations performed by an access node in response to an update assignment channel message according to some embodiments of the invention. At block 1410, it is determined whether an available path of the service level topology structure for the service level of the assigned path includes one or more links of the selected path. If not, control passes to block 1415 where the flowchart ends. If so, control is passed to block 1420. In some embodiments of the invention, block 1410 is performed by analyzing the appropriate service level topology structure to determine if a link on the selected path is represented there.

  As indicated by block 1420, it is determined whether the set (s) of path service level channels of the available path identified in block 1410 includes the assigned wavelength. . If not, control passes to block 1415. If so, control is passed to block 1425. In some embodiments of the invention, block 1420 is performed with the set of path service level channel (s) for the identified path (s) to determine the assigned wavelength. It is determined whether it exists.

  As indicated by block 1425, the selected service level topology structure is updated and control is passed to 1430. In some embodiments of the invention, block 1425 marks the wavelength assigned as assigned in the set of path service level channel (s) identified in block 1420. Executed.

  As indicated at block 1430, nodes not identified in the received assigned channel message are selected from the service level topology structure and control is passed to block 1435. In some embodiments of the invention, block 1430 “1” identifies all of the nodes in the service level topology structure that are not included in the destination set in the received update assignment channel message (1405). Identified as “new set”, and 2) performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received update assignment channel message (1405).

  At block 1435, an update assigned channel message is sent to the selected node. In some embodiments of the invention, block 1435 is performed by sending an update assignment channel message that includes the updated destination set to all nodes in the new set of block 1430.

Allocation Release In response to a request to deallocate a channel (eg, communication received by a source node of a path using a channel), signaling is used to disconnect an existing cross-connect on the path. FIG. 15 is a flow diagram illustrating operations performed by a source node of a path in response to the path allocation being released according to some embodiments of the present invention. As part of the flow diagram of FIG. 15, the source node of the path sends an update allocation release channel message (s) to several other nodes. FIG. 16 is a flow diagram illustrating operations performed by an access node in response to receiving an update allocation release channel message according to some embodiments of the invention.

  At block 1510, the service level of the channel to be deallocated is determined and control is passed to block 1515. In some embodiments, the service level is determined by analyzing the link state database and the channels to be deallocated are identified.

  As indicated by block 1520, the service level topology structure is updated and control is passed to block 1525. Block 1520 is performed similar to block 1245 except that the channel is marked unassigned. In particular, deallocated channels are marked as unallocated in the set of all path service level channels downstream of links in the deallocated path. That is, a deallocated channel is within the set (s) of path service level channels of one or more available paths that include one or more links of the deallocated path. Marked as unassigned. To illustrate, assume that paths N1: N2: N4 are deallocated with lambda 1 in FIG. Referring to FIG. 3A, lambda 1 has N1: N2, N1: N2: N4, N1: N2: N4: N3, N1: N2: N4: N5, N1: N3: N4: N2 path service level It needs to be marked unassigned from the set of channels because each contains one or more links on the deallocated path.

  As indicated by block 1525, the update allocation release channel message is sent to a node in the service level topology structure and control is passed to block 1530. A set of nodes to which this message is sent is called a destination set. In some embodiments of the present invention, the update deallocation channel message includes the source node ID, neighbor node ID, path, deallocated channel, update ID, service level, destination set. Although the node to which a message is sent can be determined in various ways, some embodiments of the present invention analyze the service level topology structure and analyze all of the nodes away from the source node (excluding duplicates). ).

  As indicated by block 1530, the routing database is updated and control is passed to block 1540. Referring to the exemplary embodiment of FIG. 7, the routing database 720 is modified to remove the deallocated channel connection. Whether the path in the opposite direction is also deallocated depends on whether the implementation needs to make all paths bidirectional and / or if the path being deallocated was bidirectional.

  At block 1540, the updated routing database message (s) is transmitted to the nodes on the selected path. In some embodiments of the invention, each update routing database message includes channel and port information related to the receiving node of the message as well as the update ID. The receiving access node receives such a message by modifying its routing database to reflect the disconnection of the incoming channel: port and outgoing channel: port as specified in the message. Respond to.

  In this way, nodes along that path are updated to reflect the deallocation, and in addition, an update deallocation channel message has already been sent to initiate the necessary updates at such nodes. Yes.

  FIG. 16 is a flow diagram illustrating operations performed by an access node in response to an update allocation release channel message according to some embodiments of the invention. At block 1610, it is determined whether an available path of the service level topology structure for the service level of the deallocated path includes one or more links of the deallocated path. If not, control passes to block 1615 where the flowchart ends. If so, control is passed to block 1620. In some embodiments of the present invention, block 1610 is performed by analyzing the appropriate service level topology structure to determine if a link on the deallocated path is represented there.

  As indicated at block 1620, it is determined whether the set (or sets) of path service level channels of the available path identified at block 1610 includes deallocated wavelengths. The If not, control passes to block 1615. If so, control is passed to block 1625. In some embodiments of the present invention, block 1620 is performed with the set of path service level channel (s) for the identified path (s) to determine the assigned wavelength. It is determined whether it exists.

  As indicated by block 1625, the selected service level topology structure is updated and control is passed to 1630. In some embodiments of the invention, block 1625 marks unassigned and deallocated wavelengths in the set of path service level channel (s) identified in block 1620. It is executed by.

  At block 1630, a node in the service level topology structure not identified in the received update allocation release channel message is selected and control is passed to block 1635. In some embodiments of the invention, block 1630 includes 1) all nodes in the service level topology structure that are not included in the destination set in the received update allocation release channel message (1605). Identified as a “new set” 2) performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received update deallocation channel message (1605).

  As indicated by block 1635, the update allocation release channel message is sent to the selected node and control is passed to block 1615. In some embodiments of the invention, this update allocation release channel message is the new destination set determined in block 1630 as opposed to the destination set in the received update allocation release channel message (1605). including.

Dynamic provisioning As already mentioned, a request to change the demand criteria for a given provisioned service (eg a request to modify the service level of a given provisioned service) Addressed by the embodiment. In particular, some such embodiments are successful in assigning a new path and, if necessary, migrating traffic from the old path to the new assigned path and releasing the old path assignment. To respond to such a request.

  Because the network topology database is small (compared to the physical network topology database) and the source-based scheme is distributed, optical circuit provisioning is real-time (or on-the-fly, or demand) Does not need to be known in advance). Furthermore, according to the QoS base standard, the type of traffic can be distinguished in the optical layer. Thus, for example, the service provided to a customer can be a high service level during the day and lowered to a low service level during the night. Of course, it is possible to increase the frequency of such switching.

  Furthermore, the implementation can push SONET to the edge of the network. For example, instead of placing a stack of network layers (IP over ATM over SONET) on the light, the network layer can be placed directly on the light (eg, IP or ATM, or SONET).

Adding and Removing Channels FIGS. 17 and 18 are flow diagrams illustrating operations performed when a channel is added or a channel with no active traffic is removed according to some embodiments of the invention. is there. The operations of FIG. 17 are performed by an access node connected by a link to which a channel is added or removed (also referred to as an adjacent node or an access node that has become adjacent by that link). As part of these operations, add / remove channel messages are sent to several other nodes. The operations of FIG. 18 are performed by the access node in response to such an update add / remove channel message.

  FIG. 17 is a flow diagram illustrating operations performed by access nodes connected by links to which channels are added / deleted according to some embodiments of the present invention. At block 1710, the service level of the channel is determined and control is passed to block 1715. When a channel is added, block 1710 is performed by comparing the wavelength parameter of the channel with the service level parameter according to some embodiments of the invention, where the channel is one of the service levels. are categorized. Once the channel is removed, block 1710 is performed by accessing the link state database of FIG. 7 in accordance with some embodiments of the present invention.

  As indicated by block 1715, a connectivity request message is sent on the link carrying the channel and control is passed to blocks 1720 and 1725. Block 1715 is performed in the same manner as block 1515 of FIG.

  At block 1720, the service level topology structure is updated in response to receiving the connectivity response message. In some embodiments of the invention, block 1720 is performed in a manner similar to block 1015 of FIG. 10 that includes modifications. Because certain data already exists in the service level topology structure, the received data in the connectivity response message is used to update (add, remove, and / Or change). In the case of channel removal, in some embodiments of the invention, the channel on each path that includes the link is removed from the service level topology structure or marked as broken.

  As indicated by block 1725, the update add / remove channel message is sent to a node in the service level topology structure that is not on the link on that channel. In some embodiments of the invention, each update add / remove channel message includes an update ID, a wavelength, whether it is an add or remove, source node ID, source neighbor node ID, service level, destination set. Including. A source neighbor identified as a source node is an access node connected by a link whose channel has been added / removed. The destination set includes the nodes in the service level topology structure of the destination to which the message is sent at block 1725 (all nodes in the service level topology structure other than the source node and the source neighbor node).

  FIG. 18 is a flow diagram illustrating operations performed by an access node in response to receiving an update add / delete channel message according to some embodiments of the invention. As indicated by block 1810, it is determined whether the service level topology structure includes one or more paths with links added / removed channels. If not, control passes to block 1815 where the flowchart ends. If so, control is passed to block 1820. In some embodiments of the present invention, block 1810 includes (based on the received update) for the link identified in the received update add / remove channel message (based on the included source node ID and source neighbor node ID). This is done by searching the service level topology structure (for the service level identified in the add / remove channel message).

  At block 1820, one or more connectivity request messages are transmitted on one or more links of these paths, and control is passed to blocks 1825, 1830. In particular, access nodes send connectivity request messages on each of their links that are part of their path.

  At block 1825, the service level topology structure is updated in response to receiving the connectivity response message. Block 1825 is performed in the same manner as block 1720 of FIG.

  As indicated by block 1830, a node not identified in the received update add / remove channel message is selected from the service level topology structure and control is passed to block 1835. In some embodiments of the invention, block 1830 includes 1) all of the nodes in the service level topology structure that are not included in the destination set in the received update add / remove channel message (1805). As a “new set” and 2) executed by forming an updated version of the destination set that is a union of the new set and the destination set in the received update add / remove channel message (1805). The

  At block 1835, an update add / remove channel message is sent to the selected node. As before, this update add / remove channel message 1) identifies whether this is an add or remove, and 2) the destination set in the received update add / remove channel message (1805). In contrast, the updated destination set is included.

  With respect to the removal of channels with active traffic, some changes follow the flow of FIGS. In particular, each participating access node (the access node connected to the channel by the link and the access node receiving the update add / remove channel message) uses the removed channel, including the link To determine whether it is the source node of one or more assigned paths to be If so, the access node implements a redundancy (protection) scheme.

Link removal When a link between two nodes in a network is removed (eg, failed or permanently removed), all channels on that link are lost. Some embodiments perform the channel removal operations of FIGS. 17 and 18 for each such channel, but in other embodiments of the present invention, the message generated by handling the link as a whole. Reduce the number. In particular, FIGS. 19 and 20 are flow diagrams illustrating operations performed when a link is removed according to some embodiments of the present invention. The operations of FIG. 19 are performed by an access node connected by the link (also referred to as an adjacent node or an access node that has become adjacent by the link). As part of these operations, link removal messages are sent to several other nodes. The operation of FIG. 20 is performed by the access node in response to such a link removal message.

  FIG. 19 is a flow diagram illustrating operations performed by an access node connected by a delete link according to some embodiments of the invention. Block 1910 is used to indicate that the following blocks are executed for each service level.

  As indicated by block 1915, it is determined whether the service level topology structure includes one or more paths with removed links. If not, control passes to block 1930. If so, control passes to block 1925. In some embodiments of the invention, block 1915 is performed by searching the service level topology structure for the presence of removed links.

  At block 1925, the service level topology structure is updated and control is passed to block 1930. In some embodiments of the invention, the channels in the set of one or more path service level channels of these paths that are common to the set of link service level channels of the removed link are All are marked as broken (indicating that one or more channels cannot be used). In some embodiments of the invention, a channel marked as broken is retained indefinitely, but in other embodiments of the invention such a mark is marked after a period of time unless the link is re-established. Delete attached channel (and corresponding path). In other embodiments of the present invention, these one or more paths and channels are simply removed immediately and added back if reestablished (see link addition section).

  As indicated at block 1930, the link removal message is sent to a node in the service level topology structure that is not on the link. In some embodiments of the invention, each link removal message includes a set of link service level channels, source node ID, source neighbor node ID, update ID, service level, transmission of the removed link. Includes destination set. The source neighbor identified as the source node is the access node that connects to the removed link. The destination set includes the nodes in the service level topology structure of the destination to which the message is sent (all nodes in the service level topology structure other than the source node and the source source neighbor node).

  FIG. 20 is a flow diagram illustrating operations performed by an access node in response to receiving a link delete message according to some embodiments of the invention. As indicated by block 2010, it is determined whether the service level topology structure includes one or more paths with removed links. If not, control passes to block 2020. If so, control is passed to block 2015. In some embodiments of the present invention, block 2010 identifies (based on the received link removal message) for the link identified in the received link removal message (based on the included source node ID and source neighbor node ID). This is done by searching the service level topology structure (for the service level specified).

  At block 2015, the service level topology structure is updated and control is passed to block 2020. In some embodiments of the present invention, one or more path service levels (identified in block 2010) of these paths that are common with the set of link service level channels of the removed link. Any channel in the set of channels is marked as broken (indicating that one or more channels cannot be used). In some embodiments of the invention, a channel marked as broken is retained indefinitely, but in other embodiments of the invention such a mark is marked after a period of time unless the link is re-established. Delete attached channel (and corresponding path). In other embodiments of the present invention, these one or more paths and channels are simply removed immediately and added back if reestablished (see link addition section).

  As indicated at block 2020, a node not identified in the received link removal message is selected from the service level topology structure and control is passed to block 2025. In some embodiments of the invention, block 2020 includes 1) “new set” of all of the nodes in the service level topology structure that are not included in the destination set in the received link removal message (2005). 2) performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received link removal message (2005).

  At block 2025, a link removal message is sent to the selected node. As before, this link removal message includes the updated destination set as opposed to the destination set in the received link removal message (2005).

  With respect to the removal of links with active traffic, some changes follow the flow of FIGS. In particular, each participating access node (the access node connected by the removed link and the access node receiving the source link removal message) is the source of one or more assigned paths containing that link. Determine if it is a node. If so, the access node implements a redundancy (protection) scheme.

Link addition When a link is added, the LSD is updated in the access node connected to that link (eg, in some embodiments of the invention, the LMP recognizes the new link). If a link is added between two nodes in the network, multiple channels on that link can be made available all at once. Some embodiments perform the add channel operation of FIGS. 17 and 18 for each such channel, but in other embodiments of the present invention, the message generated by handling the link as a whole Reduce the number. In particular, FIGS. 21 and 22 are flow diagrams illustrating operations performed when a link is added according to some embodiments of the present invention. The operation of FIG. 21 is performed by an access node connected by the link (also called an adjacent node or an access node that has become adjacent by the link). As part of these operations, an add link message is sent to some other node. The operation of FIG. 22 is executed by the access node in response to such a link addition message.

  FIG. 21 is a flow diagram illustrating operations performed by access nodes connected by additional links according to some embodiments of the present invention. At block 2107, the one or more wavelengths on the added link are classified according to service level parameters to form a set of one or more link service level channels, and control is performed at block 2110. Passed to. In some embodiments of the invention, block 2107 is performed in the same manner as brick 915, except that only added links are processed.

  Block 2110 is used to indicate that the following blocks are executed for each service level where a new channel is added (a service level where the set of link service level channels for the added link is not empty): Is done.

  As shown, at block 2115, one or more connectivity request messages are transmitted to one or more modifying neighboring nodes, and control is passed to blocks 2120 and 2125. In some embodiments, block 2115 is performed in the same manner as block 1010.

  At block 2120, the service level topology structure is updated. In some embodiments of the present invention, block 2120 is performed in the same manner as blocks 1005 and 1015 of FIG. With respect to the changes on block 1005, the service level topology structure is placed with access nodes made adjacent by the added link (the service level topology structure is already placed by other neighboring nodes). Have been). With respect to the modification on block 1015, since the specific data already exists in the service level topology structure, the received data in the connectivity response message is used to create the existing service level topology structure. Update (add anything that doesn't already exist).

  At block 2125, an add link message is sent to the selected node in the service level topology structure. In some embodiments, each add link message includes a service level and a destination set (all of the nodes in the service level topology that are remote from the source node).

  FIG. 22 is a flow diagram illustrating operations performed by an access node in response to receiving an add link message according to some embodiments of the invention.

  As shown, at block 2210, one or more connectivity request messages are transmitted to one or more modifying neighboring nodes, and control is passed to blocks 2215 and 2220. In some embodiments, block 2210 is performed in the same manner as block 1010.

  At block 2215, the service level topology structure is updated in response to receiving the connectivity response message. Block 2215 is performed in the same manner as block 1015 of FIG. With respect to the modification on block 1015, since the specific data already exists in the service level topology structure, the received data in the connectivity response message is used to create the existing service level topology structure. Update (add anything that doesn't already exist).

  As indicated at block 2220, a node not identified in the received link addition message is selected from the service level topology structure and control is passed to block 2225. In some embodiments of the invention, block 2220 includes 1) “new set” of all of the nodes in the service level topology structure that are not included in the destination set in the received link addition message (2205). 2) is performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received link addition message (2205).

  At block 2225, an add link message is sent to the selected node. As before, this link addition message includes the updated destination set as opposed to the destination set in the received link addition message (2205).

Node removal When a node is removed, the LSD is updated in one or more adjacent access nodes (eg, in some embodiments of the invention, the LMP recognizes the removal of a new link). When a node is removed, the channels on that link or links are no longer available at the same time. Some embodiments perform the link removal operations of FIGS. 19 and 20 for each such link, but in other embodiments of the present invention, the message generated by handling the node as a whole Reduce the number. In particular, FIGS. 23 and 24 are flow diagrams illustrating operations performed when a node is removed according to some embodiments of the present invention. The operations of FIG. 23 are performed by one or more adjacent access nodes. As part of these operations, a node removal message is sent to several other nodes. The operation of FIG. 24 is performed by the access node in response to such a node removal message.

  FIG. 23 is a flow diagram illustrating operations performed by one or more access nodes in the vicinity of a deletion node according to some embodiments of the invention. Block 2310 is used to indicate that the following blocks are executed for each service level.

  At block 2315, the service level topology structure is updated and control is passed to block 2320. In some embodiments of the present invention, block 2315 is performed by removing the branch having the removed node as the first hop, if any, from the service level topology structure.

  As indicated by block 2320, it is determined whether the service level topology structure includes one or more paths with removed nodes. If not, control passes to block 2325 where the flow chart ends. If so, control is passed to block 2330. In some embodiments of the invention, block 2320 is performed by searching the service level topology structure for the presence of the removed node.

  As shown, at block 2330, one or more connectivity request messages are transmitted on one or more links of these paths, and control is passed to blocks 2335 and 2340. In particular, access nodes send connectivity request messages on each of their links that are part of their path.

  At block 2335, a new service level topology structure is instantiated and updated in response to receiving the connectivity response message. In some embodiments, block 2325 is performed in the same manner as blocks 1005, 1015 that include modifications. In particular, the new service level topology structure preserves the channel state from the current service level topology structure (held until the new service level topology structure is complete).

  At block 2340, a remove node message is sent to the selected service level topology structure node. In some embodiments of the invention that instantiate a new service level topology structure as in block 2325, the service level topology structure used in block 2330 is the current service level topology structure. It is. In some embodiments, each link removal message includes the removed node ID, service level, and destination set (services away from the removed node and nodes adjacent to the removed node. All of the nodes in the level topology).

  FIG. 24 is a flow diagram illustrating operations performed by an access node in response to receiving a delete node message according to some embodiments of the invention.

  At block 2410, it is determined whether the service level topology structure includes one or more paths with removed nodes. If not, control passes to block 2415 where the flowchart ends. If so, control is passed to block 2420. In some embodiments of the invention, block 2410 is performed by searching the service level topology structure for the presence of removed nodes.

  As shown, at block 2420, one or more connectivity request messages are transmitted on one or more links of these paths, and control is passed to blocks 2425 and 2430. In particular, access nodes send connectivity request messages on each of their links that are part of their path.

  At block 2425, a new service level topology structure is instantiated and updated in response to receiving the connectivity response message. In some embodiments, block 2425 is performed similarly to blocks 1005, 1015 that include modifications. In particular, the new service level topology structure preserves the channel state from the current service level topology structure (held until the new service level topology structure is complete).

  As indicated at block 2430, nodes not identified in the received node removal message are selected from the service level topology structure and control is passed to block 2435. In some embodiments of the invention, block 2430 “1” identifies all of the nodes in the current service level topology structure that are not included in the destination set in the received node removal message (2405). Identified as a “new set” and 2) performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received node removal message (2405).

  At block 2435, a node removal message is sent to the selected node. As before, this node removal message includes the updated destination set as opposed to the destination set in the received node removal message (2405).

  In some embodiments of the invention, nodes and their paths are immediately deleted and added back when re-established (see node addition section), although alternative embodiments include other mechanisms ( For example, in some embodiments of the invention, the path is marked as broken and retained indefinitely, and in other embodiments of the invention, the path is marked as broken and after a period of time if the node is not re-established Deleted, etc.)

Add Node When a node is added, the added node executes the flow charts in FIGS. Further, the LSD is updated in one or more neighboring access nodes (eg, in some embodiments of the invention, the LMP recognizes node removal). In addition, for each adjacent node, one or more links are effectively added to the new node. Thus, each of the one or more adjacent nodes performs the flow of FIG. 21 except that block 2125 is replaced with a different operation. In particular, instead of block 2125, an add node message is sent to the selected service level topology structure node. In some embodiments, each node addition message includes an additional node ID, a service level, and a destination set. Messages are sent to nodes in the service level topology that are remote from the source nodes, and the destination set includes those nodes.

  FIG. 25 is a flow diagram illustrating operations performed by an access node in response to receiving an add node message in accordance with some embodiments of the present invention.

  As shown, at block 2510, one or more connectivity request messages are transmitted to one or more qualifying neighboring nodes, and control is passed to blocks 2515, 2520. In some embodiments, block 2510 is performed in the same manner as block 1010.

  At block 2515, a new service level topology structure is instantiated and updated in response to receiving the connectivity response message. In some embodiments, block 2515 is performed in the same manner as blocks 1005, 1015 that include modifications. In particular, the new service level topology structure preserves the channel state from the current service level topology structure (held until the new service level topology structure is complete).

  As indicated by block 2520, a node not identified in the received node addition message is selected from the service level topology structure and control is passed to block 2525. In some embodiments of the present invention, block 2520 “1” displays all of the nodes in the current service level topology structure that are not included in the destination set in the received node addition message (2505). Identified as a “new set” 2) performed by forming an updated version of the destination set that is a union of the new set and the destination set in the received node addition message (2505).

  At block 2525, an add node message is sent to the selected node. As before, this node addition message includes the updated destination set as opposed to the destination set in the received node addition message (2505).

Changing Service Level Parameters In some embodiments of the invention, on the service provider side, the service level parameters can be updated and a new copy can be pushed on each node. If new QoS criteria are added, some embodiments of the invention do the following:
1. The contents of the service level parameter database are copied and kept in memory.
2. Populates the service level parameter database with new data.
3. Blocks 915, 920 are executed to create a new service level topology structure and keep the existing service level topology structure for each service level.
4). A new service level topology structure is used for new connections.
5. By comparing the parameters, the previous service level is mapped to the current service level.
6). The connection status from the previous service level topology is mapped to the new service level topology structure to match the associated service level.
7). The previous service level topology is deleted.

Similarly, if one or more existing service level parameters are changed, some embodiments of the invention do the following:
1. A particular level of content in the service level parameter database is copied and held in memory.
2. Populates the service level parameter database with new data.
3. A new service level topology structure is built for the updated level, retaining the previous service level topology structure.
4). A new service level topology structure is used for new connections.
5. By comparing the parameters, the previous service level is mapped to the current service level.
6). The connection status from the previous service level topology is mapped to the new service level topology structure to match the associated service level.
7). The previous service level topology is deleted.

  Of course, alternative embodiments can handle such changes in other ways.

Exemplary Load Balancing Load balancing issues arise when multiple shortest paths are available. For example, some embodiments of the present invention implement load balancing and make several options available to service providers. In particular, when demand is received, there are either 1) a set of multiple shortest paths, or 2) a single shortest path. If there is a set of multiple shortest paths, the wavelength is selected from each component of the set in a round robin fashion. However, if there is a single shortest path, one of two schemes is used. In the first scheme, a threshold is specified for any link in the network (eg, specified by a service provider). When the number of channels for a particular service level crosses the threshold on that link, that link becomes unavailable for future demand. This allows the service provider to rework the traffic flow on the network. In the second scheme, a composition ratio system is used. In particular, this ratio is the ratio of the number of new paths “assigned to non-shortest paths” to the number of “shortest paths”.

Exemplary contention resolution Since requests for paths by different access nodes may overlap, contention resolution is necessary. Some embodiments of the invention solve the contention problem by prioritizing a source with a high IP number. However, this leads to a special case where a particular source node may receive demand requests more frequently than other source nodes. As a result, it may not reach other source nodes.

  In other embodiments of the invention, one of the following contention resolution schemes is used to overcome this deficiency.

  1. In one such scheme, a light path for the next demand is allocated in advance. As a result, each access node pre-assigns a light path to each accessible node at each service level. As such, this scheme can place a relatively high burden on network resources.

  2. Another such scheme is referred to herein as highest service level pre-allocation. Instead of pre-assigning light paths to each accessible node at each service level, this is done only for the highest service level. If there is an adverse contention resolution at the time of demand allocation, this demand is allocated on the light path pre-allocated at the highest service level. Thus, while this scheme places a relatively small burden on network resources, it may use the highest service level light path fastest.

  3. Yet another such scheme is referred to herein as default service level pre-assignment. In particular, an indication of the default service level is maintained for each source and destination pair (eg, the most common service level for demand received so far). Instead of pre-assigning light paths to each accessible node for each service level, or pre-assigning light paths to each accessible node at the highest service level, the source and destination Pre-assignment is performed only for the default service level for the pair. If there is an adverse contention resolution at the time of demand allocation, this demand is allocated on a light path that is pre-allocated at the default service level. Thus, this scheme places a relatively low burden on network resources than Scheme 1, which attempts to avoid using the highest service level fastest by predicting the most common service level.

Aggregation Although several embodiments have been described in which separate messages are sent, alternative embodiments aggregate different messages of such messages. For example, in some embodiments, messages for different service level topologies are aggregated at startup.

Alternative Embodiments While various embodiments of the invention have been described, some alternative embodiments of the invention may operate differently. For example, while the flowcharts in these figures illustrate that operations are performed in a particular order according to some embodiments of the present invention, it will be understood that such order is exemplary. (For example, in an alternative embodiment, operations may be performed in a different order, some operations may be combined, some operations may be overlapped, etc.). In addition, it operates to reduce the number of communications between nodes by sending messages only to selected nodes (eg, blocks 1250, 1320, 1430/1435, 1525, 1630/1635, 1725, and 1830/1835). Although some embodiments are described, alternative embodiments may be implemented using different schemes and sending to many, few, or different nodes that send such messages (e.g., some In an alternative embodiment, each such message is broadcast to all nodes). As another example, although some embodiments of the invention have been described with respect to a distributed search technique for building / maintaining a network topology database and with respect to a source-based scheme, different methods or methods are described. Alternative embodiments can be implemented with a combination of (eg, centralized network topology database construction / maintenance, centralized provisioning, hybrid, etc.).

  While the invention has been described in terms of several embodiments, those skilled in the art will not be limited to the described embodiments and modifications and variations within the spirit and scope of the appended claims. It will be understood that The description is thus to be regarded as illustrative instead of limiting.

1 is a block diagram illustrating an exemplary optical network according to one embodiment of the invention. FIG. FIG. 2 is a block diagram illustrating an example QoS-based logical network diagram of the example optical network of FIG. 1 according to some embodiments of the present invention. FIG. 3 illustrates service level A, B, and C untranslated service level topologies for N1 of the optical network of FIG. 2 according to some embodiments of the present invention. FIG. 6 is a block diagram illustrating a hierarchy of terms according to some embodiments of the invention. 2 is a flow diagram for building and maintaining a network topology database with a set of connectivity constraints according to some embodiments of the present invention. 5 is a flow diagram illustrating write path provisioning according to some embodiments of the invention. FIG. 3 is a block diagram illustrating an exemplary access node according to some embodiments of the present invention. FIG. 3 is an exemplary data flow diagram for forming a service level A service level topology for N1 of the optical network of FIG. Figure 5 is a flow diagram performed by each access node in joining an optical network according to some embodiments of the invention. 5 is a flow diagram illustrating service level topology enhancement over a single service level according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by a node in response to a connectivity request message received via a link according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node to assign a path according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node in response to an update routing database message according to some embodiments of the invention. 4 is a flow diagram illustrating operations performed by an access node in response to an update assignment channel message according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by a source node of a path in response to the path allocation being released according to some embodiments of the present invention. 6 is a flow diagram illustrating operations performed by an access node in response to an update allocation release channel message according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by access nodes connected by links to which channels are added / deleted according to some embodiments of the present invention. 6 is a flow diagram illustrating operations performed by an access node in response to receiving an update add / delete channel message according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node connected by a delete link according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node in response to receiving a link delete message according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node connected by additional links according to some embodiments of the present invention. 6 is a flow diagram illustrating operations performed by an access node in response to receiving an add link message according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by one or more access nodes in the vicinity of a deletion node according to some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node in response to receiving a delete node message in accordance with some embodiments of the invention. 6 is a flow diagram illustrating operations performed by an access node in response to receiving a node addition message according to some embodiments of the invention.

Claims (219)

  An optical network comprising a plurality of wavelength division multiplexed access nodes that builds a network topology database based on a set of one or more connectivity constraints using a distributed search-based scheme.   The optical network of claim 1, wherein the set of connectivity constraints includes one or more of a quality of service (QoS) based criterion and a no-conversion constraint.   The optical network of claim 1, wherein each of the access nodes stores its own network topology database. further,
The optical network of claim 1, further comprising a centralized network management server communicatively coupled to each of the plurality of access nodes to store the network topology database.   The set of one or more connectivity constraints includes both QoS criteria and no-conversion constraints that divide the optical network into separate service levels, and the network topology database includes the service level The optical network of claim 1, storing a representation of each untranslated service level topology of the access node for each of the access nodes.   6. The optical network of claim 5, wherein the network topology database includes a separate topology structure for each of the non-conversion service level topologies for each access node.   In order to find possible end-to-end paths that meet the set of connectivity constraints, each of the optical network devices acting as access nodes starts with each of the links and is linked to the link. A device comprising a wavelength division multiplexed optical network comprising said optical network device for propagating connectivity request messages, wherein said end-to-end path can use a set of wavelengths to establish a light path The apparatus being a series of two or more of the optical network devices connected by a simple link.   The apparatus of claim 7, wherein the set of connectivity constraints includes one or more of QoS-based criteria and no-transform constraints.   The optical network device functioning as an access node builds its own topology database to store discovered possible end-to-end paths with that access node as a source node. 8. The apparatus according to 7. further,
8. The apparatus of claim 7, comprising a centralized network management server communicatively coupled to each of the access nodes that builds a topology database that stores the discovered possible end-to-end paths.   The set of one or more connectivity constraints includes both QoS-based criteria and no-conversion constraints that divide the optical network into different service levels, and different connectivity request messages are sent at the service level. The apparatus of claim 7 used for each. A wavelength division multiplexing optical network including a plurality of nodes, each of the plurality of nodes including a plurality of access nodes each generating and transmitting an adjacent node connectivity request message, wherein the access node is designated as a source node by the message. Determine the possible end-to-end path that you have as a node and meet one or more sets of connectivity constraints, where the set of wavelengths is used to establish a light path Comprising a wavelength division multiplexed optical network, a series of two or more nodes connected by an available link;
The plurality of nodes are responsive to respective received connectivity request messages to indicate the possible end-to-end paths that the connectivity request message has already collected, the outgoing access nodes to the source Determine the neighbor nodes that can be extended to form possible additional end-to-end paths that have as nodes and conform to the set of connectivity constraints and incorporate them into the topology database Send a message with the information of its determination to determine the possible additional end-to-end paths that have the originating access node as the source node and that meet the set of connectivity constraints An apparatus for propagating the connectivity request message to any adjacent node capable of   The apparatus of claim 12, wherein the set of connectivity constraints includes one or more of QoS-based criteria and no-transform constraints.   13. The apparatus of claim 12, wherein a plurality of messages carrying discrimination information are each sent back to the connectivity access message origination access node in response to the message being generated.   Each of the access nodes has received the message with discriminating information about possible additional end-to-end paths that have the access node as the source node and that meet the set of connectivity constraints 13. The apparatus of claim 12, in response to constructing a separate topology database.   13. The apparatus of claim 12, wherein each of the separate topology databases includes a no conversion service level topology for that access node. further,
13. The apparatus of claim 12, comprising a centralized network management server coupled to receive each of the messages with the determination and to communicate with each of the plurality of access nodes to build the topology database.   The set of one or more connectivity constraints includes both QoS-based criteria and no-conversion constraints that divide the optical network into different service levels, and different connectivity request messages are sent at the service level. 13. Apparatus according to claim 12, used for each.   Each of the propagated connectivity request messages is a computed set identifying a request ID, a source node ID, a forwarding node ID, and the possible end-to-end paths already gathered by the connectivity request message. 13. The apparatus of claim 12, wherein the set of wavelengths is usable on each.   Each of the messages carrying the discrimination information is either a connectivity response message type or a connectivity stop message type, and the message of the connectivity response message type is a possible end-to-end 13. The apparatus of claim 12, wherein the apparatus identifies a path and the message of the connectivity stop message type identifies that the path is complete. An apparatus comprising an access node coupled within a wavelength division multiplexed optical network, the access node comprising:
For each link connected to the access node, a link state database storing a link state structure storing the ports of the access node to which the link is connected and wavelengths available on the link;
A start-up module that generates and sends a neighbor node connectivity request message, the message having the access node as a source node and a possible end-end that conforms to a set of one or more connectivity constraints A start-up module that determines a two-end path, where the end-to-end path is a series of two or more nodes connected by a link whose set of wavelengths can be used to establish a light path When,
Responsive to each received connectivity request message, having the possible end-to-end path that the connectivity request message has already collected with the originating access node as the source node and connectivity Determine the adjacent nodes that can be extended to form a possible additional end-to-end path that conforms to the set of constraints and determine that for inclusion in the topology database. Send a message with information to determine the possible additional end-to-end paths that have the connectivity request message with the originating access node as the source node and that meet the set of connectivity constraints An access node comprising a connectivity request module that propagates to any adjacent node that can No apparatus.   The apparatus of claim 21, wherein the set of connectivity constraints includes one or more of QoS-based criteria and no-transform constraints.   The apparatus of claim 21, wherein each of the plurality of messages carrying discrimination information is sent back to the connectivity access message origination access node in response to the message being generated.   The startup module has received the message with the discriminating information of possible additional end-to-end paths that have the access node as the source node and conform to the set of connectivity constraints The apparatus according to claim 21, wherein the apparatus constructs a topology database in response to.   25. The apparatus of claim 24, wherein each of the topology databases includes an untranslated service level topology for its access node for each of a plurality of service levels.   The set of one or more connectivity constraints includes both QoS-based criteria and no-conversion constraints that divide the optical network into different service levels, and different connectivity request messages are sent at the service level. The apparatus of claim 21 used for each. The set of one or more connectivity constraints includes a QoS-based criterion that divides the optical network into a plurality of different service levels, and the access node is further received by the access node And a service level connectivity database storing a service level topology structure for each of the plurality of service levels into which data is entered in response to the plurality of messages carrying the determined discrimination information. 27. The apparatus according to 26.   Each of the propagated connectivity request messages is a computed set identifying a request ID, a source node ID, a forwarding node ID, and the possible end-to-end paths already gathered by the connectivity request message. The apparatus of claim 21, wherein the set of wavelengths is usable on each.   At least some of the messages carrying the discrimination information are connectivity response message types or connectivity stop message types, and the messages of the connectivity response message types are possible end-to-end The apparatus of claim 21, wherein the apparatus identifies a path and the message of the connectivity stop message type identifies that the path is complete. A node of the wavelength division multiplexing optical network receives a connectivity request message including a path and a set of path channels, and the path is connected to two or more nodes connected by a link on which the connectivity request message is transmitted. A set of path channels identifying the wavelengths available on that path to establish a light path;
Determining a set of intersections between the set of path channels and the set of link channels for the port of the node, wherein the set of link channels is a write on a link connected to the port Said determining is a set of one or more wavelengths that can be used to establish a path;
Sending a connectivity response message with an updated path to include the adjacent node connected to the port and the source set of common parts if the common set is not empty. further,
If the set of common parts is not empty,
Updating the connectivity request message so that the path includes the adjacent node connected to the port, and the set of path channels is the set of common parts;
31. The method of claim 30, comprising sending the updated connectivity request message from the port. further,
The method of claim 30, comprising sending a connectivity stop message back to the source node of the connectivity request message if the set of intersections is empty.   32. The method of claim 30, wherein the sending the connectivity response message includes sending the connectivity response message back to a source node of the connectivity request.   The method of claim 30, wherein each propagated connectivity request message further includes a request ID, a source node ID, and a forwarding node ID.   The method of claim 30, wherein the optical network is divided into a plurality of service levels, and different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels. .   36. The method of claim 35, wherein the set of path channels includes only wavelengths that are suitable for a first level of the plurality of service levels.   The method of claim 36, wherein the set of link channels includes only wavelengths suitable for the first level of the plurality of service levels. When executed by a processor, the processor
In response to receiving a connectivity request message including a path and a set of path channels at a node of the wavelength division multiplexing optical network, the set of path channels and the set of link channels for the port of the node; The path is a sequence of two or more nodes connected by a link carrying the connectivity request message, and the set of path channels is Identifying the wavelengths available on the path to establish a light path;
Performing an operation including sending a connectivity response message with a path updated to include the adjacent node connected to the port and a source the common set if the common set is not empty A machine-readable medium comprising instructions to be executed. The operation further comprises:
If the set of common parts is not empty,
Updating the connectivity request message so that the path includes the adjacent node connected to the port, and the set of path channels is the set of common parts;
39. The machine-readable medium of claim 38, comprising sending the updated connectivity request message from the port. The operation further comprises:
39. The machine-readable medium of claim 38, comprising sending a connectivity stop message back to the source node of the connectivity request message if the set of intersections is empty.   40. The machine readable medium of claim 38, wherein the sending the connectivity response message comprises sending the connectivity response message back to a source node of the connectivity request.   The machine-readable medium of claim 38, wherein each propagated connectivity request message further includes a request ID, a source node ID, and a forwarding node ID.   40. The machine of claim 38, wherein the optical network is divided into a plurality of service levels, and different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels. A readable medium.   44. The machine-readable medium of claim 43, wherein the set of path channels includes only wavelengths that are suitable for a first level of the plurality of service levels.   45. The machine readable medium of claim 44, wherein the set of link channels includes only wavelengths suitable for the first level of the plurality of service levels. End-to-end possible to each adjacent node of an access node of a wavelength division multiplexing optical network having the access node as a source node and conforming to a set of one or more connectivity constraints Sending a connectivity request message propagated to determine the path, where the end-to-end path is connected by a link where a set of wavelengths can be used to establish a light path 2 Said transmitting, which is a sequence of one or more nodes;
Receiving a response destination message identifying a set of path channels that respectively identify the collected possible end-to-end paths and usable wavelengths;
Building a topology database of the access node based on the response connectivity message.   The method of claim 46, wherein the set of connectivity constraints includes one or more of QoS-based criteria and no-transform constraints.   47. The method of claim 46, wherein the topology database includes a no-conversion service level topology for each of a plurality of service levels for its access node.   The set of one or more connectivity constraints includes both QoS-based criteria and no-conversion constraints that divide the optical network into different service levels, and different connectivity request messages are sent at the service level. 49. The method of claim 46 used for each. The set of one or more connectivity constraints includes a QoS-based criterion that divides the optical network into a plurality of separate service levels;
The method of claim 46, wherein the building comprises building a separate service level topology structure for each of the plurality of service levels.   Each of the connectivity request messages includes a request ID, a source node ID, a forwarding node ID, a computed set that identifies the possible end-to-end paths already collected by the connectivity request message; 47. The method of claim 46, wherein the set of wavelengths is usable on each.   At least some of the response connectivity messages are a connectivity response message type or a connectivity stop message type, and the message of the connectivity response message type identifies a possible end-to-end path 47. The method of claim 46, wherein the message of the connectivity stop message type identifies that the path is complete.   The method of claim 46, wherein the optical network is divided into a plurality of service levels, and different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels. . When executed by a processor, the processor
End-to-end possible to each adjacent node of an access node of a wavelength division multiplexing optical network having the access node as a source node and conforming to a set of one or more connectivity constraints Sending a connectivity request message propagated to determine the path, where the end-to-end path is connected by a link where a set of wavelengths can be used to establish a light path 2 Said transmitting, which is a sequence of one or more nodes;
Responsive to receiving a response connectivity message identifying a set of possible end-to-end paths and path channels that respectively identify the usable wavelengths, based on the response connectivity message A machine readable medium comprising instructions for performing operations including building a topology database for the access node.   55. The method of claim 54, wherein the set of connectivity constraints includes one or more of QoS based criteria and no transformation constraints.   55. The method of claim 54, wherein the topology database includes a no-conversion service level topology for each of a plurality of service levels for its access node.   The set of one or more connectivity constraints includes both QoS-based criteria and no-conversion constraints that divide the optical network into different service levels, and different connectivity request messages are sent at the service level. 55. The method of claim 54 used for each. The set of one or more connectivity constraints includes a QoS-based criterion that divides the optical network into a plurality of separate service levels;
55. The method of claim 54, wherein the building includes building a separate service level topology structure for each of the plurality of service levels.   Each of the connectivity request messages includes a request ID, a source node ID, a forwarding node ID, a computed set that identifies the possible end-to-end paths already collected by the connectivity request message; 55. The method of claim 54, wherein the set of wavelengths is usable on each.   At least some of the response connectivity messages are a connectivity response message type or a connectivity stop message type, and the message of the connectivity response message type identifies a possible end-to-end path 55. The method of claim 54, wherein the message of the connectivity stop message type identifies that the path is complete.   The method of claim 54, wherein the optical network is divided into a plurality of service levels, and different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels. . Apply a set of one or more connectivity constraints, including quality of service (QoS) based criteria, to the physical network topology of a wavelength division multiplexed optical network to bring the optical network to different service levels Splitting,
Determining a service level topology for each of said service levels.   64. The method of claim 62, wherein the QoS-based criterion includes one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority.   64. The method of claim 62, wherein the determining includes determining a service level topology of the network for each service level.   64. The method of claim 62, wherein the determining includes determining a service level topology for each node of the optical network for each service level.   The method of claim 62, wherein the set of connectivity constraints also includes a set of one or more transformation criteria.   64. The method of claim 62, wherein the set of connectivity constraints also includes no conversion connectivity constraints. Maintaining a classification according to a QoS standard of wavelengths for each link of a wavelength division multiplexed optical network, wherein the QoS standard is to determine a plurality of service levels;
Maintaining service level connectivity for each of the plurality of service levels based on a conversion criterion.   69. The method of claim 68, wherein the QoS based criterion comprises one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority. further,
69. The method of claim 68, comprising tracking the wavelength for each of the links by manipulating a link management protocol at each of the nodes of the optical network.   The maintaining the classification includes comparing each parameter of the wavelength to a service level parameter, one service level for each of the plurality of service levels for each of the QoS criteria. 69. The method of claim 68, wherein there are parameters.   72. The method of claim 71, wherein the maintaining the classification includes a respective node of the optical network that performs the comparing.   69. The method of claim 68, wherein the service level connectivity for each of the plurality of service levels includes the usable wavelength and a status as assigned or unassigned.   69. The method of claim 68, wherein the conversion criteria represents the number of wavelength conversions that are acceptable for a given optical circuit. A wavelength division multiplexed optical network supporting a plurality of service levels, wherein different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels A division multiplexing optical network;
At least one other network for each of the plurality of service levels that represents connectivity between a plurality of nodes of the optical network using some of the wavelengths suitable for the service level; A device comprising a topology database.   76. The apparatus of claim 75, wherein the connectivity is no conversion connectivity.   77. The apparatus of claim 76, wherein the network topology database is stored on a centralized network server.   77. The apparatus of claim 76, wherein each access node of the optical network stores an independent one of the network topology databases for each of the plurality of service levels. For each wavelength on each link of the wavelength division multiplexed optical network, a wavelength parameter for each of the set of QoS-based criteria;
Multiple services. For each of the levels, service level parameters for each of the set of QoS-based criteria,
The link for each of the plurality of service levels that also represents, for each link, several wavelengths of the wavelength on the link that have parameters that match the service level parameters of the service level for each link A set of service level channels;
For each access node of the optical network, using the wavelength from the set of link service level channels for that service level, the access node and other nodes of the access node And a service level topology structure for each of the plurality of service levels representing connectivity.   80. The apparatus of claim 79, wherein the QoS based criterion comprises one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority.   80. The apparatus of claim 79, wherein each access node of the optical network stores the set of link service level channels for some of the links connected to the access node. .   80. The apparatus of claim 79, wherein the service level topology structure is stored in a centralized network server.   80. The apparatus of claim 79, wherein each access node stores a number of service level topology structures of the service level topology structure that represent connectivity of the access node.   80. The apparatus of claim 79, wherein each of the service level topology structures stores a path where a common part of the set of link service level channels of the link of the path is not empty. An apparatus comprising an access node coupled in a wavelength division multiplexing optical network, wherein for each link to which the access node is connected to the access node, the access node to which the link is connected A link state structure storing ports, wavelengths available on the link, link state database storing parameters of those wavelengths, and
A service level parameter database that stores service level parameters for each of the set of QoS-based criteria for each of one or more supported sets of service levels;
An apparatus comprising: for each of said set of service levels, a service level topology structure storing a representation of said service level topology of that service level for said access node.   86. The apparatus of claim 85, wherein the QoS based criterion comprises one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority.   Each of the service level topology structures is the other access nodes of the optical network that are reachable at some of the wavelengths that are suitable for the service level of the service level topology structure. 86. The apparatus of claim 85, storing paths to a number of access nodes.   86. The apparatus of claim 85, wherein each of the service level topology structures stores available paths to other access nodes in the optical network.   90. The apparatus of claim 88, wherein each of the paths is a sequence of two or more nodes connected by a link that has several wavelengths at the service level of the path.   90. The apparatus of claim 88, wherein each of the paths is a set of one or more links and a wavelength that is available on all links of the set of links and at the service level of the path.   The access node further responds to a request to change the service level of a given provisioned service with a new communication path at a service level different from the previous communication path of the service level. 86. The apparatus of claim 85, further comprising a set of one or more modules that allocates, initiates routing traffic for the service on the new communication path, and releases allocation of the previous communication path. A method for an access node of a wavelength division multiplexed optical network comprising:
For each link to an adjacent node of the wavelength division multiplexed optical network, the access node wavelength on that link according to a set of one or more service level parameters for each of a plurality of service levels. Classifying
Instantiating a service level topology structure for each of the plurality of service levels;
In response to receiving information regarding connectivity to other access nodes in the optical network at each of the plurality of service levels, the service level topology structure for such service levels And adding to the method.   94. The method of claim 92, wherein the classifying is based on one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority. further,
94. The method of claim 92, comprising tracking the wavelength by manipulating a link management protocol for each link to an adjacent node.   94. The method of claim 92, wherein the classifying comprises comparing each parameter of the wavelength to the set of service level parameters.   The adding comprises, for each of the service level topology structures, an optical network reachable at some of the wavelengths suitable for the service level of the service level topology structure. 94. The method of claim 92, comprising storing paths to some of the other access nodes.   97. The method of claim 96, wherein each of the paths is a sequence of two or more nodes connected by a link that has several wavelengths at the service level of the path. When executed by a processor, the processor
Classifying wavelengths on that link according to a set of one or more service level parameters for each of a plurality of service levels for each link to adjacent nodes of a wavelength division multiplexed optical network;
Instantiating a service level topology structure for each of the plurality of service levels;
In response to receiving information regarding connectivity to other access nodes in the optical network at each of the plurality of service levels, the service level topology structure for such service levels A machine-readable medium comprising instructions for performing an operation comprising:   99. The machine-readable medium of claim 98, wherein the classifying is based on one or more of bandwidth, bit error rate, optical signal to noise ratio, peak noise level, rerouting priority. The operation further comprises:
99. The machine-readable medium of claim 98, comprising tracking the wavelength by manipulating a link management protocol for each link to an adjacent node.   99. The machine-readable medium of claim 98, wherein the classifying includes comparing each parameter of the wavelength to the set of service level parameters.   The adding comprises, for each of the service level topology structures, an optical network reachable at some of the wavelengths suitable for the service level of the service level topology structure. 99. The machine-readable medium of claim 98, comprising storing paths to some of the other access nodes.   103. The machine-readable medium of claim 102, wherein each path is a series of two or more nodes connected by a link that has several wavelengths at the service level of the path. Receiving a request for a communication path starting from a source node in a wavelength division multiplexed optical network;
A first wavelength of a plurality of service levels is selected, and different wavelengths on at least some links of the optical network that are suitable for different service levels of the plurality of service levels are each of the optical network Forming different service level topologies for each of the plurality of service levels for a plurality of access nodes;
For each of the multiple service levels, select a path and wavelength on the path using a database that stores representations of available paths from the source node to other access nodes in the optical network Said selecting, wherein each path is a sequence of two or more nodes connected by a link having one or more sets of wavelengths of the same service level;
Causing the allocation of the selected wavelength within the sequence of nodes of the selected path to be performed.   105. The method of claim 104, wherein the communication path is a light path.   105. The method of claim 104, wherein the communication path is an optical circuit.   105. The method of claim 104, wherein the selection and assignment of the path is performed in real time.   105. The method of claim 104, wherein the database stores a separate service level topology structure for each of the service level topologies of the source node.   109. The method of claim 108, wherein the database includes the available wavelengths and either assigned or unassigned status.   105. The method of claim 104, wherein the database stores a representation of usable untranslated paths from the source node to other access nodes in the optical network for each of the plurality of service levels. When executed by a processor, the processor
In response to receiving a request for a communication path starting from a source node in a wavelength division multiplexing optical network, selecting a first level of a plurality of service levels, the plurality of services Different wavelengths on at least some links of the optical network that are suitable for several different service levels of the level are different for each of the plurality of service levels for each access node of the optical network Forming the service level topology, said selecting;
For each of the multiple service levels, select a path and wavelength on the path using a database that stores representations of available paths from the source node to other access nodes in the optical network Said selecting, wherein each path is a sequence of two or more nodes connected by a link having one or more sets of wavelengths of the same service level;
A machine readable medium comprising instructions for performing operations including performing allocation of the selected wavelength within the sequence of nodes of the selected path.   The machine-readable medium of claim 111, wherein the communication path is a light path.   The machine-readable medium of claim 111, wherein the communication path is an optical circuit.   112. The machine readable medium of claim 111, wherein the selection and assignment of the path is performed in real time.   112. The machine readable medium of claim 111, wherein the database stores a separate service level topology structure for each of the service level topologies of the source node.   116. The machine readable medium of claim 115, wherein the database includes the available wavelengths and a status of either assigned or unassigned.   112. The machine-readable medium of claim 111, wherein the database stores a representation of usable untranslated paths from the source node to other access nodes in the optical network for each of the plurality of service levels. . Receiving a request to change a provisioned service to a different one of the plurality of service levels over a communication path established in a wavelength division multiplexing optical network at one of the plurality of service levels Different wavelengths on at least some links of the optical network that are suitable for different levels of the plurality of service levels are said to be different for each access node of the optical network. Said receiving to form a different service level topology for each of a plurality of service levels;
For each of the plurality of service levels, a path and a wavelength on the path are selected using a database that stores representations of available paths from a source node to other access nodes in the optical network. Said each path being a series of two or more nodes connected by a link having one or more sets of wavelengths of the same service level;
Causing the selected wavelength allocation within the sequence of nodes of the selected path to be performed to form a new communication path;
Transitioning the service to the new communication path.   119. The method of claim 118, wherein the communication path is a light path.   119. The method of claim 118, wherein the communication path is an optical circuit.   119. The method of claim 118, wherein the selection and assignment of the path is performed in real time.   119. The method of claim 118, wherein the database stores a separate service level topology structure for each of the service level topologies of the source node.   119. The database stores a representation of usable untranslated paths from the source node of the communication path to other access nodes in the optical network for each of the plurality of service levels. the method of. The transition is
Migrating traffic from the previous communication path to the new communication path;
119. The method of claim 118, comprising deallocating the previous communication path. When executed by a processor, the processor
A request to change a provisioned service of a communication path established in a wavelength division multiplexing optical network at one of a plurality of service levels to a different one of the plurality of service levels. In response to receiving, for each of the plurality of service levels, use a database that stores representations of available paths from the source node of the communication path to other access nodes in the optical network. Different paths on at least some links of the optical network that are suitable for different service levels of the plurality of service levels. The wavelength is the plurality of services for each access node of the optical network. Form different service level topologies for each of the levels, each path being a sequence of two or more nodes connected by links with one or more sets of wavelengths of the same service level Said selecting, and
Causing the selected wavelength allocation within the sequence of nodes of the selected path to be performed to form a new communication path;
A machine readable medium comprising instructions for performing an operation including migrating the service to the new communication path.   126. The machine-readable medium of claim 125, wherein the communication path is a light path.   126. The machine-readable medium of claim 125, wherein the communication path is an optical circuit.   126. The machine-readable medium of claim 125, wherein the selection and assignment of the path is performed in real time.   126. The machine-readable medium of claim 125, wherein the database stores a separate service level topology structure for each of the service level topologies of the source node.   126. The database stores a representation of usable untranslated paths from the source node of the communication path to other access nodes in the optical network for each of the plurality of service levels. Machine-readable media. The transition is
Migrating traffic from the previous communication path to the new communication path;
126. The machine-readable medium of claim 125, comprising deallocating the previous communication path. A machine-readable medium on which data is stored,
A wavelength-division multiplexed optical network access node service level connectivity database, wherein each link of the optical network includes a set of zero or more lambdas for each of a plurality of service levels, Each of the service levels includes a set of zero or more possible end-to-end paths consisting of a series of one or more links that include one or more lambdas of that service level. Comprising the service level connectivity database, the service level connectivity database comprising:
For each of the possible end-to-end paths ending with the access node:
Data to represent,
The sequence of links of the path;
A machine-readable medium comprising a service level connectivity database including the lambdas in the path. further,
A link state database including a link state structure for each node of the optical network adjacent to the access node, the link state structure each including a set of zero lambdas for each of the plurality of service levels. 132. The machine-readable medium according to 132.   The service level connectivity database comprises a separate service level topology structure for each of the plurality of service levels, each of the service level topology structures ending with the access node. 135. The machine-readable medium of claim 132, storing data for each of the possible end-to-end paths.   135. The machine-readable medium of claim 134, wherein each of the service level topology structures is a table.   135. The machine-readable medium of claim 134, wherein each service level topology structure is a tree.   A plurality of wavelength division multiplexed access nodes in an optical network that employs a source-based scheme to establish a communication path, each based on one or more sets of connectivity constraints A device that includes an access node that stores a set of topology databases.   140. The apparatus of claim 137, wherein the communication path includes a light path.   138. The apparatus of claim 137, wherein the communication path includes one or more of an optical circuit, a light path, and an end-to-end unidirectional path.   138. The apparatus of claim 137, wherein the set of one or more network topology databases in each of the plurality of access nodes stores an untransformed topology for that access node.   138. The apparatus of claim 137, wherein the plurality of access nodes establish communication paths in real time.   The set of one or more connectivity constraints includes a quality of service (QoS) -based criterion that divides the optical network into different service levels, each of the plurality of access nodes. 138. The apparatus of claim 137, wherein the set of one or more network topology databases stores a service level topology without translation of that access node for each of the service levels.   143. The set of network topology databases in each of the plurality of access nodes includes a separate network topology database for each of the untranslated service level topologies for that access node. The device described.   138. The apparatus of claim 137, wherein the set of network topology databases in each of the plurality of access nodes is constructed and maintained by the access node. further,
A centralized network management server communicatively coupled to each of the plurality of access nodes to build and maintain the set of network topology databases within each of the plurality of access nodes. 137. The apparatus according to 137. A wavelength division multiplexed optical network including a plurality of access nodes, each of the nodes comprising:
For each link connected to the access node, a set of link channels representing at least some wavelengths on that link that can be used to establish a light path, where the light path is a wavelength The path of a given light path is a sequence of two or more nodes and links interconnecting them used to carry traffic by the wavelength of the light path, The set of link channels each starting with a source node and ending with a destination node;
A database representing unconverted connectivity from the access node to other access nodes of the access node using the wavelengths in the set of link channels, wherein the unconverted connectivity is the access An apparatus comprising the database containing the paths and wavelengths of possible light paths having a node as the source node and another node in the access node as the destination node.   Each of the plurality of access nodes further has an assignment module that selects and assigns a light path having the access node as the source node in real time in response to a request for the communication path received by the access node. 146. The apparatus of claim 146.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is the service level 148. The apparatus of claim 146, wherein the apparatus represents unconverted connectivity using a wavelength that is suitable for only one of the levels of service.   Each of the plurality of access nodes uses the wavelength suitable for that service level for each of the other service levels of the plurality of service levels from the access node to the access node. 149. The apparatus of claim 148, further comprising another database representing unconverted connectivity to other nodes.   147. The apparatus of claim 146, wherein the database in each of the plurality of access nodes is constructed and maintained by the access node. further,
147. The apparatus of claim 146, comprising a centralized network management server communicatively coupled to each of the plurality of access nodes to build and maintain the database in each of the plurality of access nodes. A plurality of access nodes of the wavelength division multiplexing optical network each tracking a wavelength for each link of the wavelength division multiplexing optical network connected to the access node;
Each of the plurality of access nodes maintains a topology based on unconverted connectivity of the plurality of access nodes to other nodes;
In response to a request for a communication path received by any one of the plurality of access nodes, the access node
Selecting both a path through the set of one or more links of the optical network and a single wavelength available in all of the set of links based on the topology held in the access node;
Assigning the selected path and wavelength.   153. The method of claim 152, wherein the communication path is a light path.   153. The method of claim 152, wherein the communication path is an optical circuit.   153. The method of claim 152, wherein the selecting and assigning are performed in real time.   The topology maintained by each of the plurality of access nodes is further based on one level of connectivity among a plurality of service levels, where different wavelengths on at least some links of the optical network are: 153. The method of claim 152, suitable for several different service levels of the plurality of service levels.   153. The method of claim 152, wherein the tracking includes manipulating a link management protocol for each of the plurality of access nodes.   153. The method of claim 152, wherein the maintaining includes each of the plurality of access nodes communicating with other nodes of the plurality of access nodes.   153. The method of claim 152, wherein the maintaining comprises each of the plurality of access nodes communicating with a centralized network management server.   153. The method of claim 152, wherein the topology for each of the plurality of access nodes includes the usable wavelength and a status as assigned or unassigned. An apparatus comprising an access node coupled within a wavelength division multiplexed optical network, the access node comprising:
For each link connected to the access node, a link state structure storing a port of the access node to which the link is connected and a link state database storing wavelengths available on the link;
A database storing a representation of available paths from the access node to other nodes of the access node using the wavelengths in the link state database, wherein the path is one or more wavelengths The database, wherein the common set is a sequence of two or more nodes connected by links available to establish one or more light paths;
In response to a request for a communication path received by the access node, an unassigned path of the available paths, a path selected from the common set of wavelengths and a module for selecting a wavelength; Equipment provided.   163. The apparatus of claim 161, wherein the module performs the selection and causes the selected path and wavelength assignment to be performed in real time.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is the plurality of service levels. 163. The apparatus of claim 161, storing an untranslated service level topology structure for each of the service levels.   163. The apparatus of claim 161, wherein the access node comprises an additional module that builds and maintains the database within the access node. further,
164. The apparatus of claim 161, comprising a centralized network management server communicatively coupled to the access node to build and maintain the database.   163. The apparatus of claim 161, wherein the access node comprises a link management protocol that populates a database with data in the link state. Receiving a demand criterion representing a request for a communication path at an access node of a wavelength division multiplexing optical network;
The path and wavelength on the path using a database stored in the access node and storing a representation of the available path from the access node to other nodes of the access node in the optical network Each path consists of two or more nodes connected by a link where a common set of one or more wavelengths can be used to establish one or more light paths Said selecting, which is a series;
The access node communicating with an access node on the selected path of the access nodes to assign the selected wavelength on the selected path.   166. The method of claim 167, wherein the communication path is a light path.   166. The method of claim 167, wherein the communication path is an optical circuit.   166. The method of claim 167, wherein the selection and the assignment are performed in real time.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is the plurality of service levels. 166. The method of claim 167, storing a non-translated service level topology structure for each of the service levels.   166. The method of claim 167, wherein the database includes the available wavelengths and either assigned or unassigned status. When executed by a processor, the processor
In response to receiving a demand criterion representing a request for a communication path at an access node of a wavelength division multiplexing optical network, the access node is stored in the access node and from the access node to the access node in the optical network. Selecting a path and a wavelength on the path using a database that stores representations of available paths to the other nodes of the node, each path having a common one or more wavelengths Said selecting, wherein the set is a sequence of two or more nodes connected by links available to establish one or more light paths;
Communicating the access node with an access node on the selected path of the access nodes to assign the selected wavelength on the selected path. A machine-readable medium comprising instructions to be executed.   178. The machine-readable medium of claim 173, wherein the communication path is a light path.   178. The machine-readable medium of claim 173, wherein the communication path is an optical circuit.   178. The machine-readable medium of claim 173, wherein the selection and the assignment are performed in real time.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is the plurality of service levels. 178. The machine-readable medium of claim 173, storing a non-translated service level topology structure for each of the service levels.   178. The machine readable medium of claim 173, wherein the database includes the available wavelengths and either assigned or unassigned status.   A wavelength division multiplexed optical network comprising a plurality of nodes, each comprising an optical cross-connect, and each storing a database representing unconverted connectivity from the node to other nodes of the plurality of nodes; Each of the nodes uses a messaging scheme that propagates notifications of changes in the optical network to other nodes of the node to maintain the database, the messaging scheme at each of the nodes being at least in part An apparatus that transmits a message only to a selected node of the plurality of nodes based on the non-conversion connectivity, and minimizes the number of communication between the nodes.   179. The apparatus of claim 179, wherein the changes include assigning and deallocating paths and their wavelengths, adding and deleting wavelengths on a link, adding and deleting links, and adding and deleting nodes.   179. For at least a given one of the nodes for at least one type of change, the selected node of the plurality of nodes includes a node that has no conversion connectivity with the given node. Equipment.   At least some of the messages contain a message indicating a destination node to which the message has already been transmitted for the given one of the nodes, and such messages of the nodes 179. The apparatus of claim 179, wherein a node that receives a message propagates only to the plurality of nodes that have no conversion connectivity and are not identified by the received message.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is respectively 180. The apparatus of claim 179, comprising a separate structure representing untranslated connectivity for each of the plurality of service levels for each of the service levels.   180. The apparatus of claim 179, wherein the optical network uses a source-based scheme to establish a communication path according to the non-conversion connectivity.   185. The apparatus of claim 184, wherein the communication path includes one or more of an optical circuit, a light path, an end-to-end unidirectional path.   180. The apparatus of claim 179, wherein at least some of the messages include a destination set that identifies other nodes of destinations for which notification of change has already been sent. An apparatus comprising an access node coupled within a wavelength division multiplexing optical network, wherein the access node is an optical cross-connect,
A database storing a non-transformed topology for the access node;
A set of one or more modules involved in a messaging scheme that propagates notification of changes in the optical network to other access nodes in the optical network and maintains the database of the access nodes; The messaging scheme comprises: a set of modules including sending a message to the set of access nodes selected from only the nodes of the no-conversion topology of the access nodes in the optical network. The access node further includes:
For each link connected to the access node, a link state database that stores a link state structure that stores an indication of the port of the access node to which the link is connected and a wavelength that can be used on the link. The path is a series of two or more nodes connected by a link where a common set of one or more wavelengths can be used to establish one or more light paths 188. The apparatus of claim 187, wherein the access node of the no-conversion topology is an access node in the optical network that is reachable using a path with the wavelength in the link state database.   188. The apparatus of claim 187, wherein the change includes assigning and deallocating a path and its wavelength, adding and removing wavelengths on a link, adding and removing links, and adding and removing nodes.   187. The apparatus of claim 187, wherein the set of one or more modules sends a message to all of the access nodes of the no conversion topology in response to detecting a change in the network.   The set of one or more modules updates the database as needed in response to messages from other nodes of the node in the optical network indicating a change in the network and already sends the message Select any of the access nodes of the no-conversion topology that is not identified in the message as being updated, update the message to include the new set, and update the message to the new set 188. The apparatus of claim 187, which propagates to the access node within.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is 189. The apparatus of claim 187, comprising a separate structure that represents the untransformed topology for each of the service levels.   188. The apparatus of claim 187, wherein at least some of the messages include a destination set that identifies other nodes of destinations for which notification of change has already been sent.   The set of one or more modules determines, for at least some types of the change, whether any of the available paths of the no-conversion topology contain affected network resources; 188. The apparatus of claim 187, wherein the selected set of nodes does not include any nodes if none of the available paths of the untransformed topology includes affected network resources.   The set of one or more modules sends a connectivity request message for at least some types of the changes, to which a connectivity response message is used by the receiving access node to update the database. 188. The apparatus of claim 187, wherein: In response to changes in a wavelength division multiplexed optical network that includes multiple access nodes,
The plurality of accesses in the optical network based on a database in which at least a first node of the plurality of access nodes represents unconverted connectivity from the first node to other nodes of the plurality of access nodes Sending a message about the change only to a selected set of nodes;
In response to receiving the message, the access nodes in the selected set of access nodes each have an updated version of the message with no conversion connectivity among the plurality of access nodes. Transmitting to an access node not already identified in the received message. Sending the message comprises:
196. The method of claim 196, comprising selecting an access node with no conversion connectivity among the plurality of access nodes as the selected set. Sending the message comprises:
196. The method of claim 196, wherein the selected set comprises selecting an access node that has unconverted connectivity and is not connected to the network resource among the plurality of access nodes. further,
196. The method of claim 196, wherein the first node includes updating the database. further,
The first node sending a connectivity request message to one or more sets of neighboring nodes of the plurality of access nodes to discover the non-conversion connectivity change;
Some of the plurality of access nodes receive a connectivity request message;
Sending a connectivity response message to the first access node;
Propagating the connectivity request message to neighboring nodes that are likely to be able to determine non-additional connectivity having the first access node as the source node of the plurality of access nodes When,
196. The method of claim 196, comprising updating the database in response to the first access node receiving the connectivity response message. In response to at least a second access node of the plurality of access nodes receiving the message or the updated version of the message, a change in connectivity of the second access node without conversion To send a connectivity request message to a set of one or more neighboring nodes of the plurality of access nodes;
Some of the plurality of access nodes receive a connectivity request message;
A connection without additional conversion having a connectivity response message transmitted to the second access node and the connectivity request message having the second access node as the source node among the plurality of access nodes. Perform propagation to adjacent nodes that may be able to determine gender,
196. The method of claim 196, wherein the second access node updates a database in response to receiving the connectivity request message.   The change is one of assignment of path and its wavelength, release of assignment of path and its wavelength, addition of wavelength to link, deletion of wavelength on link, addition of link, deletion of link, addition of node, deletion of node 196. The device of claim 196, wherein:   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is 196. The apparatus of claim 196, comprising a separate structure that represents the unconverted connectivity for each service level.   196. The apparatus of claim 196, wherein the optical network uses a source-based scheme to establish a communication path according to the non-conversion connectivity.   205. The apparatus of claim 204, wherein the communication path includes one or more of an optical circuit, a light path, and an end-to-end unidirectional path. In response to indicating a change in network resources and identifying another of the plurality of access nodes that have already received the message, the plurality of access nodes in the wavelength division multiplexed optical network Our first access node is
Updating a database storing a no-translation topology for the first access node;
Selecting any of the access nodes in the no-translation topology not identified by the message as a set of access nodes;
Transmitting the updated version of the message to the set of access nodes. The updating is
The first node sends a connectivity request message to one or more sets of neighboring nodes of the plurality of access nodes to discover an update to the no-conversion connectivity;
206. The method of claim 206, comprising: the first access node receiving a connectivity response message. The updating is
206. The method of claim 207, comprising instantiating a new no conversion connectivity database based on the connectivity response.   The change is one of assignment of path and its wavelength, release of assignment of path and its wavelength, addition of wavelength to link, deletion of wavelength on link, addition of link, deletion of link, addition of node, deletion of node 206. The method of claim 206, wherein:   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is 207. The method of claim 206, comprising a separate structure representing the untransformed topology for each service level.   207. The method of claim 206, wherein the optical network uses a source-based scheme to establish a communication path according to the non-conversion connectivity.   223. The method of claim 211, wherein the communication path includes one or more of an optical circuit, a light path, an end-to-end unidirectional path. When executed by a machine, the machine
In response to indicating a change in network resources and identifying another of the plurality of access nodes that have already received the message, the plurality of access nodes in the wavelength division multiplexed optical network Our first access node is
Updating a database storing a no-translation topology for the first access node;
Selecting any of the access nodes in the no-translation topology not identified by the message as a set of access nodes;
A machine-readable medium comprising instructions for performing an operation comprising sending an updated version of the message to the set of access nodes. The updating is
The first node sends a connectivity request message to one or more sets of neighboring nodes of the plurality of access nodes to discover an update to the no-conversion connectivity;
213. The machine readable medium of claim 213, wherein the first access node receives a connectivity response message. The updating is
224. The machine-readable medium of claim 214, comprising instantiating a new no-conversion connectivity database based on the connectivity response.   The change is one of assignment of path and its wavelength, release of assignment of path and its wavelength, addition of wavelength to link, deletion of wavelength on link, addition of link, deletion of link, addition of node, deletion of node 213. The machine-readable medium of claim 213.   The optical network is divided into a plurality of service levels, different wavelengths on at least some links of the optical network are suitable for different levels of the plurality of service levels, and the database is 213. The machine readable medium of claim 213, comprising separate structures that represent the untransformed topology for each of the service levels.   213. The machine readable medium of claim 213, wherein the optical network uses a source-based scheme to establish a communication path according to the non-conversion connectivity.   219. The machine-readable medium of claim 218, wherein the communication path includes one or more of an optical circuit, a light path, and an end-to-end unidirectional path.

Images