Patent Application
20050074029
This is the first application filed for the present invention.
Not Applicable.
The present invention relates to call control in a transport network, and in particular to call control using a layered call model.
The modern transport network space is characterised by a large number of heterogenous networks which are interconnected by gateway servers. Each network may implement any one or more of a variety of transport protocols (e.g. Ethernet, ATM, TCP/IP; SONET/SDH). Where interconnected networks employ different transport protocols, adaptation services provided by the gateway servers facilitate transport of client data between the involved networks. In response to the demand for seamless communications across this “network of networks”, much effort has been made to develop generic connection management schemes that can be utilized by network service providers independently of the transport technology used. This work has resulted in the International Telecommunications Union (ITU) standard Recommendation G.805, which defines a layered functional architecture for transport networks. Based on the G.805 layer architecture two models have been developed: G.8080 for the control plane (Automatically Switched Optical Network—ASON); and Tele-Management Forum (TMF) model TMF608 for the management plane. TMF814 is an interface specification of TMF that instantiates the TMF608 model. G.8080 defines two types of call controllers: calling/called party call controllers (CCCs); and network call controllers (NCCs). Calling/called party call controllers initiate and terminate calls, whereas network call controllers interact with the CCCs and with connection controllers to manage connections within the network. Within the management plane, TMF814 provides an interface that uses the G.805 model of connections. The relationship between connections at different layers is known in the TMF814 interface. Thus it can be used to manage multi-layer connections across the network.
G.8080 uses the existing definition of a logical separation between calls and connections. A “call” may be understood as an “agreement” or an “intent” to communicate. Typically, the call will be represented by a call object which contains metadata concerning the call. Such metadata may, for example, include billing information; characteristic information (CI) of the call (e.g., transport protocol, bit rate, etc.); security parameters; and quality of service (QoS) requirements. While a call represents an agreement to communicate, a “connection” provides the actual means for communication. Thus, for example, instantiation of a call object indicates a client's intent to communicate, and provides the information required by network service providers to facilitate and manage the communications. On the basis of the call object, the network service provider(s) can then provide (usually through signalling protocols) the physical network resources (that is, the connections) required to transport the client traffic.
In the example of
However, a limitation of G.8080 (that is, the Automatically Switched Optical Network—ASON) is that it is defined for a single-layered control plane. Thus, in the example of
In particular, when an NCC determines that there are no (or insufficient) resources in its own layer, it responds by merely indicating that the connection attempt has failed. It cannot automatically determine whether or not adequate resources exist in another layer. Even if such server-layer resources are present, the NCC has no means of instantiating the required connection or the necessary adaptation services. This situation is usually remedied by manual configuration of server layer trails that are then used by client connections. It is an operational procedure to the network operator. Thus, following the above example, the VC-3 call segment 24 and connection 26 must be manually provisioned by the network service provider, and presented to each NCCMAC 12 and 20 as a MAC trail. Once this has been done, the MAC call segment 18 can direct a connection controller to instantiate a MAC connection for it. This requirement for manual provisioning of server-layer connections effectively precludes automated instantiation of connections across heterogenous networks. Furthermore, network service providers in each network require a very high level of understanding of subscriber traffic requirements in order to properly provision network resources. Furthermore, a network failure in a server layer cannot be readily correlated with client-layer calls, because the server-layer call has no awareness of the client-layer calls that it is actually supporting.
An efficient technique for overcoming these limitations is highly desirable.
Accordingly, an object of the present invention is to provide a layered call model that enables a call segment in a client layer to be aware of associated calls/connections in a server layer.
An aspect of the present invention provides a method of controlling a multi-layer transport network. In accordance with the invention a determination is made as to whether a connection supporting a performance requirement of a call can be established within a first layer of the network. If the connection cannot be established, an association is defined between the call and a second call instantiated within a respective server layer of the network.
A further aspect of the present invention provides a method of establishing a call across a multi-layer transport network. A first call controller is instantiated within a respective first layer of the network. A determination is made as to whether a respective connection for the call can be supported by the first layer. If the connection can be supported by the first layer, the call is set up through the first layer of the network. Otherwise, a second call controller is instantiated in a respective server layer to the first layer, and an association defined between the first and second call controllers.
Further features and advantages of the present invention will become apparent from the following detailed description, taken in combination with the appended drawings, in which:
a and 1b schematically illustrate equipment and control plane views, respectively, of a connection mapped across a multi-layer transport network, using the conventional G.8080 architecture;
a and 2b schematically illustrate equipment and control plane views, respectively, of connection mapped across a multi-layer transport network, in accordance with the present invention; and
It will be noted that throughout the appended drawings, like features are identified by like reference numerals.
The present invention provides a generic layered call model that enables a call segment in a one layer to be associated with calls in other layers. In one view, the present invention extends the ITU standard G.8080 to enable multi-layer call associations that mirror the TMF814 connection model. Another view of the present invention is that it extends TMF608 and TMF814 to support layer relationships between calls.
In order to simplify the following description, the present invention will be described below with reference to an example in which a singal call in a client layer is supported by a single server layer call. However, those of ordinary skill in the art will appreciate that the present invention also encompasses embodiments in which a client layer call is supported by two or more server-layer calls, some of which may, or may not, be themselves supported by calls on still lower layers of the network. Similarly, a server layer call may support two or more client layer calls, some of which may, or may not, be themselves supporting client layer calls on still higher layers of the network.
In accordance with the present invention, upon instantiation of a call (or call segment), the involved G.8080 call controller (either an NCC or a CCC) determines whether or not the characteristic information (CI) of the call can be supported by a connection at that layer. If the CI cannot be supported at that layer, the call controller formulates and sends a request to a server layer call controller, so as to thereby trigger instantiation of one or more calls within the server layer. This process can be repeated, recursively, across multiple network layers in a manner directly analogous to the TMF608 and TMF814 connection model.
Implementation of the present invention involves extending the ITU standard G.8080 call controllers to control both connection instantiation in the client layer and call initiation in a server layer, and implementing the corresponding functionality in the TMF814 object that holds the call. A G.8080 call controller is typically designed interact with a connection controller (CC) object in its own layer to instantiate a connection. In the present invention, an enhanced call controller is defined, in which the conventional functionality is extended such that, if the connection cannot be instantiated within that layer, the enhanced call controller can formulate and send a “call request” message to an enhanced server layer call controller. The call request message includes suitable parameters to enable the server layer call controller to meet the service level requirements defined in the CI of the (client layer) call. These parameters do not need to be identical to the client layer parameters, as the server layer may choose to satisfy the call request with a more stringent service. In addition to instantiating a call, the enhanced server-layer call controller can also repeat the process of determining whether sufficient resources exist at that layer, and, if necessary, request one or more calls at a still lower layer of the network.
In order to support the extended functionality of enhanced call controllers, the connection controller (CC) remains unchanged in that the CC object will attempt to instantiate a connection only in response to a connection request from a call controller at that layer. In the event that the performance requirements of the call cannot be met within that layer, the CC will return a “connection failed” and/or “insufficient resource” indication to the call controller. With this arrangement, the enhanced call controller can be designed to formulate and send the call request message in response to the “Connection Failed” indication from the CC object.
Because the instantiation of server-layer calls (and connections) is triggered by a call request from a client layer enhanced call controller, an association between the client and server-layer call controllers is created. This association, which serves as a substitute for the “missing” client layer connection, means that the client and server layer call controllers are aware of each other. They can therefore cooperate to automatically use appropriate adaptation services, as required. They can also pass status and call control information between layers. Thus, for example, connection failures in a server layer can readily be co-related with client layer call failures.
At the ingress NE 14, the ingress e-NCCMAC 44 determines that the Characteristic Information (CI) of the call cannot be supported within the MAC layer (since there is no MAC layer in the network at all). The ingress e-NCCMAC can also access a directory service (not shown) in a conventional manner to determine the SNPP for the destination address in the call request. In this case, the SNPP is a network element identifier where the destination e-NCCMAC 46 resides. The ingress e-NCCMAC then formulates an NCCMAC call request 56 (see
In response to the ingress NCCMAC call request 56, the ingress e-NCCVC-3 48 determines that the performance requirements of the MAC flow can be satisfied within the VC-3 layer using a Generic Framing Protocol (GFP) adaptation. Thus, the ingress e-NCCVC-3 48 formulates an NCCVC-3 call request 58, which is sent to the egress e-NCCVC3 50 in order to instantiate a server-layer call 60 at the VC-3 layer between the ingress e-NCCVC-3 48 and the egress e-NCCVC-3 50 resident in the destination Network element (NE-2) 22. The NCCVC-3 call request 58 encapsulates the NCCMAC call request 56, which is opaque in the sense that it is conveyed through the VC-3 control plane without being extracted or examined at intermediate nodes. Transport of the NCCVC-3 call request 58 can be accomplished using the conventional connection controller messaging protocol in the control plane. The conventional GFP adaptation to VC-3 payload is implemented between the MAC and VC-3 connection controllers within NE-1, in a conventional manner.
When the egress e-NCCVC-3 50 receives the NCCVC-3 call request 58, it processes the VC-3 call in a conventional manner, and then extracts (at 62) the NCCMAC call request 56. This call request is passed to the MAC layer control plane which instantiates a call in the enhanced egress NCCMAC 46, and then passes the NCCMAC call request to it (at 64). In response to the NCCMAC call request, the egress e-NCCMAC 46 forwards the call request (at 66) to the destination e-CCC 42 in order to set up the MAC-layer call 68 between the egress e-NCCMAC 46 and the destination e-CCC 42. Additionally, a conventional GFP adaptation to VC-3 payload is implemented between the MAC and VC-3 connection controllers within NE-2, in a conventional manner.
The destination e-CCC 42 processes the call request, and sends a MAC call confirmation 70 back to the egress e-NCCMAC 46. When the egress e-NCCMAC receives the MAC call confirmation, the egress e-NCCMAC processes the call confirmation (at 72) to set up the MAC connection 74, and then passes the call confirmation to the egress e-NCCVC-3 50 (at 76). The egress e-NCCVC-3 50 places the MAC call confirmation into its own VC-3 call confirmation, which is sent (at 78) to the ingress e-NCCVC-3 48. The egress e-NCCVC-3 50 also processes the MAC call confirmation (at 80) to set up a GFP adaptation to VC-3 payload 32 between the MAC and VC-3 layer connection controllers. Upon receipt of the VC-3 call confirmation, the ingress e-NCCVC-3 48 processes the call confirmation (at 82) to set up the VC-3 connection 84 between the ingress and egress network elements 14, 22. The ingress e-NCCVC-3 then extracts (at 86) the MAC call confirmation, and passes it back to the ingress e-NCCMAC 44 (at 88). A GFP adaptation to VC-3 payload 33 is also set up (at 90) between the MAC and VC-3 layer connection controllers in the ingress network element 14. The MAC call confirmation can then be forwarded (at 92) to the originating e-CCC 40 by the ingress e-NCCMAC 44, and the accompanying MAC connection 94 set up (at 96).
In accordance with the second view of the present invention, a management model contains objects that correspond to calls in a multi-layer network. Where a call in a client layer and a call in a server layer have an association in G.8080 control planes, the corresponding management objects for the calls also have an association. This enables call dependencies to be managed by network management applications. An interface such as TMF814 can be extended to reflect these call object relationships.
In this model, the association of calls also includes the case where multiple client calls are associated with a common call in a server layer. For example, two Ethernet calls from separate ingress links could be supported by a single VC-4 connection using Ethernet multiplexing over GFP
The embodiment(s) of the invention described above is(are) intended to be exemplary only. The scope of the invention is therefore intended to be limited solely by the scope of the appended claims.
