WO2002019616A2

WO2002019616A2 - Cost/performance resource handler for an aggregation network

Info

Publication number: WO2002019616A2
Application number: PCT/CA2001/001201
Authority: WO
Inventors: Yves Lemieux
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2000-08-30
Filing date: 2001-08-30
Publication date: 2002-03-07
Anticipated expiration: 2003-02-28
Also published as: WO2002019616A3; AU2001293531A1

Abstract

An aggregation network is provided that includes a plurality of interconnected access routers (150). This network is configured in accordance with a partially meshed (111) physical network topology. An individual subscriber makes a request for provision of a certain bursty traffic service. Responsive thereto, a resource handler functionality (140) determines which one of a plurality of logical network topologies (ring (160), duplicated tree (162) or partially meshed (164)) is best suited to support the provision of communications for that requested bursty traffic service. Factors that are evaluated in making this determination include quality of service (QoS) entitlement, communication network resource availability, and cost constraints. The plurality of interconnected access routers are then configured by the resource handler functionality to implement the determined logical network topology on the certain physical network topology and admit the flow of bursty traffic related to the subscriber request for handling by the logical network topology.

Description

COST/PERFORMANCE RESOURCE HANDLER FOR AN AGGREGATION NETWORK

BACKGROUND OF THE INVENTION Technical Field of the Invention

The present invention relates to an aggregation network and, in particular, to an operation for managing a logical network topology configuration of the aggregation network in accordance with user cost/performance considerations.

Description of Related Art

Wireless access for communications is advantageously not constrained by any cable limitations. However, this does not mean that it is free from any problems, concerns or limitations. Wireless communications network market penetration depends to a high degree on factors like: subscriber density, range/coverage of wireless equipment, and management of subscriber databases. Availability requirements also have an impact on wireless coverage. It is accordingly recognized that one must design a wireless radio access network with appropriate consideration being given to costs and availability. Current solutions for wireless access are typically fixed in nature in that they cannot necessarily adapt to demands and changes in the nature of the communicated bursty traffic. A typical network topology utilizes a tree structure that provides satisfactory basic service, but still has some limitations relating to service availability and expected performance. Delays in bursty traffic handling must be set to accommodate the nature of the provided service. For example, video-on-demand services are more sensitive to delays than file transfer services. A problem with conventional networks, however, is an inability to effectively deal with these distinctions and tailor the network to provide the necessary quality of service (QoS) for bursty communications. There would thus be an advantage to a network implementation, especially in the wireless context, that better and more efficiently supported the handling of bursty traffic relating to different types of services.

SUMMARY OF THE INVENTION An aggregation network including a plurality of interconnected access routers is configured in accordance with a certain physical network topology. Responsive to individual subscriber requests for a bursty traffic service, a resource handler functionality determines which one of a plurality of logical network topologies is best suited to support the provision of communications for that bursty traffic service. The plurality of interconnected access routers are then configured by the resource handler functionality to implement that logical network topology on the certain physical network topology and admit the flow of bursty traffic related to the subscriber request for handling by the logical network topology.

BRIEF DESCRIPTION OF THE DRAWINGS

A more complete understanding of the method and apparatus of the present invention may be acquired by reference to the following Detailed Description when taken in conjunction with the accompanying Drawings wherein:

FIGURE 1 is a functional block diagram of an exemplary wireless local loop arrangement using a plurality of wireless Digital Subscriber Line Access Multiplexer

(W-DSLAM) modules physically interconnected in partially meshed network topology and supporting a logical configuration driven by user cost/performance considerations;

FIGURE 2 is a schematic diagram illustrating the physical interconnection of an aggregation network in a partially meshed network topology; and FIGURES 3A-3C are schematic diagrams illustrating the implementation of different logical network topologies using the partially meshed physical network topology for the aggregation network of FIGURE 2. DETAILED DESCRIPTION OF THE DRAWINGS

Reference is now made to FIGURE 1 wherein there is shown a functional block diagram of a wireless local loop arrangement 100 using a wireless Digital Subscriber Line Access Multiplexer (W-DSLAM or a Aradio head@) module 102. A high-speed common carrier network 104 is connected from one of its Asynchronous

Transfer Mode ATM switches 106 (comprising a point of presence) to at least one of a plurality of W-DSLAM modules 102 over high-speed wireless links (aggregate paths) 108. The aggregate throughput provided via each wireless link 108 is approximately 40.0 Mbps. Each W-DSLAM 102 functions to multiplex this aggregate throughput into a plurality of wireless distribution channels 110. Each wireless distribution channel 110 is capable providing a throughput of between (approximately) 384 Kbps and (approximately) 2 Mbps. At these operating throughput rates, the W- DSLAM 102 is capable of concentrating a substantial number of channels into the single high speed wireless aggregate path link 108. The plurality of W-DSLAM modules 102 provide a corresponding plurality of wireless local loop arrangements

100, and are interconnected with each other to form an aggregation network 111. The physical interconnection of the W-DSLAM modules 102 and the point of presence ATM switch 106 is made through the links 108 implementing a partially meshed physical network topology (see, also, FIGURE 2). The aggregation network 111 may utilize any one of a number of technologies including a local multi-distribution system

(LMDS) or a multi-point to multi-point distribution system (MMDS) with respect to the physically interconnecting links 108. Each of the wireless distribution channels 110 is terminated at one or more network termination (NT) nodes or modules 112, each of which being provided at a user=s (or subscribers) home, office, business, or any other facility. The network termination module 112 may further operate in a multiplexing fashion to allow plural devices (D) 114, each requiring a portion of the bandwidth provided by the wireless distribution channel 110, to be connected for communication. Examples of such devices 114 include: a computer, a telephone, home appliances, office devices, and microprocessor controlled components and the like.

The high-speed common carrier network 104 is provided as a Synchronous Optical Network (SONET) or related Synchronous Digital Hierarchy (SDH) ring 120 formed from a plurality of cable interconnected ATM Add/Drop Multiplexers (ADM)

122. Other network elements, such as a server 124, routers 126, and ATM switches 106 are interconnected to the ring 120. As mentioned above, the ATM switch 106 connecting with the W-DSLAM module 102 over the link 108 comprises a point of presence (POP) node for providing wireless local loop arrangement 100 access to the common carrier network 104. This POP ATM switch 106 is preferably connected to one of the ring ATM/ ADM modules 122 via a high-speed link 130 capable of throughput in the order of several hundred megabits per second. As one example, the high-speed link 130 may be implemented to transport an Optical Carrier 12 (OC-12) optical signal with a throughput of around 622.080 Mbps. The multi-user shareable wireless local loop distribution channels 110 may advantageously utilize Digital Subscriber Line (DSL) modem technology to support transmission operating rates that are capable of delivering a host of broadband applications such as multimedia, video-teleconferencing, video-on-demand, games-on- demand, and the like. Preferably, a suitable Time Division Multiple Access (TDMA) or Code Division Multiple Access (CDMA) scheme may be utilized for the air interface required for implementing these point to multi-point distribution channels 110. Further, the air interface may be provided in accordance with standard protocols such as ANSI- 136 or GSM, using a common carrier signal, for example, a 200 KHz (for a 384 Kbps data rate) or 1.6 MHz (for data rates up to 2 Mbps) signal. Enhanced data rates for global evolution (EDGE) or wireless local area network (W-LAN) technology may be used on this Adistribution@ side of the network.

A network management system 134 is connected to the W-DSLAM module(s) 102 via a communications link 136 that may comprise a network signaling connection. The system 134 maintains data concerning not only users and their services but also the configuration of the wireless local loop arrangements 100, the aggregation network 111 and the carrier network 104. The system 134 further performs some active management functions concerning ensuring proper operation and configuration of the wireless local loop arrangements 100, the aggregation network 111 and the carrier network 104. More specifically, with respect to configuration of the aggregation network 111, a resource handler functionality 140 is provided to manage the logical configuration of the aggregation network 111 in view of quality of service (QoS), resource availability and cost concerns. The resource handler 140 includes two components. A first component 140a, comprising the intelligence part of the resource handler functionality 140, resides in the network management system 134. A second component 140b, comprising the configuration part of the resource handler functionality 140, resides in each of the W- DSLAM modules 102. The first and second components of the resource handler functionality may communicate with each other using control signaling based on

IETF-SNMP (simple network management protocol) or ITU-CMIP (common management information protocol). The first component 140a operates to evaluate each subscriber request for bursty communications services in view of QoS, resource availability and cost factors in order to choose a certain logical network configuration (or topology) that meets those needs. This first component 140a then communicates with each of the plural W-DSLAM modules 102 of the aggregation network 111 to coordinate network-wide implementation of that logical network configuration/topology. Any suitable decision algorithm, expert system or artificial intelligence process may be implemented by the first component 140a to evaluate service requests in terms of QoS, resource availability, and cost to select an appropriate logical topology. The second component 140b in each W-DSLAM module 102 then responds to the communications by setting up the logical paths within the network 111 that are necessary to provide the chosen logical network configuration for the given subscriber request for bursty communications service. The resource handler functionality 140 accordingly serves to tailor the physical topology of the network 111 to implement, for each subscriber bursty communications service, a logical network topology that provides the desired communications service within the subscribers specified QoS and cost constraints and the network=s resource availability.

Resource allocation in connection with implementing a logical network topology on the physical network topology of the links 108 for the network 111 is accomplished is accordance with the principles of multi-protocol label switching for traffic engineering (MPLS-TE) as covered by IETF RC-2702 (the disclosure of which is incorporated by reference). These MPLS-TE principles advantageously define a trunk as a generic connection without a source-destination label edge router (LER) that describes a certain QoS and a set of specific connection constraints/control. Given this description, a trunk may be thought of as a logical link. What must be done next is to map that logical link to an LSP set (or multiple LSP sets) . This operation is performed by the second component 140b of the resource handler functionality. Reacting to a request from the first component 140a (i.e., the network management system 134), the second component 140b in an access router/W-DSLAM 102 uses the node=s resource handler functionality to create a logical specific flow arrangement. The second component 140b has access to network topology information (from the topology database stored by the network management system), and uses this information to map a trunk to a label switched path (LSP) or paths (LSPs) set, wherein a set comprises one flow arrangement, and there is one class of service provided per trunk definition. The mapping operation is accomplished by means of IP unique addressing. The foregoing may be better understood in connection with some examples.

A first LSP(s) set corresponds to a partially meshed logical network topology. A first received packet to be communicated over the partially meshed logical topology in accordance with, for example, QoS requirements, goes to a certain trunk definition which is mapped to that first LSP(s) set based on unique IP addressing pre-established on the longest prefix-match but forwarded in terms of MPLS labels. A second LSP(s) set, on the other hand, corresponds to a ring logical network topology. A second received packet to be communicated over the ring logical topology in accordance with, for example, QoS requirements, goes to a different trunk definition which is mapped to that second LSP(s) set based on unique IP addressing pre-established on the longest prefix-match but forwarded in terms of MPLS labels. This process for individually handling the packets, determining the logical topology per packet, identifying the implicated LSP(s) set in accordance with trunk definition mapping, and sending based on the unique IP addressing forwarded in terms of MPLS labels, is repeated by the second component 140b for each packet handled at the access router/W-DSLAM 102.

Reference is now made to FIGURE 2 wherein there is shown a schematic diagram illustrating the physical interconnection of an aggregation network 111 in a partially meshed network topology. The network 111 includes a plurality of access routers 150 (which may comprise W-DSLAM modules 102 as shown in FIGURE 1, or any other type of wired or wireless router) interconnected with each other and a point of presence (POP) node 152 (which may comprise an ATM switch 106 as shown in FIGURE 1) by aggregate links 154 (which may comprise the wireless aggregate links 108 as shown in FIGURE 1 or any other wired link) in a partially meshed physical network topology. The aggregation network 111 further includes a network management system 134 that is connected to the access router(s) 150 via a communications link 136. The network management system 134 operates, through a resource handler functionality 140, to tailor the partially meshed physical topology of the network 111 to implement certain logical network topologies for subscriber bursty communications service. An intelligence component 140a residing in the network management system 134 receives a subscriber request for bursty communication service, evaluates a number of factors, and selects a logical topology configuration to be implemented in the network 111 in providing that bursty communications service. The factors taken into account by the intelligence component 140a include: quality of service (QoS) entitlement; communication network resource availability; and, cost constraints. Once a logical topology configuration is selected for the communication, this selection is communicated to a configuration component 140b resident in each of the access routers 150. The configuration component 140b within each access router

150 responds to the logical network topology selection of the intelligence component 140a by configuring its communications resources to support and implement (on a network-wide basis) the specified logical topology configuration defined for that subscriber bursty communication request. The access routers then admit the flow of bursty traffic related to the subscriber request for handling by the specified logical network topology.

It should be noted that the logical topology configuration is provided on an individual subscriber communication by subscriber communication basis. By this it is meant that the physical topology of the aggregation network 111 may be configured to provide multiple logical topologies at the same time to handle the needs of different subscriber bursty communications service requests. Examples of logical topologies supported by the partially meshed physical topology of FIGURE 2 are illustrated in FIGURES 3A-3C. In FIGURE 3 A, a logical ring topology (indicated by the connections 160) is shown implemented on the physical partially meshed network 111 by certain ones of the access routers 150 and POP node 152 (as well as certain ones of the links 154). Furthermore, in FIGURE 3B, a logical duplicated tree topology (indicated by the connections 162) is shown implemented on the physical partially meshed network 111 by certain ones of the access routers 150 and POP node 152 (as well as certain ones of the links 154). Finally, in FIGURE 3C, a logical partially meshed topology (indicated by the connections 164) is shown implemented on the physical partially meshed network 111 by certain ones of the access routers 150 and POP node 152 (as well as certain ones of the links 154). Any one or more of these logical topologies, or some other logical topology, may be implemented on the physical network 111 at one time in response to subscriber communications requests and the actions of the resource handler functionality 140.

A more complete understanding of the operation of the present invention may be obtained from consideration of the handling of some exemplary subscriber communications requests. In a first example, assume the existence of a subscriber who has subscripted to a premium grade of service that reflects the subscribers insensitivity to cost. In this context, for example, the higher the grade of service (premium versus executive versus basic) subscripted to, the less the subscriber cares about how much the communications service will cost as long as he/she receives the best possible QoS. The subscriber then makes a request for a video-conference and contacts the network 111 to inquire about the availability of network resources to support this bursty traffic. Responsive thereto, the network management system 134, through the intelligence component 140a of the resource handler functionality 140, checks the stored profile for the requesting subscriber and determines that the subscriber is entitled to premium grade service. Given knowledge of the types of logical network topologies that are available, the intelligence component implements a cost/performance analysis in view of the requested service and selects the partially meshed logical network topology because this will provide the best QoS for the video conference and the monetary expense of providing the communication is not a concern. The intelligence component 140a then communicates with the configuration components 140b of the network 111 and instructs them to implement the selected partially meshed logical network topology with respect to supporting this requested video conference. Each configuration component 140b then responds by setting up the requisite resources and connections to implement the selected partially meshed logical network topology. Once established, the network 111 is informed, through its connection admission control (CAC) functionality, to open a connectionless connection oriented flow arrangement to support admission of the flow of bursty video conference traffic to the network 111 using the established logical network topology. As another example, take the same scenario except assume that the subscriber has subscripted to an executive grade of service that reflects some sensitivity on the part of the subscriber to cost. Put another way, this subscriber is willing to trade lower (but still acceptable) QoS for lower cost. The subscriber then makes a request for a video-conference and contacts the network 111 to inquire about the availability of network resources to support this traffic. Responsive thereto, the network management system 134, through the intelligence component 140a of the resource handler functionality 140, checks the stored profile for the requesting subscriber and determines that the subscriber is entitled to executive grade service. Given knowledge of the types of logical network topologies that are available, the intelligence component implements a cost/performance analysis in view of the requested service and selects the duplicated tree logical network topology because this will provide an acceptable QoS for the video conference at a slightly reduced cost. The intelligence component 140a then communicates with the configuration components 140b of the network 111 and instructs them to implement the selected duplicated tree logical network topology with respect to supporting this requested video conference. Each configuration component 140b then responds by setting up the requisite resources and connections to implement the selected duplicated tree logical network topology. Once established, the network 111 is informed, through its connection admission control (CAC) functionality, to open a connectionless connection oriented flow arrangement to support admission of the flow of bursty video conference traffic to the network 111 using the established logical network topology.

Consider next a scenario where a subscriber has subscripted to a basic grade of service that reflects a greater concern on the part of the subscriber with respect to cost. Put another way, this subscriber desires the pay the least amount for service, even if this adversely affects quality. The subscriber then makes a request for a good quality video-on-demand service and contacts the network 111 to inquire about the availability of network resources to support this traffic. Responsive thereto, the network management system 134, through the intelligence component 140a of the resource handler functionality 140, checks the stored profile for the requesting subscriber and determines that the subscriber is entitled to basic grade service. A conclusion is reached that the subscriber is looking primarily for a good economic deal in connection with providing the video-on-demand service. Given knowledge of the types of logical network topologies that are available, the intelligence component implements a cost/performance analysis in view of the requested service and selects the duplicated tree logical network topology because this will provide an acceptable QoS for the video-on-demand service at low cost. It should be recognized here that an option existed for choosing the less expensive logical ring topology, but this was not selected because video-on-demand requires a certain level of QoS to be acceptable and the ring topology cannot satisfy this demand. The intelligence component 140a then communicates with the configuration components 140b of the network 111 and instructs them to implement the selected duplicated tree logical network topology with respect to supporting this requested video-on-demand service. Each configuration component 140b then responds by setting up the requisite resources and connections to implement the selected duplicated tree logical network topology. Once established, the network 111 is informed, through its connection admission control (CAC) functionality, to open a connectionless connection oriented flow arrangement to support admission of the flow of bursty video-on-demand traffic to the network.

Next, consider a subscriber that has subscripted to a premium grade of service. The subscriber then makes a request for a file transfer (FTP) and contacts the network 111 to inquire about the availability of network resources to support this traffic. Responsive thereto, the network management system 134, through the intelligence component 140a of the resource handler functionality 140, checks the stored profile for the requesting subscriber and determines that the subscriber is entitled to premium grade service. Given knowledge of the types of logical network topologies that are available, the intelligence component implements a cost/performance analysis in view of the requested service and selects the ring logical network topology because this will provide a completely acceptable QoS for the file transfer. It is recognized that the premium grade subscriber is not necessarily sensitive to cost concerns, however, the requested file transfer service does not require any stringent levels of QoS that cannot adequately be met by the logical ring topology and it does not make economic sense

(even for a premium grade subscriber) to pay for more capability and capacity than you need. The intelligence component 140a then communicates with the configuration components 140b of the network 111 and instructs them to implement the selected ring logical network topology with respect to supporting this requested file transfer. Each configuration component 140b then responds by setting up the requisite resources and connections to implement the selected ring logical network topology. Once established, the network 111 is informed, through its connection admission control (CAC) functionality, to open a connectionless connection oriented flow arrangement to support admission of the flow of bursty file transfer traffic to the network Finally, consider a scenario where a subscriber has subscripted to a premium grade of service. The subscriber then makes a request for a good quality video-on- demand service and contacts the network 111 to inquire about the availability of network resources to support this traffic. Responsive thereto, the network management system 134, through the intelligence component 140a of the resource handler functionality 140, checks the stored profile for the requesting subscriber and determines that the subscriber is entitled to premium grade service. A conclusion is reached that the subscriber is willing to pay whatever cost is required to receive the best necessary QoS for the requested video-on-demand service. Given knowledge of the types of logical network topologies that are available, the intelligence component implements a cost/performance analysis in view of the requested service and selects the partially meshed logical network topology. However, the network management system further recognizes that some of the access routers in the network are already loaded with handling other subscriber service requests and that this load precludes the establishment of the partially meshed logical network topology. Responsive thereto, the intelligence component downgrades its decision to the next best topology (e.g., the duplicated tree logical network topology) whose implementation would not be restricted by the current loading conditions. The intelligence component 140a then communicates with the configuration components 140b of the network 111 and instructs them to implement the selected duplicated tree logical network topology with respect to supporting this requested video-on-demand service. Each configuration component 140b then responds by setting up the requisite resources and connections to implement the selected duplicated tree logical network topology. Once established, the network 111 is informed, through its connection admission control (CAC) functionality, to open a connectionless connection oriented flow arrangement to support admission of the flow of bursty video-on-demand traffic to the network.

Although preferred embodiments of the method and apparatus of the present invention have been illustrated in the accompanying Drawings and described in the foregoing Detailed Description, it will be understood that the invention is not limited to the embodiments disclosed, but is capable of numerous rearrangements, modifications and substitutions without departing from the spirit of the invention as set forth and defined by the following claims.-

Claims

WHAT IS CLAIMED IS:

1. A network, comprising: a plurality of access routers interconnected to form an aggregation network in accordance with a certain physical network topology, wherein that certain physical network topology is capable of being configured by the access routers into a plurality of different logical network topologies; and a network management system responsive to a user request for provision of a bursty communications service to be carried over the aggregation network for selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service and for instructing the plurality of access routers to implement that selected certain logical network topology on the certain physical network topology of the aggregation network in connection with the provision of the user requested bursty communications service.

2. The network of claim 1 wherein the certain physical network topology comprises a physical partially meshed network topology.

3. The network of claim 1 wherein the plurality of different logical network topologies comprises a logical ring topology, a logical duplicated tree topology and a logical partially meshed topology.

4. The network of claim 1 wherein the network management system comprises: a centralized resource handler intelligence portion located external to the aggregation network and operable to select the logical network topology and provide the instructions for implementation; and a resource handle configuration portion located within each access router and operable to implement the selected logical network topology on the certain physical network topology of the aggregation network.

5. The network of claim 4 wherein the resource handle configuration portion located within each access router is further operable to admit the flow of bursty traffic related to the user requested bursty communications service for handling by the logical network topology.

6. The network of claim 4 wherein the resource handler configuration portion located within each access router implements the selected logical network topology in accordance with multi-protocol label switching for traffic engineering techniques.

7. The network of claim 1 wherein the network management system operation for selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service takes into account quality of service concerns.

8. The network of claim 1 wherein the network management system operation for selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service takes into account cost of service provision concerns.

9. The network of claim 1 wherein the network management system operation for selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service takes into account network resource availability concerns.

10. The network of claim 1 wherein the network management system is further capable of simultaneously implementing a first logical network topology on the certain physical network topology of the aggregation network in connection with the provision of a first user requested bursty communications service and implementing a second logical network topology on the certain physical network topology of the aggregation network in connection with the provision of a second user requested bursty communications service.

11. A method for providing bursty communication service to users connected for communication over an aggregation network that includes a plurality of access routers interconnected in accordance with a certain physical network topology, wherein that certain physical network topology is capable of being configured by the access routers into a plurality of different logical network topologies, the method comprising the steps of: responding to a user request for provision of a bursty communications service to be carried over the aggregation network by: selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service; and implementing that selected certain logical network topology on the certain physical network topology of the aggregation network in connection with the provision of the user requested bursty communications service.

12. The method of claim 11 wherein the certain physical network topology comprises a physical partially meshed network topology.

13. The method of claim 11 wherein the plurality of different logical network topologies comprises a logical ring topology, a logical duplicated tree topology and a logical partially meshed topology.

14. The method of claim 11 further including the step of admitting the flow of bursty traffic related to the user requested bursty communications service for handling by the logical network topology.

15. The method of claim 11 wherein the step of selecting comprises the step of taking into account quality of service concerns in selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service.

16. The method of claim 11 wherein the step of selecting comprises the step of taking into account cost of service provision concerns in selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service.

17. The method of claim 11 wherein the step of selecting comprises the step of taking into account network resource availability concerns in selecting a certain one of the plurality of different logical network topologies that is best suited for handling the user requested bursty communications service.

18. The method of claim 11 wherein the step of implementing comprises the step of simultaneously implementing a first logical network topology on the certain physical network topology of the aggregation network in connection with the provision of a first user requested bursty communications service and implementing a second logical network topology on the certain physical network topology of the aggregation network in connection with the provision of a second user requested bursty communications service.

19. The method of claim 11 wherein the step of implementing comprises using multi-protocol label switching for traffic engineering techniques to provide the user requested bursty communications service.

20. The method of claim 11 wherein the step of implementing comprises the steps of: individually handling each user packet; determining on a per packet basis which of the logical network topologies is to be used; identifying an implicated label switched path(s) set in accordance with trunk definition mapping that meets the determined topology; and sending the packet based on unique IP addressing forwarded in terms of multiprotocol label switching labels over the identified label switched path(s).