Traitement en cours

Veuillez attendre...

Paramétrages

Paramétrages

Aller à Demande

1. GB2537084 - Coordinating telecommunications networks

Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]
Intellectual Property
Office 
Application No. GB 1421167.6 RTIv! Date :2 August 2016
The  following terms are registered trade marks and should be read as such wherever they occur in this document:
Cisco  (page 1)
Hewlett-Packard  Company (page 1)
VMWare  (page 1)
Juniper  (page 1)
FIFA  (page  11, 12)
COORDINATING TELECOMMUNICATIONS NETWORKS
FIELD  OF THE INVENTION
[1]         This invention relates to the management of telecommunications networks. In particular, the invention relates to coordinating two or more telecommunications networks in real time to ensure that data traffic flows across them efficiently.
BACKGROUND
[2]         There is a need to manage the use of telecommunications networks for ensuring that network resources are available when required and for ensuring that network resources are used efficiently.
[3]         A range of tools has been developed by network equipment vendors for modelling the flow of traffic across networks and for managing the use of network resources. For example, Cisco and Alcatel Lucent have developed tools for managing access networks including modelling switched traffic in cities. Data centres may be modelled as networks and as a result enterprise-based custom tools for managing data centre resources have been created, for example by Hewlett-Packard Company (HP), International Business Machines Corporation (IBM) and VMWare. In order to manage traffic across nationwide core networks, Juniper and Infinera have developed tools for modelling aggregated, long haul traffic.
[4]         Each of the custom tools orchestrate the use of network resources in the networks they have been designed for. Network orchestration is the automated arrangement, coordination and management of complex computing and network infrastructure for automatically controlling computing and network resources. This approach is, however, associated with limitations because the different custom built tools use different processes and descriptions. As a result, the different network management tools do not speak to each other, and although individual networks can be managed in real time it is difficult to coordinate different networks.
[5]         Since different networks cannot easily be coordinated in real time, it is generally necessary to reserve resources in different networks in advance. This approach is often taken for broadcasting television coverage of large sporting events such as the Olympics and the FIFA World Cup. For example, in such cases it is known to build an expensive bespoke access network for media studios covering the sporting event and to reserve communication lines in a core network for distributing the media data to viewers. Network resources reserved in this way are usually unavailable to other end-users of the networks during the course of the sporting event and end-users have even been asked not to upload videos using their smartphones inside the Olympic Village. Referring to Figure 1, known approaches are summarised schematically by a bespoke data pipe 2 built or reserved for a sporting event for transporting data from a data source 4 such as a television studio to a data centre 6 from where it is distributed to nearby viewers 8.
[6]         Following another approach, a known network capacity management tool is described in WO 2013/024269 A1 , the contents of which is incorporated herein by reference.
[7] With the limitations of known techniques i an improved technique for coordinating telecomm
SUMMARY  OF THE INVENTION
[8]         In a first aspect of the invention there is telecommunications networks is provided. The s first telecommunications network and a s telecommunications network. The plurality of tel data model that is readable by the first and seco representing connectedness and one or more network. The first network orchestrator is confi terms of the data model for network resources to network orchestrator is configured to receive the f determine that the requested network resources the first network orchestrator a confirmation e network resources of the first request can be prov [9] Preferably, at least two of the teleco telecommunications networks.
[10]       Preferably, the data model comprises telecommunications network.
[11]       Preferably, the data-driven network m describing properties of the network.
[12]       Preferably, the data-driven network connectedness defined in terms of a connectedn node identification in which matrix cells are popul [13] Preferably, the data-driven network m defined in terms of a cost of links matrix of unidir cells are populated by cost values.
[14]       Preferably, the data model comprises capacity of a telecommunications network.
[15]       Preferably, the data-driven capacity mod defined in terms of a capacity matrix of unidire matrix cells are populated by bandwidth values.
[16]       Preferably, the data model comprises a demands on a telecommunications network.
[17]       Preferably, the data-driven service model defined in terms of a demand matrix of start nod in which matrix cells are populated by bandwidth in mind, it is an object of the invention to provide unications networks.
provided a system for coordinating a plurality of ystem comprises a first network orchestrator of a econd network orchestrator of a second ecommunications networks is described using a nd network orchestrators and that is capable of of capacity and demands for service on the igured to transmit a first request expressed in the second network orchestrator. The second first request for network resources, automatically can be provided, and automatically transmit to xpressed in terms of the data model that the ided.
mmunications   networks are different types of
a data-driven network model for describing a
odel comprises vectors and/or matrices for
model comprises a description of network ess matrix of unidirectional link identification by ated by binary values.
odel   comprises a description of network cost ectional link identification by cost in which matrix
a data-driven capacity model for describing
el comprises a description of network capacity ctional link identification by bandwidth in which
data-driven  service model for describing service
l  comprises a description of demands for service e identification by destination node identification values.
[18] Preferably, the second network orchestra network resources can be provided by testing matrix.
[19]       Preferably, the orchestrators are compo computer apparatus.
[20]       Preferably, the first request for network between the first telecommunications network and [21] Preferably, the request for connectivity in [22] Preferably, the first request for network resources in the second telecommunications netw [23] Preferably, the first request for network r resources in the second telecommunications netw [24] Preferably, the data storage resources co [25] Preferably, the data storage resources co [26] Preferably, the data storage resources co [27] Preferably, the confirmation comprises a first request will be provided.
[28]       Preferably, the indication defines how co network and the second telecommunications netw [29] Preferably, the first request for netwo orchestrator determining an optimum mode of op [30] Preferably, the determining an optimu utilisation of the first telecommunications networ network could be used more efficiently; and deter first telecommunications network.
[31]     Preferably, the system comprises a telecommunications network, wherein the second second request for network resources to the third [32] Preferably, the first telecommunications n telecommunications network is a core network, backhaul of data from the access network to the c [33] Preferably, the first network orchestrator is network resources of the first telecommunications [34] Preferably, the first telecommunications telecommunications network is an access netwo comprises a request for data storage resources facilitating distribution of data to end-consumers.
[35]     Preferably, the one or more requests for coverage of an event to viewers globally.
[36]     Preferably, the system is configured to au based on the confirmation.
tor is configured to determine that the requested a network configuration based on the demand
nents of a global orchestrator implemented in
resources comprises a request for connectivity d the second telecommunications network.
dicates  a minimum bandwidth requirement.
resources comprises a request for data transit ork.
resources  comprises a request for data storage ork.
mprise data centre resources.
mprise a virtual data centre.
mprise a virtual cache.
n  indication of how the network resources of the
nnectivity between the first telecommunications ork will be arranged.
rk resources is based on the first network eration of the first telecommunications network. m mode of operation comprises: monitoring k; determining that the first telecommunications rmining a more efficient mode of operation of the
third network orchestrator of a third network orchestrator is configured to transmit a network orchestrator.
etwork is an access network and the second and the connectivity comprises connectivity for ore network.
configured to create a virtual data centre from network for storing data before backhaul.
network is a core network and the second rk, and the first request for network resources in the second telecommunications network for
network resources are for broadcasting media
tomatically transmit data through the networks
[37] In a second aspect of the invention telecommunications network, the orchestrator bei resources to an orchestrator of another tel orchestrator of the other telecommunications net of the request can be provided; and automati telecommunications network for transmitting confirmation.
[38]     In a third aspect of the invention there is pr network, the orchestrator being configured to: re orchestrator of another telecommunications netw network resources can be provided; automatic telecommunications network a confirmation that provided; and automatically control one or more for transmitting data through the networks based
[39]     In a fourth aspect of the invention there is telecommunications networks, the method com telecommunications network transmitting a first r network resources to a second network orchestr the second network orchestrator receivin automatically determining that the requested ne network orchestrator automatically transmitting t expressed in terms of the data model that the provided, wherein the data model describes the readable by the first and second orchestrators an one or more of capacity and demands for service
[40]     Preferably, at least two of the telecom telecommunications networks.
[41]     Preferably, the data model comprises a telecommunications network.
[42]     Preferably, the data-driven network model c properties of the network.
[43]     Preferably, the data-driven network m connectedness defined in terms of a connectedn node identification in which matrix cells are popula [44] Preferably, the data-driven network model in terms of a cost of links matrix of unidirectional are populated by cost values.
[45]     Preferably, the data model comprises a dat of a telecommunications network.
n   there is provided an orchestrator of a ing configured to: transmit a request for network lecommunications network; receive from the twork a confirmation that the network resources ically control one or more components of the data through the networks based on the
rovided an orchestrator of a telecommunications ceive a request for network resources from an ork; automatically determine that the requested ally transmit to the orchestrator of the other the network resources of the request can be components of the telecommunications network on the confirmation.
provided a method of coordinating a plurality of prising: a first network orchestrator of a first request expressed in terms of a data model for rator of a second telecommunications network; g the first request for network resources and twork resources can be provided; the second to the first network orchestrator a confirmation network resources of the first request can be plurality of telecommunications networks and is d is capable of representing connectedness and on the networks.
munications networks are different types of
data-driven network model for describing a
comprises  vectors and/or matrices for describing
odel comprises a description of network ess matrix of unidirectional link identification by ated by binary values.
comprises a description of network cost defined l link identification by cost in which matrix cells
ta-driven capacity model for describing capacity
[46] Preferably, the data-driven capacity mode defined in terms of a capacity matrix of unidire matrix cells are populated by bandwidth values.
[47]     Preferably, the data model comprises a d demands on a telecommunications network.
[48]     Preferably, the data-driven service model defined in terms of a demand matrix of start nod in which matrix cells are populated by bandwidth
[49]     Preferably, determining that the requested testing a network configuration based on the dem
[50]     Preferably, the orchestrators are compon computer apparatus.
[51]     Preferably, the first request for network r between the first telecommunications network and [52] Preferably, the request for connectivity indi [53] Preferably, the first request for network r resources in the second telecommunications netw [54] Preferably, the first request for network r resources in the second telecommunications netw [55] Preferably, the data storage resources com [56] Preferably, the data storage resources com [57] Preferably, the data storage resources com [58] Preferably, the confirmation comprises an first request will be provided.
[59]     Preferably, the indication defines how con network and the second telecommunications netw [60] Preferably, the first request for network res determining an optimum mode of operation of the [61] Preferably, the determining an optimum mo of the first telecommunications network; determi could be used more efficiently; and determining telecommunications network.
[62]     Preferably, the method comprises a telecommunications network, wherein the second second request for network resources to the third [63] Preferably, the first telecommunications n telecommunications network is a core network, backhaul of data from the access network to the c [64] Preferably, the first network orchestrator is network resources of the first telecommunications el comprises a description of network capacity ctional link identification by bandwidth in which
ata-driven service model for describing service
comprises a description of demands for service e identification by destination node identification values.
network resources can be provided comprises and matrix.
nents  of a global orchestrator implemented a
resources  comprises a request for connectivity d the second telecommunications network.
cates a minimum bandwidth requirement.
resources  comprises a request for data transit ork.
esources  comprises a request for data storage ork.
prise data centre resources.
prise a virtual data centre.
prise a virtual cache.
indication of how the network resources of the
nnectivity  between the first telecommunications ork will be arranged.
ources is based on the first network orchestrator first telecommunications network.
de of operation comprises: monitoring utilisation ining that the first telecommunications network a more efficient mode of operation of the first
a    third network orchestrator of a third network orchestrator is configured to transmit a network orchestrator.
etwork is an access network and the second and the connectivity comprises connectivity for ore network.
configured to create a virtual data centre from network for storing data before backhaul.
[65] Preferably, the first telecommunications telecommunications network is an access netwo comprises a request for data storage resources facilitating distribution of data to end-consumers.
[66]     Preferably, the one or more requests for coverage of an event to viewers globally.
[67]     Preferably, the method comprises automa based on the confirmation.
[68]     In a fifth aspect of the invention there is pr on a computer causes the computer to perform a
[69]     In a sixth aspect of the invention there is readable code which when run on a compute according to the fourth aspect.
[70]     In a seventh aspect of the invention th comprising computer readable code according to t [71] In an eighth aspect of the invention there perform a method according to the fourth aspect.
[72]     In a ninth aspect of the invention there is pr selected mode of household use, the article co and executable program instructions embodied i when executed by a programmable system cause the fourth aspect.
[73]     In a tenth aspect of the invention there is p of household use, the device comprising: a mac program instructions embodied in the machine re a programmable system causes the system to per [74] In an eleventh aspect of the invention there first telecommunications network coordinating us use of a second telecommunications network, the transmitting a request for network resources to an network; the first orchestrator receiving from the network a confirmation that the requested netw orchestrator automatically controlling one or network for transmitting data through the networks [75] In a twelfth aspect of the invention there i first telecommunications network coordinating us use of a second telecommunications network, the receiving a request for network resources from an network; the first orchestrator determining provided; the first orchestrator transmitting to the network a confirmation that the requested netw network is a core network and the second rk, and the first request for network resources in the second telecommunications network for
network resources are for broadcasting media
atically   transmitting data through the networks
ovided computer program code which when run method according to the fourth aspect.
provided a carrier medium carrying computer r causes the computer to perform a method
ere is provided a computer program product the fifth aspect.
is provided an integrated circuit configured to
rovided an article of manufacture for detecting a mprising: a machine-readable storage medium; in the machine readable storage medium that es the system to perform a method according to
provided  a device for detecting a selected mode hine-readable storage medium; and executable adable storage medium that when executed by rform a method according to the fourth aspect.
e  is provided a method of a first orchestrator of a se of the first telecommunications network with method comprising: the first orchestrator n orchestrator of the second telecommunications orchestrator of the second telecommunications ork resources can be provided; and the first more components of the telecommunications s based on the confirmation.
is  provided a method of a first orchestrator of a se of the first telecommunications network with method comprising: the first orchestrator n orchestrator of the second telecommunications that the requested network resources can be orchestrator of the second telecommunications ork resources can be provided; and the first orchestrator automatically controlling one or more components of the telecommunications network for transmitting data through the networks based on the confirmation.
DESCRIPTION  OF THE DRAWINGS
[76]       The invention will now be described in detail with reference to the following drawings of which:
Figure 2 is a schematic diagram of telecommunications networks connected together, each network having an orchestrator according to an embodiment of the invention;
Figure 3 is a schematic diagram of the orchestrators of Figure 2 showing lines of communication between them;
Figure 4 is a block diagram of two of the orchestrators of Figure 2 showing a network data model being exchanged between them and modules for network management inside each of them;
Figure 5 is a flow chart showing a method of coordinating a plurality of networks for transporting data according to an embodiment of the invention;
Figure 6 is a schematic diagram of the telecommunications networks of Figure 2 showing an example use case in which the orchestrators are coordinated for transporting data relating to a sporting event to viewers according to the method of Figure 5;
Figure 7 is a flow chart summarising the implementation of the use case of Figure 6;
Figures 8A, 8B and 8C are, respectively, a network diagram, a matrix and a vector illustrating an example data structure for describing a telecommunications network in accordance with an embodiment of the invention;
Figure 9 is a flow chart showing a method of computing an optimum set of routes across a network in accordance with an embodiment of the invention;
Figure 10 is a flow chart showing a method of evolving a routing solution towards an optimum set of routes in accordance with the method of Figure 9; and
Figure 11 is a functional block diagram of a computer system suitable for implementing one or more of the orchestrators of Figure 2.
[77] Throughout the drawings, like reference symbols refer to like features or steps.
DETAILED  DESCRIPTION OF THE INVENTION
[78]       Referring to Figure 2, a core network 202 is connected to three access networks 204, 206 and 208 for transporting data between them in accordance with an embodiment of the invention. In this example, the core network 202 is the core part of a telecommunication network of a service provider, and each of the access networks 204, 206 and 208 connects subscribers to the service provider.
[79]       The access network 204 has an orchestrator, Orchestrator 1 , which performs network management functions in relation to the access network 204, as will be described in more detail below. Similarly, network management functions for the core network 202 and the access network 206 are provided by Orchestrator 2 and Orchestrator 3, respectively. The access network 208 has a different arrangement because it includes three data centres 210, 212 and 214, each of which is housed in a respective data centre building in a different location. The access network 208 is managed by Orchestrator 4, but each of the data centres 210, 212 and 214 located within the access network 208 is independently managed by Orchestrator 5, Orchestrator 6 and Orchestrator 7, respectively.
[80]       The access network 204 has connectivity 216 to the core network 202, as shown in Figure 2. The access network 206 similarly has core connectivity 218. Each of the data centres 210, 212 and 214 of the access network 208 have core connectivity 220, 222 and 224, respectively, provided by the access network 208.
[81]       The core connectivity of each of the access networks 204, 206 and 208 enables data traffic to be transported from one network to another, and enables the seven orchestrators to exchange data between them.
[82]       Each of the orchestrators is connected to the network it orchestrates, and through that network may communicate with other orchestrators. For example, Orchestrator 3 of the access network 206 may communicate with Orchestrator 2 of the core network 202 via the core connectivity 218 of the access network 206. Similarly, Orchestrator 4 may communicate with Orchestrator 7 of data centre 214 by communications within the access network 208 and may also communicate with Orchestrator 2 through core connectivity of the access network 208. In an alternative arrangement, the orchestrators may communicate with each other through a separate control plane. In either case, the ability of the orchestrators to communicate with each other may be expressed as the orchestrators being networked together. Such a networking of the seven orchestrators is shown in Figure 3 in which orchestrator connections 302 represent the ability of two orchestrators being able to communicate with each other.
[83]       As indicated in Figure 3, Orchestrator 2 may be referred to as a core orchestrator because it is an orchestrator of the core network 202. Similarly, Orchestrators 5, 6 and 7 may be referred to as data centre orchestrators because they manage the data centres 210, 212 and 214. Orchestrators 1 , 3 and 4 may be referred to because they manage the access networks 202, virtual data centres that are established within tho below.
[84]       Each of the seven orchestrators perform functions in respect of its own network. For exam 2 and 3 performs network management functio network, one of the access networks, or one management functions are reflected by the diffe module plays a distinct role in the overall manage [85] Referring to Figure 4, example orchestra to 7. The example orchestrators 402 and 404 ar an orchestrator and the roles that each of those m [86] The example orchestrators 402 and 404 networks using data-driven network models.
communicate  with each other by exchanging dat As shown in Figure 4, each orchestrator 402, 4 configuration management module 412, a perfor assurance module 416.
[87]       The capacity management module 41 resources so that available resources are used 410 has access to an inventory of network equi resources. The physical resources may include p and optical fibre. Logical resources are functio defined in software and may be geographically d include a virtual private network (VPN), a virtual d [88] The configuration management module module 410 and manages how resources are management module 412 may manage how a network links it is connected to, how much band many wavelengths or channels it is configured to
[89]       The performance management module logical resources on the network. Thus, the perf much traffic is travelling across a particular link of performance management module 414 may also such as how many copies of a data set are being [90] The service assurance module 416 uses checks that current arrangement and use of the n to be met. SLAs comprise agreed quality of servi bandwidth.
as access and virtual data centre orchestrators 206 and 208, respectively, but may also manage se access networks, as will be described further
s  a range of network and capacity management ple, each of the Orchestrators 1 to 7 of Figures ns for managing its network - either the core of the data centres. The different network rent modules of the orchestrators, where each ment of the network.
tors 402 and 404 represent the Orchestrators 1 e presented to illustrate the different modules of odules plays.
each comprise four modules for managing their The orchestrators 402 and 404 may also ta 406 through an orchestrator connection 408. 04 has a capacity management module 410, a mance management module 414 and a service
0   is used for optimising the use of network efficiently. The capacity management module ipment including physical resources and logical hysical network items such as routers, switches ns created from physical resources which are istributed. For example, logical resources may ata centre (VDC), and a virtual cache.
412 is related to the capacity management configured. For example, the configuration router is configured - for example how many width it provides, or, if is an optical device, how provide.
414  is for monitoring the usage of physical and ormance management module may monitor how f the network - i.e. the utilisation of the link. The monitor other aspects of the use of the network transmitted across the network.
s  data from the other three modules to perform network enable service level agreements (SLAs) ice (QoS) metrics such as packet loss, jitter and
g
[91] The orchestrators 402, 404 coordinate optimum operating mode for their own network a orchestrator's network is acceptable. Thus, a ne 402, 404 for optimising the mode of operation of of one network to facilitate optimum operation of negotiations by exchanging information using a d driven model provide generic definitions of the n the networks.
[92]       For example, the orchestrators 402, 404 a sporting event to multiple scheduled viewers. coverage of the sporting event provides data to Scheduled viewers of the sporting event are orchestrator 404. As a result, in order to optimise data from the television studio to the scheduled negotiation to coordinate their networks.
[93]       In this approach, the orchestrator 402 operation its network would transmit just a single than multiple copies, and let the other networ distribution to the scheduled viewers. In this insta the other orchestrator 404 to propose this arran that a predetermined quantity of data is to be connection is to be reserved between the netwo other network for storing the data before it is distri [94] The orchestrator 404 receives the reques of its network that delivers the requested servic data transit resources and data storage resourc order to provide the storage function, the orche may be provided using the resources available in the determined optimum mode of operation still 402 that its request can be met.
[95]       In this example, the orchestrator 402 re 404 because a need to broadcast the sporti anticipated In other examples, a need may be i reacting to provide a more efficient mode of oper and arranging the networks appropriately in adva [96] In either case, a negotiation between tw involves the following steps. Referring to Figur resources for data transport from a respondi orchestrator determines whether the requested s can, confirms to the requesting orchestrator that t the use of their networks by calculating an nd checking that the impact of this on the other gotiation takes place between the orchestrators both networks, including arranging the operation the other. The orchestrators 402, 404 perform ata driven model they can both read. The data etworks, their capacity and demands placed on
may coordinate their networks for broadcasting In this example, a television studio providing the network managed by the orchestrator 402. connected to the network managed by the the use of the two networks for transporting the viewers, the orchestrators 402, 404 perform a
may determine that in an optimum mode of copy of the data from the television studio rather k generate multiple copies and manage their ance, the orchestrator 402 transmits a request to gement. In particular, the request may indicate provided, that a predetermined bandwidth of rks, and that a cache should be provided in the ibuted to the scheduled viewers.
t and determines an optimum mode of operation es. The orchestrator 404 determines optimum es for delivering the request. For example, in strator 404 may determine that a virtual cache n its network. The orchestrator 404 checks that meets its SLAs and confirms to the orchestrator
quests  network resources from the orchestrator ting event efficiently across two networks is identified by monitoring traffic on a network and ration, rather than anticipating a scheduled need nce.
o orchestrators for coordinating their networks e 5, a requesting orchestrator requests network ng orchestrator (step 802). The responding ervices can be provided (step 804) and, if they he request can be met (step 806).
[97] This negotiation process is used by the Orchestrators 1 to 7 of Figures 2 and 3 for coordinating their respective networks and data centres. For example, the Orchestrators 1 to 7 may negotiate network and data centre resources for providing an optimum solution for global data transport. Large sporting events provide an example of this situation because the event takes place in a single location but there are viewers all over the world.
[98]       Referring to Figures 6 and 7, in an example the Orchestrators 1 to 7 coordinate network resources for broadcasting the FIFA World Cup 602. The event 602 is held in an originating country and there are scheduled viewers 604 and ad hoc viewers 606 around the world. When the event starts (step 702 in Figure 7), television studios in the originating country film the sporting event 602 and provide media data to the access network 204.
[99]       Orchestrator 1 arranges the establishment of a virtual data centre 608 using resources of the access network 204 (step 704) so that the media data can be stored in the virtual data centre 608 before being backhauled to the core network 202 for onward transit. In a virtual data centre, the instances of components of a data centre are created in software on generic computer equipment including storage and compute resources. The components may, for example, include firewalls, load balances, routers, encryption functionality, and session management functionality constructed created in software using units of available generic storage and compute resources. Thus, the virtual data centre 608 is constructed in software from physical resources in the originating country and may be geographically distributed within the originating country.
[100]     The backhaul of the media data to the core network 202 is coordinated by Orchestrators 1 and 2 (step 706). Orchestrator 1 determines an optimum mode of operation of the access network 204 for transmitting the media data from the virtual data centre 608 to the core network 202, and determines that core connectivity of a sufficient bandwidth is required for transporting the media data to the core network 202. Orchestrator 1 generates a request for core connectivity and onward transit and forwards the request to Orchestrator 2. Part of the request is for coordinating onward transit of the media data (step 706) and as such Orchestrator 1 provides an indication in the request of the final destination of the media data where known. For the scheduled viewers 604 the final destination is known so this is possible. An indication that the media data is destined for access network 206 for distribution to the scheduled viewers 604 is included in the request. The request may be expressed in terms of a demand matrix. In the following example demand matrix, a source of traffic is indicated as being a 'router 1' of a virtual data centre (VDC) in the access network 204, the destination is indicated as being the access network 206, and the SLA parameters are indicated as being 120Gb of bandwidth together with other SLA parameters such as packet loss and jitter (not shown).
[101]     Orchestrator 2 receives the request. In this example, there are three questions Orchestrator 2 answers in response to the request. Firstly, core connectivity is provided for the media data to be sent from access network 204 to the core network 202. Secondly, optimum available transit resources are identified and rese core network 202 from the originating access net Thirdly, connectivity between the core network reserved for providing the media data to the facilities are established in the access network data to the scheduled viewers 604.
[102]     Reserving core connectivity for backhaul 202 is performed by Orchestrator 2. In contrast network 206 and storage facilities in the acces scheduled viewers 604 is performed as a collabor [103] As part of the collaboration, Orchestr connectivity, storage facilities and transit resource expressed in terms of a demand matrix. Th bandwidth, the locations and numbers of schedul requested. A virtual data centre 610 may be req of the media data to the virtual data centre 610 a multiple copies into and across the access netwo optimum mode of operation for access network 2 these resources and a response, which may com resources have been reserved, is sent from Orch request can be met. If the full resources cannot b [104] Having received confirmation from Orc distribution resources have been reserved in t optimal backhaul and transit resources in the co original request from Orchestrator 1 confirming scheduled viewers 604.
[105]     During the course of the event 602 the m transported to viewers around the world using the scheduled viewers 604, transmission resources a be supported by the reserved resources. Howe want to watch certain matches of the FIFA Worl this instance, orchestrators of the affected netwo coordinate the use of their networks to improv distributed to the ad hoc viewers.
[106]     For example, referring to Figure 6, the access network 208. Multiple copies of the m through connections 220, 222 and 224 to data access network 208. The transit across the co access network 208 is detected by Orchestra improvement to send fewer copies to the ac generates and sends a request to Orchestrator 4 rved for transporting the media data across the work 204 to the destination access network 206. 202 and the destination access network 206 is destination access network 206, and storage 206 for facilitating the distribution of the media
l and transit resources within the core network t, the coordination of connectivity to the access s network 206 for facilitating distribution to the ration (step 708) between Orchestrators 2 and 3. rator 2 generates and sends a request for es to Orchestrator 3. Again, the request may be e request indicates the required connectivity led viewers, and that a virtual data centre 610 is uested if it is cheaper to transport a single copy nd distribute from there, rather than transporting ork 206 for multiple scheduled viewers 604. An 06 is determined by Orchestrator 3 for providing prise an 'Ack Message' indicating that the data hestrator 3 to Orchestrator 2 confirming that the e met then those available would be returned.
hestrator 3 that onward transit, storage and he access network 206, and having reserved re network 202, Orchestrator 2 responds to the that the media data can be transported to the
edia data generated in the originating country is multiple telecommunications networks. For the re reserved and the amount of traffic is likely to ver, significant numbers of ad hoc viewers may ld Cup in a pattern that was not anticipated. In rks may react to the demands in real time and e the efficiency with which the media data is
re are ad hoc viewers 606 connected to the edia data are sent from the core network 202 centres 210, 212 and 214 respectively of the re network 202 of multiple copies going to the tor 2 which determines that it would be an cess network 208. Orchestrator 2 therefore for caching functionality in access network 208 (step 710). Orchestrator 4 receives the request would involve a virtual cache 612 created in soft distribution to the ad hoc viewers 606, and a singl the virtual cache 612. A reply is generated b confirming that the caching request can be met an [107] In the described embodiments, the orche their networks. The data-driven model comprise example the connectivity of the networks, the cap demands being placed on the networks. For e
Figure  8A may be described by numbering the n 25 (shown as straight lines connecting the circl network may be expressed in the form of vectors as inputs for routing computations. For example, 82 describing the network 80 has columns for lin each unidirectional link. As a result, each bidirecti and 'Bwd' - and each unidirectional link, such a 'Fwd'. The cells of the matrix take values of zer column is connected to the node of the row. For 1 (i.e. the first column), the cells taking a value indicates that the forward direction of link 1 conne of link 3 as another example, a value of 1 is tak There are no other cells in this column that take a the forward direction of link 3 is connected to no matrix 82. Thus, the matrix 82 fully represents number of rows that is one fewer than the number [108] Other properties of the network 80 may vectors. For example, the cost of each of the link in Figure 8C. In the vector 84 the values taken b (ID) followed by an indication of the cost of the ide the cost of that other link, and so on. Alternativel values of the links in a predefined order - thus ta the links, such as capacity and utilisation, may sim [109] Vectors are also used to represent the pr a node, the physical location coordinates of a nod and whether a node is exempt from protection.
[110]     Properties of a set of services S1 to Sn m listing the values of properties such as the maxim of each of the services in a predefined order. T network may be represented by matrices either wi nodes of the network. For example, a matrix wit t and determines that an optimal arrangement ware from data centres 212 and 214 for onward le connection 614 from the core network 202 to by Orchestrator 4 and sent to Orchestrator 2 nd the improved arrangement is implemented.
strators use a data-driven model for describing s a predefined data structure for describing for acity available on the networks, and the service xample, properties of the network 80 shown in odes 1 to 7 (shown as circles) and the links 1 to les). Using this numbering, properties of the and matrices which the routing engines can use , referring to Figure 8B, a connectedness matrix nks and rows for nodes. There is a column for tional link is represented by two columns - 'Fwd' s link 23, is represented by a single column -ro or one depending on whether the link of the example, referring to the forward direction of link of 1 are in the rows for nodes 5 and 6. This cts nodes 5 and 6. Taking the forward direction ken by the cell in the row representing node 2. a value of 1 but this implies that the other end of ode 1 which is not represented by a row in the the connectedness of the network 80 using a r of nodes.
be similarly represented by other matrices or ks may be represented by a vector 84 as shown by the elements of the vector are a link identifier entified link, followed by another link ID and then ly, a similar link cost vector could simply include aking the form [¾, c 2,c 3, ...]. Other properties of ilarly be represented by vectors.
roperties  of nodes - for example, the capacity of de, the Internet Protocol (IP) address of a node,
ay be represented by a vector - for example by um delay, required bandwidth and network layer he routes taken by services S1 to Sn across a ith reference to the links of the network or to the h a row for each service and a column for each link will have a 1 in a cell if that service uses that with a row for each service and a column for ea services by indicating for each service whether o routes, as well as point-to-point services and p represented by matrices in this way.
[111]     A vector may also be used to represen whole. For example, properties such as whet adjacency limit (i.e. the maximum number of link total network utilisation may be listed in a pred network. A modelling vector may be similarly d For example, parameters used in evolving an o below) could for example include parameters f predetermined weights for mutations, and a par defined below). It will be apparent that any sui candidate routes leading to an optimised solution.
[112]     This data structure can be used for calcu S1 to Sn. For example, a dot product of a link ut of the capacity of each link and a link capacity ve produce a scalar value representing the total ne elements of the vectors and matrices are binary, storing and computing properties of the network also makes it very straight forward to add further and nodes because the data structure is predefin [113] The data-driven model can provide a g network that is common for different data transpo instantiated for the technology mix under consi applied for computing routing solutions for the approach brings together information tech telecommunications - e.g. radio and fibre optics across the technology mix.
[114]     Not only can different technology type modelled using the data structure, but also diffe modelled. For example, at a highest level of abst can be represented in which each node represe (POP) and each link represents an intercity right of detail, a separate network model can be used and optical links of a city. At the next level of represent a data centre as a network of aggrega balancers and other data centre devices.
[115]     It will be appreciated that by simply chan and easily configured and reconfigured. This is p link, and a zero if it does not. Similarly, a matrix ch node can define routes taken by each of the r not the node is used. Primary and secondary oint-to-multipoint services and IP flows can be
t  a series of properties of the network 80 as a her load balancing is allowed in the network, s that can be connected to a single node), and etermined order to form a vector describing a efined to describe rules for network modelling. ptimal routing solution (described in more detail or defining an asymptote, parameters defining ameter defining genepool size (these terms are table routing engine could be used to generate .
lating properties of the network 80 and services ilisation vector indicating the fractional utilisation ctor indicating the total capacity of each link will twork utilisation. Since the values taken by the the data structure requires very little memory for and services. The nature of the data structure information such as other properties of the links ed.
eneric representation of a telecommunications rt technologies, in which case the model can be deration and a generic routing product can be data transport technologies of the mix. This nology (IT) - e.g. data centres - and - for network management that is coordinated
s, such as different layers of a network, be rent geographical aspects of a network can be traction, an intercity telecommunications network nts a large city or other large point of presence of way, duct or existing cabling. At the next level to represent the routers, optical devices, and IP detail, another network model can be used to tion switches, top of rack routers, firewalls, load
ging the data, the network model can be rapidly ossible because the different networks (IP layer, optical layer, intercity, city, data centre, etc) shar data needed to instantiate them. This enables c different scales of interest, as well as computing i and intercity networks. The data structure is also geographically distributed, software defined netwo [116] Described embodiments of the invention this context, a capacity is defined as a unit on example, a capacity could be a unit of bandwidt power, and so on. Consumable units can also be as a virtual firewall or virtual router. Since capaci for example with the consumption of one typ consumption of another type of capacity, cap relationships such as parent-child relationshi relationships. To represent this, capacity groups capacities on the network and for describing how Some capacity groups are node capacity groups items in an IP/optical rack with multiple ports, in groups are link capacity groups and similarly des link, such as power for signal boosting, channels and so on. This embodiment also uses 'capacity capacity group instances including rules for how represent a capacity group, a suitable capacity The language for defining a capacity group temp capacity relationships. As capacity or capacity ty (relative or real), economic modelling can also be
[117]     Demands for services on the network are templates by instance data. A 'service' is defined necessarily an end user service. For example subscriber line (DSL) load, a virtual machine (VM) link capacity, or service chains. A 'service cons how this service type consumes capacity from th type consumes capacity from multiple capaci responsible for associating other relative capaciti may require 100Mb link capacity but to provide thi total link capacity, e.g. 1Gb, 1G client port, 10G etc.
[118]     In the described embodiment of the inven structure including capacity and service consum instance of the data structure for modelling its orchestrators for communicating with each other data schema among the orchestrators. It is with re the same structural principles but differ in the ommon modelling of different technologies and interactions between the network layers and city o suitable for modelling a virtual network - i.e. a ork (SDN).
use a data driven representation of capacity. In n the network which may be consumed. For h, a port, a wavelength (i.e. channel), a unit of e provided by virtual SDN enabled devices such ities are consumed according to usage patterns, e of capacity depending on or implying the acities are related to each other by various ips, consequential relationships and disjoint are defined for representing a related group of they are related and how they are consumed. and could for example describe the consumable nput cables and output cables. Other capacity cribe the consumable items associated with the s or wavelengths, capacity on certain channels, y group templates' comprising rules for creating the different capacity types are consumed. To group template is populated by instance data. plate is rich enough to describe all the possible pe in a capacity group template can include cost performed.
represented by populating service consumption here as a vehicle to consume capacity and not e, services may include an aggregated digital ), a virtual network function (VNF), access trunk umption template' describes a service type and e available capacity groups. Usually a service ity groups. The Capacity group template is ies from that the service requests, e.g. a service is the Capacity Group Template will provide the WAN port, shelves, chasses, rack space, power
ntion,  instance data is used to populate the data ption templates, and each orchestrator uses an network. The data model is also used by the r and thus the data model provides a common the use of this common data schema that the orchestrators transmit requests for services or ne network resources can be provided.
[119]     An example of a capacity and service r data for describing a node B of a network is hierarchical structure with a representation 94 of rack representation 94 indicates a rack having 4 space meaning 1.75 inches of height within the r running general rack facilities such as electrical fa [120] The template also includes a representati 3U of space and 200W of power, and having indicate client cards occupying the two used slot 1G client card having five free ports and one used [121] The template 92 can also be expressed i Figure 9. Here, different items in the hierarchical a network model. For example, a rack node repr of free power. Similarly, the five free ports of the 916.
[122]     By representing a capacity and service r capacity and service requirements of a node suc and matrices of the data scheme described above [123] Each orchestrator determines an optimu more generic routing engines that perform compu above-described data structure. Referring to Fig process that starts with an initial set of candidate and evolves the set of solutions towards an opti determining an optimum set of routes, a set of generated at step 1002. Each candidate solutio network being modelled, their being one path in e of a demand matrix. Thus, each candidate soluti of the demand matrix and can be evaluated with r to determine the quality of the solution. Each of t 1004 by computing a fitness function whose v meets the requirements of the network and of th solutions to be compared with each other, for e establish which candidate solutions are strongest [124] The initial set of candidate solutions, candidate solutions, is evolved at step 1006 towa an iterative approach that directs the developme optimum. In each iteration, the weakest candid function values - are replaced by new candid solutions have higher fitness functions than th etwork resources to each other and confirm that
requirement  template 92 populated by instance shown in Figure 9. The template 92 has a a network rack at the top of the hierarchy. The 0U of free space, where 'Ll' is a standard unit of rack. The rack also has 10kW of free power for ns to control the temperature inside the rack.
ion 96 of a P1 chassis inside the rack, requiring two used slots. Representations 98 and 910 ts. For example, representation 98 illustrates a port, and requiring 20W power.
in  the form of a network model 912, as shown in l structure are represented by different nodes of resents the rack having 40U of space and 10kW 1G client card are represented by the five nodes
equirement  template using a network model, the h as a rack may be described using the vectors e.
m  mode of operation of its network using one or tations on a network model expressed using the ure 10, the optimisation process is an iterative solutions for routing traffic through the network mised solution. According to a method 1000 of candidate solutions for starting the process is n comprises a set of routes or paths across the ach set for each respective service requirement on offers a potential routing plan for the services respect to the network and to the demand matrix he initial candidate solutions is evaluated at step alue indicates how well the candidate solution he demand matrix. This enables the candidate xample by ranking them in order of fitness, to and which are weakest.
which may for example contain five hundred rds an optimum solution. This is achieved using nt of the set of candidate solutions towards an ate solutions - i.e. those with the lowest fitness ate solutions provided that the new candidate e weakest candidate solutions. This tends to increase the quality of the set of candidate so thereby evolving the set so that it becomes incre solution.
[125]     The fitness function may provide a mea such as bandwidth, or alternatively may take acc example bandwidth, protection, latency and cost. be evaluated, the fitness function may take acco using weighting coefficients.
[126]     To evaluate cost, the fitness function ma the total cost of all routes of a candidate solutio To evaluate latency, the fitness function may com delivered latency and requested maximum laten these could also be just minimised or if the prop better, e.g. optical signal to noise ratio (OSNR performance parameter that may be evaluated is how many nodes or links of a candidate soluti candidate solution. For example, to evaluate di function of the proportion of the nodes of a cand the candidate solution.
[127]     In general, it is suitable for the value o performing candidate solution and lower for a fitness function may take values between 0 and 1 indicates a very poor candidate solution.
[128]     There is no guarantee that the initial s contains the optimum solution. This would be v services, the candidate solutions are evolved at optimum solution as follows.
[129]     The initial candidate solutions are used process involving repeated selection of stronger the average quality of the candidate solutions inc pool contains the optimum solution.
[130]     Referring to Figure 11, the evolution candidate solutions at step 1102 using a gen candidate solutions from the initial gene pool. reproducing the candidate solutions either by mat [131] For example, a pair of child candidat randomly selected parent candidate solution corresponding portions of the parent candidate solutions. For example, the routes provided particular source node and a particular destin lutions as the number of iterations increases, asingly likely that the set includes the optimum
sure of one particular performance parameter, ount of two or more performance parameters, for If there are multiple performance parameters to unt for their relative importance, for example by
y comprise a function of the difference between n and the total budget indicated by a customer. prise a function of the mean difference between cy of all routes of a candidate solution. Both erty was something that needed to bigger to be ), could be maximised. Another example of a disjointedness. Disjointedness is a measure of ion are shared between different paths of the sjointedness, a fitness function may comprise a idate solution that are shared between paths of
f the fitness function to be higher for a better weaker candidate solution. For example, the 1 , where 1 indicates an optimised solution and 0
et of randomly generated candidate solutions ery unlikely. In order to optimise the routing of step 1006 using a genetic algorithm towards an
as a gene pool that is evolved by an iterative candidate solutions. Thus, with each iteration, reases and it becomes more likely that the gene
process (step 1006) involves reproducing the etic algorithm to produce one or more 'child' . The genetic algorithm comprises rules for ing, mutating or a combination of both.
e  solutions could be created by 'mating' two s. The mating process involves swapping solutions to create two new child candidate by the parent candidate solutions between a ation node could be swapped. Alternatively, portions of routes could be swapped, or any oth carried out. Three or more candidate solutions c candidate solutions.
[132]     Another method of reproducing is to mut parent candidate solution, which may be random random, predetermined, or partially random man could undergo a random change in functionalit through). It will be appreciated that by mutating candidate solution is created.
[133]     The creation of child candidate solutions i solutions. For example, if the initial set of ca candidate solutions and two child candidate s contains five hundred and two candidate solution evaluated, for example by calculating a fitnes candidate solutions of the enlarged gene pool are function, and if n child candidate solutions we weakest n of the enlarged gene pool are discard five hundred initial candidate solutions and two two candidate solutions of the enlarged gene po weakest two of the enlarged set are discarded. constant overtime but the quality of the populatio an iteration may produce child candidate soluti candidate solutions. In that case, the new child c and the quality of the population, which may for e value, will be unchanged. In any other case an it thereby increasing the likelihood that the gene po [134] The evolution process cycles through it (step 1104), and discarding the weakest candida population size with each cycle and gradually in pool may be said to evolve towards an optimum performing solutions introduced by discarding the [135] The evolution is stopped when a suffici reach a predetermined stopping condition 1108. reaching a solution sufficiently close to the opti likely to contain the optimum solution.
[136]     In the first approach, a fitness function o optimum solution is estimated, and the best can the optimum to determine how close the best solution. If the fitness function or performance pa sufficiently close to that of the optimum value, the is stopped.
her  scrambling of the parent solutions could be ould be scrambled to produce one or more child
tate  a parent candidate solution. In this case, a ly selected from the gene pool, is changed in a ner. For example, every fifth node of each route y (e.g. from an optical cross connect to glass a single parent candidate solution, a single child
increases the size of the gene pool of candidate ndidate solutions contains five hundred initial olutions are created, the enlarged gene pool s. At step 1104 the child candidate solutions are s function of the type described above. The now ranked in order of the value of their fitness re created in the reproducing step 1102, the ed (step 1106). Following the example above of child candidate solutions, the five hundred and ol are ranked by fitness function value and the Thus, the total size of the gene pool remains n improves with cycles of the iteration. At worst, ons that are no better than any of the existing andidate solutions will be discarded immediately xample be expressed as a mean fitness function teration will improve the quality of the population ol contains the optimum routing solution.
erations  by reproducing (step 1102), evaluating te solutions (step 1106) to maintain a constant crease the quality of the population. The gene solution as a result of the bias towards better weakest candidate solutions in each cycle.
ient  number of iterations have been applied to Two suitable stopping conditions are based on mum solution and reaching a gene pool that is
r a performance parameter such as cost of the didate solution in the gene pool is compared to candidate solution to date is to the optimum rameter of the best candidate solution to date is stopping condition is satisfied and the evolution
[137] For example, if the only performance pa date will go down with each iteration, gradually a cost of the optimum solution. At some point the c a value within a predetermined percentage, such stopping condition is satisfied. After each round becomes more reliable with every round of evolu can generally be calculated in a meaningful way order to determine an asymptote, the cost impro is approximated as a ratio of two polynomials of t be expressed as follows.
rameter is cost, the cost of the best solution to pproaching an asymptote which represents the ost of the best solution to date will have reached as 5%, of the optimum cost, at which point the of evolution an asymptote is determined. This tion as it is based on more and more data, and from, for example, the tenth round onwards. In vement as a function of the number of iterations he same order. Such a rational function, R, may
[138] where X is the number of iterations and to be evaluated to fit the data. As the number o asymptote to a value equal to the ration of the hi determine the asymptote, the constants A 0, A^ must be computed which can for example be example, a best cost to date reaches within 5% 300 and 500 iterations.
[139]     In the second approach, the distribution such as cost is approximated and the value of t the best candidate solution to date is compared t the best candidate solution to date is the optimum [140] For example, the performance paramete possible routing solutions is approximated based normal distribution and 2) that the mean and st solutions is representative of the mean and stand Assumption 2 is justified because the set of ini Thus, the distribution of the cost of routing solu having a mean and standard deviation the same of the initial candidate solutions. The probability t gene pool contains the optimum solution is calc and the value of the cost of the best solution to da
P (we have optimum) =
[141]     Thus, the stopping condition is satisfied the optimum routing solution reaches a predeterm A 0, A^ A 2. A„, B 0, B^ B 2. B„ are constants f iterations becomes large the cost tends to the ghest term coefficients, A n/B n. Thus, in order to A 2. A n, B 0, B^ B 2. Β π of the polynomials done using a method of least squares. In an of an optimum cost (asymptote) after between
of  a fitness function or a performance parameter he fitness function or performance parameter of o the distribution to determine the likelihood that solution.
r  may be cost. The distribution of the cost of all on two assumptions: 1) that the distribution is a tandard deviation of the initial set of candidate ard deviation of all possible candidate solutions. itial candidate solutions is generated randomly. utions is approximated as a normal distribution as the mean and standard deviation of the costs hat, after any particular number of iterations, the ulated by iterating the distribution between -∞ te:
/■best to date
I distribution
J—oo
when the likelihood of the gene pool containing ined percentage, such as 95%.
[142] However, it will be appreciated that a res small number of initial candidate solutions) and/or reliable that a result of 95% based on a large Therefore, a safeguard may be put in place by solution has not yet been found (where P(opti The minimum probability may comprise a functi iterations that have taken place, the standard devi as the genepool evolves), or any combination probability that the optimum solution has not yet b
P (optimum  not yet found)  min =
yj ge
[143]     According to this definition of the mini reached, further iterations will increase the likeliho this will happen much more slowly than on previo small gene pool size, a higher number of iteration can reached.
[144]     Regardless of how the stopping condition in the gene pool is selected (step 1112) and ou Figure 11.
[145]     Referring to Figure 12, a device 1202 fo embodiment of the invention is shown. The devic element 1204, a database 1206 for storing an i including capacity information as well as service portal 1208, a controller 1210, ROM 1212 and capacity management module 1216, a performan management module 1220, and a service assuran [146] In some embodiments the orchestrators acting as a global orchestrator. In other embo separate hardware and may be geographically dis [147] Functions relating to coordinating telecom computers connected for data communication vi Although special purpose devices may be used, one or more hardware platforms intended to repr commonly used so as to implement the event ide an appropriate network connection for data comm
[148]     As known in the data processing and co typically comprises a central processor or other bus, various types of memory or storage media ult of 95% based on a small sample size (e.g. a r a small number of iterations is likely to be less sample size and a large number of iterations. way of a minimum probability that the optimum imum not yet found) = 1 — P(we have optimum)). ion of the size of the genepool, the number of iation of the current genepool (which decreases n of the above. For example, the minimum een found could be defined as:
1
-nepool size + number of iterations
mum  probability, once the minimum has been ood that the gene pool contains the optimum but us iterations. This ensures that for a sufficiently s may be required before the stopping condition
n  is defined, once it is reached the best solution utputted as the optimised solution as shown in
r implementing an orchestrator according to an ce 1202 comprises an input and output interface instance of a data model describing a network e information such as SLAs, a communications RAM 1214. The controller 1210 includes a nce management module 1218, a configuration nce module 1222.
are implemented as functions of a computer diments the orchestrators are implemented in stributed.
munications networks may be implemented on ia the components of a packet data network. , such devices also may be implemented using esent a general class of data processing device ntification functions discussed above, albeit with unication.
mmunications arts, a general-purpose computer r processing device, an internal communication (RAM, ROM, EEPROM, cache memory, disk drives etc.) for code and data storage, and on communication purposes. The software functional code as well as associated stored data, e.g. e already elapsed. The software code is executable as the server or terminal device used for coordina the code is stored within the general-purpose c software may be stored at other locations and/ general-purpose computer system. Execution o platform or by a number of computer platfor methodology for coordinating telecommunications in the implementations discussed and illustrated h
[149]     Those skilled in the art will be familiar hardware platforms. As will be appreciated, s computer with user interface elements, as may other type of work station or terminal device. A may also be arranged to provide a network or hos implement a server.
[150]     For example, a server includes a da communication. The server also includes a centr more processors, for executing program instruct internal communication bus, program storage processed and/or communicated by the server, a and data via network communications.
[151]     A user terminal computer will include u addition to elements generally similar to those of size, capacity, etc. of the respective elements terminal computers. The hardware elements, ope such servers are conventional in nature, and it adequately familiar therewith. Of course, the distributed fashion on a number of similar
[152]     Hence, aspects of the methods of coor above may be embodied in programming. Progra as "products" or "articles of manufacture" typi associated data that is carried on or embodied in plurality of such media. "Storage" type media in computers, processors or the like, or associated memories, tape drives, disk drives and the like, w e or more network interface cards or ports for lities involve programming, including executable nergy usage measurements for a time period by the general-purpose computerthat functions ting telecommunications networks. In operation, omputer platform. At other times, however, the /or transported for loading into the appropriate f such code by a processor of the computer ms enables the platform(s) to implement the s networks, in essentially the manner performed erein.
with  the structure of general purpose computer uch a platform may be arranged to provide a be used to implement a personal computer or general purpose computer hardware platform t computer platform, as may typically be used to
ata   communication interface for packet data ral processing unit (CPU), in the form of one or tions. The server platform typically includes an and data storage for various data files to be although the server often receives programming
ser interface elements for input and output, in the server computer, although the precise type, will often different between server and client erating systems and programming languages of t is presumed that those skilled in the art are server functions may be implemented in a platforms, to distribute the processing load.
rdinating  telecommunications networks outlined m aspects of the technology may be thought of ically in the form of executable code and/or a type of machine readable medium and/or in a clude any or all of the tangible memory of the modules thereof, such as various semiconductor hich may provide non- transitory storage at any time for the software programming. All or portions through the Internet or various other telecomm example, may enable loading of the software fro example, from a management server or host com telecommunications networks services into th computer platform. Thus, another type of media optical, electrical and electromagnetic waves, su local devices, through wired and optical landli physical elements that carry such waves, such as also may be considered as media bearing the sof transitory, tangible "storage" media, terms such a to any medium that participates in providing instru
[153]     Hence, a machine readable medium may tangible storage medium, a carrier wave medium storage media include, for example, optical or ma in any computer(s) or the like, such as m telecommunications networks, etc. shown in t dynamic memory, such as main memory of su media include coaxial cables; copper wire and fi bus within a computer system. Carrier-wave tran electromagnetic signals, or acoustic or light w frequency (RF) and infrared (IR) data communi media therefore include for example: a floppy dis other magnetic medium, a CD-ROM, DVD or DV paper tape, any other physical storage medium EPROM, a FLASH-EPROM, any other memory c or instructions, cables or links transporting such a a computer can read programming code and/or d media may be involved in carrying one or mor processor for execution.
[154]     While the foregoing has described what a examples, it is understood that various modificati matter disclosed herein may be implemented i teachings may be applied in numerous applicati herein. It is intended by the following claims to cl variations that fall within the true scope of the pres
[155]     Although the present invention has be embodiments, it will be appreciated that various of the software may at times be communicated unication networks. Such communications, for m one computer or processor into another, for puter of the organisation providing coordinating e coordinating telecommunications networks that may bear the software elements includes ch as used across physical interfaces between ine networks and over various air- links. The s wired or wireless links, optical links or the like, ftware. As used herein, unless restricted to nons computer or machine "readable medium" refer ctions to a processor for execution.
take many forms, including but not limited to, a or physical transmission medium. Non-volatile gnetic disks, such as any of the storage devices ay be used to implement the coordinating the drawings. Volatile storage media include ch a computer platform. Tangible transmission ibre optics, including the wires that comprise a smission media can take the form of electric or aves such as those generated during radio ications. Common forms of computer-readable k, a flexible disk, hard disk, magnetic tape, any D-ROM, any other optical medium, punch cards with patterns of holes, a RAM, a PROM and hip or cartridge, a carrier wave transporting data a carrier wave, or any other medium from which ata. Many of these forms of computer readable e sequences of one or more instructions to a
re considered to be the best mode and/or other ions may be made therein and that the subject in various forms and examples, and that the ions, only some of which have been described laim any and all applications, modifications and sent teachings.
en described in terms of specific exemplary modifications, alterations and/or combinations of features disclosed herein will be apparent to those skilled in the art without departing from the spirit and scope of the invention as set forth in the following claims.