WO2015167372A1

WO2015167372A1 - Identification of suitable network service points

Info

Publication number: WO2015167372A1
Application number: PCT/SE2014/050527
Authority: WO
Inventors: Fetahi WUHIB
Original assignee: Telefonaktiebolaget L M Ericsson (Publ)
Priority date: 2014-04-29
Filing date: 2014-04-29
Publication date: 2015-11-05

Abstract

It is disclosed a network node (304, 306) and a network service point (302, 308) and methods therein, for identifying a service point for a service. Based on parameter information (314, 316) such as bandwidth and latency from service points of the network, and on link information (319, 320) between directly connected network nodes, each network can maintain (322, 324) updated parameter information about which network service point is most suitable (68) for a communication device having a certain demand according to a quality of service. Hence, for a given delay and bandwidth demand, the methods alloweach router in the network to learn (68) about the most suitable service point in that network.

Description

IDENTIFICATION OF SUITABLE NETWORK SERVICE POINTS

TECHNICAL FIELD

This disclosure relates to network service points. In more particular, it relates to a network node and a network service point, and methods therein for identification of a suitable network service point for a communication device connecting to a network via the service point.

BACKGROUND

In a distributed cloud environment where data centers can be located anywhere in a network, there is a need for a so-called 'location service' that returns, given a point of attachment of an end-user to the network, the data center that is most suitable for that user. Such a service can be generalized as a service that locates the most suitable service point in a network, whereby service points are distributed inside the network.

Today, two types of services are commonly used on the Internet when it comes to choosing a suitable service point. The first and the most common one is a service that is used to locate a service point that is geographically close to the end-user. The second is a service that is used to locate a service point that has minimal communication latency to the end-user.

Geolocation services, such as those provided through the W3C Geolocation application programming interface (API) can be used to implement a location service based on geographical proximity. This API estimates the geographical location, for example in terms of latitude and longitude, of an end-user device to the accuracy allowable by the underlying location mechanism of the end-user device. A location service have information about the geographical location of the service points, chooses, among the service points, the service point that is geographically closest to the end user.

Geolocation takes into account only geographical closeness. Therefore it can correctly be used only for selecting service points that are geographically close to the end user. Since network delay is roughly correlated with geographical distance, geographical closeness has been used as approximation for network delay. However, such an approach has low precision and cannot be efficiently used in situations where multiple service points are placed in a close geographical proximity. In addition, geolocation has little use when estimating available bandwidth.

Direct measurement- based systems have been proposed to implement a location service that selects network service points that have minimal communication latency. In such a service, the infrastructure itself performs measurements in reaction to a user query and chooses a network service point that is most suitable to the end user.

Direct measurement systems have however a number of drawbacks. First, they work based on the assumption that the service points can reliably measure their suitability metric for a specific end-user. For the purpose of estimating the latency, utilities such as ping may be used, as long as they are not filtered by a network device between the service points and end-user. However, estimating available bandwidth using measurement systems requires specialized software to be available at the end-user side. In addition, in order to identify the most suitable service point, service points will have to perform measurements towards the end-user. Depending on the specific implementation, this may result in excessive traffic to be sent to the end user and affect its performance, or even worse, affect the measurement results themselves.

The use of network coordinates has also been proposed to be used in implementing a generalized location service whereby k-dimensional coordinates are assigned to routers as well as service points. The Euclidean distance between two points in this coordinate space is an indication of the suitability of two nodes. When an end user requests for a service point, the end user's coordinate is first computed, and then, the service point with the shortest Euclidean distance to the end user is selected.

Network coordinates do not have the drawback of the geolocation-based approach since the coordinates of nodes in the network, i.e. network routers and service points, are computed entirely based on the selected metric of each service point. However, such approaches are known to have accuracy problems. For example, consider a three-node system where latency is the measure of suitability. In this setting, if the delay between two nodes is larger than the sum of their delays to the third node, then the coordinate-based approach fails to correctly represent such a network. In addition, this approach also has little use in estimating available bandwidth.

There is hence a need for a solution addressing these issues as discussed above. SUMMARY

It is an object of exemplary embodiments to address at least some of the issues outlined above, and this object and others are achieved by a network node, a network service point, and methods therein, according to the appended independent claims, and by embodiments of the exemplary embodiments according to the dependent claims.

According to an aspect, the exemplary embodiments provide a method in a network node for establishing parameters of available network resources for connecting to a network via a respective service point. The method comprises receiving from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The method also comprises storing said identity information of said node and said received parameter information. The method also comprises calculating updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the method comprises forwarding said updated parameter information, to nodes that are directly connected to the network node, except for said node.

According to another aspect, the exemplary embodiments provide a method in a network node for identifying a suitable service point for a communication device connecting to said network node, the device being associated with a QoS, criterion. The method comprises receiving from a communication device a request for a network connection associated with a QoS criterion. The method also comprises retrieving parameter information of available network resources for connecting to the network via alternative service points. The method also comprises identifying one or more service points via which the communication device is connectable, and for which the available network resources fulfil the QoS criterion of the device. In addition, the method comprises selecting a service point out of said identified one or more service points, based at least on a second criterion.

According to another aspect, the exemplary embodiments provide a method in a network service point for providing parameter information of available network resources to network nodes directly connected to the network service point. The method comprises monitoring parameters of available network resources to the network service point, and sending parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

According to another aspect, the exemplary embodiments provide a network node that is adapted to receive from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The network node is also adapted to store said identity information of said node and said received parameter information of network resources available to the node. The network node is also adapted to calculate updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and the node. In addition, the network node is also adapted to forward said updated parameter information, to nodes that are directly connected to the network node, except for said node, for establishing parameters of available network resources for connecting to a network via respective service points.

According to another aspect, the exemplary embodiments provide a network node for establishing parameters of available network resources for connecting to a network via respective service points. The network node comprises a processor and a memory, said memory comprising instructions executable by said processor whereby said network node is operative to receive from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The memory also comprises instructions executable by said processor whereby said network node is operative to store said identity information of said node and said received parameter information of network resources available to the node, and to calculate updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the memory comprises instructions executable by said processor whereby said network node is operative to forward said updated parameter information, to other nodes that are directly connected to the network node.

According to another aspect, the exemplary embodiments provide a network node for establishing parameters of available network resources for connecting to a network via respective service point. The network node comprises means that is adapted to receive from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The means is also adapted to store said identity information of said node and said received parameter information. The means is also adapted to calculate updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the means is also adapted to forward said updated parameter information, to other nodes that are directly connected to the network node.

According to another aspect, the exemplary embodiments provide a computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method for establishing parameters of available network resources for connecting to a network via a respective service point. The method comprises receiving from a node that is directly connected to a network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The method also comprises storing said identity information of said node and said received parameter information. The method also comprises determining updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the method comprises forwarding said updated parameter information, to nodes that are directly connected to the network node, except for said node.

According to yet another aspect, the exemplary embodiments provide a network node that is adapted to receive from a communication device a request for a network connection associated with a QoS criterion. The network node is also adapted to retrieve parameter information of available network resources for connecting to a network via alternative service points. The network node is also adapted to identify one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the network node is adapted to select a service point out of said identified one or more service points, based at least on a second criterion, for identifying a suitable service point for the communication device connecting to said network node.

According to yet another aspect, the exemplary embodiments provide a network node for identifying a suitable service point for a communication device connecting to said network node, the device being associated with a QoS criterion. The network node comprises a processor and a memory, said memory comprising instructions executable by said processor whereby said network node is operative to receive from a communication device a request for a network connection associated with a QoS criterion. The memory comprises instructions executable by said processor whereby said network node is operative to also retrieve parameter information of available network resources for connecting to a network via alternative service points, and to identify one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the memory also comprises instructions executable by said processor whereby said network node is operative to select a service point out of said identified one or more service points, based at least on a second criterion.

According to yet another aspect, the exemplary embodiments provide a network node for identifying a suitable service point for the communication device connecting to said network node. The network node comprises means adapted to receive from a communication device a request for a network connection associated with a QoS criterion. The means is also adapted to retrieve parameter information of available network resources for connecting to a network via alternative service points. The means is also adapted to identify one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the means is also adapted to select a service point out of said identified one or more service points, based at least on a second criterion.

According to yet another aspect, the exemplary embodiments provide a computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method for identifying a suitable service point for a communication device connecting to a network node, the device being associated with a QoS, criterion. The method comprises receiving from a communication device a request for a network connection associated with a QoS criterion. The method also comprises retrieving parameter information of available network resources for connecting to the network via alternative service points. The method also comprises identifying one or more service points via which the communication device is connectable, and for which the available network resources fulfil the QoS criterion of the device. In addition, the method comprises selecting a service point out of said identified one or more service points, based at least on a second criterion.

According to still yet another aspect, the exemplary embodiments provide a network service point that is adapted to monitor parameters of available network resources. The network service point is also adapted to send parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point, for providing said parameter information of available network resources to the network nodes directly connected to the network service point.

According to still yet another aspect, the exemplary embodiments provide a network service point for providing parameter information of available network resources to network nodes directly connected to the network service point. The network service point comprises a processor and a memory. The memory comprises instructions executable by said processor whereby said network node is operative to monitor parameters of available network resources. The memory also comprises instructions executable by said processor whereby said network node is operative to send parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

According to still yet another aspect, the exemplary embodiments provide a network service point for providing parameter information of available network resources to network nodes directly connected to the network service point. The network service point comprises means adapted to monitor parameters of available network resources. The means is also adapted to send parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

According to still yet another aspect, the exemplary embodiments provide a computer program, that comprises instructions which, when executed on at least one processor, cause the at least one processor to carry out the method for providing parameter information of available network resources to network nodes directly connected to the network service point. The method comprises monitoring parameters of available network resources to the network service point, and sending parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

Advantages of some embodiments of the present exemplary embodiments are the following:

For a given delay and bandwidth demand, the methods allow each router in the network to learn about the most suitable service point in that network.

It is also advantages that other metrics or parameters than latency and bandwidth can be taken into account such as availability of a specific feature at the network service point or different price classes or a service.

It is also advantageous that the methods require only very limited computation on the network routers.

It can be noted that the method for identifying a suitable service point for a communication device connecting to said network node, can stabilize in 0(d) steps, where d is a diameter of the network. After stabilization, the algorithm has only very limited overhead as long as the network remains stable.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments will now be described in more detail, and with reference to the accompanying drawings, in which:

Figures 1 and 2 schematically illustrate networks having network routers and service points, related to embodiments of the present disclosure; Figure 3 presents a handshake diagram of signalling according to embodiments of the present disclosure;

Figures 4 to 6 present flow charts of methods according to embodiments of the present disclosure;

- Figures 7 and 8 schematically present network nodes according to embodiments of the

present disclosure; and

Figures 9 and 10 schematically present service according to embodiments of the present disclosure. DETAILED DESCRIPTION

In the following description, different embodiments of the exemplary embodiments will be described in more detail, with reference to accompanying drawings. For the purpose of explanation and not limitation, specific details are set forth, such as particular examples and techniques in order to provide a thorough understanding.

This disclosure considers the scenario where a user of a network service may specify, as part of the quality of service (QoS) requirements for the network service, parameter values of required resources, for instance in terms of bandwidth and latency between a service point and the user. Based on the specified parameter values a suitable network service point that is capable of fulfilling the QoS requirements, is identified.

This disclosure comprises a method that is performed in network nodes, such as network routers, whereby each network node learns about the identity of the most suitable service point in the network for a specific combination of two or more parameters. This disclosure also comprises a method for updating parameter values of available network resources. In addition, this disclosure also comprises

Providing parameter information of available network resources to network nodes directly connected to network service points.

Among the number of parameters or metrics, the most important ones that can determine the suitability of a network service point may be the available bandwidth and latency along a network route passing the network service point.

This disclosure provides to a network node a method with which the network node can return a network service point for a required parameter combination, which service point will be able to fulfil the required parameter combination.

Figure 1 schematically illustrates a network having network service points in the form of service point A, 102, service point B, 104 and service point C, 106, as well as a plurality network routers 108 i, i=l, .. , N. A device that is associated with a service with a quality of service (QoS) requiring network resources may connect to any of the network nodes 108i. It is the task to determine a service point to which the device can be connected with fulfilled QoS. Figure 2 schematically illustrates a network comprising a service points A, 202, a network router 1, 204, a network router 2, 206, a network router 3, 208, as well as a service point B, 210. When a device connects to any one of the network routers 1, 2 or 3, this router as one example of a network node, will identify a service point based on the QoS as required by the service of the device. In this simply example it is thus, either service point A, 202, or service point B, 210, which are candidates via which a connection to, for instance, the Internet shall pass.

The present disclosure comprises methods that may be executed in each network node of the network. The present disclosure also comprises a method that may be executed in each service point of said network. These methods may together be considered to form a distributed overall method of the network.

Figure 3 presents a handshake diagram of signalling according to embodiments of the present disclosure. Signalling is herein performed between a service point A, 302; a network router 1, 304; a network router 2, 306; and a service point B, 308. This signalling illustrates an example of establishing parameters of available network resources for connecting to a network via a respective service point.

This example illustrates a scenario when parameter information comprises bandwidth and latency of connections across the network.

In 310, the service point A, 302, monitors the bandwidth and latency of network resources that are available to the service point A. In 312, the service point B, 308, monitors the bandwidth and latency of network resources that are available to the service point B.

Down below, it will be discussed when each service point has a reason to report parameter information.

In 314, the service point A, 302, sends its identity, or identity information with which the identity of service point A, 302 can be revealed, together with information about the bandwidth and latency available to the service point A, 302 to the directly connected network router 1, 304.

In 316, the service point B, 308, sends its identity, or identity information with which the identity of service point B, 308 can be revealed, together with information about the bandwidth and latency available to the service point B, 308 to the directly connected network router 2, 306.

In 314, 316, each service point sends identity information and information about available resources to all directly connected network routers. However, in this example there is only one directly connected network router to each service points A, 302 and B, 308.

In 317 the network router 1, 304 stores the received information on the identity of the service point

A, 302 and the information on the available resources.

In 318 the network router 2, 306 stores the received information on the identity of the service point

B, 308 and the information on the available resources.

In 319, the network router 1, 304, retrieves the bandwidth and latency property of the link between the service point A, 302 and the network router 1, 304. In 320, the network router 2, 306, retrieves the bandwidth and latency property of the link between the service point B, 308 and the network router 2, 306.

It is noted that each network router, or network node is assumed to have a priori knowledge of the properties of all links to other network nodes directly connected said network router.

In 322, the network router 1, 304 calculates updated parameter information of the resources available said network router 1, 304, given the retrieved link information. The bandwidth and latency information as received from the service point A, 302 reflects the resources available to said service point

A, 302. In 322, the network router 1, 304 calculates the resources available to the network router 1, 304 itself.

In 324, the network router 2, 306 calculates updated parameter information of the resources available said network router 2, 306, given the retrieved link information. The bandwidth and latency information as received from the service point B, 308 reflects the resources available to said service point

B, 308. In 324, the network router 2, 306 calculates the resources available to itself.

In 326, the network router 1, 304 forwards information about resources available to itself to directly connected network router 2, 306, as well as identity information of the service point for which this applies; here service point A, 302.

The calculated information which comprises updated latency information can be calculated as the sum of the received latency information and the latency of the link directly connecting the node from which the latency information was received.

Updated bandwidth information can be calculated by the determining the minimum bandwidth of the bandwidth available to the node from which information was received and the bandwidth of the link directly connecting the node from which the information was received. Here, network router 1, 304 sends updated latency information comprising the sum of the latency of the service point A, 302 and the link between the network router 1, 304 and the service point A, 302. Network router 1, 304 also sends updated bandwidth information comprising the minimum of the bandwidth available to the service point A, 302 and the bandwidth of the link between the service point A, 302 and the network router 1, 304.

In addition, in 326, also the identity information of service point A, 302 is also forwarded to the network router 2, 306.

Similarly, in 328, updated parameter information is forwarded by the network router 2, 306 to the network router 1, 304. Here, network router 2, 306 sends updated latency information comprising the sum of the latency of the service point B, 308 and the link between the network router 2, 306 and the service point B, 308. The network router 2, 306 hence also sends updated bandwidth information comprising the minimum of the bandwidth available to the service point B, 308 and the bandwidth of the link between the service point B, 308 and the network router 2, 306.

In addition, in 328, also the identity information of service point B, 308 is also forwarded to the network router 1, 304. In 330, the network router 1, 304 retrieves the bandwidth and the latency property of the link between the network router 1, 304 and the network router 2, 306.

In 332, the network router 2, 306 retrieves the bandwidth and the latency property of the link between the network router 2, 306 and the network router 1, 304.

In 334, the network router 1, 304 can hence calculate updated parameter information.

In 336, the network router 2, 308 can hence calculate updated parameter information.

The network router 1, 304 and network router 2, 306 have thus both parameter information about the available resources in the form of bandwidth and latency for connection to two different service points, service point A, 302 and service point B, 308, respectively.

Down below it will be described how information of parameters, here bandwidth and latency, will be used to select how to establish a connection for a communication device to a network.

Figure 4 presents a flow chart of a method for in a network service point for providing parameter information of available network resources to network nodes directly connected to the network service point. The method comprises monitoring 42 parameters of available network resources to the network service point, and sending 44 parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

The method for providing parameter information of available network resources may further comprise determining whether parameter information of the monitored parameters has changed during a pre-determined time, and wherein sending 44 the parameter information is performed when the parameter information has changed.

The method may comprise monitoring an internally available bandwidth, for example the one of a back bone network of a data center, as well as an internal latency of the network service point.

Initially, i.e. the first time the method is started in the network service point, the method may send parameter information in a message, comprising its identity, an internal latency and an internal bandwidth, to network nodes, i.e. network routers, which the network service point is directly connected to.

Then, whenever these metrics or parameters are found to be change significantly, for example changed by x%, for example by 5 or 10%, which can specified by an administrator, the method will again perform sending the new or updated metrics or parameter information to its neighboring, i.e. directly connected, network routers.

Figure 5 presents a flow chart of a method in a network node 108i, 204, 206, 208, 304, 306 for establishing parameters of available network resources for connecting to a network via a respective service point. The method comprises receiving 52, 314, 316 from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The method also comprises storing 54, 317, 318 said identity information of said node and said received parameter information. The method also comprises determining 56, 322, 324 updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the method comprises forwarding 58, 326, 328 said updated parameter information, to nodes that are directly connected to the network node, except for said node.

Determining updated parameter information within the method, may comprise adding latency information from parameter information from said node, to latency information from parameter information of the link between the network node and said node.

Determining updated parameter information within the method, may comprise determine a minimum of the bandwidth information from parameter information from said node, and the bandwidth information from parameter information of the link between the network node and said node.

Parameter information within determining updated parameter information within the method, may comprise combinations of discrete levels of network resource parameters.

Determining updated parameter information may comprise disregarding inferior combinations of discrete levels of network resource parameters, in favour of superior combinations. This will be further described below.

A network node wherever mentioned here may be a network router.

Each network router may maintain a table like the one shown below in table I, in which the network router can keep track of the service points it knows of and has associated information about. The entry network router id is the identity of the router that reported the latency and bandwidth of for a specific service point. This network router is also a next-hop router, i.e. a directly connected network router, for packets destined to said service point. Reported latency and reported bandwidth are the parameter values reported by said next-hop router. The link latency and link bandwidth are that of the link connecting the current network router with the next hop, or directly connected network router, respectively.

Table I: Example of a table presenting parameter information within a network router According to Table I, illustrating one example only, the current network router is able to offer a connection to service point A, via the directly connected network router Rl . This latency between Rl and the service point A is 50 ms, and bandwidth 100 Mbps, where the link between the current network router and the network router Rl has a latency and a bandwidth of 1 ms and 500 Mbps, respectively.

Similarly, a connection to service point B can be offered via the directly connected, or next-hop, network router R5. The latency between network router R5 and service point B is 10 ms, whereas the bandwidth is 30 Mbps, where the link between the current network router and the network router R5 has a latency and a bandwidth of 1 ms and 150 Mbps, respectively.

Also, connection to service points D and G can be offered, of which both passes the directly connected network router R4. The latency of R4 when connecting to service point D is 5 ms, and the bandwidth is 25, whereas when connecting to service point G, the latency between network router R4 and service point G is 1 ms and the bandwidth is 15 Mbps.

It is assumed that the method in each network router can retrieve or knows the latency as well as the available bandwidth of all links directly connected to it. In addition, the method assumes that these parameter values do not change too frequently, for example, not more often than one or twice a week. In addition, the method assumes that the set of parameters levels used by the method is known a-priori by all network routers in the network.

An example of many possible ways of representing the sets of one dimensional parameter level, is as follows. Let a set of possible latency requirements be given by L={li< 1₂< ... < l_n} and a set of all possible bandwidth demands be given by B={bi> b₂> ... >b_m} . A combination of the two sets provides nxm distinct combinations LxB={(li,bi),(li,b₂), ... ,(l_n, b_m)} .

It is noted that this classification in parameter levels is consistent with how information and communication technology (ICT) services can be sold today. For instance, when purchasing broadband access, a user may select, from a menu of available bandwidths , for example 512 Kbps, 1 Mbps, 5 Mbps, 10 Mbps, and 100 Mbps, instead of specifying the bandwidth that is in fact required, for example 3Mbps.

A service provider may also simply provide a user with the next higher bandwidth, in this case 5 Mbps when 3 Mbps was requested by the end user. The set of combinations that are provided by a service is typically much less than the entire set L x B.

One property of combinations of parameter levels, which this disclosure exploits, is that a service point may serve more than one combination of parameter levels. For example, let latency L = { 1 ms, 5 ms, 10 ms, 50 ms} and bandwidth B = { 100 Mbps, 10 Mbps, 5 Mbps, 1 Mbps, 512 Kbps} . If a network service point satisfies the combination (1ms, 100Mbps), it also satisfies all other less demanding combinations of parameter levels, and hence can be said to serve or cover all other combinations. In this respect { 1 ms, 100 Mbps} is superior to all inferior combinations of this. This property that one combination of parameter levels can cover or serve other combinations, may be called "covering" or "serving", and is an important property for the reason that it allows to drastically reduce the overhead of identifying suitable network service points. For instance, if a network node such as a router knows of a network service point that satisfies (1 ms, 100 Mbps), it does not need to learn about any other network service point, because this service point already covers or serves all other combinations, in this example.

However, when there is no coverage or overlap between two or more combinations of parameter levels, these may be considered to be "uncovered" combinations. For instance, (1 ms, 1 Mbps) and (50 ms, 10 Mbps) are "uncovered" combinations since the first combination (1 ms, 1 Mbps) is preferable when it comes to latency, since 1 ms is a smaller latency than 10 ms, whereas the second combination (50 ms, 10 M bps) is preferred with respect to bandwidth, since 10 Mbps is higher than 1 Mbps.

The number of "uncovered" combinations of a set of combinations, reflects the number of instances that may have to be handled by a network node. For the case in which there are only two parameters, it can be shown that the maximum number of uncovered combinations is given by min(n,m), the size of the smaller of either L or B.

Identifying the number of uncovered combinations may help in minimizing overhead of embodiments of the present disclosure.

In the case of a two dimensional combination, the parameter information that can be provided from a network node to directly connected network node comprises the identity of the service point, a latency value and a bandwidth value for the path via the network to the service point.

According to some embodiments of the present disclosure, when determining whether the parameter information has changed, the method can compare a state of a table with the state of the table at the end of a previous round, i.e. before receiving reported information from any directly connected network router, or service point. The method may then comprise to determine whether a significant change has occurred in the table. Significant changes may be defined as a change that either resulted in one or more new uncovered combinations, resulted in a change of a service point identity in one or more uncovered combinations or resulted in the bandwidth/latency of the service points corresponding to uncovered combinations to change significantly, for instance by more than x%, where the value of x may be configured.

It should be mentioned that the updated parameter information to be forwarded to directly connected network nodes, may be determined as follows. The method can first determine the latency and bandwidth of each service point at the current network router. Latency is, as earlier mentioned, calculated as the sum of the reported latency and the link latency. Bandwidth is calculated as the minimum of the reported bandwidth and the link bandwidth. Then, the method may also identify uncovered service points by comparing the offered combination of each service point, and keeping superior combination to the expense of inferior combinations.

In case multiple service points correspond to the same combination, the method may selects just one service point based on a local policy, for instance, to choose the service point with smallest latency.

It is noted that a table with parameter information, as for instance table I above, can be updated if the identity of the sending network router is listed as a next-hop, or directly connected, network node in the table to reach the service point listed in the information received.

A method in a network router receiving parameter information or notifications about service points may hence comprise determining the bandwidth and latency between itself and the service point, taking into account the link characteristics, i.e. the bandwidth and link latency, between itself and network router that sent the parameter information notification.

If the determined metrics or parameter information makes a service point eligible for some uncovered combination, the method in the receiving network router will note this service point as responsible for the combination and all other combinations covered by this combination. The method in the network router then forwards this information to its directly connected network nodes. This way, information about suitable service points may be sent and learnt by all network routers in the network.

Figure 6 presents a flow chart of a method in a network node 108i, 204, 206, 208, 304, 306 for identifying a suitable service point 102, 104, 106, 202, 210, 302, 308, 90, 1000 for a communication device connecting to said network node, the device being associated with a QoS, criterion. The method comprises receiving 62 from a communication device a request for a network connection associated with a QoS criterion. The method also comprises retrieving 64 parameter information of available network resources for connecting to the network via alternative service points. The method also comprises identifying 66 one or more service points via which the communication device is connectable, and for which the available network resources fulfil the QoS criterion of the device. In addition, the method comprises selecting 68 a service point out of said identified one or more service points, based at least on a second criterion.

The second criterion within the method for identifying a suitable service point may comprise selecting the service point, among the alternative service points, for which the available resources have the highest bandwidth or the smallest latency.

The parameter information within the method for identifying a suitable service point may comprise combinations of discrete levels of network resource parameters.

The present disclosure also comprises a network node 70, 80, 108i, 204, 206, 208, 304, 306 that is adapted to receive 62 from a communication device a request for a network connection associated with a QoS criterion. The network node is also adapted to retrieve 64 parameter information of available network resources for connecting to a network via alternative service points. The network node is also adapted to identify 66 one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the network node is adapted to select 68 a service point out of said identified one or more service points, based at least on a second criterion, for identifying a suitable service point for the communication device connecting to said network node.

The present disclosure also comprises a network node 70, 80, 108i, 204, 206, 208, 304, 306 that is adapted to receive 52 from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The network node is also adapted to store 54 said identity information of said node and said received parameter information of network resources available to the node. The network node is also adapted to determine 56 updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and the node. In addition, the network node is also adapted to forward 58 said updated parameter information, to nodes that are directly connected to the network node, except for said node, for establishing parameters of available network resources for connecting to a network via respective service points.

Figure 7 schematically presents a network node 70 for identifying a suitable service point for a communication device connecting to said network node, the device being associated with a QoS criterion. The network node 70 comprises a processor 72 and a memory 74, said memory 74 comprising instructions executable by said processor 72 whereby said network node is operative to receive 62 from a communication device a request for a network connection associated with a QoS criterion. The memory 74 comprises instructions executable by said processor 72 whereby said network node is operative to also retrieve 64 parameter information of available network resources for connecting to a network via alternative service points, and to identify 66 one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the memory 74 also comprises instructions executable by said processor 72 whereby said network node is operative to select 68 a service point out of said identified one or more service points, based at least on a second criterion.

Figure 7 may alternatively present a network node 70 for establishing parameters of available network resources for connecting to a network via respective service points. The network node 70 comprises a processor 72 and a memory 74, said memory 74 comprising instructions executable by said processor 72 whereby said network node is operative to receive 52 from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The memory 74 also comprises instructions executable by said processor 72 whereby said network node is operative to store 54 said identity information of said node and said received parameter information of network resources available to the node, and to determine 56 updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the memory 74 comprises instructions executable by said processor 72 whereby said network node is operative to forward 58 said updated parameter information, to other nodes that are directly connected to the network node.

Figure 8 schematically presents a network node 80 for identifying a suitable service point for the communication device connecting to said network node. The network node 80 comprises means 82 adapted to receive 62 from a communication device a request for a network connection associated with a QoS criterion. The means 82 is also adapted to retrieve 64 parameter information of available network resources for connecting to a network via alternative service points. The means 82 is also adapted to identify 66 one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device. In addition, the means 82 is also adapted to select 68 a service point out of said identified one or more service points, based at least on a second criterion.

Figure 8 may alternatively present a network node 80 for establishing parameters of available network resources for connecting to a network via respective service point. The network node 80 comprises means 82 that is adapted to receive 52 from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The means 82 is also adapted to store 54 said identity information of said node and said received parameter information. The means 82 is also adapted to determine 56 updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the means 82 is also adapted to forward 58 said updated parameter information, to other nodes that are directly connected to the network node.

The present disclosure also comprises a computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method for establishing parameters of available network resources for connecting to a network via a respective service point. The method comprises receiving 52, 314, 316 from a node that is directly connected to a network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point. The method also comprises storing 54, 317, 318 said identity information of said node and said received parameter information. The method also comprises determining 56, 322, 324 updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node. In addition, the method comprises forwarding 58, 326, 328 said updated parameter information, to nodes that are directly connected to the network node, except for said node. The present disclosure also comprises a computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method 306 for identifying a suitable service point 102, 104, 106, 202, 210, 302, 308, 1000 for a communication device connecting to a network node, the device being associated with a QoS, criterion. The method comprises receiving 62 from a communication device a request for a network connection associated with a QoS criterion. The method also comprises retrieving 64 parameter information of available network resources for connecting to the network via alternative service points. The method also comprises identifying 66 one or more service points via which the communication device is connectable, and for which the available network resources fulfil the QoS criterion of the device. In addition, the method comprises selecting 68 a service point out of said identified one or more service points, based at least on a second criterion.

The present disclosure also comprises a network service point 90, 102, 104, 106, 202, 210, 302, 308, 1000 that is adapted to monitor 42 parameters of available network resources. The network service point is also adapted to send 44 parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point, for providing said parameter information of available network resources to the network nodes directly connected to the network service point.

Figure 9 schematically presents a network service point 90 for providing parameter information of available network resources to network nodes directly connected to the network service point. The network service point 90 comprises a processor 92 and a memory 94. The memory 94 comprises instructions executable by said processor 92 whereby said network node is operative to monitor 42 parameters of available network resources. The memory 94 also comprises instructions executable by said processor 92 whereby said network node is operative to send 44 parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

Figure 10 schematically presents a network service point 1000 for providing parameter information of available network resources to network nodes directly connected to the network service point. The network service point 1000 comprises means 1002 adapted to monitor 42 parameters of available network resources. The means is also adapted to send 44 parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

The present disclosure also comprises a computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method for providing parameter information of available network resources to network nodes directly connected to the network service point. The method comprises monitoring 42 parameters of available network resources to the network service point, and sending 44 parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point. The request or demand of an end user may comprise a two or multi-dimensional combination of parameters levels, where each parameter corresponds to one dimension of resources required for a service. A combination may accordingly define a combination of the maximum latency and minimum bandwidth that is required by a service of an end user when connecting to the service point.

It is noted that the methods as presented in this disclosure may be comprised within a method in a system.

Within a system, each router only exchanges messages with its directly connected network nodes, and no single node in the network has a global knowledge about the set of service points or network routers. The methods are robust and may dynamically adapt to changes in network condition, for instance addition/removal and failure/recovery of network routers, links as well as changing link performance metrics.

Other metrics or parameters may also be taken into account such as availability of a specific feature at the network service point or different price classes or a service.

The methods require only very limited computation on the network routers.

The amount of memory in each network router may be dependent on the number of supported or uncovered combinations.

As mentioned above, a plurality of parameters or metrics may be considered in addition to latency and bandwidth in the identification and selection of service point. This is achieved by extending the number of parameters of the combinations of parameters or metrics. Metrics or parameters that are service-point related, for instance, the availability of specific software and/or hardware, access to specific network functions, cost of using the service, etc., may be included in a straight forward manner. Metrics that change from one network router to another network router may also be included. One provision may be that these metrics can be determined in an incremental manner, and the parameter values are or can be discretized, just like bandwidth and latency are, in the above disclosure.

According to another some alternative embodiments, a single metric or parameter may be used to find the parameter value selection which the service point preferably fulfils. Each network router thus only needs to learn the identity or identity information of only one service point.

It may be further noted that the above described embodiments are only given as examples and should not be limiting to the present exemplary embodiments, since other solutions, uses, objectives, and functions are apparent within the scope of the embodiments as claimed in the accompanying patent claims.

ABBREVIATIONS

API application programming interface

bps bits per second

ICT information and communication technology

Kbps kilo bps

Mbps mega bps

QoS quality of service

Claims

A method in a network node (70, 80, 108i, 204, 206, 208, 304, 306) for identifying a suitable service point (90, 102, 104, 106, 202, 210, 302, 308, 1000) for a communication device connecting to said network node, the device being associated with a quality of service, QoS, criterion, the method comprising:

receiving (62) from a communication device a request for a network connection associated with a QoS criterion;

retrieving (64) parameter information of available network resources for connecting to the network via alternative service points;

identifying (66) one or more service points via which the communication device is connectable, and for which the available network resources fulfil the QoS criterion of the device; and

selecting (68) a service point out of said identified one or more service points, based at least on a second criterion.

The method according to claim 1, wherein the second criterion on which selecting the service point is based, comprises selecting the service point, among the alternative service points, for which the available resources have the highest bandwidth or the smallest latency.

The method according to claim 1 or 2, wherein the parameter information comprises combinations of discrete levels of network resource parameters.

A method in a network node (70, 80, 108i, 204, 206, 208, 304, 306) for establishing parameters of available network resources for connecting to a network via a respective service point, the method comprising:

receiving (52, 314, 316) from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point;

storing (54, 317, 318) said identity information of said node and said received parameter information;

determining (56, 322, 324) updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node; and

forwarding (58, 326, 328) said updated parameter information, to nodes that are directly connected to the network node, except for said node.

The method according to claim 4, wherein calculating updated parameter information, comprises adding latency information from parameter information from said node, and latency information from parameter information of the link between the network node and said node.

The method according to claim 4 or 5, wherein calculating updated parameter information, comprises determining a minimum of the bandwidth information from parameter information from said node, and the bandwidth information from parameter information of the link between the network node and said node.

The method according to any of claims 4 to 6, wherein the parameter information comprises combinations of discrete levels of network resource parameters.

The method according to any of claims 4 to 7, wherein determining updated parameter information comprises disregarding inferior combinations of discrete levels of network resource parameters, in favour of superior combinations.

A method in a network service point (100, 102, 104, 106, 202, 210, 302, 308, 1000) for providing parameter information of available network resources to network nodes directly connected to the network service point, the method comprising:

monitoring (42) parameters of available network resources, and

sending (44) parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

The method according to claim 9, further comprising determining whether parameter information of the monitored parameters has changed during a pre-determined time, and wherein sending the parameter information is performed when the parameter information has changed.

A network node (70, 80, 108i, 204, 206, 208, 304, 306) adapted to receive (62) from a communication device a request for a network connection associated with a QoS criterion; to retrieve (64) parameter information of available network resources for connecting to a network via alternative service points; to identify (66) one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device; and to select (68) a service point out of said identified one or more service points, based at least on a second criterion, for identifying a suitable service point for the communication device connecting to said network node.

12. A network node (70, 108i, 204, 206, 208, 304, 306) for identifying a suitable service point for a communication device connecting to said network node, the device being associated with a QoS criterion, comprising a processor (72) and a memory (74), said memory (74) comprising instructions executable by said processor (72) whereby said network node is operative to receive (62) from a communication device a request for a network connection associated with a QoS criterion; to retrieve (64) parameter information of available network resources for connecting to a network via alternative service points; to identify (66) one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device; and to select (68) a service point out of said identified one or more service points, based at least on a second criterion.

A network node (80, 108i, 204, 206, 208, 304, 306) for identifying a suitable service point for the communication device connecting to said network node, the network node comprising means (82) adapted to receive (62) from a communication device a request for a network connection associated with a QoS criterion; to retrieve (64) parameter information of available network resources for connecting to a network via alternative service points; to identify (66) one or more service points via which the communication device is connectable, and for which the available network resources fulfil a QoS criterion of the device; and to select (68) a service point out of said identified one or more service points, based at least on a second criterion.

A computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method according to any of claims 1 to 3.

15. A network node (70, 80, 108i, 204, 206, 208, 304, 306) adapted to receive (52) from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point; to store (54) said identity information of said node and said received parameter information of network resources available to the node; to determine (56) updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and the node; and to forward (58) said updated parameter information, to nodes that are directly connected to the network node, except for said node, for establishing parameters of available network resources for connecting to a network via respective service points.

A network node (70, 108i, 204, 206, 208, 304, 306) for establishing parameters of available network resources for connecting to a network via respective service points, the network node (70, 108i, 204, 206, 208, 304, 306) comprising a processor (72) and a memory (74), said memory (74) comprising instructions executable by said processor (72) whereby said network node is operative to receive (52) from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point; to store (54) said identity information of said node and said received parameter information of network resources available to the node; to determine (56) updated parameter information of network resources available to the node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node; and to forward (58) said updated parameter information, to other nodes that are directly connected to the network node.

17. A computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method according to any of claims 4 to 8.

18. A network node (80, 108i, 204, 206, 208, 304, 306) for establishing parameters of available

network resources for connecting to a network via respective service point, the network node comprising means (82) that is adapted to receive (52) from a node that is directly connected to the network node, identity information of said node and parameter information of network resources available to said node, for connecting to a network via a specified service point; to store (54) said identity information of said node and said received parameter information; to determine (56) updated parameter information of network resources available to the network node, based on the received parameter information from said node, and on parameter information of a link between the network node and said node; and to forward (58) said updated parameter information, to other nodes that are directly connected to the network node.

19. A network service point (90, 102, 104, 106, 202, 210, 302, 308, 1000) adapted to monitor (42) parameters of available network resources, and to send (44) parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point, for providing said parameter information of available network resources to the network nodes directly connected to the network service point.

20. A network service point (90, 102, 104, 106, 202, 210, 302, 308) for providing parameter

information of available network resources to network nodes directly connected to the network service point, the network service point (90, 102, 104, 106, 202, 210, 302, 308) comprising a processor (92) and a memory (94), said memory (94) comprising instructions executable by said processor (92) whereby said network node is operative to monitor (42) parameters of available network resources, and to send (44) parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.

21. A computer program, comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method according to claim 9 or 10.

22. A network service point (1000, 102, 104, 106, 202, 210, 302, 308) for providing parameter

information of available network resources to network nodes directly connected to the network service point, the network service point (1000, 102, 104, 106, 202, 210, 302, 308) comprising means (1002) adapted to monitor (42) parameters of available network resources, and to send (44) parameter information of the monitored parameters of available network resources to network nodes that are directly connected to the network service point.