WO2022066072A1

WO2022066072A1 - Management of network resources

Info

Publication number: WO2022066072A1
Application number: PCT/SE2020/050890
Authority: WO
Inventors: Rohit Singhal; Rakesh Pandey; Lars Angelin; Markus Persson
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2020-09-23
Filing date: 2020-09-23
Publication date: 2022-03-31
Anticipated expiration: 2023-03-23

Abstract

According to an aspect, there is provided a method of operating an analysis node. The method comprises receiving (701) customer information for a first customer of a first network operator of a communication network. The method further comprises analysing (702) the customer information to determine a customer importance score, CIS, for the first customer.

Description

Management of network resources Technical Field of the Invention This disclosure relates to methods and apparatus for the management of network resources, and in particular relates to the management of network resources in view of an importance of a customer to a network operator. Background of the Invention According to existing processes, an Enterprise Customer (e.g. a video streaming service) negotiates a contract and its terms and conditions (T&Cs) with a Communication Service Provider (CSP), which is also known as a network operator. When the negotiation is concluded, the contract is signed-off and includes a Service-Level Agreement (SLA) for one or more services that the CSP is to provide to the Enterprise Customer (also referred to herein simply as ‘customer’ or ‘enterprise’). Once the services are in operation, the performance of the service by the network operator and/or their communication network is measured and/or monitored for compliance with the SLA. Aspects of the performance that can be measured (and to which the SLA can relate) include throughput, jitter, latency or delay, packet loss, etc. The SLA can include Service Level Objectives (SLO) which provides for the measuring and monitoring of one or more of the above-mentioned service performance metrics against the defined thresholds. The SLA also provides for penalties to the network operator for breaching the agreed SLO/SLA. For example, a video streaming service can agree an SLA with a network operator that requires the network operator to provide certain service levels (e.g. a minimum throughput) to users of the streaming service. Therefore, when users are using the streaming service via the network operator’s communication network (e.g. streaming a video to their smartphones), the network operator should provide the service with at least that minimum throughput in order to comply with the SLA. In existing processes, SLA management can happen as follows: 1. The SLOs are measured and monitored, and if a SLO threshold breach occurs, the breach is reported as a ‘SLA violation’ and the network operator pays the penalty, which can be in the form of service credits. The breach can happen for several reasons. The CSP can ascertain the cause of the breach and subsequently either do nothing until the violation starts to impact their business more significantly, or take a corrective action. 2. Alternatively, the CSP can proactively measure and monitor the SLOs using soft thresholds, a static rule-based algorithm or artificial intelligence (Al)Zmachine learning (ML) based predictions, and when alarms are raised of a possible breach of a SLO threshold, the cause can be ascertained and proactive action can be taken to manage the network and avoid the breach. However if the breach still happens and the SLA is violated then the violation is reported and the penalty paid.

Summary

These two practices of SLA management can have the following problems:

1 . SLA management in point 1 above is passive and reactive in nature. The SLA can be violated without any warning.

2. The corrective action is taken manually, based on the knowledge (documented or undocumented) of operations personnel managing the network.

3. The corrective action usually entails either fixing problems if there is a network resource breakdown or adding more network resources when resource constraints are impairing the service performance. If there are no free resources available in the network and network capacity expansion is not planned in advance, then there can be a much-delayed response to breaches of the SLA.

4. Most existing CSP networks have static network resource allocation to services or customers, so they lack the ability to perform dynamic resource allocation. A dynamic resource allocation can help allocate resources from services or customers where there is ‘slack’ (i.e. resources allocated to services or customers that are currently unused) to the services or customers where there is a shortage. Hence, currently network resources are used non- optimally.

5. Some CSPs have built elasticity into their network infrastructure, however they still lack ways to determine an optimal re-provisioning of network resources from services or customers with a surplus of resources to services or customers where there is a shortage.

Therefore there is a need for improvements in the management of network resources for customers by a network operator.

According to a first aspect, there is provided a method of operating an analysis node. The method comprises receiving customer information for a first customer of a first network operator of a communication network; and analysing the customer information to determine a customer importance score, CIS, for the first customer. This aspect provides a metric that enables the importance of customers to the first network operator to be quantified, enabling appropriate decisions to be made about resource allocation to that customer or other customers.

According to a second aspect, there is provided a method of operating a recommendation node. The method comprises receiving prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtaining a customer importance score, CIS, for the first customer; and determining, based on the CIS forthe first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service. This aspect provides that recommendations for resource reallocation can be determined based on the importance of the first customer to the first network operator, and therefore resource reallocation recommendations can be determined to satisfy the business objectives of the first network operator. In addition, determining recommendations in this way (if the recommendations are subsequently adopted) can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator.

According to a third aspect, there is provided a method of operating a network orchestration node in a communication network. The method comprises receiving, from a recommendation node, a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determining whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and if it is determined to allocate the network resources, allocating those resources to the first service provided to the first customer. This aspect provides that a decision can be made on whether to follow a recommendation for resource reallocation to improve the network experience of certain customers. This is an important factor in churn reduction and can increase customer retention forthe first network operator. In addition, deciding that resources can be reallocated from one customer to another in this way can avoid the need forthe first network operator to increase the capacity of the communication network (i.e. avoid capital expenditure).

According to a fourth aspect, there is provided a method of operating a network orchestration node in a communication network. The method comprises receiving a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtaining a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determining whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and if it is determined to grant the resource request, allocating those resources to the user of the first customer. This aspect provides a mechanism to prioritise resource requests when the network does not have the capacity or capability to process all of them. Priority for a request from a particular customer can be based on the importance of that customer to the first network operator and/or a consequence of prioritising other customers over that customer, and therefore resource prioritisation can be determined to satisfy the business objectives of the first network operator. In addition, determining resource prioritisation in this way can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator.

According to a fifth aspect, there is provided a computer program product comprising a computer readable medium having computer readable code embodied therein, the computer readable code being configured such that, on execution by a suitable computer or processor, the computer or processor is caused to perform the method according to any of the first aspect, second aspect, third aspect, fourth aspect, or any embodiments thereof.

According to a sixth aspect, there is provided an analysis node. The analysis node is configured to: receive customer information for a first customer of a first network operator of a communication network; and analyse the customer information to determine a customer importance score, CIS, for the first customer. This aspect provides a metric that enables the importance of customers to the first network operator to be quantified, enabling appropriate decisions to be made about resource allocation to that customer.

According to a seventh aspect, there is provided a recommendation node. The recommendation node is configured to receive prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtain a customer importance score, CIS, for the first customer; and determine, based on the CIS for the first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service. This aspect provides that recommendations for resource reallocation can be determined based on the importance of the first customer to the first network operator, and therefore resource reallocation recommendations can be determined to satisfy the business objectives of the first network operator. In addition, determining recommendations in this way (if the recommendations are subsequently adopted) can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator.

According to an eighth aspect, there is provided a network orchestration node for use in a communication network. The network orchestration node is configured to: receive, from a recommendation node, a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determine whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and, if it is determined to allocate the network resources, allocate those resources to the first service provided to the first customer. This aspect provides that a decision can be made on whether to follow a recommendation for resource reallocation to improve the network experience of certain customers. This is an important factor in churn reduction and can increase customer retention for the first network operator. In addition, deciding that resources can be reallocated from one customer to another in this way can avoid the need for the first network operator to increase the capacity of the communication network (i.e. avoid capital expenditure).

According to a ninth aspect, there is provided a network orchestration node for use in a communication network. The network orchestration node configured to: receive a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtain a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determine whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and, if it is determined to grant the resource request, allocate those resources to the user of the first customer. This aspect provides a mechanism to prioritise resource requests when the network does not have the capacity or capability to process all of them. Priority for a request from a particular customer can be based on the importance of that customer to the first network operator and/or a consequence of prioritising other customers over that customer, and therefore resource prioritisation can be determined to satisfy the business objectives of the first network operator. In addition, determining resource prioritisation in this way can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator.

According to a tenth aspect, there is provided an analysis node. The analysis node comprises a processor and a memory, said memory containing instructions executable by said processor whereby said analysis node is operative to receive customer information for a first customer of a first network operator of a communication network; and analyse the customer information to determine a customer importance score, CIS, for the first customer. This aspect provides a metric that enables the importance of customers to the first network operator to be quantified, enabling appropriate decisions to be made about resource allocation to that customer.

According to an eleventh aspect, there is provided a recommendation node, the recommendation node comprising a processor and a memory, said memory containing instructions executable by said processor whereby said recommendation node is operative to receive prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtain a customer importance score, CIS, for the first customer; and determine, based on the CIS for the first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service. This aspect provides that recommendations for resource reallocation can be determined based on the importance of the first customer to the first network operator, and therefore resource reallocation recommendations can be determined to satisfy the business objectives of the first network operator. In addition, determining recommendations in this way (if the recommendations are subsequently adopted) can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator. According to a twelfth aspect, there is provided a network orchestration node for use in a communication network. The network orchestration node comprises a processor and a memory, said memory containing instructions executable by said processor whereby said network orchestration node is operative to receive, from a recommendation node, a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determine whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and, if it is determined to allocate the network resources, allocate those resources to the first service provided to the first customer. This aspect provides that a decision can be made on whether to follow a recommendation for resource reallocation to improve the network experience of certain customers. This is an important factor in churn reduction and can increase customer retention for the first network operator. In addition, deciding that resources can be reallocated from one customer to another in this way can avoid the need for the first network operator to increase the capacity of the communication network (i.e. avoid capital expenditure).

According to a thirteenth aspect, there is provided a network orchestration node for use in a communication network. The network orchestration node comprises a processor and a memory, said memory containing instructions executable by said processor whereby said network orchestration node is operative to: receive a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtain a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determine whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and, if it is determined to grant the resource request, allocate those resources to the user of the first customer. This aspect provides a mechanism to prioritise resource requests when the network does not have the capacity or capability to process all of them. Priority for a request from a particular customer can be based on the importance of that customer to the first network operator and/or a consequence of prioritising other customers over that customer, and therefore resource prioritisation can be determined to satisfy the business objectives of the first network operator. In addition, determining resource prioritisation in this way can improve the experience of customers that have a higher importance to the first network operator, which is an important factor in churn reduction and increase customer retention for the first network operator.

Brief Description of the Drawings

Various embodiments are described herein with reference to the following drawings, in which:

Fig. 1 is a signalling diagram illustrating the collection of information for determining a customer importance score (CIS) according to various embodiments;

Fig. 2 is a diagram illustrating a Bayesian Network for calculating a CIS;

Fig. 3 illustrates an exemplary CIS calculation;

Fig. 4 illustrates management of network resources according to various embodiments;

Fig. 5 illustrates a virtualisation environment that can be used to implement one or more of the nodes described herein according to various embodiments;

Fig. 6 is a simplified block diagram of an apparatus that can be used to implement any of the nodes described herein;

Fig. 7 is a flow chart illustrating a method of operating an analysis node according to various embodiments;

Fig. 8 is a flow chart illustrating a method of operating a recommendation node according to various embodiments;

Fig. 9 is a flow chart illustrating a method of operating a network orchestration node according to various embodiments; and

Fig. 10 is a flow chart illustrating another method of operating a network orchestration node according to various embodiments.

Detailed Description of the Preferred Embodiments

Some of the embodiments contemplated herein will now be described more fully with reference to the accompanying drawings. Other embodiments, however, are contained within the scope of the subject matter disclosed herein, the disclosed subject matter should not be construed as limited to only the embodiments set forth herein; rather, these embodiments are provided by way of example to convey the scope of the subject matter to those skilled in the art. In some instances, detailed descriptions of well-known methods, nodes, interfaces, circuits, and devices are omitted so as not obscure the description with unnecessary detail. Those skilled in the art will appreciate that the functions described may be implemented in one or more nodes using hardware circuitry (e.g., analog and/or discrete logic gates interconnected to perform a specialized function, ASICs, PLAs, etc.) and/or using software programs and data in conjunction with one or more digital microprocessors or general purpose computers. Nodes that communicate using the air interface also have suitable radio communications circuitry. Moreover, where appropriate the technology can additionally be considered to be embodied entirely within any form of computer-readable memory, such as solid-state memory, magnetic disk, or optical disk containing an appropriate set of computer instructions that would cause a processor to carry out the techniques described herein.

Hardware implementation may include or encompass, without limitation, digital signal processor (DSP) hardware, a reduced instruction set processor, hardware (e.g., digital or analog) circuitry including but not limited to application specific integrated circuit(s) (ASIC) and/or field programmable gate array(s) (FPGA(s)), and (where appropriate) state machines capable of performing such functions.

In terms of computer implementation, a computer is generally understood to comprise one or more processors, one or more processing units, one or more processing modules or one or more controllers, and the terms computer, processor, processing unit, processing module and controller may be employed interchangeably. When provided by a computer, processor, processing unit, processing module or controller, the functions may be provided by a single dedicated computer, processor, processing unit, processing module or controller, by a single shared computer, processor, processing unit, processing module or controller, or by a plurality of individual computers, processors, processing units, processing modules or controllers, some of which may be shared or distributed. Moreover, these terms also refer to other hardware capable of performing such functions and/or executing software, such as the example hardware recited above.

As noted above, there is a need for improvements in the management of network resources by a network operator for customers of the network operator. In particular, improvements in the management of network resources is required where a service level agreement (SLA) between a customer and the network operator is at risk of being breached or is being breached.

The solution described herein is provided in the context of an enterprise customercontract for a 5^th Generation (5G) network (NW) slice product. An enterprise customer (which is also referred to simply as ‘customer’ or ‘enterprise’ herein) is an entity (e.g. a company) that provides a service, e.g. media streaming or video conferencing, with that service being accessible by users via the communication network of the network operator. The enterprise customer is therefore a customer of the network operator. It will be appreciated that users (e.g. people and/or other enterprises) wishing to use the service provided by the enterprise may enter into their own contracts (e.g. a subscription) with the enterprise. Some of those people and/or other enterprises may also be customers of the network operator (e.g. they may have smartphone usage plans with the network operator), and therefore those people and/or enterprises can use their devices (e.g. smartphones, tablets, etc.) to access the service provided by the enterprise via the network operator’s communication network. The term “users” is used herein to refer to at least these people and/or other enterprises that access the service provided by the enterprise via the network operator’s communication network.

It is assumed that the NW slice is not designed to have fixed network resources for the duration of the contract as any fixing of network resources for the contract duration may create a conflict situation for managing SLAs. Assuring service performance in line with SLAs and the techniques described herein requires network resources to be allocated in an elastic manner.

In a particular embodiment of the proposed solution, it is suggested that when there is a possibility of SLA violation (i.e. breach of a term or condition of the SLA) for a customer or set of customers, network resource re-provisioning or re-balancing is performed in such a way that the CSP’s (network operator’s) business objectives can be taken into account, and in some cases, maximised.

The solution accounts for the network operator’s business objectives by establishing the importance of certain customers, or relative importance of certain customers to the network operator and/or to the network operator’s business.

The customer importance can be ascertained based on a number of different factors or parameters. The factors or parameters can include, for example: a customer life-time value, which includes annual revenue per unit or customer (ARPU) and APRU growth rate, the number of users the customer has, the growth rate of the number of users of customer’s service in the communication network, the amount of service used by the users of the customer’s service the possibility of the customer establishing a contract with a different network operator, or otherwise not renewing the service contract with the network operator, and the length of time that the customer has been a customer of the network operator; the customer’s current contract and T&Cs, including the softness or harshness of SLAs, i.e. the level of thresholds and the associated value or cost of penalties for breaching the thresholds, information about previous contracts between the network operator and the customer; the network capacity used in relation to generated invoices, i.e. the price per unit of network resource used by users of the enterprise; the customer’s invoice payment behaviour; customer journeys and interactions data in a customer relationship management (CRM) system and other relevant systems, such as enquiries, negotiations, purchases and complaints, etc.; other customer data such as information about the industry in which the customer operates (including the growth rate of that industry), the customer’s market ranking, financial information about the customer (e.g. including revenue, profits, growth rate, funding models such as debt/equity ratio, venture capital (VC) level funding, credit rating, etc.), social media information about the customer, e.g. popularity on social media of the customer and its products/services.

The customer importance can be expressed as a customer importance score (CIS), and can be calculated at the time the customer is onboarded with the network operator (e.g. when the SLA is first established). The CIS can be refreshed periodically as more data about the customer is available or collected.

The above data about a customer can be analysed to determine a CIS forthat customer. In some embodiments, AI/ML technology can be used to correlate the above data to determine a CIS. The CIS can subsequently be used in SLA management. The CIS for a customer can be a value from a discrete range of values (e.g. there could be values representing ‘high importance’, ‘medium importance’, ‘low importance’, etc.), or the CIS can be a value in a continuous numerical range.

The calculated CIS can be stored by the network operator as a customer attribute in their customer information database.

Some embodiments of SLA management using a CIS are outlined below. As noted above, the re-provisioning of network resources is required to be done dynamically. Certain embodiments also require that it is possible for the system to have the capability to predict that a SLA violation may occur, which can trigger the network to consider re-balancing or reprovisioning network resources.

A recommendation node (also referred to herein as a recommendation engine (RE)) can provide a recommendation for a new network state to manage SLAs based on the CIS of any customers whose SLA might be violated based on the prediction(s); the CIS of all other customers of the network operator; SLA thresholds and the penalties for breaching those thresholds for all customers; a weightage factor of the geography of the communication network (e.g. that takes into account geographical factors that affect the network, such as (relative) popularity of the customer in a particular area); a current state of the communication network and network resource utilisation by all services and customers; any available network resource inventory (i.e. an inventory of network resources that are available for allocation); any previous violations that have occurred and the associated penalties.

In addition to providing a recommendation (or not) to allocate network resources that are in use by other customers or users of the communication network to the service provided to the customer whose SLA might be breached, the RE may also provide an indication of cost savings with respect to new state of network vis-a-vis the penalty cost for violating the SLA if no action is taken.

It will be appreciated that the new network state (i.e. if resources were reallocated) may result in SLA violations for other customers of the network operator due to those other customers now having a lack of sufficient network resources. In this case, the RE, or a network orchestration node that determines whether to follow the recommendation, can determine or take into account other penalties that may be incurred by following the recommendation.

It may be the case that cost savings are a negative value in some SLA measurement periods, however the RE provides a positive aggregated value in the financial period (assuming SLA measurement period to be either equal or less than the financial period).

The RE may also provide an indication of a cost of re-provisioning the network resources, including the risks. The cost of re-provisioning the network resources which impacts only a small number of resources may be less than a cost of a large number of resources. One way to ascertain the cost is to estimate the risk-mitigating actions that are required and probability of risk realisation weighted-costs.

In some embodiments, the RE can use Al or ML technology to determine the recommendations.

A network orchestration node (also referred to herein as a “service orchestrator”) can receive the recommendation of the new network state (new resource configuration), the cost savings due to the new network state and the cost of re-provisioning the network resources, and determine whether to follow the recommendation, and if so, to implement or effect the reprovisioning of resources. In some embodiments, the network orchestration node can make use of Al or ML technology to take the decision on whether to go ahead with the recommendation from the RE or not.

The above approach to network resource management can provide a number of advantages. In particular, the approach enables the network to be in a state that maximises the business objectives of the network operator in the event that there are limited network resources available. For example the network can strive to manage the network in a dynamic way in line with business objectives to reduce or delay capital expenditure (CAPEX) to expand network capacity; to maximise the service experience of customers having a higher CIS, which is an important factor in churn reduction (i.e. the customers contracting with another network operator) and thereby increasing the customer ‘stickiness’ and value to the network operator over the lifetime of the service contract; and minimising the overall penalties paid out by the network operator due to SLA violations across all customers, reduces costs and thereby increases profits.

Figs. 1-3 relate to embodiments of determining a CIS according to the techniques described herein. Fig. 1 is a signalling diagram showing the signalling for a number of nodes operated or managed by the network operator in determining a CIS. Fig. 1 shows a CRM and Order Management (OM) node 101 , a CIS node 102 (also referred to as an “analysis node” herein), a charging and billing node 103, a social media information database 104, a product and service catalogue database 105, a service orchestrator node 106, a service assurance node 107 and an external source 108 (i.e. external to the communication network of the network operator). The CIS node 102 determines the CIS using information obtained from one or more of the other nodes and databases shown in Fig. 1. Although nodes/databases 101- 107 are shown in Fig. 1 as separate entities, it will be appreciated that multiple ones of the nodes/databases can be implemented by or within a single node.

Signals/steps 121-127 in Fig. 1 represent the derivation of the CIS for a customer when they first establish a contract which includes a SLA with the network operator. Thus the CIS can be calculated while ‘onboarding’ the customer, which means that operational data for the customer or their users (such as charging and billing data or service performance data) is not available, and so is not included in the CIS calculation. Signals/steps 128-133 in Fig. 1 represent the updating of the CIS for a customer once operational data about the customer and their users is available. The refreshing or updating of the CIS is calculated according to a configured frequency where operational data is now available. For example, payment data relating to the customer can be available from the 2^nd month of operations. SLA violation data will only be available if there is a prediction that there will be a SLA violation, or if SLAs have been violated. When SLAs are violated, penalty data is also available.

The data or information that can be used to determine the CIS can be divided into three groups based on how quickly that data or information can change during the life of the contract between the customer and the network operator). For example there is immutable data such as contracts or products/services, slow moving data such as invoices or payments by the customer to the network operator, and fast moving data such as usage data by users.

Thus, in Fig. 1 , the CIS node 102 can receive information about the customer and the contract information from the CRM node 101 (signal 121). The CIS node 102 can then calculate the CIS (step 122) using any of: information about the popularity of the customer on social media (social media information 123 received from social media information database 104), information about the product or services provided by the customer (product/service information 124 received from the product and service catalogue database 105), information about the customer received from external source(s) (customer external data 125 received from the external source 108, and/or network capacity/resource information (network capacity information 126 received from the service orchestrator node 106.

Once determined, the CIS can be stored as an attribute of the customer in the CRM 101 (signal 127).

The CIS node 102 can receive a report from the service assurance node 107 (signal 128) indicating the status of the SLA with the customer, and information about any violations of the SLA. When the CIS is to be updated (step 129), the CIS node 102 can receive further information about the customer and their users that can be taken into account when updating the CIS. Thus, the CIS node 102 can receive any of: charging and billing information 130 for the customer and their users from the charging and billing node 103, updated social media popularity information 131 from social media information database 104, updated network capacity/resource information 132 from the service orchestrator node 106, and updated external information 133 about the customer received from external source 108.

Once the CIS is updated, the updated CIS can be stored in the CRM 101 (signal 134).

The process in Fig. 1 can be performed for each customer of the network operator.

In some embodiments, the CIS node 102 can use an Al or ML model to determine the CIS for each customer. As noted above, the CIS can be value from a discrete set of values, or a value in a continuous range of values. There are multiple ways in which AI/ML models can be used to derive a CIS, and an example is provided below.

In this approach, a Bayesian Network (BN) is used to determine the CIS. In this approach each variable has assigned states (which in the illustrated example is high, medium, or low for simplicity, although other labels or states can be used). An inference algorithm runs over the BN with the observed data value (the actual value received as an input) and computes the edge value (i.e. the CIS).

Fig. 2 illustrates an exemplary Bayesian Network for calculating a CIS. In order to further simplify the inputs to the BN, the number of variables can be reduced by pre-computation based on prior knowledge using several methods. For instance:

Businesscredentials 201 (for example credit risk rating (CRR)) relates to the (financial) strength of the customer and can be derived from on existing knowledge of CRR. CRR data is available from various rating agencies for rating companies, and is used to train a classification algorithm that is used to determine a Businesscredential classification when each customer is onboarded and for subsequent updates of the CIS. Some exemplary features or parameters that can be taken into account in the CRR/BusinessCredentials calculation can be the past revenues of the customer, the profits, the growth in revenue and profits, and debt-equity ratio.

FYRevenuelmportance 202 relates to the financial importance derived from the revenue from the customer to the network operator and can be calculated using domain knowledge as a combination of existing Average revenue per account (ARPA) and growth in ARPA. For example, the revenue importance 202 can be calculated as shown in Table 1 below:

Table 1

LifeTimeDuration 203 relates to the amount of time that the customer has been, and might continue to be, associated with the network operator (i.e. how long an agreement has been in place for). Thus, LifeTimeDuration 203 can be determined from AgeOnNetwork (i.e. how long the customer has been on the network) and the ChurnRate (i.e. how long is it before the customer is forecasted to churn).

SocialNetworkPopularity 204, UsageGrowth 205 and UserGrowth 206 are all direct value categories, i.e. they are not calculated or derived. The classification value is directly preassigned based on observation. Thus, these values can be directly received by the CIS node 102, such as in signal 123 (social network popularity 204/231) and in signals 128 and 130.

Some other observed variables include InvoicePaymentBehaviour 207, which relates to the timeliness of invoice payments by the customer and other behaviour relating to financial transactions with the network operator, and is, e.g., received in signal 130, a SLAViolationAndComplaints 208, which relates to violations or potential violations of the SLA by the network operator, and is, e.g., received in signal 128, and a SLAThresholdPenaltyRatio 209, which relates to a ratio of a threshold used to trigger a penalty and the penalty itself, and is, e.g., received in signal 128. As an example of the SLAThresholdPenaltyRatio, each potential SLA violation has a threshold, e.g. for jitter the jitter threshold should be less than 10 ms. If the jitter is greater than 10 ms than there may be a penalty of a 5% discount applied to an invoice from the network operator to the customer. The lower the threshold the higher the more stringent the SLA. The higher the penalty the higher the more stringent the SLA. Therefore the threshold/penalty ratio can provide useful information on the importance of a customer.

The affinity of nodes can be used to create subsequent layers and thus build the entire BN incrementally. The layers are created for convenience to handle various combinations of observed values to compute intermediate derived variables. For example, the Businesscredentials 201 and the FYRevenuelmportance 202 can be combined to determine a measure of the strength of the customer’s business, Businessstrength 210; the Businessstrength 210 and the InvoicePaymentBehaviour 207 can be combined to determine a measure of the financial importance of the customer to the network operator, Financiallmportance 211 ; the LifeTimeDuration 203 and the SocialNetworkPopularity 204 can be combined to determine a measure of the popularity and loyalty of the customer to the network operator, LoyaltyAndPopularity 212; the LoyaltyAndPopularity 212 and the SLAViolationAndComplaints 208 can be combined to determine a satisfaction score that represents how satisfied the customer is with the network’s services and how satisfied the users of the customer are with the service, Satisfactionscore 213; the UsageGrowth 205 and the UserGrowth 206 can be combined to determine a measure of the growth prospects for the customer, Growth Prospects 214.

Once the BN is ready, an inference algorithm can be run to derive the value of the CIS (CustomerlmportanceScore 215).

Fig. 3 shows an exemplary CIS calculation based on the BN structure shown in Fig. 2. In this example, Businesscredentials 201 are ‘high’, the FYRevenuelmportance 202 is ‘high’, LifeTimeDuration 203 is ‘long’, the SocialNetworkPopularity 204 is ‘high’, the UsageGrowth 205 is ‘positive’, the UserGrowth 206 is ‘positive’, the InvoicePaymentBehaviour 207 is ‘good’, a ‘hardly’ for SLAViolationAndComplaints 208, and the SLAThresholdPenaltyRatio 209 is ‘soft’. These all lead to a ‘high’ Businessstrength 210, a ‘high’ Financiallmportance 211 ; a ‘high’ LoyaltyAndPopularity 212, a ‘high’ Satisfactionscore 213, a ‘positive’ for GrowthProspects 214. Finally, these combine to provide a ‘high’ CIS 215 for the customer.

The signalling diagram in Fig. 4 illustrates the management of network resources according to various embodiments of the techniques described herein. Fig. 4 shows the signalling for a number of nodes operated or managed by the network operator to manage resources in the network operator’s communication network. Fig. 4 shows a CRM node 401 , a charging and billing node 402, a service assurance (SA) and SLA Management node 403, a SA SLA Violation Prediction Service 404 (also referred to as a ‘prediction node’ herein), a SA Recommendation Engine (RE) 405 (also referred to as a ‘recommendation node’ herein), a NW inventory node 406 and a service orchestrator node 407 (also referred to as a ‘network orchestration node’ herein). Although nodes 401-407 are shown in Fig. 4 as separate entities, it will be appreciated that multiple ones of the nodes can be implemented by or within a single node.

In a first step, the SLA Violation Prediction Service 404 obtains service performance data (signal 421) for the customers of the network operator from the SA and SLA Management node 403. The SA and SLA Management node 403 is a central system that gathers all of the service performance data for the communication network and the customers of the network.

In step 422, the SLA Violation prediction service 404 analyses (e.g. measures and monitors) the service performance data. As part of the analysis, the prediction service 404 calculates if any SLAs are being violated, and predicts if any SLAs are likely to be violated soon (step 423). For example the prediction service 404 can determine if degradation of the service provided to a customer will occur that will breach or violate the SLA between that customer and the network operator. Techniques for predicting the violation of SLAs are known to those skilled in the art, and are not described in detail herein. For example, a suitable technique is described in the paper “Predicting SLA Violations in Real Time using Online Machine Learning” by Jawwad Ahmed et al. (available at: https://arxiv.org/abs/1509.01386). Other techniques are outlined in the Background section, such as the use of soft thresholds, static rule-based algorithms or AI/ML-based predictions.

If a SLA for a particular customer (denoted a ‘first customer’ herein) is being violated or is likely to be violated, a trigger event is communicated to the RE 405 (signal 424). The trigger event can indicate the service that is being degraded (e.g. a minimum throughput for a video streaming service). The RE 405 operates to determine a recommendation for a new network state to manage the SLA violation for the first customer.

The RE 405 can obtain an indication of the resources available in the network by sending a request 425 to the NW inventory 406. The RE 405 can also obtain the CIS for the first customer, and optionally also the other customers of the network operator, by sending a request 426 to the CRM 401 .

In step 427 the RE 405 determines a recommendation to address the indicated SLA violation. In some embodiments, in determining the recommendation the RE 405 can consider any or all of the following three options.

Firstly, the RE 405 can determine from the resource information obtained from the NW Inventory node 406 if the resources required to address or prevent the SLA violation are available for the service of the first customer being degraded. Resources that are ‘available’ in the present case are resources that are not allocated to any services or to users of a particular customer. If sufficient resources are available, then the RE 405 sends a recommendation (signal 428) to the service orchestrator 407 for a new network state to be adopted in which the additional resources are provided to the service provided by the first customer. Thus, in this case, the recommendation is straightforward to determine as there are additional resources available in the network. The cost of re-provisioning the network is also minimal, and no SLA violation will occur. It will be noted that there is no need to use the CIS in this situation.

Secondly, if the RE 405 determines from the resource information obtained from the NW Inventory node 406 that the resources required to address or prevent the SLA violation are not available, then the RE 405 considers whether there are network resources already allocated to other services or other customers of the network operator that are not being used (or not being used yet), and determines whether those resources can be used to prevent the SLA violation for the first customer. These allocated but unused resources can be referred to as the slack in the network (where slack = the allocated network resources - the utilised network resource). If there are sufficient resources available in the slack, then the RE 405 sends a recommendation (signal 428) to the service orchestrator 407 for a new network state to be adopted in which additional resources are taken from the slack of one or more other customers or users of the communication network and are provided to the service provided by the first customer. In this situation the cost of re-provisioning the network is also minimal, and no SLA violation should occur. Again, there is no need to use the CIS in this situation.

Thirdly, if the RE 405 determines that the resources required to address or prevent the SLA violation are not available from the pool of unallocated resources or from the slack, then the RE 405 uses the CIS fetched in signal 426 from the CRM 401 to prioritise the allocation of network resources. In particular, the RE 405 determines if network resources that are in use by other customers or users of the communication network should be allocated to the service provided to the first customer to address the predicted SLA violation. For example, the RE 405 can determine from the CIS of the first customer (and optionally also from the CISs of the other customers) that the first customer is sufficiently important to the network operator (or the consequences of breaching the SLA are sufficiently severe) that resources should be prioritised for the first customer, even if this is at the expense of the level of service provided to another customer (who may be less important to the network operator and/or for whom the penalties for breaching the SLA are less severe). Assuming that this recommendation is subsequently adopted by the network, this approach may lead to an SLA violation for customers other than the first customer (i.e. other than the customer that was the subject of the initial SLA violation prediction. Table 2 below provides a simplified illustration of priority recommendation by the RE 405 for the allocation of network resources to a particular customer, based only on two input parameters, the CIS and size of the penalty for breaching the SLA.

Thus, according to above illustration, customers or services with the lowest SLA penalties and the lowest CIS would always have the lowest priority in terms of the allocation of network resources to that customer or service. On the other hand, customers or services with the highest SLA penalties and the highest CIS would always have the highest priority in terms of the allocation of network resources to that customer or service.

In some embodiments, in step 427 the RE 405 can evaluate a set of rules or conditions to determine an appropriate recommendation in response to a degradation prediction. However, in other embodiments, in step 427 the RE 405 can use a trained AI/ML model to determine the revised allocation of network resources to services and customers.

In some embodiments, the RE 405 can comprise a knowledge base (e.g. similar to a Bayesian network of belief) and a reasoner (an inference algorithm).

The reasoner analyses the prediction received from the prediction node 404 and builds a path to transition from the current network state to a desired network state for each prediction of service degradation. The RE 405 provides a probability of success for each of the recommendations. The RE 405 can also calculate the cost aspects, i.e. a cost saving and a cost of implementation of the desired network’s state in-line with a technical and business perspective. To arrive at the recommended set of actions, the RE 405 can filter out not required or unfeasible recommendations by consulting the knowledge base.

The knowledge base can comprise any of: information relating to predictions of service degradation, information relating to root causes of service degradation, CIS values of all customers, information about the current network state in the current geography, information about past SLA violations for the customers, information about target solution options for each root cause, information relating to past recommendations and actions on past recommendations, information relating to SLA violation pay-outs based on the CIS-based process (e.g. in the year to date/the current financial year), and information relating to SLA violation “possible" pay-outs based on old/existing process (e.g. in the year to date/the current financial year).

Various techniques for implementing knowledge bases and inference algorithms will be known to those skilled in the art, and further details are not provided herein. Nevertheless, exemplary techniques that can be used include first order logic, forward chaining, backward chaining, etc.

Once the recommendation is determined, it is communicated to the service orchestrator 407 (signal 428), together with an indication of the cost of re-provisioning the network resources, and/or an indication of cost savings by avoiding the SLA violation.

Based on the received recommendation from the RE 405, the cost of re-provisioning, and the cost savings by avoiding the SLA violation, the service orchestrator 407 determines in step 429 whether to follow the recommendation and re-provision the network resources from one or more other customers to the first customer. The service orchestrator 407 therefore provides a ‘go’ or ‘no go’ decision with respect to the received recommendation.

In some embodiments, in step 429 the service orchestrator 407 can evaluate a set of rules or conditions to determine whether a received recommendation can be followed. However, in other embodiments, in step 429 the service orchestrator 407 can use a trained AI/ML model to determine whether to follow the received recommendation.

In embodiments where the RE 405 comprises a knowledge base and inference algorithm, as part of the knowledge base and inference logic the recommendation engine 405 can provide a recommendation of a new desired network state for affected customers. The service orchestrator 407 is required to take a decision on whether to implement the recommendation given the cost savings and cost of implementation. The above decision can be made statically, for example as a ratio of cost-saving to cost of implementation - the higher the ratio the better it is, with a threshold value below which the recommendation is automatically rejected. Alternatively, the decision can be made using Al technology, such as reinforced learning. For example, a Markov decision process (MDP) can be used to make the most optimal decisions.

If the recommendation is followed and there are still SLA violations (e.g. due to the network re-provisioning putting constraints on resources of other customers), any violation of SLAs is reported to the SA and SLA Management node 403 (signal 430).

The SA and SLA Management node 403 in turn reports the SLA violation to the Charging and Billing node 402 (signal 431) and the Charging and Billing node 402 calculates the penalty associated with the violation (step 432). Thus, the techniques described above establish a service assurance system in such a way as to obtain optimal utilisation of network resources in a way that maximises the business objectives of the network operator, which are driven by the importance of the customer to the network operator.

The techniques described herein relate to using the CIS concept to effect an optimal network state for providing services to customers who have higher CISs. The CIS concept is a general concept which can be used in any constrained resource problems while managing network operations. When it comes to race condition or conflicts in network resource allocation to the customers, the CIS concept can be used to help prioritise the allocation of network resources. For example, the allocation of network resources to customers (or the users associated with the customers) can be based on the CIS of the customers, with the more important customers being a higher priority than customers with lower CISs. Another example is related to executing a bulk order from an enterprise customer. There may be several enterprises for whom an order requires fulfilment under an SLA of order delivery. For example, in the case of a video streaming service company A, 1000 users have signed up for the service, and there is a second video streaming service company B for which 2000 users have signed up for the service. Each video streaming service company sends user activation requests for service usage in the network to the network service provider (network operator) at the end of the day. The SLAs for order fulfilment are 2000 user activations in 24 hours for each of the video streaming service companies. The network service provider now has 3000 orders (3000 user activations) that it is required to fulfil. Given that the network operator has received all the orders at the same time and it has potential to process only 2500 orders in a day, the network operator needs to determine how to prioritise the order fulfilment. If the first video streaming service company A has a higher CIS then its 1000 users order will be fulfilled first before the orders of the second video streaming service company B are fulfilled. By the end of 24 hours there may still be 500 users of the second video streaming service company B whose orders need to be fulfilled, hence this may lead to violation of SLA with the second video streaming service company B. Of the two video streaming service companies, the violation of SLAs with the second video streaming service company is justified as it has a lower CIS.

Fig. 5 is a schematic block diagram illustrating a virtualisation environment 500 in which functions implemented by some embodiments may be virtualized. In the present context, virtualising means creating virtual versions of apparatuses or devices which may include virtualising hardware platforms, storage devices and networking resources. As used herein, virtualisation can be applied to any of the nodes shown in Figs. 1 and 4 and as described above, or to components thereof and relates to an implementation in which at least a portion of the functionality is implemented as one or more virtual components (e.g., via one or more applications, components, functions, virtual machines or containers executing on one or more physical processing nodes in one or more networks).

In some embodiments, some or all of the functions described herein may be implemented as virtual components executed by one or more virtual machines implemented in one or more virtual environments 500 hosted by one or more of hardware nodes 530.

The functions may be implemented by one or more applications 520 (which may alternatively be called software instances, virtual appliances, network functions, virtual nodes, virtual network functions, etc.) operative to implement some of the features, functions, and/or benefits of some of the embodiments disclosed herein. Applications 520 are run in virtualisation environment 500 which provides hardware 530 comprising processing circuitry 560 and memory 590. Memory 590 contains instructions 595 executable by processing circuitry 560 whereby application 520 is operative to provide one or more of the features, benefits, and/or functions disclosed herein.

Virtualisation environment 500, comprises general-purpose or special-purpose network hardware devices 530 comprising a set of one or more processors or processing circuitry 560, which may be commercial off-the-shelf (COTS) processors, dedicated Application Specific Integrated Circuits (ASICs), or any other type of processing circuitry including digital or analog hardware components or special purpose processors. Each hardware device may comprise memory 590-1 which may be non-persistent memory for temporarily storing instructions 595 or software executed by processing circuitry 560. Each hardware device may comprise one or more network interface controllers (NICs) 570, also known as network interface cards, which include physical network interface 580. Each hardware device may also include non-transitory, persistent, machine-readable storage media 590-2 having stored therein software 595 and/or instructions executable by processing circuitry 560. Software 595 may include any type of software including software for instantiating one or more virtualisation layers 550 (also referred to as hypervisors), software to execute virtual machines 540 as well as software allowing it to execute functions, features and/or benefits described in relation with some embodiments described herein.

Virtual machines 540, comprise virtual processing, virtual memory, virtual networking or interface and virtual storage, and may be run by a corresponding virtualisation layer 550 or hypervisor. Different embodiments of the instance of virtual appliance 520 may be implemented on one or more of virtual machines 540, and the implementations may be made in different ways. During operation, processing circuitry 560 executes software 595 to instantiate the hypervisor or virtualisation layer 550, which may sometimes be referred to as a virtual machine monitor (VMM). Virtualisation layer 550 may present a virtual operating platform that appears like networking hardware to virtual machine 540.

As shown in Fig. 5, hardware 530 may be a standalone network node with generic or specific components. Hardware 530 may comprise antenna 5225 and may implement some functions via virtualisation. Alternatively, hardware 530 may be part of a larger cluster of hardware (e.g. such as in a data centre or customer premise equipment (CPE)) where many hardware nodes work together and are managed via management and orchestration (MANO) 5100, which, among others, oversees lifecycle management of applications 520.

Virtualisation of the hardware is in some contexts referred to as network function virtualisation (NFV). NFV may be used to consolidate many network equipment types onto industry standard high volume server hardware, physical switches, and physical storage, which can be located in data centers, and customer premise equipment.

In the context of NFV, virtual machine 540 may be a software implementation of a physical machine that runs programs as if they were executing on a physical, non-virtualised machine. Each of virtual machines 540, and that part of hardware 530 that executes that virtual machine, be it hardware dedicated to that virtual machine and/or hardware shared by that virtual machine with others of the virtual machines 540, forms a separate virtual network elements (VNE).

Still in the context of NFV, Virtual Network Function (VNF) is responsible for handling specific network functions that run in one or more virtual machines 540 on top of hardware networking infrastructure 530 and corresponds to application 520 in Fig. 5.

Fig. 6 is a block diagram of an apparatus 600 that can be used to implement any of the nodes described herein, for example any of the nodes described above with reference to Figs. 1 and 4, and any of the methods described below with reference to Figs. 7-10. It will be appreciated that the apparatus 600 may comprise one or more virtual machines running different software and/or processes. The apparatus 600 may therefore comprise one or more servers, switches and/or storage devices and/or may comprise cloud computing infrastructure that runs the software and/or processes.

The apparatus 600 comprises processing circuitry 601 that controls the operation of the apparatus 600 and can implement the methods described herein, including any of the methods described below with reference to Figs. 7-10. The processing circuitry 601 can comprise one or more processors, processing units, multi-core processors or modules that are configured or programmed to control the apparatus 600 in the manner described herein. In particular implementations, the processing circuitry 601 can comprise a plurality of software and/or hardware modules that are each configured to perform, or are for performing, individual or multiple steps of any of the methods described herein.

In some embodiments, the apparatus 600 may optionally comprise a communications interface 602. The communications interface 602 can be for use in communicating with other nodes, such as other virtual nodes. For example, the communications interface 602 can be configured to transmit to and/or receive from other nodes, network nodes or network functions requests, resources, information, data, signals, or similar. The processing circuitry 601 may be configured to control the communications interface 602 of the apparatus 600 to transmit to and/or receive from other nodes, network nodes, or network functions requests, resources, information, data, signals, or similar.

Optionally, the apparatus 600 may comprise a memory 603. In some embodiments, the memory 603 can be configured to store program code that can be executed by the processing circuitry 601 to perform any of the methods described herein. Alternatively or in addition, the memory 603 can be configured to store any requests, resources, information, data, signals, or similar that are described herein. The processing circuitry 601 may be configured to control the memory 603 to store any requests, resources, information, data, signals, or similar that are described herein.

Figs. 7-10 are flow charts illustrating methods of operating various nodes according to the embodiments described herein. These methods are described with reference to a customer of a network operator. The network operator provides services via a communication network. In the following, the customer is referred to as a ‘first customer’ and the network operator is referred to as a ‘first network operator’. The first customer has an agreement or contract with the first network operator. This agreement or contract can be, or include, an SLA. The agreement or contract may require the first network operator to provide a service for the first customer according to some defined service levels, e.g. minimum throughput, maximum latency, etc. The first customer can be an enterprise customer, for example a media streaming and/or download service. More generally, the service provided by the first customer is referred to as a ‘first service’. In a 5G network, the agreement between the first network operator and the first customer can be for the first network operator to provide a network slice for the first customer. The first customer may have a plurality of users of its service, some of which may access the first customer’s service via the first network operator’s network. It will be appreciated that the first network operator can have a plurality of customers, that may each provide their own service. It will also be appreciated that more than one network operator can operate on the same communication network. Fig. 7 is a flow chart illustrating a method of operating an analysis node (CIS service) 102 according to various embodiments. The analysis node 102 may be operated by the first network operator. The method in Fig. 7 relates to the method illustrated in Fig. 1 above.

In step 701 , the analysis node 102 receives customer information for the first customer.

The customer information can comprise any information relating to the first customer. In some embodiments, the customer information can comprise information relating to a service contract (also referred to as an SLA) between the first customer and the first network operator. The service contract information can include information about the SLA violation history (i.e. whether, when and/or how the SLA has previously been violated or breached), information about SLA threshold(s) and penalty(ies) (i.e. the thresholds against which the service is measured and the penalties for breaching those thresholds), and/or information about the current terms and conditions in the agreement. In some embodiments, the customer information can comprise information relating to a number of users associated with the first customer over time, including any of a current number of users, a change of user numbers over time and a projected number of users. In some embodiments, the customer information can comprise information relating to usage of resources of the communication network by users associated with the first customer. In some embodiments, the customer information can comprise information relating to a relationship between the first customer and the first network operator. The relationship information can include information about the propensity or likelihood of the first customer ending the agreement and/or entering into an agreement with another network operator, information about the intent of the first customer in renewing the contract, information about the length of time that the customer has had an agreement with the first network operator first network operator, information about an ARPU and/or ARPU growth rate, information about the invoice payment behaviour of the first customer (e.g. are invoices paid on time), and/or information about customer journeys and/or a complaints history for the first customer. In some embodiments, the customer information can comprise enterprise information relating to the first customer (in embodiments where the first customer is an enterprise). The enterprise information can include any of financial information (such as average revenue per account, credit risk rating, past year revenue, revenue growth, etc.), social media information (such as the popularity of the first customer on social media platforms) and corporate information (such as the size of enterprise, number of users, etc.).

In step 702 the analysis node 102 analyses the customer information to determine a CIS for the first customer. The CIS for the first customer can represent an importance of the first customer to the first network operator. In some embodiments, step 702 can comprise inputting the received customer information into a trained Al model. The trained Al model provides the CIS as an output.

In some embodiments, the analysis node 102 can store the CIS in a customer information database (e.g. CRM 101).

In some embodiments, steps 701 and 702 can be performed when an agreement between the first customer and the first network operator is established (this is also referred to as the ‘onboarding’ stage). In alternative embodiments, steps 701 and 702 can be performed once the agreement has been in place for a while, which allows time for customer information to be generated and collected.

In some embodiments, the analysis node 102 can receive further or updated customer information for the first customer. In that case, the analysis node 102 can update the CIS for the first customer based on the received further customer information. The updating step can be performed in a similar way to analysing step 702.

In some embodiments, the analysis node 102 can receive customer information for a plurality of other customers of the first network operator, analyse the customer information for each of the plurality of other customers to determine respective CISs for those customers.

Fig. 8 is a flow chart illustrating a method of operating a recommendation node (SA recommendation engine) 405 according to various embodiments. The recommendation node 405 may be operated by the first network operator. The method in Fig. 8 relates to the method illustrated in Fig. 4 above.

In step 801 , the recommendation node 405 receives prediction information for the first customer of the first network operator. The prediction information can be received from a prediction node 404 (e.g. SA SLA Violation Prediction Service 404). The prediction information indicates that degradation of the first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator.

In step 802, the recommendation node 405 obtains a CIS for the first customer. The CIS can be obtained from a customer information database (e.g. CRM 101/401).

Based on the CIS for the first customer, in step 803 the recommendation node 405 determines whether network resources in the communication network in use by other customers or non-first customer users of the communication network should be allocated to the first service to address the predicted degradation of the first service.

In step 804, if it is determined that network resources in use by other customers or users should be allocated to the first service, the recommendation node 405 sends a first recommendation to a network orchestration node (e.g. service orchestrator) 407. The first recommendation can recommend the allocation of network resources that are in use by other customers or users to the first service. The network orchestration node 407 can be operated by the first network operator.

In some embodiments, the first recommendation can further comprise an indication of a consequence to the first network operator of the predicted degradation of the first service. In this case the consequence to the first network operator may be a cost (e.g. a financial cost) to the first network operator due to the degradation, for example the amount of penalty payable by the first network operator due to the breach according to the agreement. In some embodiments, the first recommendation can further comprise an indication of a consequence to the first network operator of following the first recommendation. In this case the consequence to the first network operator may be a cost (e.g. a financial cost) to the first network operator due to degrading the service provided to other customers or users of the first network operator (and, e.g. incurring penalties due to breaching the agreements with those other customers).

In some embodiments, the decision in step 803 can be based on the CIS for the first customer and a cost to the first network operator associated with allocating the network resources in use by other customers or users to the first service provided to the first customer. This enables the first network operator to balance the importance of the first customer to the first network operator and the cost implications of avoiding the degradation of the first service.

In some embodiments, the decision in step 803 can be further based on a CIS for one or more other customers of the first network operator. In these embodiments the recommendation node 405 can compare the CISs to evaluate the relative importance of the customers to the first network operator. As a guide, the recommendation node 405 may recommend that network resources are prioritised for the first customer (and therefore taken from other customers) if the first customer is more important to the first network operator than the other customers.

In some embodiments, before determining if resources can be reallocated from other customers/users based on the CIS of the first customer in step 803, the recommendation node 405 may attempt to prevent the predicted degradation of the first service in other ways, using resources in the network that are either not yet allocated to any service, or that are allocated to another customer but are not yet being used.

Thus, in some embodiments, prior to step 802, the method can comprise the recommendation node 405 determining if there are additional network resources available in the communication network that can be allocated to the first service. These additional network resources are resources that are not yet allocated to any particular customer. If there are additional network resources available, the recommendation node 405 can send a recommendation to the network orchestration node 407 to allocate the available additional network resources to the first service provided to the first customer. In some embodiments, the recommendation node 405 can determine if there are additional network resources available by consulting with a network resource management node (NW inventory) 406.

In further embodiments, if it is determined that there not any additional network resources available forthe first service, the recommendation node 405 may determine if there are unused network resources available in the communication network. Unused network resources are network resources allocated to other services and/or customers that are not being used. If unused network resources are available, then the recommendation node 405 can send a recommendation to the network orchestration node 407 to allocate the unused network resources to the first service provided to the first customer.

In some embodiments, if degradation of the first service provided to the first customer breaches the service contract, the recommendation node 405 can report to service management node 403 (e.g. SA and SLA Management node 403) that a breach of the service contract has occurred.

Fig. 9 is a flow chart illustrating a method of operating a network orchestration node (e.g. service orchestrator) 405 according to various embodiments. The network orchestration node 407 may be operated by the first network operator. The method in Fig. 9 relates to the method illustrated in Fig. 4 above.

In step 901 the network orchestration node 407 receives a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator. This recommendation is received from recommendation node 405. In some embodiments, the recommendation comprises any of a CIS for the first customer, an indication of a consequence to the first network operator of the predicted degradation of the first service and/or an indication of a consequence to the first network operator of following the first recommendation. Thus the consequence may be a cost (e.g. a financial cost) to the first network operator due to the degradation, for example the amount of penalty payable by the first network operator due to the breach according to the agreement. The consequence may also or alternatively be a cost (e.g. a financial cost) to the first network operator due to degrading the service provided to other customers or users of the first network operator.

In step 902, the network orchestration node 407 determines whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer. In embodiments where the first recommendation comprises an indication of a consequence, the decision in step 902 is based on the CIS for the first customer and/or the indication of the first consequence.

If the network orchestration node 407 decides to allocate the network resources to the first service, the network orchestration node 405 allocates those resources to the first service provided to the first customer (step 903).

In some embodiments, the network orchestration node 407 can obtain a respective CIS for the other customers of the communication network and/or an indication of a consequence to the first network operator of allocating the network resources that are in use by those other customers to the first service provided to the first customer. The network orchestration node 407 can use this information in step 902 to take the decision on whether to allocate the network resources to the first customer.

Fig. 10 is a flow chart illustrating a method of operating a network orchestration node (e.g. service orchestrator) 405 according to various embodiments. The network orchestration node 407 may be operated by the first network operator. The method in Fig. 10 relates to the use of CIS in handling network resource allocation when there are conflicting resource requests.

Thus, in step 1001 , the network orchestration node 407 receives a plurality of resource requests from users of a plurality of customers of the first network operator. The resources requested by the plurality of resource requests exceed a resource capacity in the communication network of the first network operator.

In step 1002, the network orchestration node 407 obtains a CIS for at least a first customer in the plurality of customers, and/or a consequence indication. The consequence indication can be an indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or an indication of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers. In some embodiments, the network orchestration node 407 obtains a CIS for each of the plurality of customers.

In step 1003, the network orchestration node 407 determines whether to grant a resource request from a user of the first customer. The decision can be based on the CIS for the first customer and/or the consequence indication. For example, the network orchestration node 407 can decide to prioritise the granting of resource requests from users of customers that are the most important to the first network operator (as indicated by the CISs of the customers).

If it is determined to grant the resource request, the network orchestration node 407 allocates those resources to the user of the first customer (step 1004). The foregoing merely illustrates the principles of the disclosure. Various modifications and alterations to the described embodiments will be apparent to those skilled in the art in view of the teachings herein. It will thus be appreciated that those skilled in the art will be able to devise numerous systems, arrangements, and procedures that, although not explicitly shown or described herein, embody the principles of the disclosure and can be thus within the scope of the disclosure. Various exemplary embodiments can be used together with one another, as well as interchangeably therewith, as should be understood by those having ordinary skill in the art.

Claims

1 . A method of operating an analysis node, the method comprising: receiving (701) customer information for a first customer of a first network operator of a communication network; and analysing (702) the customer information to determine a customer importance score, CIS, for the first customer.

2. A method as claimed in claim 1 , wherein the CIS for the first customer represents an importance of the first customer to the first network operator.

3. A method as claimed in claim 1 or 2, wherein the customer information comprises any one or more of: information relating to a service contract between the first customer and the first network operator; information relating to a number of users associated with the first customer over time; information relating to usage of resources of the communication network by users associated with the first customer; information relating to a relationship between the first customer and the first network operator; and enterprise information relating to the first customer when the first customer is an enterprise.

4. A method as claimed in any of claims 1-3, wherein the step of analysing (702) the customer information comprises inputting the received customer information into a trained artificial intelligence model, and the trained artificial intelligence model providing the CIS as an output.

5. A method as claimed in any of claims 1-4, wherein the method further comprises: storing the CIS in a customer information database.

6. A method as claimed in any of claims 1-5, wherein the method further comprises: receiving further customer information for the first customer; and updating the CIS for the first customer based on the received further customer information.

7. A method as claimed in any of claims 1-6, wherein the method further comprises: receiving customer information for a plurality of other customers of the first network operator; and for each of the plurality of other customers, analysing their customer information to determine a respective CIS.

8. A method of operating a recommendation node, the method comprising: receiving (801) prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtaining (802) a customer importance score, CIS, for the first customer; and determining (803), based on the CIS for the first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service.

9. A method as claimed in claim 8, wherein if it is determined that network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer, the method further comprises: sending a first recommendation to a network orchestration node in the communication network to allocate network resources that are in use by other customers or users of the communication network to the first service provided to the first customer.

10. A method as claimed in claim 9, wherein the first recommendation further comprises an indication of a consequence to the first network operator of the predicted degradation of the first service and/or an indication of a consequence to the first network operator of following the first recommendation.

11. A method as claimed in any of claims 8-10, wherein the step of determining (803) whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer is based on the CIS for the first customer and a cost to the first network operator associated with allocating the network resources in use by other customers or users to the first service provided to the first customer.

12. A method as claimed in any of claims 8-11 , wherein the step of determining (803) whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer is further based on a CIS for one or more other customers of the first network operator.

13. A method as claimed in any of claims 8-12, wherein the step of obtaining (802) the CIS comprises obtaining the CIS from a customer information database in the communication network.

14. A method as claimed in any of claims 8-13, wherein the method further comprises: prior to obtaining the CIS, determining if there are additional network resources available in the communication network that can be allocated to the first service; and if it is determined that there are additional network resources available, sending a second recommendation to a network orchestration node in the communication network to allocate the available additional network resources to the first service provided to the first customer.

15. A method as claimed in claim 14, wherein if it is not determined that there are additional network resources available in the communication network for the first service, the method further comprises: determining if there are unused network resources available in the communication network that are network resources allocated to other services and/or customers that are not being used; and if it is determined that unused network resources are available, sending a third recommendation to the network orchestration node to allocate the unused network resources to the first service provided to the first customer.

16. A method as claimed in claim 15, wherein the step of determining if there are unused network resources available is performed if it is not determined that there are additional network resources in the communication network that can be allocated to the first service provided to the first customer.

17. A method as claimed in any of claims 8-16, wherein the method further comprises: if degradation of the first service provided to the first customer breaches the service contract, reporting, to a service management node in the communication network, that a breach of the service contract has occurred.

18. A method of operating a network orchestration node in a communication network, the method comprising: receiving (901), from a recommendation node, a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determining (902) whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and if it is determined to allocate the network resources, allocating (903) those resources to the first service provided to the first customer.

19. A method as claimed in claim 18, wherein the first recommendation further comprises a customer importance score, CIS, for the first customer, and/or an indication of a first consequence to the first network operator of the predicted degradation of the first service.

20. A method as claimed in claim 19, wherein the step of determining (902) whether to allocate the network resources is based on the CIS for the first customer and/or the indication of the first consequence.

21 . A method as claimed in claim 20, wherein the method further comprises: obtaining a respective CIS for the other customers of the communication network and/or a second indication of a consequence to the first network operator of allocating the network resources that are in use by other customers to the first service provided to the first customer; wherein the step of determining whether to allocate the network resources is further based on the obtained CISs for the other customers and/or the second indication.

22. A method of operating a network orchestration node in a communication network, the method comprising: receiving (1001) a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtaining (1002) a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determining (1003) whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and if it is determined to grant the resource request, allocating (1004) those resources to the user of the first customer.

23. A computer program product comprising a computer readable medium having computer readable code embodied therein, the computer readable code being configured such that, on execution by a suitable computer or processor, the computer or processor is caused to perform the method of any of claims 1-22.

24. An analysis node (102), the analysis node (102) configured to: receive customer information for a first customer of a first network operator of a communication network; and analyse the customer information to determine a customer importance score, CIS, for the first customer.

25. An analysis node (102) as claimed in claim 24, wherein the CIS for the first customer represents an importance of the first customer to the first network operator.

26. An analysis node (102) as claimed in claim 24 or 25, wherein the customer information comprises any one or more of: information relating to a service contract between the first customer and the first network operator; information relating to a number of users associated with the first customer over time; information relating to usage of resources of the communication network by users associated with the first customer; information relating to a relationship between the first customer and the first network operator; and enterprise information relating to the first customer when the first customer is an enterprise.

27. An analysis node (102) as claimed in any of claims 24-26, wherein the analysis node (102) is configured to analyse the customer information by inputting the received customer information into a trained artificial intelligence model, with the trained artificial intelligence model providing the CIS as an output.

28. An analysis node (102) as claimed in any of claims 24-27, wherein the analysis node (102) is further configured to: store the CIS in a customer information database (101).

29. An analysis node (102) as claimed in any of claims 24-28, wherein the analysis node (102) is further configured to: receive further customer information for the first customer; and update the CIS for the first customer based on the received further customer information.

30. An analysis node (102) as claimed in any of claims 24-29, wherein the analysis node (102) is further configured to: receive customer information for a plurality of other customers of the first network operator; and for each of the plurality of other customers, analyse their customer information to determine a respective CIS.

31 . A recommendation node (405), the recommendation node (405) configured to: receive prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtain a customer importance score, CIS, for the first customer; and determine, based on the CIS for the first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service.

32. A recommendation node (405) as claimed in claim 31 , wherein the recommendation node (405) is further configured to: if it is determined that network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer, send a first recommendation to a network orchestration node (407) in the communication network to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer.

33. A recommendation node (405) as claimed in claim 32, wherein the first recommendation further comprises an indication of a consequence to the first network operator of the predicted degradation of the first service and/or an indication of a consequence to the first network operator of following the first recommendation.

34. A recommendation node (405) as claimed in any of claims 31-33, wherein the recommendation node (405) is configured to determine whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer based on the CIS for the first customer and a cost to the first network operator associated with allocating the network resources in use by other customers or users to the first service provided to the first customer.

35. A recommendation node (405) as claimed in any of claims 31-34, wherein the recommendation node (405) is configured to determine whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer based on a CIS for one or more other customers of the first network operator.

36. A recommendation node (405) as claimed in any of claims 31-35, wherein the recommendation node (405) is configured to obtain the CIS from a customer information database in the communication network.

37. A recommendation node (405) as claimed in any of claims 31-36, wherein the recommendation node (405) is further configured to: prior to obtaining the CIS, determine if there are additional network resources available in the communication network that can be allocated to the first service; and if it is determined that there are additional network resources available, send a second recommendation to a network orchestration node (407) in the communication network to allocate the available additional network resources to the first service provided to the first customer.

38. A recommendation node (405) as claimed in claim 37, wherein the recommendation node (405) is further configured to: if it is not determined that there are additional network resources available in the communication network for the first service, determine if there are unused network resources available in the communication network that are network resources allocated to other services and/or customers that are not being used; and if it is determined that unused network resources are available, send a third recommendation to the network orchestration node (407) to allocate the unused network resources to the first service provided to the first customer.

39. A recommendation node (405) as claimed in claim 38, wherein the recommendation node (405) is configured to determine if there are unused network resources available if it is not determined that there are additional network resources in the communication network that can be allocated to the first service provided to the first customer.

40. A recommendation node (405) as claimed in any of claims 31-39, wherein the recommendation node (405) is further configured to: if degradation of the first service provided to the first customer breaches the service contract, report, to a service management node (403) in the communication network, that a breach of the service contract has occurred.

41. A network orchestration node (407) for use in a communication network, the network orchestration node (407) configured to: receive, from a recommendation node (405), a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determine whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and if it is determined to allocate the network resources, allocate those resources to the first service provided to the first customer.

42. A network orchestration node (407) as claimed in claim 41 , wherein the first recommendation further comprises a customer importance score, CIS, for the first customer, and/or an indication of a first consequence to the first network operator of the predicted degradation of the first service.

43. A network orchestration node (407) as claimed in claim 42, wherein the network orchestration node (407) is configured to determine whether to allocate the network resources based on the CIS for the first customer and/or the indication of the first consequence.

44. A network orchestration node (407) as claimed in claim 43, wherein the network orchestration node (407) is further configured to: obtain a respective CIS for the other customers of the communication network and/or a second indication of a consequence to the first network operator of allocating the network resources that are in use by other customers to the first service provided to the first customer; wherein the network orchestration node (407) is configured to determine whether to allocate the network resources based on the obtained CISs for the other customers and/or the second indication.

45. A network orchestration node (407) for use in a communication network, the network orchestration node (407) configured to: receive a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtain a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determine whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and if it is determined to grant the resource request, allocate those resources to the user of the first customer.

46. An analysis node, the analysis node comprising a processor and a memory, said memory containing instructions executable by said processor whereby said analysis node is operative to: receive customer information for a first customer of a first network operator of a communication network; and analyse the customer information to determine a customer importance score, CIS, for the first customer.

47. An analysis node as claimed in claim 46, wherein the CIS for the first customer represents an importance of the first customer to the first network operator.

48. An analysis node as claimed in claim 46 or 47, wherein the customer information comprises any one or more of: information relating to a service contract between the first customer and the first network operator; information relating to a number of users associated with the first customer over time; information relating to usage of resources of the communication network by users associated with the first customer; information relating to a relationship between the first customer and the first network operator; and enterprise information relating to the first customer when the first customer is an enterprise.

49. An analysis node as claimed in any of claims 46-48, wherein the analysis node is operative to analyse the customer information by inputting the received customer information into a trained artificial intelligence model, with the trained artificial intelligence model providing the CIS as an output.

50. An analysis node as claimed in any of claims 46-49, wherein the analysis node is further operative to: store the CIS in a customer information database.

51 . An analysis node as claimed in any of claims 46-50, wherein the analysis node is further operative to: receive further customer information for the first customer; and update the CIS for the first customer based on the received further customer information.

52. An analysis node as claimed in any of claims 46-51 , wherein the analysis node is further operative to: receive customer information for a plurality of other customers of the first network operator; and for each of the plurality of other customers, analyse their customer information to determine a respective CIS.

53. A recommendation node, the recommendation node comprising a processor and a memory, said memory containing instructions executable by said processor whereby said recommendation node is operative to: receive prediction information for a first customer of a first network operator of a communication network, wherein the prediction information indicates that degradation of a first service provided to the first customer by the first network operator may occur that will breach a service contract between the first customer and the first network operator; obtain a customer importance score, CIS, for the first customer; and determine, based on the CIS for the first customer, whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer to address the predicted degradation of the first service.

54. A recommendation node as claimed in claim 53, wherein the recommendation node is further operative to: if it is determined that network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer, send a first recommendation to a network orchestration node in the communication network to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer.

55. A recommendation node as claimed in claim 54, wherein the first recommendation further comprises an indication of a consequence to the first network operator of the predicted degradation of the first service and/or an indication of a consequence to the first network operator of following the first recommendation.

56. A recommendation node as claimed in any of claims 53-55, wherein the recommendation node is operative to determine whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer based on the CIS for the first customer and a cost to the first network operator associated with allocating the network resources in use by other customers or users to the first service provided to the first customer.

57. A recommendation node as claimed in any of claims 53-56, wherein the recommendation node is operative to determine whether network resources in the communication network in use by other customers or users of the communication network should be allocated to the first service provided to the first customer based on a CIS for one or more other customers of the first network operator.

58. A recommendation node as claimed in any of claims 53-57, wherein the recommendation node is operative to obtain the CIS from a customer information database in the communication network.

59. A recommendation node as claimed in any of claims 53-58, wherein the recommendation node is further operative to: prior to obtaining the CIS, determine if there are additional network resources available in the communication network that can be allocated to the first service; and if it is determined that there are additional network resources available, send a second recommendation to a network orchestration node in the communication network to allocate the available additional network resources to the first service provided to the first customer.

60. A recommendation node as claimed in claim 59, wherein the recommendation node is further operative to: if it is not determined that there are additional network resources available in the communication network for the first service, determine if there are unused network resources available in the communication network that are network resources allocated to other services and/or customers that are not being used; and if it is determined that unused network resources are available, send a third recommendation to the network orchestration node to allocate the unused network resources to the first service provided to the first customer.

61. A recommendation node as claimed in claim 60, wherein the recommendation node is operative to determine if there are unused network resources available if it is not determined that there are additional network resources in the communication network that can be allocated to the first service provided to the first customer.

62. A recommendation node as claimed in any of claims 53-61 , wherein the recommendation node is further operative to: if degradation of the first service provided to the first customer breaches the service contract, report, to a service management node in the communication network, that a breach of the service contract has occurred.

63. A network orchestration node for use in a communication network, the network orchestration node comprising a processor and a memory, said memory containing instructions executable by said processor whereby said network orchestration node is operative to: receive, from a recommendation node, a recommendation to allocate network resources that are in use by other customers or users of the communication network to a first service provided to a first customer of the first network operator; determine whether to allocate the network resources that are in use by other customers or users of the communication network to the first service provided to the first customer; and if it is determined to allocate the network resources, allocate those resources to the first service provided to the first customer.

64. A network orchestration node as claimed in claim 63, wherein the first recommendation further comprises a customer importance score, CIS, for the first customer, and/or an indication of a first consequence to the first network operator of the predicted degradation of the first service.

65. A network orchestration node as claimed in claim 64, wherein the network orchestration node is operative to determine whether to allocate the network resources based on the CIS for the first customer and/or the indication of the first consequence.

66. A network orchestration node as claimed in claim 65, wherein the network orchestration node is further operative to: obtain a respective CIS for the other customers of the communication network and/or a second indication of a consequence to the first network operator of allocating the network resources that are in use by other customers to the first service provided to the first customer; wherein the network orchestration node is operative to determine whether to allocate the network resources based on the obtained CISs for the other customers and/or the second indication.

67. A network orchestration node for use in a communication network, the network orchestration node comprising a processor and a memory, said memory containing instructions executable by said processor whereby said network orchestration node is operative to: receive a plurality of resource requests from users of a plurality of customers of a first network operator, wherein the resources requested in the plurality of resource requests exceed a resource capacity in the communication network; obtain a customer importance score, CIS, for at least a first customer in the plurality of customers, and/or a first indication of a consequence to the first network operator of failing to meet a resource request from users of the first customer and/or of a consequence to the first network operator of prioritising the users of the first customer over users from other customers in the plurality of customers; determine whether to grant a resource request from a user of the first customer based on the CIS for the first customer and/or the first indication; and if it is determined to grant the resource request, allocate those resources to the user of the first customer.