WO2022008082A1

WO2022008082A1 - Methods and apparatus for network control

Info

Publication number: WO2022008082A1
Application number: PCT/EP2020/069623
Authority: WO
Inventors: Saurabh Singh; Alexandros NIKOU; Lackis ELEFTHERIADIS; Pedro BATISTA; Rafia Inam
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2020-07-10
Filing date: 2020-07-10
Publication date: 2022-01-13
Anticipated expiration: 2023-01-10

Abstract

Aspects of embodiments provide network operation methods and core network nodes for controlling components of a physical network infrastructure using a high priority logical network. The method may comprise: activating a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component of the physical network infrastructure, wherein the step of activating the high priority logical network comprises generating an energy symbol, and wherein the energy symbol is included in transmission frames to control component configuration adjustment; and controlling the operation of the high priority logical network on the physical network infrastructure.

Description

METHODS AND APPARATUS FOR NETWORK CONTROL

Technical Field

Embodiments of the present disclosure relate to methods and apparatus for network control, and particularly methods and apparatus for controlling components of a physical network infrastructure using a high priority logical network.

Background

Recent developments in telecommunication network technology have allowed increased separation between the control of physical network infrastructure and logical networks. Logical networks are a form of Software Defined Networks (SDN), and may also be referred to as virtual networks. SDNs essentially decouple the network control functions (the control plane) from the data forwarding functions (the data plane), introducing a degree of separation between control of the physical components forming the network infrastructure (nodes, cables, etc.) and the overall network control. In SDN, data transfer services can be used to provide a user with a data connection between two points, without requiring the user to have detailed knowledge of exactly which components of the network are responsible for providing the connection. As such, a data transfer service can be used to satisfy the data traffic requirements of a user, such as transferring a given volume of data traffic between two points at a given rate, with a given reliability, and so on.

The implementation of the 3^rd Generation Partnership Project (3GPP) 5^th Generation (5G) technology has facilitated the implementation of network slices. Network slices are a category of logical network, which provide specific network capabilities and network characteristics. Using network slicing allows different applications with different requirements to co-exist while keeping their respective Quality of Service (QoS) guarantees; the QoS guarantees may contain threshold standards relating to latency, throughput, reliability, security, and so on, where penalties may be imposed upon an infrastructure provider that fails to satisfy an agreed QoS guarantee with a Mobile Virtual Network Operator, MVNO, also referred to herein as an operator. Different QoS guarantees may be in place with different operators, as agreed in Service Level Agreements (SLAs) with the operators. Document [1] includes a discussion of 5G system architecture, including the implementation of network slices in 5G systems. The generation and managements of NSIs is discussed in greater detail in Document [2] As explained in Document [2] the lifecycle of an NSI comprises a number of different phases; a typical NSI lifecycle is illustrated schematically in Figure 1.

Figure 1 can be found in Document [2] as Figure 4.1.1. As shown in Figure 1 , Network slice template(s) is (are) created during the preparation phase; the NSI does not exist during the preparation phase. The NSI is created during the instantiation/configuration phase, in which all shared/dedicated resources to the NSI are created and configured, i.e. to a state where the NSI is ready for operation. The activation step includes any actions that makes the NSI active, for example, diverting traffic to the NSI. During the run-time phase the NSI handles traffic and supports communication services of certain type(s). The run-time phase includes supervision/reporting of, for example Key Performance Indicators, KPI, as well as activities related to modification (potentially based on reported KPI). The term modification encompasses various processes such as: upgrades, reconfiguration,

NSI scaling, changes of NSI capacity, changes of NSI topology, association and disassociation of network functions with NSI, and so on. Finally, when the NSI is no longer required, the decommissioning phase is initiated. The decommissioning phase includes deactivation of the NSI (taking the NSI out of active duty and removing any remaining traffic from the NSI) as well as the reclamation of dedicated resources and configuration of shared/dependent resources. The NSI does not exist after decommissioning, and must be recreated (instantiated) if required again at a later time.

Different types of network slice may satisfy different roles, and a User Equipment (UE) may be served by a plurality of network slices at any given time. Typically, a UE may be served by up to 8 Network Slice Instances (NSI), which may include standardised types mapped to “enhanced mobile broadband” (eMBB), “ultra-reliable low-latency communication” (URLLC)”, “massive loT” (mloT), critical Machine Type Communication (cMTC), and so on, as well as non-standardized ones, which may be defined by the Communication Service Providers (CSPs).

Document [3] discusses how mobile traffic requirements may be forecast to help maximise the usage of 5G network resources, while managing the risk of slice SLA violation. The complexity of the slice management tends to grow drastically as the number of users, applications and the slice instances increase. The agreed QoS requirements should be met on an end to end (e2e) basis throughout the lifecycle of the slice to avoid, for example, breaching a SLA. Further, the SLA may be determined at least in part by the priority level of a network slice (relative to other slices on the network).

Documents

[1] “System architecture for the 5G System (5GS)”, TS 23.501 V16.4.0, by 3GPP, available at https://portal.3gpp.org/desktopmodules/Specifications/Specification Details. aspx?specificationld=3144 as of 19 June 2020.

[2] “Telecommunication management; Study on management and orchestration of network slicing for next generation network”, TR 28.801 V15.1.0, by 3GPP, available at https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails. aspx?specificationld=3091 as of 19 June 2020.

[3] “Mobile traffic forecasting for maximizing 5G network slicing resource utilization”, Sciancalepore, V. et al. IEEE INFOCOM 2017 - IEEE Conference on Computer Communications, available at https://ieeexplore.ieee.org/document/8057230 as of 19 June 2020.

[4] “Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation”, TS 36.211 V13.0.0, by 3GPP, available at https://portal.3gpp.org/ desktopmodules/Specifications/SpecificationDetails.aspx?specificationld=2425 as of 19 June 2020.

Summary

It can be envisioned that in some situations shortages of available network resources due to, for example, damage to network infrastructure or atypically high traffic levels, may mean that it is not possible to simultaneously satisfy all QoS requirements. It is desirable to avoid QoS failures, particularly in the case of high priority logical networks which may have high importance roles such as providing communication capabilities to emergency services, control of automated vehicles, and so on. Current systems may not adequately ensure QoS requirements for high priority logical networks, which may result in QoS failures and potentially severe consequences.

In situations where it is likely one or more QoS requirements for different NSIs may be compromised, it is desirable to ensure that high priority or critical logical networks are provided with guaranteed e2e QoS while any potentially necessary compromising of QoS is focussed on lower priority slices. An example of a high priority logical network (such as a slice) would be one serving critical infrastructure, such as a hospital. In the event of a need to compromise QoS for one or more slices, the slice serving the hospital should be prioritized over the slice serving residential area irrespective of any SLAs. In a factory, a production line slice may have higher priority than other slices, such as those serving a staff recreation area.

This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. For the avoidance of doubt, the scope of the claimed subject matter is defined by the claims.

It is an object of the present disclosure to provide increased protection and improved reliability and resource availability to enhance high priority slice QoS.

Embodiments of the disclosure aim to provide methods and apparatus that alleviate some or all of the problems identified above.

An aspect of the disclosure provides a network operation method. The network operation method comprises activating a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component of the physical network infrastructure. The step of activating the high priority logical network comprises generating an energy symbol, wherein the energy symbol is included in transmission frames to control component configuration adjustment. The network operation method further comprises controlling the operation of the high priority logical network on the physical network infrastructure. The ability of the high priority logical network to adjust component configurations using the energy symbol facilitates improved reliability and QoS protection for the high priority logical network. The configuration adjustment may comprise altering a power level of a base station, the base station forming part of the physical network infrastructure. In this way, the method may ensure that the base station can support the high priority logical network.

The configuration adjustment may comprise controlling the scheduling of data on further logical networks, other than the high priority logical network, using the physical network infrastructure. By adjusting the scheduling of data of further logical networks that share the physical network infrastructure with the high priority logical network, data capacity may be made available for the use of the high priority logical network.

The high priority logical network may cover a certain geographical area, which may vary with time and which may be dictated by the position of a user of the high priority logical network. Accordingly, the coverage provided by the high priority logical network may be tailored to meet the needs of users.

The high priority logical network may be activated by a core network node, which may use a machine learning agent to determine an adjustment of a component of the physical network to perform, wherein the machine learning agent may receive as an input a current status of physical network infrastructure components and may output a suggested component configuration adjustment. Use of a machine learning agent to determine component adjustments may allow the high priority logical network to be supported in an efficient way, with resource wastage minimised.

A further aspect of the disclosure provides a Core Network Node, CNN, configured to perform a network operation method. The CNN comprises an activation module configured to activate a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component of the physical network infrastructure. The activation of the high priority logical network comprises generation of an energy symbol, to be included in transmission frames to control physical network component configuration adjustment. The CNN further comprises a controller configured to control the operation of the high priority logical network on the physical network infrastructure. A still further aspect of the disclosure provides a Core Network Node, CNN, configured to perform a network operation method. The CNN comprises processing circuitry and a memory containing instructions executable by the processing circuitry. The CNN is operable to activate a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component of the physical network infrastructure. The activation of the high priority logical network comprises generation of an energy symbol, to be included in transmission frames to control physical network component configuration adjustment. The CNN is further operable to control the operation of the high priority logical network on the physical network infrastructure.

The CNNs may provide advantages as discussed in the context of the network operation method.

Further aspects provide systems and computer-readable media comprising instructions for performing the methods set out above.

Brief Description of Drawings

For a better understanding of the present disclosure, and to show how it may be put into effect, reference will now be made, by way of example only, to the accompanying drawings, in which:

Figure 1 is a schematic diagram of a typical NSI lifecycle;

Figure 2A is a diagram of a resource grid for a slot;

Figure 2B is a diagram of a resource grid illustrating an energy symbol in accordance with aspects of embodiments;

Figure 3 is a flowchart illustrating a method in accordance with aspects of embodiments;

Figure 4A is a schematic diagram of a CNN in accordance with aspects of embodiments;

Figure 4B is a schematic diagram of a CNN in accordance with further aspects of embodiments;

Figure 5A is a schematic diagram of a network node in accordance with aspects of embodiments;

Figure 5B is a schematic diagram of a network node in accordance with further aspects of embodiments;

Figure 6 is a conceptual diagram of a system for implementing reinforcement learning;

Figure 7 is a diagram illustrating activities at different levels of a telecommunications system; and

Figures 8A, 8B and 8C are a scheduling diagram indicating steps that may be performed in accordance with aspects of embodiments; Detailed Description

For the purpose of explanation, details are set forth in the following description in order to provide a thorough understanding of the embodiments disclosed. It will be apparent, however, to those skilled in the art that the embodiments may be implemented without these specific details or with an equivalent arrangement.

Examples of high priority logical networks (which may be, for example, 3GPP 5G network slices) serving critical infrastructure include a logical network serving a hospital and a logical network serving a production line in a factory, as referred to above. Both of these examples of logical networks cover essentially constant geographical areas; the coverage area of a logical network serving a hospital would not be expected to vary on a short timescale as the geographical extent of the hospital would remain the same. As such, the base stations (which may be 4^th Generation, 4G, Evolved Node Bs, eNB, or 5^th Generation, 5G, next Generation Node Bs, gNBs, for example) used to provide coverage may be selected from a small, constant set; any time the logical network is active, the physical infrastructure used to support the logical networks may be selected from a small and constant group of components. Aspects of embodiments of the invention relate to a further type of a high priority logical networks for which it is desirable to provide a guaranteed QoS; logical networks serving users that are moving with respect to the physical network infrastructure, for example, emergency vehicles. Logical networks serving moving users may be referred to as dynamic logical networks (as contrasted to the essentially static logical networks serving hospitals, for example), as the geographical area covered by the logical network and hence the physical infrastructure used to provide coverage will vary rapidly over time. This can be understood by considering the example of an ambulance attending an emergency call; the ambulance may travel at a significant speed relative to physical network infrastructure (such as base stations) and accordingly may transition between the coverage areas of different base stations rapidly. An ambulance attending an emergency call is an example of a user moving rapidly relative to physical network infrastructure, however dynamic slices may also cover geographical areas that vary at a slower rate. As an example of a geographical area with a slower rate of variation; a slice may be used to provide coverage to an area of a city that is on fire, such that emergency personnel and equipment are provided with connectivity. If the fire expands, contracts or moves, the dynamic geographical area covered by the slice can also be altered accordingly.

Current systems may not adequately provide guaranteed e2e QoS for high priority slices, including high priority slices serving essentially constant geographical areas but with particular reference to dynamic high priority slices.

Aspects of embodiments may provide proactive configuration for a high priority (potentially critical) logical network, which may be dynamic in nature.

A given logical network may be referred to as “high priority” when it is determined to be more important to maintain the QoS for the given logical network than the average importance of logical network using a particular physical network, that is, the logical network QoS is of above average importance among the logical networks using a physical network. Additionally or alternatively, each logical network priority may be indicated using an alphanumeric indicator, for example, logical networks of priority level “3” may be of higher priority than those of priority level “2”, and so on. Where alphanumeric indicators are used, a logical network which is of “high” priority may be identified as one having a priority level of at least a predetermined indicator, for example, at least a priority level of 2. In some aspects of embodiments, this may result in a majority of the logical networks using a physical infrastructure being considered to be of high priority. As will be appreciated by those skilled in the art, alternative systems may also be used to determine and/or indicate logical network priority. A network controller (which may form part of a 3GPP 5G network, for example) may deem it of more importance to maintain the QoS for a high priority logical network that to maintain the QoS for a low priority (less important) logical network. The relative priorities of logical networks operating using physical network infrastructure may be allocated by any suitable source, for example, a network controller, a component of the physical network infrastructure, a CNN, a MVNO, and so on.

The high priority slice may have a higher priority level than other slices using the physical network infrastructure (in some cases the slice may be a critical slice and/or may have the joint or sole highest priority of all slices using the physical infrastructure). In order to provide resources for the high priority slice, configurations of components forming part of the physical infrastructure may be adjusted. As an example of a physical network component adjustment, redundant power systems of base stations may be placed into a higher state of readiness to ensure service is provided for the slice (for example, to increase reliability of coverage in the slice path). In this way, the QoS of the e2e downstream may be improved throughout the lifecycle of the high priority slice. The components of the physical infrastructure that are adjusted are not necessarily themselves physical components, as will be appreciated by those skilled in the art software based or virtual components may also form part of a physical network infrastructure, and these components may be adjusted in addition or alternatively to physical components. In some aspects of embodiments adjustments of the physical network components can comprise ensuring that the physical network components are at or above a pre-set minimum level, for example, a pre-set minimum power level, a pre-set available capacity, and so on. If the review determines that the components are already at or above the minimum level, no further adjustment is required. By contrast, if the review determines that the components are not already at or above the minimum level, further adjustments (activating backup power supplies, data traffic management) are implemented.

In some aspects of embodiments, the adjustments to physical network component configurations may be effected using one or more energy symbols, for example, 5G energy symbols. Where an energy symbol is referred to herein in the singular, this should be understood to mean one or more energy symbols unless stated otherwise.

As will be appreciated by those skilled in the art, where an energy symbol is used in the adjustment of physical network component configurations, it is necessary to generate the energy symbol and propagate the energy symbol through the physical network. The energy symbol may therefore be generated on activation of the high priority logical network, and included in transmission frames to propagate the energy symbol through the physical network. In aspects of embodiments where the high priority logical network is activated by a Core Network Node (CNN, which may be a 3GPP 5G CNN), the CNN may also be responsible for generating the energy symbol. Alternatively, the energy symbol may be generated in another physical component, such as a base station.

Where present, a Network Function Virtualisation (NFV) or Service Level Agreement (SLA) orchestrator within a CNN may send an energy symbol based on SLA requirements for the high priority logical slice. The orchestrator may be active during the lifecycle of the high priority logical network. Where the high priority logical network provides dynamic coverage (from a start point to an end point), the energy symbol may contain information about component configurations along an end to end (e2e) route path of the high priority logical network, encompassing radio access network (RAN), core network and transport level aspects. The energy symbol may enable and improve the physical infrastructure along a route to be taken by a user (for example, an ambulance), as discussed in greater detail below.

Where the high priority logical network is implemented in a 3GPP network, such as a 5G network, there may be available (that is, empty or partially empty) transmission slots that are not used for variable signalling or reference signals (for example). Where slot capacity is available, this capacity may be used for the energy symbol. Figure 2A is a diagram showing an example of a resource grid for a slot; Figure 2A can be found in Document [4] as Figure 5.2.1. As can be seen from Figure 2A, each transmission slot comprises a number of resource elements (identified by indexes k and I in Figure 2). A number of the resource elements form a symbol (specifically a Single Carrier Frequency Division Multiple Access, SC-FDMA, symbol in the example illustrated in Figure 2A). The energy symbol may be transmitted using available resource elements. The energy symbol may consist of several bits; the number of bits may be determined by a modulation scheme in use in a given system in which the energy symbol is used. In non-limiting examples of the modulation schemes influence on the number of bits used for the energy symbol: where Quadrature Phase-Shift Keying (QPSK) is in use, the energy symbol may use 2 bits; where 16 point Quadrature Amplitude Modulation (16QAM) is in use, the energy symbol may use 4 bits; and where 64 point Quadrature Amplitude Modulation (64QAM) is in use, the energy symbol may use 6 bits. Further, the information included in the energy symbol may vary depending on the specific needs of the system, including the modulation scheme used. Figure 2B is a modified version of Figure 2A, which illustrates the energy symbol occupying resource elements. In the example illustrated in Figure 2B, the energy symbol occupies a single resource element in the resource grid; the energy symbol is transmitted using available resource elements and may occupy larger numbers of resource elements. In further aspects of embodiments, the energy symbol may be propagated through a network (including a 3GPP network, such as a 5G network) using any other suitable transmission means.

Figure 3 is a flowchart of a method in accordance with aspects of embodiments.

The method may be performed by any suitable apparatus. Examples of suitable apparatus for performing the method shown in Figure 3 are the CNNs 40A and 40B shown schematically in Figure 4A and Figure 4B respectively; the CNNs 40A and 40B may collectively be referred to using reference sign 40. As discussed above, the method may also be performed by any other suitable component or components, such as a further network component. The CNN 40A as shown in Figure 4A may execute steps of the method in accordance with a computer program stored in a memory 42, executed by a processor 41 in conjunction with one or more interfaces 43. The CNN 40B may execute steps of the method using activation module 44 and controller 45. The CNNs 40A and 40B may also be configured to execute the steps of other aspects of embodiments, as discussed in detail below.

The high priority logical network may be activated when required, and may remain active until no longer required. Clearly the need for the high priority logical network will vary depending on the particular configuration of a given system, for example, the user or users for which the high priority logical network is intended to provide coverage. In some aspects of embodiments, such as the aspect of an embodiment illustrated by the flowchart of Figure 3, the high priority logical network is used to provide coverage in a critical situation. Examples of critical situations can include localised emergencies, including those relating to emergency services as discussed above (ambulance journeys, fires, etc.; the term localised encompasses emergencies occurring in the geographical area for which coverage is provided by the physical network infrastructure), but can also include other situations where a primary coverage mechanism for a required system (providing coverage for a hospital or factory, for example) fails and it is necessary to ensure coverage is provided in another way. A failure of a primary coverage mechanism may occur when one or more components of a physical network infrastructure are damaged or fail, and it is necessary to prioritise coverage for some users over others. In aspects of embodiments the high priority logical network may activated not in response to a critical situation; returning to the example of a hospital, the slice providing coverage may be a high priority logical network. Further examples include automotive critical applications (as may arise through increase use of connected automobiles), factory production lines (where disruption of coverage may result in damage or losses), UE critical applications, and so on. Typically, the high priority logical network is activated in an end-to-end manner, that is, in RAN, core and transport portions of the network.

In step S301 , the CNN 40 receives a notification of a critical situation. The notification may be received directly from a local user, from a control centre (which may be a fire brigade, hospital, or police control centre), or from another source such as an operator. Returning to the example of an ambulance journey, a notification from a local user could be triggered by the crew of the ambulance indicating a need for coverage (provided by a high priority slice) for a journey to a patient or to hospital. The notification could be received by the CNN 40 via a control centre (such as an ambulance control centre); and the notification could also be triggered by a control centre responsible for dispatching the ambulance rather than triggered by the ambulance directly. Examples of users may include human users (police officers, doctors, fire services, etc.), autonomous vehicles such as ambulances, or fire engines, and so on. Allowing a local user to directly trigger activation of the high priority logical network may ensure that the network can promptly be made available when required by the local user, but may also increase the number of inadvertent or inappropriate activations of the high priority logical network; in some aspects of embodiments local users may therefore not have the ability to trigger the activation of the high priority logical network and this ability may be reserved for control centres.

In some aspects of embodiments, the high priority logical network may not be activated in response to a received notification of a critical situation, for example, the CNN 40 may activate the high priority logical network after detecting a need for the high priority logical network, potentially through monitoring of QoS values.

When the CNN 40 becomes aware of the need for a high priority logical network (either because of receiving a notification as shown in S301 , or in another way), the CNN 40 then activates the high priority logical network, as shown in S302 of Figure 3. The activation of the high priority logical network may comprise generation of an energy symbol, as discussed above. In addition to generating the energy symbol, the CNN may cause the energy symbol to be included in transmission frames, thereby allowing the energy symbol to propagate through the physical network. In this way, the CNN may control the high priority logical network (see S303).

The control of the high priority logical network may comprise adjustments to components of the physical network infrastructure. Two examples of physical network infrastructure component adjustment are shown in S303A and S303B of Figure 3. The components of the physical network infrastructure that are caused to execute adjustments may be network nodes, such as network nodes 50 as shown in Figure 5A and Figure 5B. Network nodes 50 may be further core network nodes, base stations (including distributed base stations), low power nodes serving micro or pico cells, or other types of network node, and may form part of a 3GPP network such as a 4G or 5G network. The network node 50A as shown in Figure 5A may execute steps of the method in accordance with a computer program stored in a memory 52, executed by a processor 51 in conjunction with one or more interfaces 53. The network nodes 50B may execute steps of the method using power controller 55 and data traffic controller 56, based on transmissions received by receiver 54.

The network nodes 50 may also be configured to execute the steps of other aspects of embodiments, as discussed in detail below

As shown in S303A, the energy symbol may cause the power level of one or more base stations to be altered. Where the power level of a network node 50B as shown in Figure 5B is to be adjusted, the adjustment may be effected by the power controller 55. Although the energy symbol may cause the power level of a given base station to be reduced; the adjustments to components are generally used to increase resources available to the high priority logical network, so more commonly the power level of a given base station is caused to be increased. The increase in the power level of a base station may comprise drawing additional power from a mains power grid, activating a backup power supply (such as a generator, battery reserves, solar power unit, and so on). Raising the power level of a given base station may comprise switching the base station from a dormant to active state. This may be of particular relevance where base station in question is a reserve base station, which may typically be used to provide additional capacity in support of other base stations during periods of high data traffic. The energy symbol may cause such a reserve base station to enter an active state, to ensure QoS for the high priority logical network.

Although a given base station may be caused to remain in a higher power state permanently, typically the raised power state is provide only when required by the needs of the high priority logical network; when the raised power level is no longer required by the high priority logical network, the base station may return to a lower power level.

A further example of an adjustment to components of the physical network infrastructure is shown in S303B; the energy symbol may alter the scheduling of data traffic. Where the data traffic scheduling of a network node 50B as shown in Figure 5B is to be adjusted, the adjustment may be effected by the data traffic controller 56. Where physical network infrastructure components that may be used to provide coverage for the high priority logical network do not have capacity available due to the data traffic requirements of one or more other logical networks, a CNN 40 may control the scheduling of the data traffic of the one or more other logical networks such that capacity for the high priority logical network is made available, such that QoS for the high priority logical network may be guaranteed. The scheduling control may comprise various different options, one or more of which may be implemented for other logical networks using the physical network infrastructure components. Examples of scheduling control include postponing transmission of data for other logical networks, deprioritizing the data of the other logical networks in favour of the data for the high priority logical network, shifting data between other logical networks, and discarding data of other logical networks. The selection of exactly how to control the data scheduling may be taken based on the specific status of physical network infrastructure, as discussed in more detail below.

In order to determine what actions to take to ensure QoS for a high priority logical network, for example, what adjustments of components of the physical network infrastructure should be performed, the CNN 40 may use modelling of the predicted response of the physical network infrastructure to potential adjustments. The modelling may be performed using a machine learning agent (MLA), which may be located within the CNN 40 or within a further network component connected to the CNN 40 (for example, a further CNN). The modelling of the system may use a Markov Decision Process (MDP) to optimise the network model. A MDP uses a tuple that contains states, rewards, actions, and an unknown transition probability matrix. The probabilities for the actions are learned by applying actions to the environment and receiving a feedback reward. The MDP is used to optimize over a sequence of state and action pairs by maximising the cumulative reward over a period of time..

Any suitable machine learning agent may be used; in some aspects of embodiments a neural network (NN), optionally a deep NN, trained using a reinforcement learning (RL) training mechanism may be implemented. RL allows a MLA, which may be based on a neural network, to learn by attempting to maximise a reward for a series of actions utilising trial-and-error. Examples of reinforcement learning techniques which may be used in aspects of embodiments include Quality Learning (Q-Learning) and Deep Q-learning (DQN) . Q-Learning is based on finding the policy that maximize a cumulative reward obtained in successive steps, starting from an initial state. DQN is an algorithm that combines RL with deep NNs in order to allow the RL agent to learn over complex and high-dimensional environment. The basic idea is to approximate the Q function through deep NNs and, after collecting data and storing it a memory, to learn the optimal actions over a period of time.

Figure 6 is a conceptual diagram of a system for implementing RL. In the architecture shown in Figure 1 , an agent receives data from, and transmits actions to, the environment which it is being used to model/control. For a time t, the agent receives information on a current state of the environment S_t. The agent then processes the information S_t, and generates an action to be taken A_t. This action is then transmitted back to the environment and put into effect. The result of the action is a change in the state of the environment with time, so at time t+1 the state of environment is S_t+i. The action also results in a (numerical) reward R_t+i, which is a measure of effect of the action A_t. The changed state of the environment S_t+i is then transmitted from the environment to the agent, along with the reward R_t+i. Figure 6 shows reward R_t being sent to the agent together with state S_t; reward R_t is the reward resulting from action An, performed on state Sn. When the agent receives state information S_t+i this information is then processed in conjunction with reward R_t+i in order to determine the next action A_w, and so on. The actions are selected by the agent from a number of available actions with the aim of maximising the cumulative reward. In the context of the present modelling, the state of the environment may be the current configuration of the physical network components (congestion levels, available connections, data to be sent, component power levels, and so on). The action may be a determination of an adjustment to make (shifting traffic between logical networks, for example), and the reward may be determined based on KPI variation for a high priority logical network. Higher reward values may represent more positive effects (such as increased QoS reliability for a high priority logical network).

In aspects of embodiments, the RL trained MLA may receive as inputs the status of some or all of the physical network components. The RL trained system is used as a predictor, while a high priority logical slice is active, of optimal actions to be performed. The system may be used to predict, for example, when one or more network components are likely to become overloaded, and may then be used to determine traffic movement or power system configuration adjustments of network components. At each time step of the model (when an action is selected), an observation is made and reward received based on the KPI change. The rewards may be normalized in the interval [-1 ,1], with -1 indicating undesired performance of the system (for example, QoS failure) and 1 desired performance, respectively. The measured reward is determined according to the corresponding measured KPI. Available actions may include: network level actions, increasing reliability and availability by choosing a higher MCS (Modulation Coding Scheme), requesting a UE to change its operator, in case one operator is becoming overloaded; change configuration of a power system, and so on. RL may be used to “catch” failed actions for prediction in the slice, based on negative outcomes such as communication link failures. The RL enables the MLA to suggest a the network path and “catch” all actions that are not successful actions, based on quality of transitions.

The data from the network for training the MLA may be received over a pre-defined horizon (for example every 24 hours). The data are composed of sequences of (s, a, r, s’) tuples which are stored in a connected memory; s is the current state configuration of the system, a is the action on that state, r is the measured reward according to the corresponding measured KPI; and s’ denotes the next state. The data are defined over samples of each time step that the measurement are taken (time sequences). Once trained using RL training based on the received data, the MLA may be used to predict the optimal actions for different network configurations, thereby ensuring the stability of the requested KPI of the slice. Upon receiving the current status of the network (preferably including information on the status of individual network components) that are currently used to support a high priority logical network, or which may be used to do so, the MLA may output suggested configuration adjustments for one or more components and/or of the network in general. The suggested adjustments from the MLA may be based a predicted response of the physical network components (and potentially also of the physical network as a whole) to the adjustment, and the adjustment may ensure a given QoS threshold is met for the high priority logical network.

In some aspects of embodiments, particularly where the high priority slice is dynamic and provides coverage for a geographical area that varies with time, energy symbols may be used to adjust configurations of physical network components in advance of the arrival of the high priority logical network. In order to provide coverage for a geographical area, a coverage area of a high priority logical network may be dictated by the geographical area. Where the high priority logical network covers an essentially static geographical area, this may easily be achieved by having the high priority logical network hosted by physical network infrastructure components providing coverage for that area (such as 3GPP 4G or 5G base stations). Where the high priority logical network covers a dynamic geographical area, this may require the physical network infrastructure components used to vary. In particular, where a high priority logical network provides coverage for a moving user, the geographical area covered by the high priority logical network may be dictated by the position of the user, for example, may be centred on the user so as to move with the user. Where the high priority logical network is used to provide coverage for a moving user (eg a vehicle such as an ambulance which may be an autonomous vehicle, or a person), the energy symbol may be used to adjust configurations of physical network components such as base stations prior to those network components being used to provide coverage for the high priority logical network, essentially preparing the components in advance to support the high priority logical network.

The route to be taken by the user may be known or predicted in advance, for example, because the user is a medical vehicle or personnel responding to a distress call, the location of the medical vehicle or personnel and the location from which the distress call originated are both known (based on Global Positioning System, GPS, data or wireless triangulation, for example), and a route between the user and destination has been plotted (potentially using navigational software). Where the route is known (predicted), the network components providing coverage along the route may be activated in advance. Alternatively, where a route to be taken by a user is not known, either because the route was not plotted in advance or because a user has deviated from a plotted route, the configurations of components within a certain distance (which may be set by the user, by a control centre, within the network, and so on) of the user may be reviewed for adjustment on a predetermined frequency.

Where the route to be taken by a user (and hence high priority logical network providing coverage for the user) is known, the configurations of all physical network components along the route may be adjusted as soon as the high priority logical network is activated and route information available. Alternatively, physical network components within a certain distance of the user along the route and/or a certain number of physical network components may be adjusted, but not necessarily all of the physical network components along the route may be adjusted as soon as the high priority logical network is activated and route information available. Further, once the high priority logical network has passed out of the range of physical network components, these components may either return to a previous configuration or retain the adjusted configuration.

If aspects of embodiments are configured such that a certain number of physical network components (such as base stations, for example) in advance of a moving user are adjusted, a RL trained MLA may be used to determine an optimal number of components to adjust. As adjusting the configuration of components may require additional use of resources (for example, because a backup power supply has been activated), or may make resources unavailable for other logical networks (for example, because other data traffic has been suspended to ensure capacity is available for the high priority logical network), determining the minimum number of components to adjust such that the QoS for the high priority logical network is ensured may result in the least disruption to the remainder of the logical networks using the physical network infrastructure. In aspects of embodiments wherein the components of the physical network under consideration are base stations, the optimal number of base stations to reconfigure may be determined using the RL methods discussed above wherein the state of the environment encompasses the capacity of the next H base stations (where H may be the number of base stations between the user and the destination) and the physical speed of the user (which determines how quickly the user is likely to reach the coverage area of a give base station). The reward is determined using QoS equation 1 , where QOSR is the required Quality of Service and the equation uses KPIs measuring throughput and latency:-

Equation

The value of QOSR is 2; QoS is equal to or greater than 2 when the required resources are available, and less when they are not. The action is a decision as to how many base stations in advance of the user along the route to reconfigure; a value between 0 and H. Training the MLA using RL based on the above information allows a decision as to the adjustment of base station configurations to be made by the network.

As discussed in detail above, aspects of embodiments may be active at different levels of a telecommunications system to ensure that a QoS for a high priority logical network is protected. The different levels can include the network level, radio link level, energy source level and operator level, as illustrated in Figure 7.

On a network level, a MLA trained using network data from a given number n days of network activity may be used to model network responses, communications between different network components (such as base stations and back haul components, for example), and so on. The MLA may then be used to suggest what adjustments of components of the physical network infrastructure should be performed.

On a radio link level, the energy symbol may be propagated throughout the network, causing component reconfigurations to be implemented. The energy symbol may consist of, for example, 2, 3 or 6 bits of information (wherein the number of bits in the energy symbol may be determined by the modulation scheme used, as discussed above). The energy symbol may be included in available transmission slots of various carriers and/or frequency bands, and may be sent using different routes including macro or small cell networks. As discussed above, the energy symbol may be sent in advance of the high priority logical network, particularly in cases where the high priority logical network provides dynamic (geographically varying) coverage.

On a component level, individual components may be reconfigured to ensure QoS for the high priority logical network. As illustrated in Figure 7, this can include activating reserve energy sources for components. Taking the example of a radio base station (RBS), the RBS controller is aware of what hardware entities are activated and not activated as this activity is managed inside the RBS controller. If the energy symbol arrives (from the radio link level), and requires reconfiguration, then the RBS controller may activate a further power source such as a diesel generator.

On an operator level, the high priority logical network may change to another operator if one operator becomes overloaded or is at risk of overload, or where operator equipment becomes inoperable. Based on the SLA requirements, the high priority logical network receives priority scheduling, so less important data/or traffic relating to lower priority logical networks or non SLA important scheduling will be postponed or deprioritized, or moved to another operator. The high priority logical network may also be switched between operators, for example, in the event of interference.

Accordingly, QoS for the high priority logical network may be reliably ensured.

The following example explains how aspects of embodiments may be implemented, using the situation of a user that is an ambulance travelling from a starting point (for example, a hospital) to a destination point (a patient location). Accordingly, the high priority logical network (network slice instance, NSI) in this example is dynamic, covering a geographical area that varies with time. A scheduling diagram indicating steps corresponding to the example that may be performed is shown in Figure 8A. Figure 8B and Figure 8C; the steps are numbered in Figure 8.

An emergency call centre may receive an emergency call, and transfer the data about a location where an ambulance is required into an ambulance GPS system. The ambulance system enables the route path for the ambulance. A NSI is enabled, to provide coverage for the ambulance along the route. In steps 1-4, the emergency call centre operator sends a request to a mobile network operator (MNO) to activate the NSI, the call centre operator also sends the location information for the start and end of the ambulance journey to the MNO (in step 3).

The RAN network (in this example a 3GPP 5G network) knows the position of the ambulance so the starting point is known, also the destination point (the patient location) is known. In between the two points, there are a number of network components, such as 5G radio base stations, that provide coverage. The MNO processes the request from the operator in step 5, and activates the NSI. Based on the SLA and on the position of the ambulance, the energy symbol is activated along the route to the destination (steps 6-8, indicating base stations 1 , 2 and n+1 receiving the activation). The energy symbol activates different component configurations along the route. Once activation is complete, the base stations send an acknowledgement back to the MNO (step 9), which informs the operator that the NSI is granted (step 10).

The ambulance (on which its path will have a critical service) may have a transmitter, to enable service from the NSI. The NSI may follow the ambulance along its route, adjusting configurations in advance of all RBSs along the ambulance path until the destination is reached (steps 11 to 26). The NSI may then be disabled (steps 27 and 28). The NSI may later be enabled on the return journey (for example, to a hospital), in which case some or all of the process of Figures 8A to 8C may be repeated. Alternatively, in order to reduce the number of NSI activations and deactivations, the NSI may remain active while the ambulance is at the destination, and then provide coverage for the return journey. Activation and deactivation of the NSI may be dependent on the ambulance position, starting point, ending point, waiting time at destination, and preparing the person in need to get into the ambulance. The ambulance service people may enable and disable the NSI, via an in ambulance control.

In general, the various exemplary embodiments may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. For example, some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device, although the disclosure is not limited thereto. While various aspects of the exemplary embodiments of this disclosure may be illustrated and described as block diagrams, flow charts, or using some other pictorial representation, it is well understood that these blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

As such, it should be appreciated that at least some aspects of the exemplary embodiments of the disclosure may be practiced in various components such as integrated circuit chips and modules. It should thus be appreciated that the exemplary embodiments of this disclosure may be realized in an apparatus that is embodied as an integrated circuit, where the integrated circuit may comprise circuitry (as well as possibly firmware) for embodying at least one or more of a data processor, a digital signal processor, baseband circuitry and radio frequency circuitry that are configurable so as to operate in accordance with the exemplary embodiments of this disclosure.

It should be appreciated that at least some aspects of the exemplary embodiments of the disclosure may be embodied in computer-executable instructions, such as in one or more program modules, executed by one or more computers or other devices. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types when executed by a processor in a computer or other device. The computer executable instructions may be stored on a computer readable medium such as a hard disk, optical disk, removable storage media, solid state memory, RAM, etc. As will be appreciated by one of skill in the art, the function of the program modules may be combined or distributed as desired in various embodiments. In addition, the function may be embodied in whole or in part in firmware or hardware equivalents such as integrated circuits, field programmable gate arrays (FPGA), and the like.

References in the present disclosure to “one embodiment”, “an embodiment” and so on, indicate that the embodiment described may include a particular feature, structure, or characteristic, but it is not necessary that every embodiment includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to implement such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described. It should be understood that, although the terms “first”, “second” and so on may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and similarly, a second element could be termed a first element, without departing from the scope of the disclosure. As used herein, the term “and/or” includes any and all combinations of one or more of the associated listed terms.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the present disclosure. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises”, “comprising”, “has”, “having”, “includes” and/or “including”, when used herein, specify the presence of stated features, elements, and/or components, but do not preclude the presence or addition of one or more other features, elements, components and/ or combinations thereof. The terms “connect”, “connects”, “connecting” and/or “connected” used herein cover the direct and/or indirect connection between two elements.

The present disclosure includes any novel feature or combination of features disclosed herein either explicitly or any generalization thereof. Various modifications and adaptations to the foregoing exemplary embodiments of this disclosure may become apparent to those skilled in the relevant arts in view of the foregoing description, when read in conjunction with the accompanying drawings. However, any and all modifications will still fall within the scope of the non-limiting and exemplary embodiments of this disclosure. For the avoidance of doubt, the scope of the disclosure is defined by the claims.

Claims

1. A network operation method, comprising:

Activating (S302) a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component (50) of the physical network infrastructure, wherein the step of activating the high priority logical network comprises generating an energy symbol, and wherein the energy symbol is included in transmission frames to control component configuration adjustment; and controlling (S303) the operation of the high priority logical network on the physical network infrastructure.

2. The network operation method of claim 1 , wherein the high priority logical network is activated in response to receiving (S301) a notification of a critical situation.

3. The network operation method of claim 2, wherein the critical situation results from a localised emergency.

4. The network operation method of any of claims 2 and 3, wherein the high priority logical network is activated in response to a notification from a control centre, or wherein the high priority logical network is activated in response to a notification from a local user.

5. The network operation method of any preceding claim, wherein the configuration adjustment comprises altering (S303A) a power level of a base station (50), the base station (50) forming part of the physical network infrastructure.

6. The network operation method of claim 5, wherein the base station (50) is a reserve base station, and wherein the altering (S303A) of the power level comprises switching between a dormant state and an active state.

7. The network operation method of any preceding claim, wherein the configuration adjustment comprises controlling (S303B) the scheduling of data on further logical networks, other than the high priority logical network, using the physical network infrastructure.

8. The network operation method of claim 7, wherein the controlling (S303B) of the scheduling of the data on further logical networks comprises one or more of: postponing data transmission, deprioritizing the data, discarding the data and switching the data to a different logical network.

9. The network operation method of any preceding claim, wherein the high priority logical network covers a certain geographical area.

10. The network operation method of claim 9, wherein the position of the certain geographical area is dictated by the position of a user of the high priority logical network.

11. The network operation method of any of claims 9 and 10, wherein the certain geographical area varies with time.

12. The network operation method of any of claims 9 to 11 , wherein the configurations of one or more physical network infrastructure components (50) along a predicted route of the certain geographical area are adjusted.

13. The network operation method of any of claims 11 and 12, wherein the configurations of one or more physical network infrastructure components (50) within a certain distance of the position of the user are adjusted.

14. The network operation method of any of claims 12 and 13, wherein the configurations of a certain number of physical network infrastructure components (50) are adjusted.

15. The network operation method of any preceding claim, wherein the high priority logical network is activated by a Core Network Node, CNN (40).

16. The network operation method of claim 15, wherein the CNN (40) uses a machine learning agent, MLA, to determine an adjustment of a component (50) of the physical network infrastructure to perform, wherein the MLA receives as an input a current status of physical network infrastructure components (50) and outputs a suggested component configuration adjustment.

17. The network operation method of claim 16, wherein the MLA determines the suggested component configuration adjustment based on a predicted response of the physical network infrastructure to the adjustment.

18. The network operation method of any of claims 16 and 17, wherein the suggested component configuration adjustment is predicted to cause the high priority logical network to satisfy a particular Quality of Service, QoS, threshold.

19. The network operation method of any of claims 16 to 18, wherein the MLA is trained using reinforcement learning, the reinforcement learning utilising data on physical network infrastructure adjustments and results as training data.

20. The network operation method of any preceding claim, wherein the high priority logical network is allocated a priority that is equal to or higher than the priority of every other logical network operating using the physical network infrastructure.

21. The network operation method of any preceding claim, wherein the physical network is a 5^th Generation, 5G, network and the high priority logical network is a 5G network slice.

22. A Core Network Node, CNN, (40B) configured to perform a network operation method, the CNN (40B) comprising: an activation module (44) configured to activate (S302) a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component (50) of the physical network infrastructure, the activation of the high priority logical network comprising generation of an energy symbol, to be included in transmission frames to control physical network component configuration adjustment; and a controller (45) configured to control the operation of the high priority logical network on the physical network infrastructure.

23. The CNN (40B) of claim 22, further configured to activate the high priority logical network in response to receiving (S301) a notification of a critical situation.

24. The CNN (40B) of claim 23, wherein the critical situation results from a localised emergency.

25. The CNN (40B) of any of claims 23 and 24, wherein the high priority logical network is activated in response to a notification from a control centre, or wherein the high priority logical network is activated in response to a notification from a local user.

26. The CNN (40B) of any of claims 22 to 25, wherein the configuration adjustment comprises altering (S303A) a power level of a base station (50), the base station (50) forming part of the physical network infrastructure.

27. The CNN (40B) of claim 26, wherein the base station (50) is a reserve base station, and wherein the altering (S303A) of the power level comprises switching between a dormant state and an active state.

28. The CNN (40B) of any of claims 22 to 27, wherein the configuration adjustment comprises controlling (S303B) the scheduling of data on further logical networks, other than the high priority logical network, using the physical network infrastructure.

29. The CNN (40B) of claim 28, wherein the controlling of the scheduling of the data on further logical networks comprises one or more of: postponing data transmission, deprioritizing the data, discarding the data and switching the data to a different logical network.

30. The CNN (40B) of any of claims 22 to 29, wherein the high priority logical network covers a certain geographical area.

31. The CNN (40B) of claim 30, wherein the position of the certain geographical area is dictated by the position of a user of the high priority logical network.

32. The CNN (40B) of any of claims 30 and 31 , wherein the certain geographical area varies with time.

33. The CNN (40B) of any of claims 30 to 32, wherein the configurations of one or more physical network infrastructure components (50) along a predicted route of the certain geographical area are adjusted.

34. The CNN (40B) of any of claims 32 and 33, wherein the configurations of one or more physical network infrastructure components (50) within a certain distance of the position of the user are adjusted.

35. The CNN (40B) of any of claims 33 and 34, wherein the configurations of a certain number of physical network infrastructure components (50) are adjusted.

36. The CNN (40B) of any of claims 22 to 35, further configured to use a machine learning agent, MLA, to determine an adjustment of a component of the physical network to perform, wherein the MLA is configured to receive as an input a current status of physical network infrastructure components, and to output a suggested component configuration adjustment.

37. The CNN (40B) of claim 36, wherein the machine learning agent is configured to determine the suggested component configuration adjustment to perform based on a predicted response of the physical network infrastructure to the adjustment.

38. The CNN (40B) of any of claims 36 and 37, wherein the suggested component configuration adjustment is predicted to cause the high priority logical network to satisfy a particular Quality of Service, QoS, threshold.

39. The CNN (40B) of any of claims 36 to 38, wherein the machine learning agent is trained using reinforcement learning, the reinforcement learning utilising data on physical network infrastructure adjustments and results as training data.

40. The CNN (40B) of any preceding claim, further configured to allocate to the high priority logical network a priority that is equal to or higher than the priority of every other logical network operating using the physical network infrastructure.

41. The CNN (40B) of any preceding claim, wherein the physical network is a 5^th Generation, 5G, network and the high priority logical network is a 5G network slice.

42. A Core Network Node, CNN, (40A) configured to perform a network operation method, the CNN (40A) comprising processing circuitry (41) and a memory (42) containing instructions executable by the processing circuitry (41), whereby the CNN is operable to: activate (S302) a high priority logical network to operate on a physical network infrastructure, wherein the high priority logical network can adjust a configuration of a component (50) of the physical network infrastructure, the activation of the high priority logical network comprising generation of an energy symbol, to be included in transmission frames to control physical network component configuration adjustment; and control (S303) the operation of the high priority logical network on the physical network infrastructure.

43. A system comprising a CNN (40) according to any of claims 22 to 42, further comprising a base station (50), wherein the base station (50) is a component of the physical network infrastructure.

44. A computer-readable medium comprising instructions which, when executed on a computer, cause the computer to perform a method in accordance with any of claims 1 to 21.