GB2576693A

GB2576693A - Cellular telecommunications network

Info

Publication number: GB2576693A
Application number: GB1809017.5A
Authority: GB
Inventors: fitch Michael; Mackenzie Richard
Original assignee: British Telecommunications PLC
Current assignee: British Telecommunications PLC
Priority date: 2018-06-01
Filing date: 2018-06-01
Publication date: 2020-03-04
Anticipated expiration: 2038-06-01
Also published as: GB2576693B; GB201809017D0

Abstract

A virtualised radio access network (also known as network slicing) includes a first virtual machine delivering a first service type and a second virtual machine delivering a second service type. Data is received indicating a performance measure of a first parameter for the first and second virtual machines. A first spectrum utilisation is determined for the first virtual machine based on the performance measure of the first parameter for the first virtual machine relative to a performance requirement for the first parameter. A second spectrum utilization is determined for the second virtual machine based on the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter. The first spectrum utilisation is compared to a target spectrum utilisation for the first service type and the second spectrum utilisation is compared to a target spectrum utilisation for the second service type. A resource allocation is configured for the first and/or second virtual machine based on the comparisons of the first and second spectrum utilisations to their respective target spectrum utilisations. A slice controller may perform the allocation steps.

Description

CELLULAR TELECOMMUNICATIONS NETWORK

Field of the Invention

The present invention relates to a cellular telecommunications network.

Background

A conventional cellular telecommunications network uses dedicated hardware components that performed specific tasks. This gave network operators fine control over its infrastructure such that it could be carefully planned and deployed. However, this was also inflexible such that all parts of the infrastructure served all forms of applications (e.g. voice, file transfer, Device-to-Device communications), rather than be tailored for providing the optimal configuration for a particular application. Furthermore, components of new generations of cellular networks were typically not backwards compatible with previous generations. To address these issues, a technique called network slicing was introduced to cellular networks.

Network slicing introduces virtualised networking infrastructure such that the functions of each hardware component of the cellular network may be run on general computing hardware (e.g. servers). This may be performed for each node in the cellular network, including the radio access network, edge network and core network. In doing so, a virtual network “slice” may be established which utilises virtual machines established on one or more nodes in the cellular network, wherein the network slice may be optimised for a particular application or to be backwards compatible with a particular technology.

For edge and core networking nodes and their respective links, a network operator may plan and deploy suitable general computing hardware to serve the expected traffic for all network slices. Furthermore, the network operator may readily upgrade the general computing hardware for these edge and core networking nodes to account for expected future traffic for all network slices. For radio access networking nodes, a network operator may plan their usage of radio spectrum across the various network slices (i.e. to allocate a first portion of radio spectrum to a first network slice and a second portion of radio spectrum to a second network slice). However, it may not be possible (e.g. due to regulatory constraints) to simply increase the amount of radio spectrum if the demands of the network slices increase. It is therefore important to allocate resources, particularly radio spectrum, between network slices carefully.

Summary of the Invention

According to a first aspect of the invention, there is provided a method of allocating a resource in a virtualised radio access network, wherein the virtualised radio access network includes a first virtual machine delivering a first service type and a second virtual machine delivering a second service type, the method comprising: receiving data indicating a performance measure of a first parameter for the first virtual machine and further indicating a performance measure of the first parameter for the second virtual machine; determining a first spectrum utilisation for the first virtual machine based on the performance measure of the first parameter for the first virtual machine relative to a performance requirement for the first parameter; determining a second spectrum utilisation for the second virtual machine based on the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter; comparing the first spectrum utilisation to a target spectrum utilisation for the first service type; comparing the second spectrum utilisation to a target spectrum utilisation for the second service type; and configuring a resource allocation for the first and/or second virtual machine based on the comparisons of the first and second spectrum utilisations to their respective target spectrum utilisations.

The data may indicate a performance measure of a second parameter for the first virtual machine and further indicates a performance measure of the second parameter for the second virtual machine, the first spectrum utilisation for the first virtual machine may be further based on the performance measure of the second parameter for the first virtual machine relative to a performance requirement for the second parameter, and the second spectrum utilisation for the second virtual machine may be further based on the performance measure of the second parameter for the second virtual machine relative to the performance requirement for the second parameter.

The first spectrum utilisation for the first virtual machine may be based on a product or sum of (i) the performance measure of the first parameter for the first virtual machine relative to the performance requirement for the first parameter, and (ii) the performance measure of the second parameter for the first virtual machine relative to the performance requirement for the second parameter, and the second spectrum utilisation for the second virtual machine may be based on a product or sum of (i) the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter, and (ii) the performance measure of the second parameter for the second virtual machine relative to the performance requirement for the second parameter.

The performance measure of the first parameter may be one of a set of performance measures of the first parameter, each relating to a UE of a plurality of UEs, and the performance measure of the second parameter is one of a set of performance measures of the second parameter, each relating to a UE of the plurality of UEs, the first spectrum utilisation for the first virtual machine may be based on a sum of the product of (i) each performance measure of the first parameter for each UE for the first virtual machine relative to the performance requirement for the first parameter, and (ii) each performance measure of the second parameter for each UE for the first virtual machine relative to the performance requirement for the second parameter, and the second spectrum utilisation for the second virtual machine may be based on a sum of the product of (i) each performance measure of the first parameter for each UE for the second virtual machine relative to the performance requirement for the first parameter, and (ii) each performance measure of the second parameter for each UE for the second virtual machine relative to the performance requirement for the second parameter.

The first and/or second spectrum utilisations may be further based on one or more of the following: cost of serving the UE, bandwidth, and coverage area.

The step of configuring the resource allocation may include one or more of the following: adjusting frame time, reallocating resource blocks, and reallocating processing resources.

The first virtual machine may be one of a plurality of first virtual machines, each being deployed on a respective networking node, for delivering the first service type, and the second virtual machine may be one of a plurality of second virtual machines, each being deployed on a respective networking node, for delivering the second service type.

According to a second aspect of the invention, there is provided a computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method of the first aspect of the invention. The computer program may be stored on a computer-readable data carrier.

According to a third aspect of the invention, there is provided a network node for a virtualised radio access network, wherein the virtualised radio access network includes a first virtual machine delivering a first service type and a second virtual machine delivering a second service type, the network node comprising: a receiver adapted to receive data indicating a performance measure of a first parameter for the first virtual machine and further indicating a performance measure of the first parameter for the second virtual machine; and a processor adapted to: determine a first spectrum utilisation for the first virtual machine based on the performance measure of the first parameter for the first virtual machine relative to a performance requirement for the first parameter; determine a second spectrum utilisation for the second virtual machine based on the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter; compare the first spectrum utilisation to a target spectrum utilisation for the first service type; compare the second spectrum utilisation to a target spectrum utilisation for the second service type; and configure a resource allocation for the first and/or second virtual machine based on the comparisons of the first and second spectrum utilisations to their respective target spectrum utilisations.

Brief Description of the Figures

In order that the present invention may be better understood, embodiments thereof will now be described, by way of example only, with reference to the accompanying drawings in which:

Figure 1 is a schematic diagram illustrating a first embodiment of a cellular telecommunications network of the present invention;

Figure 2 is a schematic diagram of a server of the network of Figure 1;

Figure 3 is a schematic diagram illustrating a first and second network slice across several nodes of the network of Figure 1;

Figure 4 is a schematic diagram of a slice controller of the network of Figure 1; and Figure 5 is a flow diagram illustrating an embodiment of a method of the present invention.

Detailed Description of Embodiments

A first embodiment of a cellular telecommunications network 1 will now be described with reference to Figures 1 to 4. As shown in Figure 1, the cellular telecommunications network includes a core networking node 10, an edge networking node 20, an access networking node 30, a plurality of UEs 40 and a slice controller 50. There may be many other core, edge and access networking nodes in the cellular network 1 (wherein there are typically more edge networking nodes than core networking nodes, and there are typically more access networking nodes than edge networking nodes), but only one of each is shown for simplicity. In this embodiment, the core, edge and access networking nodes are each implemented by a server.

A schematic diagram of a server for a core networking node 10 is shown in Figure 2. The server includes a central processor unit (CPU) 12 communicatively connected to storage 14 and an input/output (I/O) interface 16 via a data bus 18. Storage 14 can be any read/write storage device such as a random access memory (RAM) or a non-volatile storage device. An example of a non-volatile storage device includes a disk or tape storage device. The I/O interface 16 is an interface to devices for the input or output of data, or for both input and output of data. In this embodiment, the I/O device is a network connection (for connection, for example, to one or more other core networking nodes or one or more edge networking nodes), but may also include a keyboard, a mouse, or a display (such as a monitor). The CPU 12, storage 14, I/O interface 16 and data bus 18 cooperate to define a Software Defined Network (SDN) operating environment in which any networking function of a core networking node (e.g. packet data network gateways, User Plane Function (UPF), Control Plane Function (CPF), etc.) may be implemented as software. In this manner, the server may be configured to operate as any one of the core networking nodes by using the corresponding software. Furthermore, the server is configured to implement a Network Function Virtualisation (NFV) architecture such that the server may implement a plurality of virtual machines, wherein each virtual machine may provide a different SDN operating environment implementing a particular networking function that is optimised for a particular service type. For example, the server may implement a first virtual machine being configured to implement a particular networking function that is optimised for a first service type (e.g. Device-to-Device (D2D) communications) and a second virtual machine being configured to implement the same networking function but optimised for a second service type (e.g. virtual reality application).

The edge and access networking nodes are substantially similar to the server illustrated in Figure 2, such that they may also implement SDN operating environments and NFV architectures. Examples of networking functions of an edge networking node that may be implemented in any one virtual machine of a server include a Multi-Access Edge Computer (MEC) platform and UPF; and examples of networking functions of an access networking node that may be implemented in any one virtual machine of a server include Radio Access Network (RAN) Centralised Unit (CU) or Remote Unit (RU) processing.

Furthermore, it is possible to establish a plurality of virtual machines across several networking nodes (e.g. across a combination of core, edge and/or access networking nodes) which cooperate to implement a virtual cellular network. This is often referred to as a “network slice”. Figure 3 is a schematic diagram of the cellular network 1, illustrating a first and second network slice being implemented by several core, edge and/or access networking nodes. Each slice is for delivering a particular service type to one or more UEs of the plurality of UEs and each virtual machine of the network slice is configured to provide a particular networking function which is optimised for that service type. In this example, the first network slice is for delivering a D2D service, such that the network slice is configured to serve the UE with relatively high latency, relatively low bit-rate communications, and the second network slice is for delivering a virtual reality service, such that the second network slice is configured to serve the UE with relatively low latency, relatively high bit-rate communications.

As noted above, the cellular network further includes a slice controller 50. The slice controller 50 is shown in Figure 4, and includes a CPU 52, storage 54, I/O interface 56 (connected via a bus 58). Storage 54 includes a first database storing data for all forms of service types that the cellular network supports, such as voice, virtual reality, D2D, autonomous vehicles, file transfer, etc. In this embodiment, this data includes, for each service type, a target Spectrum Utility (SU) metric and a threshold SU metric. Furthermore, storage 54 includes a second database having a table identifying a requirement for each service type. The following tables illustrate this data:

Service Type	Target SU	Threshold SU
D2D (ID=1)	^Target	SjjThresh
Virtual Reality (ID=2)	^Target	SfjThresh

Table 1: Table illustrating data stored in the first database in the slice controller 50

Service Type	Mean Bit-Rate (ID=1)	Latency (ID=2)	...
D2D (ID=1)	Req^	Reql
Virtual Reality (ID=2)	Reql	Reql

Table 2: Table illustrating data stored in the second database in the slice controller 50

In this embodiment, the second database includes the requirements for each service type which applies to all UEs. However, in enhanced embodiments, the second database may include separate tables for each UE or a particular subset of UEs in the cellular network, such that the requirements for each UE or each subset of UEs may be distinct.

Storage 54 of the slice controller 50 also includes a third database storing data relating to a plurality of policies for slice management in the cellular network 1. The purpose of these databases will become clear upon review of the embodiment of the method of the invention, which will now be described with reference to Figure 5.

In the example, a first subset of UEs of the plurality of UEs use the first network slice for a D2D service. Each instance of a UE of the first subset of UEs using the first network slice for the D2D service is called an “application” on the first network slice. Furthermore, a second subset of UEs of the plurality of UEs use the second network slice for a virtual reality service. Again, each instance of a UE of the second subset of UEs using the second network slice for the virtual reality service is called an application on the second network slice. It is noted that a single UE may be a member of both the first and second subset of UEs if that UE instantiates applications on the first and second network slices (for D2D and virtual reality services respectively).

In step S1 of the method, each UE of the first and second subsets of UEs performs measurements on one or more parameters in order to produce a measurement report. These may include, for example, mean bit-rate, latency, jitter, and error rate. Each UE of the first and second subsets of UEs sends their respective measurement reports to their respective access networking nodes.

In step S3, the access networking node(s) forward this data to the slice controller 50. In step S5, the CPU 52 of the slice controller 50 processes this data to determine a Satisfaction Level (SL) for each application. This is achieved by retrieving data from its second database identifying the requirements for each parameter for the particular service type (and, in enhanced embodiments, for that particular UE), and implementing the following first function:

( Y

SL(k,l)= ^Req ¹⁺⁽4f

In this first function, the CPU 52 calculates, for each application (/) and each requirement (fc), a ratio (normalised to a number between 0 and 1) of the measurement (F?) for a particular parameter (e.g. mean bit rate) compared to the stored requirement (Fleq) for that particular parameter. Integer ε is an operator defined weighting that may be used to make some parameters more influential than others.

The CPU 52 then implements a second function:

SL(l) = Π

KI

SL(k, l)^ak-,a_k k=l

In this second function, the CPU 52 calculates an SL for each application by multiplying the SL for each requirement for a particular application. Integer a is also an operator defined weighting that may be used to make some parameters more influential than others.

The slice controller 50 therefore determines an SL indicating a performance of the first network slice at delivering the first service type to each application (i.e. each UE of the first subset of UEs) and a performance of the second network slice at delivering the second service type to each application (i.e. each UE of the second subset of UEs). This performance measure takes into account each parameter type, which may be weighted by the operator.

In step S7, the slice controller 50 calculates an SU (that is, Spectrum Utility) for each network slice by summing the SLs (calculated in step S5 for each application) across all applications. In this embodiment, the SU further takes into account the value of that service to each application (i.e. to the end user of that application), the cost of delivering that service to each application, and the bandwidth used in delivering that service. The cost is derivable from the measurement report (e.g. the mean bit rate may be converted into a cost by translating the mean bit rate into a resource usage and multiplying that by a unit cost for that resource). The CPU 52 of the slice controller 50 calculates the SU for the first and second network slices based on the following third function:

St/ = |y^L SL(Z).v(Z)/c(Z)

D 4-1(=1

In the above third function, integer B is the bandwidth of the radio spectrum used by the network slice, v is the value of that service to each application, and c is the cost of delivering that service for each application. The third function therefore sums ratios of the product of the SL for each application and its value to the end user against the cost of delivering the service to each application across all applications in a network slice.

The slice controller 50 therefore determines an SU indicating a performance of all applications across a network slice relative to the bandwidth used by that network slice. This performance metric may be used by the slice controller 50 to identify if resources in the network should be reallocated and, if so, how those resources should be reallocated. Accordingly, in step S9, CPU 52 controller retrieves the threshold and target SUs for the D2D and virtual reality service types (as the first and second network slices relate to these service types). CPU 52 then compares the SU for the first network slice to the threshold SU and target SU for D2D communications, and compares the SU for the second network slice to the threshold SU and target SU for virtual reality.

In step S11, the CPU 52 retrieves policies from the third database to identify a suitable course of action following the comparison of the SUs for the first and second network slices to the relevant target and threshold SUs. In this embodiment, the policies are:

• If the SUs for the first and second network slices are both above the target SUs, then no action is taken.

• If the SU for the first network slice is below its respective target SU, and the SU for the second network slice is above its respective target SU, then a resource of the cellular network is reallocated from the second network slice to the first network slice;

• If the SU for the second network slice is below its respective target SU, and the SU for the first network slice is above its respective target SU, then a resource of the cellular network is reallocated from the first network slice to the second network slice; and • If the SUs for both the first and second network slices are below their respective target SUs, then a resource of the cellular network is reallocated from the network slice having the greater SU to the other network slice.

Following step S11, the method loops back to step S1 (via a delay timer). The above process therefore has the effect of providing more resources to network slices that are not satisfactorily serving all their respective UEs to the requirements of the service type being delivered by that network slice. These resources are reallocated from network slices that are satisfactorily serving all their respective UEs (or, at least performing better at serving all their respective UEs). The SU metric is therefore a network slice level performance measure that is used to equitably reallocate resources. In this embodiment, these resources relate to those used by access networking nodes in the cellular network, such as resource blocks allocated to each network slice, transmit time allocated to each network slice, allocation of processing power to each network slice, etc.

By operating iteratively, the above process may apply as new applications and new network slices are dynamically added or removed from the cellular network 1. That is, at each iteration, the CPU 52 may evaluate the SL for each application (including any new applications that have been added to a particular network slice since the last iteration) and SU for each network slice (including any new network slices that have been added since the last iteration). This method may then be used to reallocate resources in a dynamic cellular network in which network slices are added/removed to the network and applications are added/removed to network slices. Furthermore, the method may also be triggered upon a request for a new application or network slice in the network such that resources may be reallocated immediately upon a change in the network slices for a network or, if the new request would render the SU for any single network slice as less than their respective threshold SU, then the request may be denied or only partially fulfilled. Thus, as shown in Figure 5, step S1 may be implemented periodically or upon a trigger condition being satisfied.

The skilled person will understand that it is non-essential that the method is practiced upon a network slice acting across multiple network nodes in the cellular network. That is, the benefits of the invention may be realised upon implementation in a single node (e.g. the access networking node alone).

Furthermore, the skilled person will understand that there are many ways in which the resources in the cellular network may be reallocated based on the comparisons of the spectrum utilisations for the different network slices to their respective target spectrum utilisations, and the above is just an example.

In the second function of the above embodiment, the satisfaction level is calculated for an application using a product of the satisfaction levels for each requirement. However, this is non-essential, and alternatively a summation may be used (for example, if the satisfaction level for each application does not depend jointly on the individual satisfaction levels for each requirement).

In the above embodiment, a cost factor is used as part of the SU calculation which is based on the cost of the resource usage in providing that service. The skilled person would understand that this cost may be multi-factor, in which each factor contributing to the cost of providing that service may be a sum or a product of those factors.

Insofar as embodiments of the invention described are implementable, at least in part, using a software-controlled programmable processing device, such as a microprocessor, digital signal processor or other processing device, data processing apparatus or system, it will be appreciated that a computer program for configuring a programmable device, apparatus or system to implement the foregoing described methods is envisaged as an aspect of the present invention. The computer program may be embodied as source code or undergo compilation for implementation on a processing device, apparatus or system or may be embodied as object code, for example.

Suitably, the computer program is stored on a carrier medium in machine or device readable form, for example in solid-state memory, magnetic memory such as disk or tape, optically or magneto-optically readable memory such as compact disk or digital versatile disk etc., and the processing device utilises the program or a part thereof to configure it for operation. The computer program may be supplied from a remote source embodied in a communications medium such as an electronic signal, radio frequency carrier wave or optical carrier wave. Such carrier media are also envisaged as aspects of the present invention.

It will be understood by those skilled in the art that, although the present invention has been described in relation to the above described example embodiments, the invention is not limited thereto and that there are many possible variations and modifications which fall within the scope of the invention.

The scope of the present invention includes any novel features or combination of features disclosed herein. The applicant hereby gives notice that new claims may be formulated to such features or combination of features during prosecution of this application or of any such further applications derived therefrom. In particular, with reference to the appended claims, features from dependent claims may be combined with those of the independent claims and features from respective independent claims may be combined in any appropriate manner and not merely in the specific combinations enumerated in the claims.

Claims

1. A method of allocating a resource in a virtualised radio access network, wherein the virtualised radio access network includes a first virtual machine delivering a first service type and a second virtual machine delivering a second service type, the method comprising:

receiving data indicating a performance measure of a first parameter for the first virtual machine and further indicating a performance measure of the first parameter for the second virtual machine;

determining a first spectrum utilisation for the first virtual machine based on the performance measure of the first parameter for the first virtual machine relative to a performance requirement for the first parameter;

determining a second spectrum utilisation for the second virtual machine based on the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter;

comparing the first spectrum utilisation to a target spectrum utilisation for the first service type;

comparing the second spectrum utilisation to a target spectrum utilisation for the second service type; and configuring a resource allocation for the first and/or second virtual machine based on the comparisons of the first and second spectrum utilisations to their respective target spectrum utilisations.

2. A method as claimed in Claim 1, wherein:

the data indicates a performance measure of a second parameter for the first virtual machine and further indicates a performance measure of the second parameter for the second virtual machine, the first spectrum utilisation for the first virtual machine is further based on the performance measure of the second parameter for the first virtual machine relative to a performance requirement for the second parameter, the second spectrum utilisation for the second virtual machine is further based on the performance measure of the second parameter for the second virtual machine relative to the performance requirement for the second parameter.

3.

A method as claimed in Claim 2, wherein:

the first spectrum utilisation for the first virtual machine is based on a product of (i) the performance measure of the first parameter for the first virtual machine relative to the performance requirement for the first parameter, and (ii) the performance measure of the second parameter for the first virtual machine relative to the performance requirement for the second parameter, and the second spectrum utilisation for the second virtual machine is based on a product of (i) the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter, and (ii) the performance measure of the second parameter for the second virtual machine relative to the performance requirement for the second parameter.

4. A method as claimed in either Claim 2 or Claim 3, wherein:

the performance measure of the first parameter is one of a set of performance measures of the first parameter, each relating to a UE of a plurality of UEs, and the performance measure of the second parameter is one of a set of performance measures of the second parameter, each relating to a UE of the plurality of UEs, the first spectrum utilisation for the first virtual machine is based on a sum of the product of (i) each performance measure of the first parameter for each UE for the first virtual machine relative to the performance requirement for the first parameter, and (ii) each performance measure of the second parameter for each UE for the first virtual machine relative to the performance requirement for the second parameter, and the second spectrum utilisation for the second virtual machine is based on a sum of the product of (i) each performance measure of the first parameter for each UE for the second virtual machine relative to the performance requirement for the first parameter, and (ii) each performance measure of the second parameter for each UE for the second virtual machine relative to the performance requirement for the second parameter.

5. A method as claimed in Claim 4, wherein the first and/or second spectrum utilisations are further based on one or more of the following: cost of serving the UE, bandwidth, and coverage area.

6. A method as claimed in any one of the preceding claims, wherein the step of configuring the resource allocation includes one or more of the following: adjusting frame time, reallocating resource blocks, and reallocating processing resources.

7. A method as claimed in any one of the preceding claims, wherein the first virtual machine is one of a plurality of first virtual machines, each being deployed on a respective networking node, for delivering the first service type, and the second virtual machine is one of a plurality of second virtual machines, each being deployed on a respective networking node, for delivering the second service type.

8. A computer program product comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method of any one of the preceding claims.

9. A computer-readable data carrier having stored thereon the computer program of Claim 8.

10. A network node for a virtualised radio access network, wherein the virtualised radio access network includes a first virtual machine delivering a first service type and a second virtual machine delivering a second service type, the network node comprising:

a receiver adapted to receive data indicating a performance measure of a first parameter for the first virtual machine and further indicating a performance measure of the first parameter for the second virtual machine; and a processor adapted to:

determine a first spectrum utilisation for the first virtual machine based on the performance measure of the first parameter for the first virtual machine relative to a performance requirement for the first parameter;

determine a second spectrum utilisation for the second virtual machine based on the performance measure of the first parameter for the second virtual machine relative to the performance requirement for the first parameter;

compare the first spectrum utilisation to a target spectrum utilisation for the first service type;

compare the second spectrum utilisation to a target spectrum utilisation for the second service type; and configure a resource allocation for the first and/or second virtual machine based on the comparisons of the first and second spectrum utilisations to their respective target spectrum utilisations.