Method for assessing the energy consumption of a service unit in a communication network

Description

[0001] The present invention relates to energy consumption in communication networks.

[0002] We know, for example, on the basis of the ETSI ES 202 706-V1.4.1 standard entitled « Measurement method for power consumption and energy efficiency of wireless access network equipment », methods for evaluating energy consumption in a mobile access network. Also known, for example based on the article by W. Yoro et al. entitled "Service-oriented sharing of energy in wireless access networks using Shapley value" (Computer Networks, 2017), are methods for evaluating energy consumption in mobile networks, service by service.

[0003] These methods consist of measuring the energy efficiency of a given service based on the intensity of the associated traffic, using an analytical model of energy consumption.

[0004] For example, in 5G networks, we essentially consider three classes of services: “Enhanced Mobile Broadband (eMBB)”: these are services in which users benefit from improved access to multimedia content, services and data, with improved performance and a more seamless user experience; these use cases have various characteristics, for example: the “hotspot” use case with high user density, very high traffic capacity and relatively low user mobility, or the wide area coverage use case with seamless radio coverage providing much improved data rates compared to current rates, with medium to high user mobility; “Ultra-reliable and low latency communications (URLLC)”: these are services with stringent requirements regarding characteristics such as throughput, latency and availability;Examples include wireless control of industrial manufacturing or production processes, remote medical surgery, automation of distribution in a smart grid, or transport security; “Massive machine type communications (MMTC)”: these are services characterized by a very large number of connected devices, usually transmitting a relatively low volume of data that is not very sensitive to transmission delays; the devices concerned must be inexpensive and have a very long battery life.

[0005] It is clear that if we considered each class of service separately, and if we designed a 5G network for each of these classes of service, the result would be very different access network architectures; however, the only economically and ecologically acceptable solution is represented by a common access network capable of providing these three classes of service.

[0006] It should also be noted that 5G technology is considered not only as a new radio access and core network architecture, but also as an orchestration platform where integrators can assemble specialized services for their customers; this allows the creation of a large number of services, which belong to the three service classes (eMBB, URLLC and mMTC) defined above, but with a plethora of requirements.

[0007] The concept of "service slice" ( « slice » in English) then appeared as an effective way to be able to set up all these services in a common infrastructure.

[0008] A service slice is defined as a virtual network built on a hardware infrastructure. The service slicing technique ( « slicing » in English) consists of defining a number of service slices exploiting the same hardware infrastructure. According to ITU-T Y.3011 and Y.3012, service slicing allows for logically isolated network divisions, in which a service slice is a unit of programmable resources, such as wireless and wireline capabilities, compute, and storage. According to 3GPP TR 22.891, a service slice is associated with the communication service of a particular connection type with a specific way of handling the control and user planes for that service; for this purpose, a 5G service slice is composed of a set of specific 5G network functions and access network settings, which are combined for a specific use case or business model.Service slicing is mainly associated with different market segments corresponding to different business areas, such as medical services, transportation, cities, agriculture, manufacturing, automotive, consumer retail, construction, energy, logistics, or banking.

[0009] A service slice can describe an end-to-end system, meaning that its functionality can span both the core network and the access network. A service slice can be seen as an independent network, with associated benefits such as security and guaranteed Service Level Agreement ( « Service Level Agreement », or SLA in English). However, in contrast to the deployment of independent network infrastructures in previous generations of mobile radio networks, service slices can be implemented, completely or partially, on a common infrastructure layer, providing resources such as frequency spectrum. Therefore, one of the main challenges in implementing the service slice concept is the design and management of multiple slices on the same shared infrastructure or resources in an efficient way, guaranteeing the SLA for each of these service slices.

[0010] The service slice concept therefore helps to implement these services in a flexible and dynamic way, and helps to ensure end-to-end performance.

[0011] Specific performance indicators, such as Quality of Service (QoS) can be associated with each service segment of a communications network. « Quality of Service » , or QoS, in English) linked to characteristics such as throughput or latency. The present invention relates, more particularly, to the performance indicator linked to the energy consumption of a service slice.

[0012] While QoS performance metrics can be measured at the packet or packet flow level, the performance metric for measuring energy consumption is more problematic, as the latter is linked to the infrastructure, which is shared between the different service slices. The energy consumption induced by a service slice therefore does not only depend on the traffic related to the service slice in question, but also on other service slices and other services sharing the same infrastructure. Thus, a service slice deployed alone, for example on a base station, will not induce the same energy consumption when deployed at the same time as another service slice.

[0013] Furthermore, conventional methods for evaluating energy consumption, such as those mentioned above, are difficult to apply to service slices, given the wide variety presented by service slices in terms of traffic characteristics, target QoS and coverage, but also the complexity of the multiplexing technologies of these service slices at the infrastructure level (different numerologies, different control channel structures, use or not of multiple antennas at transmission and / or reception, and so on).

[0014] The present invention therefore relates to a method for evaluating the energy consumption of a first service slice of a reality of service slices sharing an infrastructure of a communications network, said first service slice being deployed, or intended to be deployed, end-to-end, on said infrastructure. Said method comprises the following steps: a) the energy consumption W_no measured on said infrastructure in the absence of said first service slice is determined, b) the energy consumption W_yes measured on said infrastructure in the presence of said first service slice is determined, and c) the energy consumption induced by the deployment of said first service slice is obtained by calculating the difference (W_yes - W_no).

[0015] Naturally, one can reverse the order of steps a) and b), or first determine the two classes before determining their respective energy consumptions.

[0016] Thus, the present invention proposes a method for evaluating the energy consumption induced by a service slice on a given infrastructure (for example, a radio site) and in a given context (service slices coexisting on a given infrastructure, a given radio environment, and so on). According to the invention, the share of consumption of a service slice on an infrastructure device is pragmatically evaluated by comparison with other devices having similar characteristics but not deploying the service slice in question: this takes into account the fact that, as explained above, the share of consumption of a service slice is not necessarily the same when this service slice is deployed alone, and when this service slice shares the radio resources with other service slices.

[0017] Thanks to these provisions, an integrator or network operator can estimate the energy cost associated with the deployment of a new service slice. Furthermore, a customer benefiting from a service slice can thus know the energy consumption of this service slice.

[0018] According to particular characteristics, said method comprises a preliminary phase, during which a database is constructed indicating a respective measured energy consumption for a set of respective classes of infrastructure, and: during said step a), the class C_no to which said infrastructure belongs is identified in the absence of said service slice, and the energy consumption W_no recorded for said class C_no is read from said database, and during said step b), the class C_yes to which said infrastructure belongs is identified in the presence of said service slice, and the energy consumption W_yes recorded for said class C_yes is read from said database.

[0019] Thanks to these provisions, it is easy to determine the energy consumption associated with any given infrastructure belonging to a set of infrastructures previously entered into said database.

[0020] According to even more specific characteristics, the construction of said database includes the following steps: a number of classes are determined by classifying various infrastructures on the basis of selected technical characteristics of these infrastructures, and an energy consumption is determined, on the basis of measurements, for each of the said determined classes.

[0021] This first variant has the advantage that it is simple to implement. But it has the disadvantage that care must be taken not to put radio sites with potentially very different energy consumptions in the same class.

[0022] According to other even more particular characteristics: we measure the energy consumption for a certain number of infrastructures, and we generate classes of radio sites homogeneous in terms of energy consumption, and we develop a model linking the technical characteristics of said infrastructures to their energy consumption, and therefore to a corresponding class.

[0023] This second variant is more complex to implement than the first variant, but it automatically provides homogeneous classes in terms of energy consumption.

[0024] Correlatively, the invention relates to a system for evaluating the energy consumption of a service slice deployed, or intended to be deployed, on a given infrastructure of a communications network. Said system comprises means for: determine the energy consumption W_no measured on said infrastructure in the absence of said service slice, determine the energy consumption W_yes measured on said infrastructure in the presence of said service slice, and obtain the energy consumption induced by the deployment of said service slice by calculating the difference (W_yes - W_no).

[0025] According to particular features, said system further comprises means for constructing a database indicating a respective measured energy consumption for a set of respective classes of infrastructure, as well as means for: identify the class C_no to which said infrastructure belongs in the absence of said service slice, and read from said database the energy consumption W_no recorded for said class C_no, and identify the class C_yes to which said infrastructure belongs in the presence of said service slice, and read from said database the energy consumption W_yes recorded for said class C_yes.

[0026] According to even more particular characteristics, to construct said database, said system further comprises means for: determine a number of classes by classifying various infrastructures on the basis of selected technical characteristics of these infrastructures, and determine, on the basis of measurements, an energy consumption for each of said determined classes.

[0027] According to other even more particular characteristics, to construct said database, said system further comprises means for: measure the energy consumption for a number of infrastructures, and generate classes of radio sites homogeneous in terms of energy consumption, and develop a model linking the technical characteristics of said infrastructures to their energy consumption, and therefore to a corresponding class.

[0028] The advantages offered by these systems are essentially the same as those offered by the correlative processes briefly set out above.

[0029] Said system for evaluating the energy consumption of a service slice may advantageously comprise at least one service slice management entity hosted in an access network node, in a core network node, or in an operational, administration and maintenance (OAM) center of said communications network.

[0030] It should be noted that said service slice management entity may be physical or virtual, and that it is possible to implement it in the context of software instructions and / or in the context of electronic circuits.

[0031] The invention also relates to a computer program downloadable from a communications network and / or stored on a computer-readable medium and / or executable by a microprocessor. This computer program is remarkable in that it comprises instructions for executing the steps of the energy consumption evaluation method succinctly set out above, when it is executed on a computer.

[0032] The advantages offered by this computer program are essentially the same as those offered by the said method.

[0033] Other aspects and advantages of the invention will appear on reading the detailed description below of particular embodiments, given as non-limiting examples.

[0034] We will now describe an embodiment of the method for evaluating the energy consumption of a service unit according to the invention.

[0035] This embodiment comprises a preliminary phase, during which a database is constructed indicating a respective measured energy consumption for a set of respective classes of infrastructure. In a non-limiting manner, an example of infrastructure will be a radio site, such as a base station of a cellular network.

[0036] According to a first variant, this construction implements the following steps.

[0037] In an EP-1 step, various radio sites are classified taking into account certain selected technical characteristics, such as: the transmission power, coverage, number of transmitting and receiving antennas, or the frequency band used, the radio environment, for example rural or urban, or the traffic situation, for example the types of active service slots, their traffic levels, or the measured interference level.

[0038] Naturally, other factors may be taken into account for the purposes of this classification.

[0039] In step EP-2, energy consumption is determined, based on measurements, for each of the classes determined in step EP-1.

[0040] Finally, during a step EP-3, the said database is built accordingly.

[0041] According to a second variant, this construction implements the following steps.

[0042] In step EP'-1, the energy consumption is first measured for a certain number of radio sites, and then classes of radio sites that are homogeneous in terms of energy consumption are generated.

[0043] In step EP'-2, a model is developed that links the characteristics of radio sites to their energy consumption, and therefore to a corresponding class.

[0044] Finally, during a step EP'-3, the said database is constructed accordingly.

[0045] To implement this second variant, and build a model linking the characteristics of a radio site to its energy consumption class, we can conveniently use known statistical data processing techniques, as well as artificial intelligence techniques, such as data partitioning (" data clustering » in English), in particular k-means partitioning or supervised classification. In this regard, it is recalled that k-means partitioning ( « k-means clustering » in English) is a method of data partitioning and combinatorial optimization, in which a set of elements is subdivided into k subsets (" clusters » in English), so as to minimize a certain function.

[0046] These techniques can, for example, use neural networks or naive Bayesian classifiers. In this regard, it is recalled that naive Bayesian classification constructs models associated with characteristics assumed to be statistically independent.

[0047] When, following this preliminary phase, we wish to evaluate the energy consumption of a given service slice S on a given infrastructure, we implement the following steps.

[0048] According to a step EA-1, the class C_no to which said infrastructure belongs is identified in the absence of the service slice S, and the corresponding energy consumption W_no is read from said database.

[0049] According to a step EA-2, the class C_yes to which said infrastructure belongs is identified in the presence of the service slice S, and the corresponding energy consumption W_yes is read from said database.

[0050] Naturally, one can also, as a variant, first determine the said classes during a first step, then their respective energy consumption during a second step.

[0051] Finally, according to a step EA-3, we obtain the energy consumption due specifically to the deployment of the service slice S by calculating the difference (W_yes - W_no).

[0052] The service slice management system according to the invention can be implemented, by means of software and / or hardware components, within one or more service slice management entities hosted in nodes of a communication network, such as nodes of an access network, for example in the cloud ( « cloud RAN » in English), or nodes of a core network. Such a service slice management entity can also be deployed as a specific module in an operations, administration and maintenance center ( « Operations, Administration and Maintenance », or OAM in English) of the network (or an equivalent); it is recalled in this regard that an OAM is a center responsible, in a traditional manner, for the collection and archiving of performance measurements.

[0053] Said software components may be integrated into a conventional computer program for managing a network node. Therefore, as indicated above, the present invention also relates to a computer system. This computer system conventionally comprises a central processing unit controlling a memory by signals, as well as an input unit and an output unit. In addition, this computer system may be used to execute a computer program comprising instructions for implementing any of the methods for evaluating energy consumption according to the invention.

[0054] Indeed, the invention also relates to a computer program downloadable from a communication network comprising instructions for executing the steps of a method for evaluating energy consumption according to the invention, when it is executed on a computer. This computer program can be stored on a computer-readable medium and can be executable by a microprocessor.

[0055] This program may use any programming language, and may be in the form of source code, object code, or code intermediate between source code and object code, such as in a partially compiled form, or in any other desirable form.

[0056] The invention also relates to an information medium, irremovable, or partially or totally removable, readable by a computer, and comprising instructions of a computer program as mentioned above.

[0057] The information carrier may be any entity or device capable of storing the program. For example, the carrier may include a storage medium, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording medium, such as a hard disk, or a USB flash drive (“ USB flash drive » in English).

[0058] On the other hand, the information carrier may be a transmissible carrier such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means. The computer program according to the invention may in particular be downloaded from a network such as the Internet.

[0059] Alternatively, the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of any of the methods for evaluating energy consumption according to the invention.

Claims

1. Method for evaluating the energy consumption of a first service slice of a plurality of service slices sharing an infrastructure of a communications network, said first service slice being deployed, or intended to be deployed, end-to-end on said infrastructure, said method comprising: a) determining the energy consumption W_no measured on said infrastructure in the absence of said first service slice; b) determining the energy consumption W_yes measured on said infrastructure in the presence of said first service slice; and c) obtaining the energy consumption caused by the deployment of said first service slice by calculating the difference (W_yes - W_no).

2. Evaluation method according to Claim 1, characterized in that it comprises a preliminary phase, during which a database is constructed which indicates a respective measured energy consumption for a set of respective infrastructure classes, and in that: - in said step a), the class C_no to which said infrastructure belongs in the absence of said first service slice is identified, and the energy consumption W_no recorded for said class C_no is read from said database; and - in said step b), the class C_yes to which said infrastructure belongs in the presence of said first service slice is identified, and the energy consumption W_yes recorded for said class C_yes is read from said database.

3. Evaluation method according to Claim 2, characterized in that the construction of said database comprises the following steps: - determining a certain number of classes by classifying various infrastructures on the basis of selected technical characteristics of these infrastructures; and - determining, on the basis of measurements, an energy consumption for each of said determined classes.

4. Evaluation method according to Claim 2, characterized in that the construction of said database comprises the following steps: - measuring the energy consumption for a certain number of infrastructures, and generating radio site classes that are homogeneous in terms of energy consumption; and - developing a model linking the technical characteristics of said infrastructures to their energy consumption, and therefore to a corresponding class.

5. System for evaluating the energy consumption of a first service slice of a plurality of service slices sharing an infrastructure of a communications network, said first service slice being deployed, or intended to be deployed, end-to-end on said infrastructure, said system comprising means for: - determining the energy consumption W_no measured on said infrastructure in the absence of said first service slice; - determining the energy consumption W_yes measured on said infrastructure in the presence of said first service slice; and - obtaining the energy consumption caused by the deployment of said first service slice by calculating the difference (W_yes - W_no).

6. System according to Claim 5, characterized in that it further comprises means for constructing a database indicating a respective measured energy consumption for a set of respective infrastructure classes, and means for: - identifying the class C_no to which said infrastructure belongs in the absence of said first service slice, and reading from said database the energy consumption W_no recorded for said class C_no; and - identifying the class C_yes to which said infrastructure belongs in the presence of said first service slice, and reading from said database the energy consumption W_yes recorded for said class C_yes.

7. System according to Claim 6, characterized in that, in order to construct said database, it further comprises means for: - determining a certain number of classes by classifying various infrastructures on the basis of selected technical characteristics of these infrastructures; and - determining, on the basis of measurements, an energy consumption for each of said determined classes.

8. System according to Claim 6, characterized in that, in order to construct said database, it further comprises means for: - measuring the energy consumption for a certain number of infrastructures, and generating radio site classes that are homogeneous in terms of energy consumption; and - developing a model linking the technical characteristics of said infrastructures to their energy consumption, and therefore to a corresponding class.

9. System according to any one of Claims 5 to 8, characterized in that it comprises at least one entity for managing service slices that is hosted in an access network node, in a core network node, or in an operations, administration and maintenance (OAM) centre of said communications network.

10. Non-removable, or partially or fully removable, data storage means comprising computer program code instructions for the execution of the steps of a method for evaluating the energy consumption of a first service slice according to any one of Claims 1 to 4.

11. Computer program that can be downloaded from a communication network and / or is stored on a computer-readable medium and / or is able to be executed by a microprocessor, characterized in that it comprises instructions for the execution of the steps of a method for evaluating the energy consumption of a first service slice according to any one of Claims 1 to 4 when it is executed on a computer.