TECHNICAL FIELD

[0001] Various example embodiments relate generally to apparatuses, methods, and computer programs for mapping machine learning based applications to intent logic units.

BACKGROUND

[0002] Owing to the ever-increasing complexity of communication networks, especially mobile communication networks, there is always a need for more automation and abstraction. One solution for the abstraction is through the use of intents in what may be termed Intent-Based Networking (IBN). Therein, the network should be able to respond to user's intents without the user (e.g. an operator) specifying the technical details of how his intended outcome may be realized by the underlying architecture.

[0003] The technical report 3GPP TR 28.812 V17.1.0 (dated 2020-12) (Technical Specification Group Services and System Aspects; Telecommunication management; Study on scenarios for Intent driven management services for mobile networks) provides example use cases for intent driven management of mobile networks.

[0004] The complexity of the specified intents may significantly vary. The simple intents may be fulfilled with a single command to a single network object. An intent may allow to solve specific network problem or achieve a specific network target.

[0005] In order to fulfil the complex intents, it may be necessary to perform continuous network monitoring, data processing and inferring the best command to be applied to a single or multiple network object.

[0006] SUMMARY

[0007] The scope of protection is set out by the independent claims. The embodiments, examples, and features, if any, described in this specification that do not fall under the scope of the protection are to be interpreted as examples useful for understanding the various embodiments and examples that fall under the scope of protection.

[0008] According to a first aspect, there is provided an apparatus comprising means for performing a method, the method comprising: sending a request for capability information, wherein the request is sent to at least one of a ML based application, a ML applications orchestrator and a ML capability producer; receiving capability information representative of capabilities of one or more machine learning, ML, based applications; generating one or more intent logic unit, ILUs, wherein an I LU is mapped with at least one corresponding ML based application, wherein the ILU is configured to launch an execution of the at least one corresponding ML based application mapped with this ILU; storing the one or more ILUs in an intent logic library, wherein an ILU is stored in association with capability information derived for the ILU from the capability information received for the at least one corresponding ML based application mapped with this ILU.

[0009] The capability information received for a ML based application may include information representative of a type of function performed by the ML based application for at least one communication network, wherein the type is at least one of an optimization function, a control function, an analytics function and an ML orchestration function.

[0010] The ILU may be configured to launch the execution of the at least one corresponding ML based application and I or an ML orchestration function configured to monitor the at least one corresponding ML based application.

[0011] The capability information received for a ML based application may include information representative of one or more objects or one or more object types for which the function is performed.

[0012] The capability information received for a ML based application may include information representative of at least one configuration parameter for an object or object type for which the function is performed.

[0013] The capability information received for a ML based application may include information representative of an objective and I or constraint for the execution of the function, wherein the objective and I or constraint is defined on the basis of at least one network metric.

[0014] The capability information received for a ML based application may include information representative of a configuration parameter of an ML model implemented by the ML based application.

[0015] The method may comprise: storing in the intent logic library mapping information for the mapping performed between the ILU and the at least one corresponding ML based application.

[0016] The method may comprise: receiving a request including intent information representative of an intent to be achieved for at least one communication network; identifying, based on capability information stored in the intent logic library, one or more ILUs configured to launch an execution of one or more corresponding ML based applications adapted to fulfill the intent for the at least one communication network; executing the identified one or more ILUs. [0017] The method may comprise: configuring the one or more ML based applications mapped with the identified one or more ILUs.

[0018] According to a second aspect, there is provided a method comprising: sending a request for capability information, wherein the request is sent to at least one of a ML based application, a ML applications orchestrator and a ML capability producer; receiving capability information representative of capabilities of one or more machine learning, ML, based applications; generating one or more intent logic unit, ILUs, wherein an I LU is mapped with at least one corresponding ML based application, wherein the ILU is configured to launch an execution of the at least one corresponding ML based application mapped with this ILU; storing the one or more ILUs in an intent logic library, wherein an ILU is stored in association with capability information derived for the ILU from the capability information received for the at least one corresponding ML based application mapped with this ILU.

[0019] Generally, the apparatus according to the first aspect comprises means for performing one or more or all steps of the method according to the second aspect. The means may include circuitry configured to perform one or more or all steps of the method. The means may include at least one processor and at least one memory including computer program code, the at least one memory and the computer program code being configured to, with the at least one processor, cause the apparatus to perform one or more or all steps of the method.

[0020] According to another aspect, there is provided a non-transitory computer-readable medium comprising program instructions stored thereon for performing at least the following: sending a request for capability information, wherein the request is sent to at least one of a ML based application, a ML applications orchestrator and a ML capability producer; receiving capability information representative of capabilities of one or more machine learning, ML, based applications; generating one or more intent logic unit, ILUs, wherein an ILU is mapped with at least one corresponding ML based application, wherein the ILU is configured to launch an execution of the at least one corresponding ML based application mapped with this ILU; storing the one or more ILUs in an intent logic library, wherein an ILU is stored in association with capability information derived for the ILU from the capability information received for the at least one corresponding ML based application mapped with this ILU. Generally the non-transitory computer-readable medium may comprise program instructions stored thereon for performing one or more or all steps of a method according to the second aspect.

[0021] According to a third aspect, there is provided an apparatus comprising means for performing a method, the method comprising: receiving a request for capability information from an intent logic execution platform; sending in response to the request capability information representative of capabilities of one or more machine learning, ML, based applications; executing the one or more machine learning, ML, based applications to fulfill an intent, wherein the executing is launched by at least one intent logic unit mapped with the one or more ML based application.

[0022] According to a fourth aspect, there is provided a method comprising: receiving a request for capability information from an intent logic execution platform; sending in response to the request capability information representative of capabilities of one or more machine learning, ML, based applications; executing the one or more machine learning, ML, based applications to fulfill an intent, wherein the executing is launched by at least one intent logic unit mapped with the one or more ML based application.

[0023] Generally, the apparatus according to the third aspect comprises means for performing one or more or all steps of the method according to the fourth aspect. The means may include circuitry configured to perform one or more or all steps of the method. The means may include at least one processor and at least one memory including computer program code, the at least one memory and the computer program code being configured to, with the at least one processor, cause the apparatus to perform one or more or all steps of the method.

[0024] According to another aspect, there is provided a non-transitory computer-readable medium comprising program instructions stored thereon for performing at least the following: receiving a request for capability information from an intent logic execution platform; sending in response to the request capability information representative of capabilities of one or more machine learning, ML, based applications; executing the one or more machine learning, ML, based applications to fulfill an intent, wherein the executing is launched by at least one intent logic unit mapped with the one or more ML based application. Generally the non-transitory computer-readable medium may comprise program instructions stored thereon for performing one or more or all steps of a method according to the fourth aspect.

BRIEG DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate example embodiments together with the general description given above, and the detailed description given below.

[0026] FIG. 1 is a schematic diagram illustrating an exemplary system for intent fulfilment according to an example.

[0027] FIG. 2 is a schematic diagram illustrating an exemplary system for intent fulfilment according to an example.

[0028] FIG. 3 is a schematic diagram illustrating aspects of the relationships between a ML application, an Intent Logic Execution Platform and an Intent Logic Unit according to an example.

[0029] FIG. 4 is a schematic diagram illustrating aspects of the relationships between ML applications, an Intent Logic Execution Platform and Intent Logic Units according to an example.

[0030] FIG. 5 is a flow diagram illustrating a method for intent fulfillment according to an example.

[0031] FIG. 6 is a flow diagram illustrating a method for intent fulfillment according to an example.

[0032] FIG. 7 is a flowchart illustrating a method for intent fulfillment according to an example.

[0033] FIG. 8 is a flowchart illustrating a method for intent fulfillment according to an example.

[0034] FIG. 9 is an exemplary hardware structure of an apparatus according to an example.

[0035] It should be noted that these drawings are intended to illustrate various aspects of devices, methods and structures used in example embodiments described herein. The use of similar or identical reference numbers in the various drawings is intended to indicate the presence of a similar or identical element or feature.

DETAILED DESCRIPTION

[0036] Detailed example embodiments are disclosed herein. However, specific structural and functional details disclosed herein are merely representative for purposes of describing example embodiments. Accordingly, these embodiments are shown by way of illustrative examples in the drawings and will be described herein in detail so as to provide a thorough understanding of the various aspects. However, it will be understood by one of ordinary skill in the art that example embodiments are capable of various modifications and alternative forms and may be practiced without all the specific details. In addition, systems and processes may be shown in block diagrams so as not to obscure the example embodiments in unnecessary detail. In other instances, well-known processes, structures, and techniques may be shown without unnecessary detail in order to avoid obscuring the example embodiments.

[0037] Exemplary embodiments provide apparatuses, methods, system and computer programs for mapping machine learning (ML) based applications to intent logic units. Exemplary embodiments provide apparatuses, methods, system and computer programs for intent fulfillment.

[0038] The terms “ML functionality” are used herein to designate a ML based application or a ML orchestrator. The terms “ML based application” or “ML application” are used herein to designate a software application that uses Al (Artificial Intelligence) I ML technology, e.g. by implementing one or more ML models. The terms “ML orchestration function” or “ML orchestrator” are used herein to designate a software application that is configured to orchestrate the execution of ML applications. The orchestration of ML applications may include performing at least one of synchronizing, triggering, configuring, monitoring and controlling the execution of the orchestrated ML applications.

[0039] Exemplary applications of ML comprise without limitation: voice recognition; image processing/computer vision; natural language processing; information retrieval; personalization and recommendation; robotics, data analytics including predictive and prescriptive analytics; use-cases for the design and/or planning and/or optimization and/or configuration and/or control and/or management of communication systems and I or networks.

[0040] Exemplary use-cases may be without limitation: use-cases related to the physical-layer of communication networks such as modulation, coding, decoding, signal detection, channel estimation, prediction, compression, interference mitigation; use-cases related to the medium access control layer of communication networks such as multiple access and resource allocation (e.g., power control, scheduling, spectrum management); channel modeling; network optimization; cell capacity estimation in cellular networks; routing; resource management; data traffic management; security and anomaly detection; root cause analysis; transport protocol design and optimization; user/network/application behavior analysis/prediction; transport-layer congestion control; user experience modeling and optimization; user mobility and positioning management; network slicing, network virtualization and software defined networking; non-linear impairments compensation in optical networks (e.g., visible-light communications, fiber-optics communications, and fiber-wireless converged networks), and quality-of-transmission estimation and optical performance monitoring in optical networks.

[0041] Exemplary analytics and/or decision function comprise without limitation: coverage analysis, coverage problems analysis, handover problems analysis, faults detection, interference detection, coverage optimization, capacity optimization, handover optimization, interference reduction, energy saving optimization,

[0042] In the context of the present document, an intent may correspond to an objective to be met to reach a target (e.g. target state, target position, target configuration, target quality, etc) for one or more devices I entities interacting with and I or within a communication network. An intent may for example define a network management objective or network deployment objective. An intent does not define the necessary steps to get to the target. The expressed intent may need to be transformed into executable tasks to be performed to achieve the target.

[0043] The complexity of intents may vary, from simple intents that can be fulfilled with a single command to specific network object to very complex intents that may require continuous network monitoring and inferring the best approach, i.e., by employing ML- based solutions leading to different commands and including multiple network nodes.

[0044] The aim of I BN (Intent Based Network) is to be able to respond to operator intents without the need for specifying the technical details of how the intent may be fulfilled by the underlying architecture. Example intents may for example include, without limitation:

Collect/get Carrier Aggregation statistics for all cells in city X;

Restrict/deny handovers of high mobility users to small cells;

- Allow load balancing to a cell Y or to small cells or to only urban cells; Rehome a base station from controller A to controller B.

[0045] In general, as illustrated in the system of Figure 1 , a user has access to an intent specification platform (ISP) 120 through which the user (e.g. a human, an operator of the concerned communication network or a user device) specifies the intent to be achieved. The intent may be specified using several types of user interfaces 110, including for example: a graphical user interface (GUI), Command Line Interface (CLI), a text interface like a text file, an audio interface that captures the user’s audio commands, etc.

[0046] The ISP 120 may be configured to analyze the input data expressing the intent. The ISP 120 may comprise a mechanism for enforcing syntax rules and checking for correctness and completeness of the specified intent.

[0047] The ISP 120 may for example be configured to analyze (e.g. parse) the input data expressing the intent provided by the user in order to identify the data fields fitting to a predefined intent specification syntax. Several data processing methods may be used for analyzing the specified intent and breaking down the intent into its constituent subparts to facilitate the identification of the tasks I sub-tasks to be performed for fulfillment of the intent. [0048] Patent application published under number WO2021/164878 A1 which is hereby incorporated by reference discloses an example ISP that is configured to receive the request from a user through a user interface, wherein the request uses a specific intent specification syntax and may comprise a control object of the communication network and a verb which indicates what action has to be performed for the control object.

[0049] An intent execution interface 125 is used for exchanging information between the ISP and an intent fulfillment system 130. The intent execution interface 125 may be used e.g. to send a given intent for execution to the intent fulfillment system 130 or to receive reports on the success of the execution from the intent fulfillment system 130. The intent fulfillment system 130 is itself connected to network functions and/or processes 140 that perform tasks to fulfill the received intent.

[0050] A system 200 for fulfilling user intents using Intent Logic Units (ILUs) is illustrated by FIG. 2. The system 200 may be part of a communication network. Exemplary communication networks comprise, without limitations: radio communication networks, wired or wireless communication networks, optical fiber-based communication networks, optical wireless communication networks, satellite communication networks, or any combination thereof.

[0051] The system 200 includes an Intent Specification platform (ISP) 120 and an Intent Fulfillment System (IFS) 230. Like in the system of FIG. 1 , a user has access to an Intent Specification platform (ISP) 120 through user interfaces 110 to allow the user to specify an intent.

[0052] Like in the system of FIG. 1 , the ISP 120 communicates through an intent execution interface 125 with the IFS 230. The intent execution interface 125 is used for exchanging information between the ISP 120 and the IFS 230. The intent execution interface 125 may be used e.g. to send an intent for execution to the IFS 230 or to receive reports on the success of the execution from the IFS 230.

[0053] In this embodiment, the IFS 230 is implemented as an Intent-Logic Execution Platform (ILEP) 230 and is configured to execute one or more Intent Logic Units (ILUs). The ILUs may be stored in and retrieved from an Intent Logic Library (ILL) 135 by the ILEP 230. The ILEP 230 is itself connected to network functions and processes 140 that allow to fulfill the intent. Intent execution is fulfilled via an Intent-Logic Execution Platform (ILEP) that identifies the appropriate I LU and executes the I LU onto the concerned communication network or its network objects.

[0054] An I LU includes instructions (e.g. logic or command sequences) to be executed. The I LU includes instructions that when executed by a processor cause a host apparatus or system (e.g. Intent Fulfillment System (IFS) or Intent-Logic Execution Platform (ILEP)) to launch the execution of the one or more corresponding tasks. The instructions may be in any language (any computer-programming language or script language, including, but not limited to assembly, C, C++, Basic, SQL, MySQL, HTML, PHP, Python, Java, Javascript, etc) and any format (script, binary code, executable code, pre-compiled code, interpretable code, etc).

[0055] An I LU may be seen as a wrapper around the logic or command sequences that may need to be executed to achieve a corresponding intent. An I LU may correspond to a software unit that is configured to launch and I or configure the execution of one or more tasks, that may be for example any network operation related task or jobs. These tasks may be executed to achieve a corresponding intent. Multiple ILUs can be combined to form another I LU.

[0056] An executable task may be any function and I or process or combination thereof that may be executed by an apparatus. The execution of the function and I or process may be performed using one or more software units and one or more hardware units, in any combination.

[0057] The executable task may be, without limitation:

- a network management task, including e.g. a QoS management task, taking a decision of handover, a load balancing task, configuring a network device or network cell, implementing network access control;

- Identifying, formulating, activating and implementing network management policies.

[0058] In order to fulfil complex intents, the use of a single or multiple ML models (executed in parallel or in sequence) may be needed or useful. The approach to use the ML models to fulfill a given intent may depend on for example the complexity level of the intent, the availability of ML models and their suitability to fulfil the given intent.

[0059] The present disclosure relates to the use of intents as abstraction means and their automated fulfilment by Machine Learning (ML) based solutions mapped to one or more Intent Logic Units (ILUs). A specified intent (e.g. an intent specified by an operator of a communication network) may be fulfilled by using ML functionalities that are available in the communication network. The specified intent may be fulfilled in an automated way by using ML functionalities mapped with ILUs. [0060] To be able to fulfill an intent using ML technology, there is provided a capability discovery and exposure mechanism for ML based applications available in a communication network. The capability exposure may be available as soon as the concerned ML based application is deployed.

[0061] To allow fulfilling specified intents (expressed by a user, e.g. a network operator) using ML based applications, it is proposed to leverage on exposed capabilities of ML based applications so as to derive new ILUs mapped to one or more ML based applications or a combination of ML based applications orchestrated by an ML orchestrator.

[0062] Using the capability discovery and exposure mechanism allows an intent fulfilment system to determine the right ML based applications to be used for fulfilling a given intent. Also the intent fulfilment system may chain the ILUs and I or the ML based applications in order to satisfy a given intent.

[0063] There may be different combinations/approaches for fulfilling the intents by using ML based applications:

There may be one-to-one or one-to-many mapping between ILU and ML based applications; different combinations of ILUs and ML based applications may be possible for fulfilling a given intent; complex intent may be broken up into intent components, where each intent component may be fulfilled by using one or more ILUs with a mapping with one or more ML based applications (executed either in parallel or in sequence).

[0064] Each ILU may be mapped to one or more corresponding ML based applications from which the ILU is derived. The capabilities of an ILU correspond to the capabilities of the one or more ML based applications mapped with this ILU. The newly generated ILUs may then be used to satisfy an intent (e.g. an intent specified by an operator for a communication network), using the capabilities of the new ILUs. In such a way the specified intents may be, where applicable and to the extent possible, mapped to ML based solutions via the ILUs.

[0065] In addition, the specified intents may be mapped to other ILUs that are not themselves mapped to ML based applications. For example, in order to completely fulfill the intent the non-ML based solutions (logic or commands) may be needed as well. In the present description, for simplification reasons, we consider example intents or part of an intent that is achievable by ML-based solution, whereas the non-ML based commands may still be applicable in addition for complete fulfilment of the given intent.

[0066] One or more intent logic units, ILUs, may be generated by the ILEP with a mapping between the one or more ILUs and the of one or more ML based applications. An ILU may be mapped with one or more corresponding ML based applications and the I LU is configured to launch an execution of the one or more corresponding ML based applications mapped with this ILU.

[0067] Each ILU that is mapped with one or more ML applications is configured to launch the execution of the one or more ML applications, either directly or through an ML orchestration function. Launching the execution of an ML application may include sending, e.g. by the ILEP and I or the mapped ILU, configuration parameters to the launched ML application. The configuration parameters may be sent before and I or when and I or after triggering the execution of the launched ML application. The launching may be performed using a remote procedure call, a web interface or any appropriate software technology to trigger the execution of an ML application that may be executed on a remote device.

[0068] While the ILU is executed by the ILEP, each of the mapped ML application may be executed by a host apparatus (e.g. a network entity), physically distinct and separate from the ILEP. The ML application may receive input data from the ILU, from the host apparatus or from another network entity (e.g. from or network devices like sensors, data producers, etc). The ML application may generate output data that may be used by the host apparatus or by another network entity. This host apparatus may be located within the communication network such that the ML-based application has access to the required input data and may output data for concerned network entities.

[0069] Exemplary network entities comprise without limitation: radio access network entities such as base stations (e.g., cellular base stations like eNodeB in LTE and LTE-advanced networks and gNodeB used in 5G networks, and femtocells used at homes or at business centers); relay stations; control stations (e.g., radio network controllers, base station controllers, network switching sub-systems); access points in local area networks or ad-hoc networks; gateways and radio access network entities; network management entities (e.g., Operation, Administration and Management (OAM) entity); network automation systems; distributed analytics entities such as self-autonomous systems (D-SONs); network functions (e.g., network data analytics function or NWDAF defined in current 3GPP standards).

[0070] The one or more ILUs may be stored in an intent logic library (ILL). An ILU is stored in association with ILU capability information derived for this ILU from the capability information received for the one or more corresponding ML based applications mapped with this I LU. The ILL may be configured to perform a search of ILUs on the basis of search criteria related to capability information.

[0071] The I LU may be stored in the ILL in association with at least one of an identifier, a name and a description which is understandable by a user to allow identification of the ILU. For example, each ILU has a description intended for human users who may either want to revise, reuse or remove the ILU.

[0072] In one or more embodiments, the proposed system provides a ML capability discovery process implemented between two functions: ML capability consumer and ML capability producer.

[0073] The ML capability consumer is a function that may be implemented by an Intent Logic Execution platform or an Intent Fulfillment System or any other entity configured to generate new ILUs or any other notion of execution logic. The ML capability producer is a function that may be implemented by an ML based application itself or an ML orchestrator or a dedicated network entity (e.g. a network management function, an automation function, an analytics function, or a network function like a gNB, or cell) configured to provide an interface for ML capability exposure on behalf of one or more ML based applications or one or more ML orchestrator.

[0074] Authorizations may be necessary to allow one or more ML capability consumers to request capabilities such that only an authorized ML capability consumer may request the capabilities of existing ML based applications. Likewise ML based applications may report their capabilities only to an authorized ML capability consumer.

[0075] An authorized ML capability consumer (typically the intent fulfillment system or ILEP) may be configured to request from the ILL (Intent Logic Library) for a mapping of a specified ILU to the set of applicable ML based applications. The ILL may be configured to report the mapping of a specified ILU to the set of applicable ML based applications.

[0076] The capability information may be exposed over an interface (e.g. an open interface) between a ML capability consumer and a ML capability producer. This interface may be used for various operations including at least one of: sending a request for capability information from the ML capability consumer to the ML capability producer; receiving capability information by the ML capability consumer from the ML capability producer; sending configuration data of a ML application from the ML capability consumer to the ML capability producer. [0077] Capability information representative of capabilities of one or more machine learning, ML, based applications are received by the ML capability consumer. The capability information may be received in response to a request sent from the ML capability consumer to one or more ML capability producers (e.g. to each new ML-based applications available). The capability information may also be received on the basis of a subscription scheme in which the ML capability consumer sends at least one requests (e.g. to a network entity that implements a ML capability producer function on behalf of ML based applications) to receive capability information and receives the capability information matching its request when new ML applications are available without having to send a request to each of new ML-based application.

[0078] There may be several deployment/configuration options for mapping I LU to ML- based applications and intent fulfilment using the mapped ILUs as shown in FIGS. 3 and 4. [0079] In the example of FIG. 3, a ML capability producer is a ML application 261 and the ML capability consumer is the ILEP 230. The capability information may be exposed over an interface 265 between the ML capability consumer and the ML capability producer.

[0080] In the example of FIG. 3, one ILU 251 in the ILL 235 is mapped to one ML-based application 261. The ML-based application 261 may implement one or more ML models. The ILEP 230 may load the ILU 251 from the database and execute the ILU 251 which itself launches the execution of the ML application 261.

[0081] In the example of FIG. 4, several ML capability producers are provided including: the ML applications 271 to 273, the ML applications 274 to 276 and the ML orchestrator 280. The ML capability consumer is the ILEP 230. Like in FIG. 3, each ML-based application may implement one or more ML models.

[0082] The capability information are exposed over interfaces 265-1 to 265-4 between the ML capability consumer (ILEP 230) and the ML capability producers (here the ML applications 271 to 273 and the ML orchestrator 280). An interface 265-1 to 265-3 is implemented between the ILEP 230 and each of the ML application 271 to 273. An interface 265-4 is implemented between the ILEP 230 and the ML orchestrator 280, without the need for an interface between the ILEP 230 and the ML application 274 to 276 orchestrated by the ML orchestrator 280.

[0083] In the example of FIG. 4, one ILU is mapped to several ML-based applications. There are several options to map an ILU to several ML-based applications.

[0084] In a first option, the ML capability consumer (ILEP 230) derives one ILU 252 from the ML based applications 271 to 273 without an overlooking ML based application (i.e., without ML orchestrator) such that the ILU 252 is mapped with the ML based applications 271 to 273. The ILEP 230 may load the ILU 252 from the ILL 235 and execute the ILU 252. [0085] In this first option, the ILU 252 may launch the execution of one or more of the ML applications mapped with this ILU. For example the ILU 252 launches the execution of each of the ML applications 271 to 273. Alternatively, the ILU 252 may be configured to launch one of the ML applications mapped with this ILUs (for example the ML application 271) and one of the ML applications may itself be configured to launch at least one other ML applications (for example the ML applications 272 and I or 273) mapped with this ILU. In both alternatives the ML applications 271 , 272, 273 may be executed in a sequential manner or at least partly in parallel or in parallel.

[0086] In a second option, the ML capability consumer (ILEP 230) derives one ILU 253 from the ML based applications 274 to 276 along with an overlooking ML based application (i.e., ML orchestrator 280) which is responsible for orchestrating the execution of the ML based applications such that the ILU 253 is mapped with the ML based applications 274 to 276 and also with the ML orchestrator 280.

[0087] In this second option the ILU 253 launches the execution of one or more of the ML applications mapped with this ILU, the execution and performance of these ML applications being monitored by the ML orchestrator 280. For example the ILU 252 launches the execution of the ML orchestrator 280 which itself launches the execution of the ML based applications 274 to 276 orchestrated by the ML orchestrator 280. In this second option, the ML applications 274, 275, 276 may be executed in a sequential manner or at least partly in parallel or in parallel.

[0088] The capability information that are exposed for the available ML functionalities may include various types of information. The capability information derived for an ILU may include the same type of information.

[0089] The capability information of an ML based application may include a description describing the functionality or ML model implemented by the concerned ML based application. The description may include at least one of a text (e.g. “predicting the QoS level within a QoS scope”), keywords, an identification a functionality within a list of predefined functionalities, etc.

[0090] The capability information of an ML based application may include information representative of a type of function performed by the ML based application for the communication network. The type may be at least one of: an optimization function, a control function, an analytics function and an ML orchestration function. Other relevant types may be used. The type of function may be identified by a name or an identifier within a list of identifiers or using another type of identification method.

[0091] The capability information of an ML based application may include information representative of entities for which the functionality implemented by ML based application is applied. These entities may be one or more objects or one or more object types (e.g. “UEs” to indicate that one object is a User Equipment). An object may be used as input or output of the ML based application. An object may correspond to any entity in the communication network: a UE, a network entity, a functional unit, a virtual function, a database, or other similar/related objects. The object or object type may be identified by a name or an identifier or using another type of identification method.

[0092] The capability information may include information representative of at least one configuration parameter for an object or object type for which the function is performed. The configuration parameter for an object or object type may be any parameter that is usable to configure the concerned object. For example, a cell needs to have an antenna tilt to determine where the antenna should face to maximize coverage and minimize interference. The antenna tilt in this example is a configuration parameter for the object cell. A configuration parameter for an object or object type is an example of a configuration parameter of a ML based application.

[0093] The capability information may include information representative of a configuration parameter for an ML model implemented by the ML based application. The configuration parameter (metaparameter, input parameter, etc) for an ML model may be any parameter that is usable to configure the concerned ML model. For example, an ML model may have inputs which the ML model uses to drive to decisions, e.g. a ML model to optimize coverage may take the antenna tilt as input data. In this case the antenna tilt is a configuration parameter of the ML model. A configuration parameter for an object or object type is an example of a configuration parameter of a ML based application.

[0094] The capability information may include information representative of an objective and I or constraint for the execution of the function, wherein the objective and I or constraint is defined on the basis of at least one network metric. Such an objective and I or constraint may be coded by a text, may be identified by a name or an identifier within a list of objectives or using another type of identification method.

[0095] The capability information of an ML based application may be coded and reported in various formats. The capability information of an ML based application may be coded using a descriptive text and I or t-uples and I or tables, etc.

[0096] The capability information may represent a decision which can be reported. For example the capability information may be in the form of a triple <x,y,z> indicating

- x: the object or object types for which the ML based application can undertake optimization or control; y: the configuration parameters of object or object types x, which the ML based application optimizes controls to achieve the desired outcomes; z: the network metrics which the ML based application optimizes through its actions. [0097] The capability information may represent an analysis which can be reported. For example in the form of tuple <x,z> indicating:

- x: the object or object types for which the ML based application can undertake analysis z: the network characteristics (on object x) for which the ML based application produces analysis.

[0098] FIG. 5 is a flow diagram illustrating an exemplary implementation of a method for intent fulfilment according to some embodiments.

[0099] The method is implemented by a ML capability consumer 550 (e.g. an Intent Logic Execution Platform) and a ML capability producer 590 and an Intent Specification Platform 520.

[0100] The ML capability consumer 590 discovers the available ML based applications and their supported capabilities provided by the ML capability producer. The ML capability consumer may send (step 501) a request for ML capability information and receive a response from the ML capability producer 590 including ML capability information (step 502). Instead of sending the ML capability information in response to a request, the ML capability consumer 530 may receive the ML capability information from one or more ML capability producer 590 when one or more new ML based applications are available, e.g. on the basis of a subscription scheme and I or without the need to send a request. This discovery phase (steps 501 and 502) may be repeated as needed.

[0101] The ML capability consumer 530 derives (step 503) one or more new ILUs from one or more ML based applications on the basis of the received ML capability information. The ML capability consumer stores (step 503) each new ILU in a library (e. g. Intent Logic Library, ILL) along with ILU capability information and mapping information. The mapping information may include an identification of the ML based applications mapped with the ILU. The ILU capability information of an ILU may be derived from the ML capabilities of the ML based applications mapped with the ILU.

[0102] The ILL may be configured to provide capability information for one or more ILUs and I or mapping information of a mapping between an ILU and one or more ML based applications. The ILL may be configured to perform a search of ILUs on the basis of search criteria related to capability information and I or mapping information of ILU(s) to be found. [0103] The ML capability consumer receives (step 504) an intent (e.g. from the operator) and identifies (step 505) in the Intent Logic Library suitable ILU(s) that are mapped with ML- based application so as to fulfill the received user intent. The ILll(s) may be identified on the basis of the I LU capability information stored with each I LU. It is to be noted that non- ML based logic and I or commands (in addition to ML-based ILUs) may still be needed to fulfill the given intent.

[0104] Patent application published under number WO2021/213632 A1 which is hereby incorporated by reference discloses embodiments of a system for fulfilling user intents using Intent Logic Units (ILUs) that are stored in an Intent Logic Library (ILL). For a given intent to be fulfilled, one or more intent logic units may be used and executed for achieving the intent.

[0105] The ILEP 230 may search in the ILL 235 for an I LU 250 that matches a specified intent. The search may be based on the capability information and I or description and I or mapping information of the ILUs. For a given ILU, the description of an ILU may include at least one of: a descriptive text, one or more functions of the ILU, one or more parameters of the ILU, etc.

[0106] If the I LU exists, the I LL 235 returns the I LU 250 to the I LEP 230 otherwise it returns a failure event. If the ILU 230 exists, the ILEP 230 may schedule the execution of the ILU 250, i.e. the ILEP may check if ILU can be immediately executed or if it has to wait and must be scheduled at a different time. After the identified ILU or sequence of ILUs are executed the ILEP may evaluate the outcomes of the ILU execution to confirm that the intent is achieved.

[0107] If the ILU does not exist, the ILEP 230 may determine which ILUs can be combined to achieve the specified intent. If a combined ILU can be generated by combining several ILUs, the ILEP 230 may schedule the execution of the combined ILU. Otherwise the ILEP 230 may return an execution failure to the ISP 220.

[0108] Once suitable ILU(s) are identified in step 205, the ML capability consumer may configure and I or execute (step 506) the identified ILU(s) that in turn launch the execution of the ML based applications that are mapped with the identified ILU(s) to fulfil the received intent.

[0109] FIG. 6 shows an example flow chart of a method for intent fulfilment using a ML based solution. The method involves several entities: an Intent Specification Platform 620, an intent Logic Library 635, an Intent Logic Execution Platform 630, and ML capability producers 690 including at least one ML orchestrator 680 and at least one ML application 660.

[0110] Step 601 : discovery phase. The Intent Logic Execution Platform discovers the available ML based applications, here for example the ML orchestrator 680 and the ML application 660.

[0111] Steps 602 and 603: requests for capability information. The Intent Logic Execution Platform 630 requests from one or more ML capability producers 690 (ML based application 660 and / or ML orchestrator 680) the details on their supported capabilities. Each request for capability information may be sent to an ML-based application 660 (step 603) or to an ML orchestrator 680 (step 602).

[0112] Steps 604 and 605: Publication of capabilities. The one or more ML capability producers 690 (ML based application 660 and / or ML orchestrator 680) may report their capabilities. In step 604 an ML orchestrator 680 may send capability information related to the ML applications orchestrated by this ML orchestrator 680. In step 605 an ML application 660 may send its own capability information.

[0113] The capability information received for an ML application 660 may be associated with an identifier (e.g. ModellDI) of an ML model implemented by the ML based application 660 or an identifier of the ML based application itself. The capability information received for an ML orchestrator 680 may be associated with an identifier (e.g. MLOrchestratorlDI) of the ML orchestrator.

[0114] For example the capability information of an ML based application includes: ModellDI, predicting the QoS level within a QoS scope, [Ues, QoS ID (QCI/5QI), QoS objective] wherein

- “ModellDI” is the identifier of the ML based application or associated ML model,

- “predicting the QoS level within a QoS scope” is the capability information related to the functionality performed by the ML application, and

- “[Ues, QoS ID (QCI/5QI), QoS objective]” defines the object types (here UE and QoS parameters) to which the functionality performed by the ML model are applied.

[0115] Similarly, as another example, the capability information of an ML based application includes:

ModellD2, “reducing the handover metrics by percentage P” wherein

- “ModellD2” is the identifier of the ML based application or associated ML model,

- “reducing the handover metrics by percentage P’ is the capability information related to the functionality performed by the ML application, and wherein the handover metrics include: ping pongs, Radio Link Failures due to too early handovers, Radio Link Failures due to too late handovers.

[0116] Similarly, as another example, the capability information of an ML based application includes: ModellD3, “move Ues from cell/frequency layer X to cell/frequency layer Y by reducing the cell or layer load below threshold T”. wherein X, Y, T are configuration parameter of the ML based application

[0117] Similarly, as another example, the capability information of an ML based application includes:

ModellD4, “predict the location, speed, trajectory of Ues at time instance X or time period Y to Z” wherein X, Y, Z are configuration parameter of the ML based application

[0118] Similarly, as another example, the capability information of an ML orchestrator includes:

MLOrchestratorlDI , “ensure orchestration of models with following IDs: ModellD3, ModellD4, supporting performance monitoring of ML QoS metrics: KPI1 , KPI2”

[0119] Step 606: Creating ILUs mapping to ML based applications. The Intent Logic Execution Platform derives one or more ILUs based on the discovered ML based applications capability information received in steps 604 and 605. In addition, the Intent Logic Execution Platform may store an identifier of the I LU in association with I LU capability information derived from the capability information of the one or more ML applications mapped with this ILU. In addition, the Intent Logic Execution Platform may store mapping information for the mapping between the ILU and the corresponding one or more ML based applications mapped with this ILU.

[0120] For example, for an ILU mapped with the ML based application identifier by ModellDI , the following information may be stored

ILU1, “Ensure target QoS is met", MAPPING (ModellDI, ILU1) where

- “ILU1” is the identifier of the ILU;

- “Ensure target QoS is met” is the capability information of the ILU;

“MAPPING (ModellDI , ILU1)” is the mapping information defining a mapping between ILU identified by ILU1 and the ML model identified by ModellDI .

[0121] Similarly, as another example, for an ILU mapped with the ML based application identifier by ModellD2, the following information may be stored:

ILU2, “Minimize RLF due to bad handover decisions", MAPPING (ModellD2, ILU2) [0122] Similarly, as another example, for an ILU mapped with the ML based application identifier by ModellD3, the following information may be stored:

ILU3, “Move Ues from technology/cell/frequency layer X to technology/cell/frequency layer Y”, MAPPING (ModellD3, ILU3)

[0123] Similarly, as another example, for an ILU mapped with the ML based application identifier by ModellD4, the following information may be stored:

ILU4, “Empty the cell layer X", MAPPING (ModellD3, ILU4)

[0124] Similarly, as another example, for an ILU mapped with the ML based application identifier by ModellD5, the following information may be stored:

ILU5, “Deduce the location of Ues at particular time instance or period", MAPPING (ModellD5, ILU5)

[0125] Step 607: the Intent Logic Execution Platform may create ILUs mapped to several ML applications so as to combine their capabilities. There are several ways to combine the capabilities of the ML applications.

[0126] In a first embodiment, the Intent Logic Execution Platform generates a new ILU which is mapped to several ML applications. For example the ILU is mapped with a specific combination of these ML applications according to which the ML application are executed (without using an ML orchestrator) in a sequential manner, such that output data of one ML application are used as input data to a subsequent ML application.

[0127] In a second embodiment, the Intent Logic Execution Platform derives combined ILUs as a combination of ILUs such that the combined ILU combines the capacities of the ML applications mapped with the ILUs that are combined. Such more complex ILUs will likewise be mapped to several ML based applications needed for the fulfillment of the complex ILU.

[0128] In a third embodiment, the Intent Logic Execution Platform derives an ILU which is mapped to an ML orchestrator such that the newly derived ILU combines the capacities of the ML applications orchestrated by the ML orchestrator.

[0129] For example, for an ILU mapped with the ML applications identified by ModellDI and ModellD3, the following information may be stored in association with this ILU:

ILU6, “Use technology/cell/frequency layer X for Ues which QoS meets certain condition", MAPPING (ModellDI, Model I D3, I LU 6)

Where

“ILU6” is the identifier of the ILU;

“Use technology/cell/frequency layer X for Ues which QoS meets certain condition” is the capability information that defines the functionality of the ILU;

“MAPPING (ModellDI , ModellD3, ILU6)” is the mapping information defining the mapping between the ILU identified by ILU6 and the ML applications identified by ModellDI and ModellD3.

[0130] Likewise, for an ILU mapped with the ML applications identified by ModellD3 and ModellD4, the following information may be stored in association with this ILU:

ILU7, “Empty the technology/cell/frequency layer X for Ues in specific area at specific time”, MAPPING (ModellD3, ModellD4, ILU7)

[0131] Likewise, for an ILU identified by ILLI8 mapped with the ML applications orchestrated by an orchestrator identified by the identifier MLOrchestratorlDI , the following information may be stored in association with this ILU:

ILU8, “Monitor and ensure accuracy of ML models are above X%”, MAPPING (MLOrchestratorlDI, I LU 8)

[0132] Step 608: for each newly derived ILU, the Intent Logic Execution Platform stores the capability information of the newly derived ILU and its mapping to corresponding ML based applications in the Intent Logic Library (ILL) in association with the respective identifier of the newly derived ILU.

[0133] The capability information of the newly derived ILU is derived from the capability information of the corresponding ML based application(s) mapped with this ILU. When the new ILU is mapped with a single ML based application, the capability information of the newly derived ILU may be the same as the capability information of this ML based application. When the new ILU is mapped with several ML based applications, the capability information of the newly derived ILU correspond to the capability information of all the mapped ML based applications. The capability information of the newly derived ILU may be generated automatically and I or be checked and I or approved by a user (operator).

[0134] Step 609: the Intent Logic Execution Platform receives a request to fulfil new intents. A user may request the fulfilment of intents of different kinds. For example:

Intent A: “Ues with QoS targets deterioration ofX% move to 5G layer”',

Intent B: “Empty the 5G layer in rural area during night in order to enable 5G layer switch off, then switch of the corresponding cells. Ensure at least 90% accuracy of the solution".

[0135] Step 610: the Intent Logic Execution Platform searches and selects in the ILL the ILUs matching the received intents. The search in the ILL may include a semantic analysis and I or keywords analysis of the capability information of the ILUs stored in the ILL on the basis of the specified intent. The selection may be checked and I or approved by a user (operator). If several matching ILUs are found in the ILL, the selection of the ILUs to be used may be finally performed and I or approved by a user (operator).

[0136] For example:

For fulfilling the intent A, the ILU identified by ILU6 mapped to (ModellDI , ModellD3) is selected;

For fulfilling the intent B, the ILU identified by ILU7 mapped to (ModellD3, ModellD4) and the ILU identified by ILU8 mapped to (MLOrchestratorlDI) are selected;

[0137] Step 611 : the Intent Logic Execution Platform launches the execution of ML based applications using one or more ILUs selected in step 610, where each selected I LU launches the one or more ML based applications mapped with the concerned I LU. Configuration parameters extracted from the intent to be fulfilled may be used by the ILEP or by the I LU selected in step 610 for the configuration of the one or more ML based applications.

[0138] For example, for fulfilling the above intent A: the configuration parameters extracted from the intent A includes the level of QoS targets decrease (X%) and the indication on frequency layers (e.g. from any frequency layer to 5G frequency layer); these configuration parameters are used by the ILEP for the configuration of the ML application ModellDI and ModellD3; the ILEP executes the selected I LU identified by ILU6 mapped to (ModellDI , ModellD3); the ILU ILU6 triggers the execution of the ML based applications implementing the ML applications ModellDI , ModellD3; the ML application ModellDI is configured to detect Ues which QoS objectives experiences X% decrease; the ML application ModellD3 is configured to move Ues detected by the ML application ModellDI from any frequency layer to 5G frequency layer.

[0139] Steps 612, 613 and 614: the Intent Logic Execution Platform launches the execution of ML based applications 660 (step 613) and ML orchestrator 680 (step 612) using one or more ILUs selected in step 610, where each selected ILU launches the one or more ML based applications mapped with the concerned ILU. The ML orchestrator 680 monitors (step 614) the performance of the launched ML based applications 660.

[0140] For example, for fulfilling the above intent B, the ILEP extracts configuration parameters from intent B including: the location (e.g. urban area coordinates) of the concerned Ues, required time (e.g. “any time instance”), the indication on frequency layers (e.g. from 5G frequency layer to any frequency layer) and accuracy (e.g. 90%). These configuration parameters shall be used by the ILEP for configuring the ML applications ModellD4, ModellD3 and ML orchestrator MLOrch estrato rlD1. Then: the ILEP executes the selected ILUs, i.e. the ILU7 mapped to ML applications (ModellD3, ModellD4) and the ILU8 mapped to ML orchestrator (MLOrchestratorlDI) for monitoring the ML applications (ModellD3, ModellD4 the ILU8 triggers the execution of the ML orchestrator MLOrchestratorlDI (step

612); the ILU7 triggers the execution of the ML applications (ModellD3, ModellD4) (step

613); the ML orchestrator MLOrchestratorlDI , once triggered by the ILLI8, monitors the execution of the ModellD3, ModellD4 (step 614); the ILEP performs the configuration of the ML application ModellD4 to cause it to identify Ues within urban area (e.g. based on urban area coordinates) for any time instance; the ILEP performs the configuration of the ML application Modelld3 to cause it to move Ues identified by ModellD4 from 5G frequency layer to any frequency layer; the ILEP performs the configuration of the ML orchestrator MLOrchestratorlDI to cause it to monitor and ensure the accuracy of monitored ML models, i.e., ModellD3 and ModellD4, does not fall below 90%.

[0141] Applying the ML based solution for Intent B enables the achievement of needed conditions for switching off the cells (emptying the frequency layer). However, in addition to ML based solution, a complete intent fulfilment may in addition require non-ML based commands, such as actual cell switch off which may be performed either automatically or may be operator-assisted.

[0142] FIG. 7 shows a flowchart of a method for intent fulfillment. The method involves a ML capability consumer as disclosed herein and ML capability producers as disclosed herein.

[0143] In step 710: the ML capability consumer may send a request for capability information to at least one ML capability producer. The request may be sent to at least one of: a ML based application, a ML applications orchestrator and another ML capability producer. The request may be a request to receive in response capability information for currently available capabilities or a subscription request to receive (e.g. on a regular basis) capability information for currently available capabilities and capabilities that may be available in the future.

[0144] In step 720, the ML capability consumer receives (e.g. from one or more at least one ML capability producers) capability information representative of capabilities of one or more machine learning, ML, based applications.

[0145] In step 730, the ML capability consumer generates one or more intent logic unit, ILUs. A newly created I LU is mapped with at least one corresponding ML based application and is configured to launch an execution of the at least one corresponding ML based application mapped with this I LU.

[0146] In step 740, the ML capability consumer stores the one or more ILUs in an intent logic library. Each I LU is stored in association with capability information derived for the I LU from the capability information received for the at least one corresponding ML based application mapped with this I LU . The ML capability consumer may store in the intent logic library, in association with the stored I LU, mapping information for the mapping performed between this I LU and the at least one corresponding ML based application.

[0147] In step 750, the ML capability consumer receives a request including intent information representative of an intent to be achieved for at least one communication network.

[0148] In step 760, the ML capability consumer identifies, based on ILU capability information stored in the intent logic library, one or more ILUs configured to launch an execution of one or more corresponding ML based applications adapted to fulfill the intent for the at least one communication network.

[0149] In step 770, the ML capability consumer executes the identified one or more ILUs. The execution of the identified one or more ILUs causes the launching of the execution of the one or more ML based applications mapped with the identified one or more ILUs. The launching may include configuring the mapped ML based application(s) with one or more configuration parameters. The one or more configuration parameters of an ML based application may be received from the ILU mapped with this ML based application or from the ML capability consumer (e.g. from the ILEP).

[0150] FIG. 8 shows a flowchart of a method for intent fulfillment. The method involves a ML capability consumer as disclosed herein and a ML capability producer as disclosed herein.

[0151] In step 810, a ML capability producer receives (e.g. from a ML capability consumer like an ILEP) a request for capability information.

[0152] In step 820, the ML capability producer sends, in response to the request, capability information representative of capabilities of one or more ML based applications. If the ML capability producer is a ML orchestrator, the ML capability producer sends the capability information representative of capabilities of one or more ML based applications monitored by the ML orchestrator. If the ML capability producer is a ML based application, the ML capability producer sends the capability information representative of capabilities this ML based application. If the ML capability producer is a network entity that implements a ML capability producer function on behalf of one or more ML based applications, the ML capability producer may send capability information representative of capabilities of these one or more or all ML based applications.

[0153] In step 830, a host apparatus executes one or more ML based applications. The execution is launched by the one or more ILUs mapped with these one or more ML based applications. These one or more ILUs are selected and executed to fulfill a given intent. The launching may include configuring, by each concerned ILU, the one or more mapped ML based applications with one or more configuration parameters. The configuration parameters may be stored in the ILU such that when selecting a given ILU the configuration parameters stored by the ILU are selected).

[0154] The launching may be performed through an ML orchestrator. In such a case, the ML orchestrator may receive (e.g. from a ML consumer or from the ILU mapped with the ML orchestrator) one or more configuration parameters for the one or more ML based applications monitored by the ML orchestrator. The ML orchestrator may itself configure the one or more ML based applications monitored by the ML orchestrator with the received one or more configuration parameters before and I or when and I or after triggering the execution of the monitored one or more ML based applications.

[0155] The proposed solution for intent fulfillment based on ML functionalities is relevant for intent driven management services used for mobile networks and provide a method to translate the received intent to executable actions. An intent driven management service may correspond to a management service that allows a user (e.g. a consumer of the intent driven management service) to express an intent. An intent driven management service allows the user to express desired intent for managing the network and services. The Intent driven management service producer paraphrases the expressed intent to transform the intent into executable tasks.

[0156] The proposed solution is also relevant for the management of ML functionalities, in particular the configuration of the ML functionalities. The proposed solution is also relevant for creating new relations between management function of ML functionalities and other entities (e.g. gNBs) within the communication network.

[0157] The ML applications may be configured to implement a given analytics and/or decision function that may be used in the context of a use-cases in a wide variety of applications and in different systems and various industries.

[0158] It should be appreciated by those skilled in the art that any functions, engines, block diagrams, flow diagrams, state transition diagrams, flowchart and I or data structures described herein represent conceptual views of illustrative circuitry embodying the principles of the invention. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in computer readable medium and so executed by a computer or processing apparatus, whether or not such computer or processor is explicitly shown.

[0159] Although a flow chart may describe operations as a sequential process, many of the operations may be performed in parallel, concurrently or simultaneously. Also some operations may be omitted, combined or performed in different order. A process may be terminated when its operations are completed but may also have additional steps not disclosed in the figure or description. A process may correspond to a method, function, procedure, subroutine, subprogram, etc. When a process corresponds to a function, its termination may correspond to a return of the function to the calling function or the main function.

[0160] Each described function, engine, block, step described herein can be implemented in hardware, software, firmware, middleware, microcode, or any suitable combination thereof.

[0161] When implemented in software, firmware, middleware or microcode, instructions to perform the necessary tasks may be stored in a computer readable medium that may be or not included in a host apparatus or system. The instructions may be transmitted over the computer-readable medium and be loaded onto the host apparatus or system. The instructions are configured to cause the host apparatus I system to perform one or more functions disclosed herein. For example, as mentioned above, according to one or more examples, at least one memory may include or store instructions, the at least one memory and the instructions may be configured to, with at least one processor, cause the host apparatus I system to perform the one or more functions. Additionally, the processor, memory and instructions, serve as means for providing or causing performance by the host apparatus I system of one or more functions disclosed herein.

[0162] The host apparatus or system may be a general-purpose computer and I or computing system, a special purpose computer and I or computing system, a programmable processing apparatus and I or system, a machine, etc. The host apparatus or system may be or include or be part of: a user equipment, client device, mobile phone, laptop, computer, network element, data server, network resource controller, network apparatus, router, gateway, network node, computer, cloud-based server, web server, application server, proxy server, etc.

[0163] FIG. 9 illustrates an example embodiment of an apparatus 1000. The apparatus 1000 may be a host apparatus or be part of a host system as disclosed herein. The apparatus may be configured to host at least one entity disclosed herein: a ML capability consumer function, a ML capability producer function, one or more ML based application, an ILEP or one or more functions of an ILEP, an ILL or one or more functions of an ILL.

[0164] As represented schematically by FIG. 9, the apparatus 1000 may include at least one processor 1010 and at least one memory 1010. The apparatus 1000 may include one or more communication interfaces 1040 (e.g. network interfaces for access to a wired / 1 wireless network, including Ethernet interface, WIFI interface, USB interfaces etc) connected to the processor and configured to communicate via wired I non wired communication link(s). The apparatus 1000 may include other associated hardware such as user interfaces 1030 (e.g. keyboard, mouse, display screen, etc) connected with the processor. The apparatus 1000 may further include one or more media drives 1050 for reading a computer-readable storage medium (e.g. digital storage disc 1060 (CD-ROM, DVD, Blue Ray, etc), USB key 1080, etc). The processor 1010 is connected to each of the other components 1030, 1040, 1050 in order to control operation thereof.

[0165] The memory 1020 may include a random access memory (RAM), cache memory, non-volatile memory, backup memory (e.g., programmable or flash memories), read-only memory (ROM), a hard disk drive (HDD), a solid state drive (SSD) or any combination thereof. The ROM of the memory 1020 may be configured to store, amongst other things, an operating system of the apparatus 1000 and I or one or more computer program code of one or more software applications. The RAM of the memory 1020 may be used by the processor 1010 for the temporary storage of data.

[0166] The processor 1010 may be configured to store, read, load, execute and/or otherwise process instructions 1070 stored in a computer-readable storage medium 1060, 1080 and I or in the memory 1020 such that, when the instructions are executed by the processor, causes the apparatus 1000 to perform one or more or all steps of a method described herein for the concerned apparatus 1000.

[0167] The instructions may correspond to computer program instructions, computer program code and may include one or more code segments. A code segment may represent a procedure, function, subprogram, program, routine, subroutine, module, software package, class, or any combination of instructions, data structures or program statements. A code segment may be coupled to another code segment or a hardware circuit by passing and/or receiving information, data, arguments, parameters or memory contents. Information, arguments, parameters, data, etc. may be passed, forwarded, or transmitted via any suitable technique including memory sharing, message passing, token passing, network transmission, etc.

[0168] When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. The term “processor” should not be construed to refer exclusively to hardware capable of executing software and may implicitly include one or more processing circuits, whether programmable or not. A processor or likewise a processing circuit may correspond to a digital signal processor (DSP), a network processor, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a System-on-Chips (SoC), a Central Processing Unit (CPU), a Processing Unit (CPU), an arithmetic logic unit (ALU), a programmable logic unit (PLU), a processing core, a programmable logic, a microprocessor, a controller, a microcontroller, a microcomputer, any device capable of responding to and/or executing instructions in a defined manner and/or according to a defined logic. Other hardware, conventional or custom, may also be included. A processor or processing circuit may be configured to execute instructions adapted for causing the host apparatus or system to perform one or more functions disclosed herein for the concerned host apparatus or system.

[0169] A computer readable medium or computer readable storage medium may be any tangible storage medium suitable for storing instructions readable by a computer or a processor. A computer readable medium may be more generally any storage medium capable of storing and/or containing and/or carrying instructions and/or data. A computer- readable medium may be a portable or fixed storage medium. A computer readable medium may include one or more storage device like a permanent mass storage device, magnetic storage medium, optical storage medium, digital storage disc (CD-ROM, DVD, Blue Ray, etc), USB key or dongle or peripheral, a memory suitable for storing instructions readable by a computer or a processor.

[0170] A memory suitable for storing instructions readable by a computer or a processor may be for example: read only memory (ROM), a permanent mass storage device such as a disk drive, a hard disk drive (HDD), a solid state drive (SSD), a memory card, a core memory, a flash memory, or any combination thereof.

[0171] In the present description, the wording "means configured to perform one or more functions" or “means for performing one or more functions” may correspond to one or more functional blocks comprising circuitry that is adapted for performing or configured to perform the concerned function(s). The block may perform itself this function or may cooperate and I or communicate with other one or more blocks to perform this function. The "means" may correspond to or be implemented as "one or more modules", "one or more devices", "one or more units", etc. The means may include at least one processor and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause an apparatus or system to perform the concerned function(s).

[0172] As used in this application, the term “circuitry” may refer to one or more or all of the following:

(a) hardware-only circuit implementations (such as implementations in only analog and/or digital circuitry) and

(b) combinations of hardware circuits and software, such as (as applicable) : (i) a combination of analog and/or digital hardware circuit(s) with software/fi rmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions); and

(c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.”

[0173] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and/or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, an integrated circuit for a network element or network node or any other computing device or network device.

[0174] The term circuitry may cover digital signal processor (DSP) hardware, network processor, application specific integrated circuit (ASIC), field programmable gate array (FPGA), etc. The circuitry may be or include, for example, hardware, programmable logic, a programmable processor that executes software or firmware, and/or any combination thereof (e.g. a processor, control unit/entity, controller) to execute instructions or software and control transmission and receptions of signals, and a memory to store data and/or instructions.

[0175] The circuitry may also make decisions or determinations, generate frames, packets or messages for transmission, decode received frames or messages for further processing, and other tasks or functions described herein. The circuitry may control transmission of signals or messages over a radio network, and may control the reception of signals or messages, etc., via one or more communication networks.

[0176] Although the terms first, second, etc. 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 this disclosure. As used herein, the term "and/or," includes any and all combinations of one or more of the associated listed items.

[0177] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. 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," "includes," and/or "including," when used herein, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

[0178] While aspects of the present disclosure have been particularly shown and described with reference to the embodiments above, it will be understood by those skilled in the art that various additional embodiments may be contemplated by the modification of the disclosed machines, systems and methods without departing from the scope of what is disclosed. Such embodiments should be understood to fall within the scope of the present disclosure as determined based upon the claims and any equivalents thereof.

[0179] LIST OF ABBREVIATIONS

5G Fifth Generation

CAN Cognitive Autonomous Networks

CF Cognitive Function

CLI Command Line Interface

CNM Cognitive Network Management

GUI Graphical User Interface

I BN Intent- Based Networking loT Internet of Things

KPI Key Performance Indicator

NE Network Element

NM Network Management

OAM Operations, Administration and Management

SON Self-Organizing Network