EP4011038A1 - Configuration d'analytique de réseau - Google Patents

Configuration d'analytique de réseau

Info

Publication number
EP4011038A1
EP4011038A1 EP19753026.4A EP19753026A EP4011038A1 EP 4011038 A1 EP4011038 A1 EP 4011038A1 EP 19753026 A EP19753026 A EP 19753026A EP 4011038 A1 EP4011038 A1 EP 4011038A1
Authority
EP
European Patent Office
Prior art keywords
analytics model
analytics
network
request
commanding
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
EP19753026.4A
Other languages
German (de)
English (en)
Inventor
Konstantinos Samdanis
Janne Tapio ALI-TOLPPA
Andreas Maeder
Yannick Lair
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nokia Technologies Oy
Original Assignee
Nokia Technologies Oy
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nokia Technologies Oy filed Critical Nokia Technologies Oy
Publication of EP4011038A1 publication Critical patent/EP4011038A1/fr
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present invention relates to network analytics, in particular to configuring a network analytics function.
  • NWDAF Network Data Analytics Function
  • SBA Service Based Architecture
  • NF Network Function
  • SBI SBA Interface
  • AFs Application Functions
  • UDR User Data Repository
  • OAM Operations Administration and Maintenance
  • NWDAF Network Automation
  • NWDAF relies on a computational logic or algorithm to process the received data in order to produce the desired statistics and/or predictions. Such logic or algorithms require a training phase that can initially take place offline but would typically need thereafter a continuous re training that takes place regularly based on observed data, i.e. online data that is collected from related NFs or AFs.
  • NWDAF logic or algorithms needs to have a service target scope related to a set of one or more application(s), a set of one or more UE(s), and/or a set of one or more cell(s), a set of input data sources, i.e.
  • 3GPP NWDAF exploits the means of enhancing the operation of NFs using analytics. However, for 3GPP, NWDAF is a “black box” with no further considerations related to its logic, i.e. analytics model, and location. I.e., according to 3GPP, NWDAF only gets input from certain NFs, AFs, UDR and OAM and provides output analytics to subscribed NFs, AFs and OAM.
  • 3GPP does not specify how to introduce a customized algorithm, i.e. beyond the specified one, e.g. mobility, load, etc., and how the algorithm training process and validation is accomplished.
  • NEF allows to expose capabilities for monitoring and for information provisioning from/to outside the operator network. NEF is also considered for exposure of particular NWDAF events towards an external AF or allows an external AF to feed the NWDAF with certain user or service related information.
  • NEF [2] [3] [4] allows network capability exposure towards an external 3 rd party AF in terms of: (i) network events monitoring, (ii) provisioning capability towards external functions, (iii) policy and charging capabilities toward external NFs/AFs and (iv) core network internal capabilities for analytics (monitoring and data correlation, i.e. AF to provide an expected UE and/or service behaviour).
  • the NEF concentrates on:
  • Expose capability/event to AF e.g. monitor mobility, load, location, etc.
  • UDR e.g. regulate Background Data Transfer (BDT), communication patterns, QoS, loss of connectivity, roaming status, etc.
  • external exposure by NEF can be categorized as monitoring, provisioning and policy/charging capability.
  • NEF The role of NEF has also been considered in the context of NWDAF in 3GPP TR 23.791 [1] considering the following solutions: • AF provides the observed service experience data, which allows operator to check, i.e. correlate, against the provisioned service quality (Solution 3).
  • NEF provides NWDAF metadata to AF in order to select the analytics services that it would subscribe to the NWDAF focusing on monitoring events (Solution 13).
  • Network conditions may change impacting the BDT policy agreed with AF; the NWDAF can notify the Policy Control Function (PCF) of potential congestion experience allowing the PCF to re-negotiate policy with the AF (Solution 28).
  • PCF Policy Control Function
  • NEF The role of NEF has also been extended in the context of NWDAF in 3GPP TS 23.288 [7] to allow collecting data from AF by 5G Core Network.
  • the role of NEF related to NWDAF concentrates on (i) providing AF data to the network, (ii) allow the AF to monitor analytics and network information, (iii) analytics metadata that allow the AF to subscribe to monitoring events and (iv) on policy re-negotiation.
  • an AF can subscribe to a particular NWDAF using the Analytics ID, indicating the target of analytics reporting i.e. UE or set of UEs, analytics filter (i.e. type of parameters), analytics reporting information (including event, maximum reports, duration, level of aggregation, etc.), observation period, level of accuracy, validity period, confidence degree and the correlation information [7]
  • the AF can provide an area of interest, which can be e.g. a geographical area, but then in the 5GC there is a need to find the NWDAF relevant for this area of interest so there needs to be a mapping between the area of interest and the Analytics serving areas [9].
  • Preprocessor entity responsible for aggregation or formatting, i.e. preparing, data for ML Model
  • ML model is the logic and/or algorithm, which needs to be trained using offline and/or online data before it can be put into action
  • Sink is target node of the distributor to forward the ML output
  • the ML-Pipeline relies on an ML Function Orchestrator (MLFO) that takes care of the ML life- cycle, i.e. composition, configuration (including ML Chaining, i.e. a logical execution sequence that realizes an ML-Pipeline) and management (scale-up/down, re-locate).
  • MLFO ML Function Orchestrator
  • An ML-Pipeline can be described in a declarative fashion by the means of ML Meta-Language, which can be used by 3 rd parties for acquisition and/or configuration.
  • An ML Application can then be configured in an existing ML-Pipeline with specific roles for the Sources, Collector, Distributor, Sinks, etc. and specific deployment details considering the capabilities of network nodes.
  • the operation of a ML Application considers a closed-loop that can capture network and service dynamics that allows the ML-Pipeline to adapt to an evolving environment.
  • 3GPP TS 23.288 “Architecture enhancements for 5G System (5GS) to support Network Analytics Services”, V16.0.0, June. 2019.
  • 3GPP SP-190557 “SID: Study on Enablers for Network Automation for 5G - Phase 2”, TSG-SA Meeting #84, June 2019.
  • an apparatus comprising means for monitoring configured to monitor if a creation request from a creator is present, wherein the creation request requests to create an analytics model in a network; means for obtaining configured to obtain a value of an analytics model identifier related to the analytics model if the creation request is present; means for ensuring configured to ensure that the value of the analytics model identifier is unique in the network; means for commanding configured to command creating the analytics model in the network if the creation request is present; means for supervising configured to supervise, if the commanding and ensuring was successful, if at least one of a verification request to perform a verification of the analytics model, an update request to perform an update of the analytics model, and a modification request to perform a modification of the analytics model from the creator is present; means for instructing configured to instruct to perform the at least one of the verification, the update, and the modification if the respective at least one of the verification request, the update request, and the modification request is present, wherein the instructing is based on
  • an apparatus comprising means for monitoring configured to monitor if a creation request to create an analytics model is received from a requestor, wherein the request comprises the analytics model and analytics service operational information as meta data, and the analytics service operational information comprises at least one of a type of a source of the analytics model, a type of a sink of the analytics model, a target user equipment to be monitored by the analytics model, an application to be monitored by the analytics model, and an area of interest of the analytics model; the source comprises at least one network function of a network; the application is applied to the network; the area of interest belongs to the network; and the meta data are independent from the network; means for inquiring configured to inquire a respective identification of the at least one of the source, the sink, the application, and the area of interest from a repository function of the network if the creation request is received, wherein the respective identification is used if the at least one of the source, the sink, the application, and the area of interest is addressed within the network; means for commanding configured
  • a method comprising monitoring if a creation request from a creator is present, wherein the creation request requests to create an analytics model in a network; obtaining a value of an analytics model identifier related to the analytics model if the creation request is present; ensuring that the value of the analytics model identifier is unique in the network; commanding creating the analytics model in the network if the creation request is present; supervising, if the commanding and ensuring was successful, if at least one of a verification request to perform a verification of the analytics model, an update request to perform an update of the analytics model, and a modification request to perform a modification of the analytics model from the creator is present; instructing to perform the at least one of the verification, the update, and the modification if the respective at least one of the verification request, the update request, and the modification request is present, wherein the instructing is based on the value of the analytics model identifier.
  • a method comprising monitoring if a creation request to create an analytics model is received from a requestor, wherein the request comprises the analytics model and analytics service operational information as meta data, and the analytics service operational information comprises at least one of a type of a source of the analytics model, a type of a sink of the analytics model, a target user equipment to be monitored by the analytics model, an application to be monitored by the analytics model, and an area of interest of the analytics model;
  • the source comprises at least one network function of a network; the application is applied to the network; the area of interest belongs to the network; and the meta data are independent from the network; inquiring a respective identification of the at least one of the source, the sink, the application, and the area of interest from a repository function of the network if the creation request is received, wherein the respective identification is used if the at least one of the source, the sink, the application, and the area of interest is addressed within the network; commanding creating the analytics model based on the respective identification.
  • Each of the methods of the third and fourth aspects may be a method of configuring network analytics.
  • a computer program product comprising a set of instructions which, when executed on an apparatus, is configured to cause the apparatus to carry out the method according to any of the third and fourth aspects.
  • the computer program product may be embodied as a computer-readable medium or directly loadable into a computer.
  • an external AF may instantiate, configure, and validate an analytics model in a network
  • Fig. 1 illustrates an overview of a data analytics system according to some example embodiments of the invention
  • Fig. 2 shows a method according to some example embodiments of the invention
  • Fig. 3 shows an apparatus according to an example embodiment of the invention
  • Fig. 4 shows a method according to an example embodiment of the invention
  • Fig. 5 shows an apparatus according to an example embodiment of the invention
  • Fig. 6 shows a method according to an example embodiment of the invention
  • Fig. 7 shows an apparatus according to an example embodiment of the invention.
  • the apparatus is configured to perform the corresponding method, although in some cases only the apparatus or only the method are described.
  • the analytics service operational information needs to be communicated to the network if the NWDAF service is customized, i.e. a newly created (analytics model) and configured, by a 3 rd party, i.e. by one or more untrusted AFs.
  • the untrusted AF(s) need to communicate a set of one or more application(s), a set of one or more UE(s), and/or Area of Interest that define a set of one or more cell(s), etc., which enables the selection of the optimal new NWDAF instance and/or the discovery and selection of the optimal sources/sinks considering the said targets.
  • 3GPP has not specified how a 3 rd party AF can communicate to the NWDAF its preferences related to
  • Some example embodiments of this invention fill these gaps by assisting a 3 rd party AF (untrusted AF) to subscribe to NWDAF analytics services.
  • untrusted AF untrusted AF
  • the mobile operator provides the “platform” for realizing NWDAF with respect to an external AF’s needs.
  • the operator may allow the untrusted AF to install the statistics and/or prediction logic or algorithm at selected locations inside the mobile operator network (on the platform provided by the operator for NWDAF) and/or to feed the necessary operational scope with respect to the analytics service operational information (e.g. an application, UE, set of UEs, cell(s), etc., and desired sources and sinks regarding input/output analytics data).
  • NWDAF may be operator-specific, but still certain parameters related to the means of introducing and configuring the analytics logic and corresponding algorithms should preferably be commonly defined in order to allow smooth interaction with external AFs.
  • external AF may perform at least one of the following activities:
  • NEF faces the following fundamental issues for allowing an external AF to exploit NWDAF analytics services:
  • NEF supports no means for allowing a 3 rd party AF to optimize the selection of the NWDAF instance location or to configure the NWDAF with the analytics model and/or algorithm.
  • NEF supports no means for allowing a 3 rd party AF to be involved in NWDAF analytics model training and validation process.
  • Some example embodiments of the invention solve at least one of these problems.
  • Some example embodiments of this invention provide a means for AFs to select the NWDAF placement and analytics model configuration. They may allow AFs to configure NWDAF specific services with user subscription enabling a NWDAF service to be used once a particular application is instantiated and the user subscription record is checked.
  • An identifier e.g. analytics model identifier
  • unique in the mobile network is assigned to such a NWDAF type with a customized Analytics Model.
  • Some example embodiments of the invention provide a method to allow a 3 rd party AF to interact with the NWDAF by means of NEF extensions and by a new subscription type related to Analytics Model Validation.
  • NEF subscription handles the analytics model training and validation as a NEF event by assigning an associated Event Id to the particular AF, taking advantage of the conventional NEF subscription paradigm of (i) Configuration (HTTP POST), (ii) Modification (HTTP PUT / HTTP PATCH) to modify the properties related to the current subscription and (iii) Cancelation (HTTP DELETE).
  • HTTP POST Configuration
  • Modification HTTP PUT / HTTP PATCH
  • Cancelation HTTP DELETE
  • Such Analytics Model Validation may allow the AF to perform one or more of the following activities but not limited to:
  • model validation data including at least one of a. Correctness/deviation of prediction information compared to real events, i.e. if the difference surpasses a predefined threshold then trigger a re-training. b. KPIs on false positives or negatives related to categorization tasks. c. KPIs of network optimizations based on analytics models.
  • Supervised validation expose validation information to AF, including at least one of a. Indication tool e.g. gate keeper, to take analytics model offline and replace it with another one.
  • a. Indication tool e.g. gate keeper
  • List data sources e.g. NFs
  • Dynamic metadata data to trace the source of data in relation with the analytic data for e.g. validation, or for the model inference, internal exception 4.
  • Configure a simulation environment that takes online data as input (e.g. on a regular basis, randomly selected, etc.) and compares it with the actual outcome
  • the NEF may check for each of the aforementioned processes if it conforms with the operator's policy for the requesting AF. If it does not, it may reject particular parameters or even the entire analytics service request.
  • Example embodiments of the invention may operate with different NWDAF deployment models.
  • only one NWDAF is deployed in the network.
  • the network operator deploys different instances of the NWDAF, each supporting certain analytics capabilities.
  • the NWDAF instances may reside at one or more distinct locations.
  • FIG. 1 An overview of a data analytics system according to some example embodiments of the invention is illustrated in Fig. 1. It depicts the components of a ML-pipeline, i.e. data sources (including NFs, AFs, UDR and OAM), data collector, and analytics model (e.g. NWDAF), and the sinks (NFs, AFs, OAM) that consume the analytics services.
  • Fig. 1 also shows how an untrusted AF interacts with the 3GPP network via NEF. NEF authenticates the untrusted AF and subscribes the untrusted AF's request to configure and consume an analytics service from the NWDAF.
  • NEF authenticates the untrusted AF and subscribes the untrusted AF's request to configure and consume an analytics service from the NWDAF.
  • An analytic service can be realized by a network operator with an NWDAF with the option of allowing a 3 rd party to install its own statistics and/or predictive logic and/or algorithms
  • the NEF subscription allows a 3 rd party AF to install an analytics model by means of a Meta-Language, indicating NWDAF location within the operator’s network, e.g. platform selection, and configuring the so called ITU ML-Pipeline, i.e. indicating the type of source(s)/sink(s) and the corresponding target(s), i.e. a set of one or more application(s), a set of one or more UE(s), and/or Area of Interest that define a set of one or more cell(s), etc.
  • the NWDAF with a new Analytics Model type will be assigned a new Analytics model ID. It is ensured that each Analytics model ID is unique within the network.
  • the check for uniqueness may be performed e.g. by the NEF or NWDAF. For example, they may compare the entries included in a database related to the existing analytics models and their analytics model IDs with the newly generated analytic model ID.
  • the database may provide such a unique analytics model ID.
  • the database may be included in NRF, UDR or OAM, or there may be a dedicated database for holding NWDAF related information including Analytics model IDs.
  • NWDAF provides analytics information with respect to specified analytics serving area [9], with the notion of an analytics service area defined geographically or via the means of Tracking Area or cell(s). Such association is known to the NRF that provides this information upon request.
  • NEF may select the location of the NWDAF instance with the assistance of NRF with respect to the target set of one or more application(s), a set of one or more UE(s), and/or Area of Interest that define a set of one or more cell(s), etc., considering any combination of the following attributes: (i) geographical location, (ii) mobility aspects with respect to speed, expected travel distance of the UE(s), etc. (iii) relative distance between sources/sinks and NWDAF, (which provides an indication of the expected amount of data that may relate to the cost of a data analytics service), and (iv) service type.
  • the NEF allows the AF to create a subscription related to the Analytics Model and allows to configure the related parameters.
  • the authorization by NEF may be omitted.
  • the authorization by NEF is not necessary if the AF is a trusted AF (inside the network). In Fig. 2, AF is external to the network (non-trusted).
  • Fig. 2 shows a method (message chart) according to some example embodiments of the invention. The method has three major portions:
  • Authentication and authorization may be omitted in some example embodiments
  • NWDAF instantiation i.e., the creation of an analytics model on NWDAF. This step is performed only once for each analytics model instance;
  • Fig. 2 comprises the following steps:
  • the AF creates an analytics model/verification authentication by issuing a provision request towards the NEF (step 1 ).
  • the request is then authorized and authenticated by NEF (step 2).
  • a response is provided to the AF (step 3).
  • an error message is returned (this applies to all the requests and messages shown in Fig. 2).
  • the AF introduces the new analytics model (step 4) by contacting NEF.
  • the NEF in turn introduces bindings for (i) installing the analytics model; and at least one of (ii) the types of source(s)/sink(s) and targets UE(s), application(s), cell(s), etc. (iii) the re-training/reporting conditions, e.g. reporting period, KPI deviation, etc. (iv) validation reporting e.g. performance deviation, performance data, etc., considering the predefined reporting conditions, and (v) model re-training indication or new Analytics Model configuration notifying the NWDAF of the need to replace current Analytics Model.
  • These processes enable the NEF to support the installation of the Analytics Model including at least one of the re-training conditions, validation reporting data and model-retraining/modification indications.
  • NEF obtains an Analytics model ID for the binding purposes.
  • the analytics model ID is unique in the network.
  • the analytics model ID may be provided to NRF for discovering the newly created NWDAF.
  • Uniqueness of the Analytics model ID inside the mobile operator domain may be ensured by consulting a database (DB) entity (an example of a network or data repository) storing all the analytics model IDs with the parameters of the respective analytics model.
  • the database may be located e.g. in NRF, UDR, OAM, or it may be a dedicated NWDAF database.
  • the process of assigning the Analytics model ID can be performed by one of the following two options: Option 1) where the check of the uniqueness of the Analytics model ID is performed in the same step as the NEF binding, and Option 2) where the NWDAF performs that check of the uniqueness of the Analytics model ID after installing the Analytics Model successfully.
  • the NEF which keeps track of the bindings towards an external AF, may check the DB (e.g. NRF, UDR, OAM or dedicated NWDAF database) and assign a unique Analytics model ID (Option 1 - step 4). Otherwise it may assign a temporary binding ID until the NWDAF performs a check for uniqueness (Option 2 - step 7).
  • the external AF can then identify the desired NWDAF based on the Analytics model ID for operations, validation, monitoring, and modification.
  • the discovery of the appropriate NWDAF platform for installing the Analytics Model is performed by the NEF by contacting the NRF (step 5) indicating the desired analytics capabilities.
  • the NEF upon receiving the available options for potential NWDAF platforms, can select the appropriate one considering - if desired or necessary - location (e.g. geographical, area of interest that defines a set of cells, target UEs and mobility, distance among the NWDAF platform with potential sources/sinks instances, service type, etc.).
  • the NWDAF configuration can then follow, where the AF triggers via the NEF, the installation of the Analytics Model acknowledging the discovery process (step 6), with the NEF carrying out then the installation and configuration of the NWDAF platform based on the created bindings (step 7).
  • the NWDAF platform is selected, the desired installations and configurations are provided.
  • the NWDAF checks a database (which can be e.g. NRF, UDR, OAM or dedicated NWDAF database) and assigns a unique Analytics model ID, which is going to be used for identifying the new NWDAF type in the mobile operator network (Option 2 - step 7).
  • the new analytics model (NWDAF type) then registers with the NRF providing the Analytics model ID (step 8), which is going to be used by other NF to discover the new NWDAF type.
  • the new Analytics model ID may be dynamic in nature in the sense that it is created and revoked according to the lifetime of the new NWDAF.
  • the NWDAF further establishes the subscription to the required sources/sinks (e.g. NF, UDR, OAM, AF) creating the producer- consumer relation according to the desired reporting/validation conditions (step 9), before responding back to the AF, with a create response (step 10).
  • the required sources/sinks e.g. NF, UDR, OAM, AF
  • step 9 introduces a dynamic metadata data to trace the source of data in relation with the analytic data for the model inference.
  • the create response of step 10 comprises the unique analytics model ID such that NEF may replace the temporary binding ID with the unique analytics model ID.
  • the create response of step 10 may comprise the unique analytics model ID in Option 1 , too.
  • the translation of the metadata into identifiers used to address the respective function in the network may be performed in NEF or in NWDAF, depending on implementation.
  • some of the analytics service operational information provided by the external AF may be translated into respective identifiers in the NEF, and the remaining part may be translated by NWDAF.
  • identifiers received from the analytics function e.g. in closed loop control
  • identifiers may be translated into meta data. For each identifier, such translation may be performed by NEF or NWDAF, depending on implementation.
  • the process of authentication and mapping/binding creation can alternatively be combined in a single step. In that case step 1 and step 2 can be merged with step 4, and step 3 and step 10 may be merged for creating a combined response considering the authentication and Analytics model create. In between these two merged steps, the remaining steps may be executed as described above.
  • AF may update, validate, or modify the analytics model.
  • the analytics model is identified by the unique analytics model ID.
  • NEF maps the analytics model ID to an identifier of the AF (or a combination of the identifier of the AF and an identifier of an application on the AF) and vice versa.
  • the AF may request performance data for validation, i.e. to examine if the Analytics Model performs as expected (step 11 ).
  • performance data i.e. to examine if the Analytics Model performs as expected (step 11 ).
  • such process can be triggered once certain conditions are met, e.g. periodic or upon a performance deviation observation at the NWDAF.
  • the NWDAF prepares and formats the validation data (step 11 ).
  • the NEF may provide an appropriate data abstraction, e.g. hiding network topology or other mobile network operator internal network or data details.
  • the NEF may also keep track of the analytics sources related to the validation data, to be able to take the appropriate action, e.g. trigger examination of the health state of a certain NF if the AF raises such suggestion after processing the validation data.
  • the NEF can then forward the validation data notification towards the AF (step
  • the AF may then determine if re-training or an installation of a new Analytics Model is needed (step 14). In case the AF identifies that the deviation of validation data compared to the expected result is beyond a pre-defined threshold (e.g. completely out of range), it can trigger a root cause analysis to explore if there is a fault with the sources of the analytics data.
  • the NEF can then notify the NWDAF to check e.g. by correlating different analytics services, whether certain indicated sources of data, (with this being the main reason for the NEF to keep track of the analytics sources related to the validation data), result in continuous performance deviations or other abnormal behavior.
  • the NWDAF may then identify if a certain data analytics source is faulty and trigger a replacement with another equivalent one, e.g. change an AMF instance in use provided that the new one can serve the same target UEs in the same location.
  • the AF determines that there is a need for re-training or a need to introduce and replace the current Analytics Model at the NWDAF, it issues an update Analytics Model request via the NEF (step 15).
  • the NWDAF executes either the instructed re-training or modification/replacement process (step 16) and then replies to the AF with an update response acknowledging the action or providing a report of potential errors (step 17).
  • the AF identifies a performance deviation, e.g. beyond a pre determined threshold (step 15)
  • it triggers a root cause analysis that is processed by the NEF indicating to the NWDAF that a certain data analytics source may be faulty.
  • the NWDAF may then determine if the assumption on which the root cause analysis is based is valid. For example, NWDAF may correlate data with other analytics processes, or run a pre-installed test procedure, or trigger the OAM to check the health state of the indicated NF (step 16). Then, NWDAF may reply to the AF with an update response acknowledging the action or providing a report of potential errors (step 17).
  • the process of deletion of the NEF subscription can follow the conventional means as described in TS 29.522 [4]
  • the NEF keeps track of the data analytics sources related to the validation data, it can notify also a different NWDAF instance to check the health of the data analytics source.
  • the NWDAF can keep track of the data analytics sources without exposing them to the NEF, with the NEF then only notifying the NWDAF that a health check is required related with the operating state of a data analytics source.
  • the Option 2 as described above may be adopted for assigning and ensuring that the new Analytics Service type has a unique Analytics model ID.
  • Fig. 3 shows an apparatus according to an embodiment of the invention.
  • the apparatus may be binding function such as a NEF, or an element thereof.
  • Fig. 4 shows a method according to an embodiment of the invention.
  • the apparatus according to Fig. 3 may perform the method of Fig. 4 but is not limited to this method.
  • the method of Fig. 4 may be performed by the apparatus of Fig. 3 but is not limited to being performed by this apparatus.
  • the apparatus comprises means for monitoring 10, means for obtaining 20, means for ensuring 30, means for commanding 40, means for supervising 50, and means for instructing 60.
  • the means for monitoring 10, means for obtaining 20, means for ensuring 30, means for commanding 40, means for supervising 50, and means for instructing 60 may be a monitoring means, obtaining means, ensuring means, commanding means, supervising means, and instructing means, respectively.
  • the means for monitoring 10, means for obtaining 20, means for ensuring 30, means for commanding 40, means for supervising 50, and means for instructing 60 may be a monitor, obtainer, insurer, commander, supervisor, and instructor, respectively.
  • the means for monitoring 10, means for obtaining 20, means for ensuring 30, means for commanding 40, means for supervising 50, and means for instructing 60 may be a monitoring processor, obtaining processor, ensuring processor, commanding processor, supervising processor, and instructing processor, respectively.
  • the means for monitoring 10 monitors if a creation request from a creator is present (S10).
  • the creation request requests to create an analytics model in a network.
  • the creator may be a network external AF (non-trusted) or a network internal AF (trusted).
  • the means for obtaining 20 obtains a value of an analytics model identifier related to the analytics model (S20).
  • the obtaining (S20) may comprise receiving the value from another entity or generating the value.
  • the means for ensuring 30 ensures that the value of the analytics model identifier is unique in the network (S30). As a consequence, the analytics model identifier identifies the analytics model unambiguously.
  • S20 and S30 may be performed one after the other or jointly in a single step. In the latter case, the means for obtaining 20 comprises the means for ensuring.
  • the means for commanding 40 commands creating the analytics model in the network (S40).
  • S40 and the combination S20+S30 may be performed in an arbitrary sequence (see e.g. Options 1 and 2 hereinabove).
  • the following actions S50 and S60 are performed only if the ensuring (S30) by the means for ensuring 30 and the commanding (S40) by the means for commanding 40 was successful:
  • the means for supervising 50 supervises, if at least one of a verification request to perform a verification of the analytics model, an update request to perform an update of the analytics model, and a modification request to perform a modification of the analytics model from the creator is present (S50).
  • the means for instructing 60 instructs to perform the at least one of the verification, the update, and the modification (S60).
  • the instructing is based on the value of the analytics model identifier. I.e., the instructing addresses the analytics model in the network (e.g. NRF, NWDAF) by the value of the analytics model identifier.
  • Fig. 5 shows an apparatus according to an example embodiment of the invention.
  • the apparatus may be an NWDAF or an element thereof.
  • Fig. 6 shows a method according to an example embodiment of the invention.
  • the apparatus according to Fig. 5 may perform the method of Fig. 6 but is not limited to this method.
  • the method of Fig. 6 may be performed by the apparatus of Fig. 5 but is not limited to being performed by this apparatus.
  • the apparatus comprises means for monitoring 110, means for inquiring 120, and means for commanding 130.
  • the means for monitoring 110, means for inquiring 120, and means for commanding 130 may be a monitoring means, inquiring means, and commanding means, respectively.
  • the means for monitoring 110, means for inquiring 120, and means for commanding 130 may be a monitor, inquirer, and commander, respectively.
  • the means for monitoring 110, means for inquiring 120, and means for commanding 130 may be a monitoring processor, inquiring processor, and commanding processor, respectively.
  • the means for monitoring 110 monitors if a creation request to create an analytics model is received from a requestor (S110).
  • the request comprises analytics service operational information as meta data.
  • the analytics service operational information comprises at least one of a source of the analytics model, a sink of the analytics model, an application to be monitored by the analytics model, and an area of interest of the analytics model; the source comprises at least one network function of a network; the application is applied to the network; the area of interest belongs to the network.
  • the meta data are independent from the network.
  • the means for inquiring 120 inquires a respective identification of the at least one of the source, the sink, the application, and the area of interest from a repository function of the network, such as a network repository function NRF (S120).
  • a repository function of the network such as a network repository function NRF (S120).
  • NRF network repository function
  • UDR unified data repository
  • the means for commanding 130 commands creating the analytics model based on the respective identification (S130).
  • Fig. 7 shows an apparatus according to an embodiment of the invention.
  • the apparatus comprises at least one processor 810, at least one memory 820 including computer program code, and the at least one processor 810, with the at least one memory 820 and the computer program code, being arranged to cause the apparatus to at least perform at least the method according to at least one of Figs. 4 and 6.
  • analytics ID analytics ID
  • NEF constitutes the interface between the external AF and NWDAF.
  • the invention is not limited to an NEF based interface.
  • the interface is provided by a dedicated function, or it is integrated with any other function such as NRF or NWDAF.
  • Example embodiments of the invention are described for 5G networks.
  • the invention is not restricted to 5G networks and may be employed in other 3GPP networks such as 3G networks, 4G networks, and upcoming 3GPP releases, too.
  • the invention may be employed in non-3GPP networks (mobile networks, fixed networks, and converged networks) provided they comprise a function corresponding to a NWDAF.
  • One piece of information may be transmitted in one or plural messages from one entity to another entity. Each of these messages may comprise further (different) pieces of information.
  • Names of network elements, protocols, and methods are based on current standards. In other versions or other technologies, the names of these network elements and/or protocols and/or methods may be different, as long as they provide a corresponding functionality.
  • each of the entities described in the present description may be based on a different hardware, or some or all of the entities may be based on the same hardware. It does not necessarily mean that they are based on different software. That is, each of the entities described in the present description may be based on different software, or some or all of the entities may be based on the same software.
  • Each of the entities described in the present description may be embodied in the cloud.
  • example embodiments of the present invention provide, for example, a mapping function such as a NEF, or a component thereof, an apparatus embodying the same, a method for controlling and/or operating the same, and computer program(s) controlling and/or operating the same as well as mediums carrying such computer program(s) and forming computer program product(s).
  • Implementations of any of the above described blocks, apparatuses, systems, techniques or methods include, as non-limiting examples, implementations as hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

Il est fourni un procédé, comprenant les étapes suivantes : surveiller si une demande de création d'un modèle analytique est présente; obtenir une valeur d'un identifiant de modèle analytique associé au modèle analytique si la demande est présente; garantir que la valeur de l'identifiant de modèle analytique est unique dans le réseau; commander la création du modèle analytique dans le réseau si la demande est présente; superviser, si la commande et la garantie ont été réussies, si au moins une demande parmi une demande de vérification pour effectuer une vérification du modèle analytique, une demande de mise à jour pour effectuer une mise à jour du modèle analytique, et une demande de modification afin d'effectuer une modification du modèle analytique est présente; ordonner d'effectuer l'au moins une de la vérification, la mise à jour, et la modification si l'au moins une de la demande de vérification respective, de la demande de mise à jour et de la demande de modification sont présentes, l'instruction étant basée sur la valeur de l'identifiant de modèle analytique.
EP19753026.4A 2019-08-08 2019-08-08 Configuration d'analytique de réseau Pending EP4011038A1 (fr)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/EP2019/071316 WO2021023388A1 (fr) 2019-08-08 2019-08-08 Configuration d'analytique de réseau

Publications (1)

Publication Number Publication Date
EP4011038A1 true EP4011038A1 (fr) 2022-06-15

Family

ID=67620440

Family Applications (1)

Application Number Title Priority Date Filing Date
EP19753026.4A Pending EP4011038A1 (fr) 2019-08-08 2019-08-08 Configuration d'analytique de réseau

Country Status (2)

Country Link
EP (1) EP4011038A1 (fr)
WO (1) WO2021023388A1 (fr)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023006212A1 (fr) * 2021-07-30 2023-02-02 Huawei Technologies Co., Ltd. Traçage analytique de réseau, et rollback pour une consommation stable de la sortie analytique
WO2023118307A1 (fr) * 2021-12-21 2023-06-29 Telefonaktiebolaget Lm Ericsson (Publ) Systèmes et procédés de commande de re-apprentissage de modèle aiml dans des réseaux de communication
WO2023213286A1 (fr) * 2022-05-05 2023-11-09 维沃移动通信有限公司 Procédé et appareil de gestion d'identifiant de modèle, et support de stockage
WO2023247060A1 (fr) * 2022-06-22 2023-12-28 Telefonaktiebolaget Lm Ericsson (Publ) Premier nœud, deuxième nœud, troisième nœud et procédés mis en œuvre par ces derniers pour gérer le trafic
WO2024094279A1 (fr) * 2022-10-31 2024-05-10 Huawei Technologies Co., Ltd. Entité orchestratrice d'apprentissage machine pour un système d'apprentissage machine

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20210176650A1 (en) * 2017-12-22 2021-06-10 Telefonaktiebolaget Lm Ericsson (Publ) Methods, Network Function Entities and Computer Readable Media for Data Collection

Also Published As

Publication number Publication date
WO2021023388A1 (fr) 2021-02-11

Similar Documents

Publication Publication Date Title
EP4011038A1 (fr) Configuration d'analytique de réseau
EP3122100B1 (fr) Système de réseau à auto-optimisation (son - self-optimizing network) pour réseaux mobiles
US20210250814A1 (en) Device and method for managing a service related to multiple network slices
WO2007103449A2 (fr) Système et procédé de génération d'un registre comptable unifié pour une session de communication
US12022576B2 (en) Cloud-based interworking gateway service
US7093010B2 (en) Operator-defined consistency checking in a network management system
JP2013168183A (ja) 権限管理に基づく端末装置管理のための方法
US20220272510A1 (en) Method and apparatus for use in communication networks having control and management planes
US7254387B2 (en) Management and control of telecommunication services delivery
CN112567684B (zh) 支持核心网络中订阅事件的方法、设备和计算机可读介质
WO2021095655A1 (fr) Système et procédé pour permettre les politiques et la facturation d'un équipement d'utilisateur présentant une ou plusieurs identités d'utilisateur
Lin Artificial intelligence in 3gpp 5g-advanced: A survey
EP4080924B1 (fr) Injection d'analytique dans une fonction de référentiel réseau (nrf) pour la gestion automatisée de la 5g
US11838311B2 (en) Systems and methods for automated quantitative risk and threat calculation and remediation
Charalambides et al. Policy conflict analysis for diffserv quality of service management
Garcia-Martin et al. Network automation and data analytics in 3gpp 5g systems
EP4309348A1 (fr) Procédés, systèmes et supports lisibles par ordinateur pour la génération de scénario d'essai de réseau autonome
Mwanje et al. Intent-driven network and service management: Definitions, modeling and implementation
US20240112087A1 (en) Ai/ml operation in single and multi-vendor scenarios
US20230409734A1 (en) Systems and methods for secure aggregating and reporting of monitored data
US12047851B2 (en) Communications network architecture
Zhou et al. A method of 5G core network service fault diagnosis
Donertasli et al. Disaggregated Near-RT RIC Control Plane with Unified 5G DB for NS, MEC and NWDAF Integration
Suárez et al. Defining a communication service management function for 5G network slices
EP2552052A1 (fr) Procédé de gestion de réseau et système de gestion de réseau

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20220309

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

DAV Request for validation of the european patent (deleted)
DAX Request for extension of the european patent (deleted)
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: EXAMINATION IS IN PROGRESS

17Q First examination report despatched

Effective date: 20230210