Feedback in alternative QOS formats to enable application statistical satisfaction calculations

Abstract

The present invention relates to a method for providing feedback information to an application, the method being implemented by at least one network function of a mobile communication network, the method comprising: - obtaining an applicable request indicating a reference quality of service profile and at least one alternative quality of service profile; providing feedback information regarding the probability of variation of the quality of service delivered by the mobile communication network for the application function over a period of time between a plurality of levels corresponding to the quality of service profile indicated in the applicable request; Includes. The invention further relates to a corresponding device and to a corresponding computer program.

Description

[Technical Field]

[0001] The present disclosure relates to the field of telecommunications. Priority is claimed to European Patent Application No. 23305582.1, filed April 17, 2023, the contents of which are incorporated herein by reference.

[0002] More particularly, the present disclosure relates to a method, a corresponding device, and a corresponding computer program for providing feedback to application functionality. [Background technology]

[0003] The industrial deployment of applications across private mobile communication networks is a key topic to ensure the success of Industry 4.0.

[0004] 3GPP® TS 23.502 details the state-of-the-art workflow for this deployment in 5G communication networks.

[0005] This state-of-the-art workflow allows 5GCN to dynamically adapt QoS according to UE mobility and radio condition variations by reducing the QoS of a data flow when the 5GCN cannot provide sufficient resources to meet the target QoS of the associated application.

[0006] There are two fundamental issues with this state-of-the-art workflow involving dynamic QoS adaptation.

[0007] Applications must dynamically react to the current QoS response, which does not prevent outages, i.e., when 5GCN responds with the lowest AQP for an extended period of time exceeding the lifetime of the application. This situation leads to outages for non-elastic applications. Furthermore, the long-term QoE of the application is not optimized by this reactive approach.

[0008] From a 5GCN perspective, there is no a priori knowledge of an application's ability to be robust to QoS fluctuations, which can result in either providing too many resources when available, or in the blind denial of QoS requests.

[0009] Therefore, there is a need for a signaling mechanism that allows for the integration of applications in industrial deployments over private mobile communication networks and overcomes the above-mentioned drawbacks. [Prior art documents] [Non-patent literature]

[0010] [Non-Patent Document 1] 3GPP® TS 23.502-Procedures for the 5G System (5GS), Version 18.0.0 and earlier.

[0011] [Non-patent document 2] 5GAA TR 23-700-81 - Study of Enablers for Network Automation for the 5G System (5GS); Phase 3, Version 18.0.0 and earlier. Summary of the Invention [Problem to be solved by the invention]

[0012] This disclosure improves the situation. [Means for solving the problem]

[0013] A method is proposed for providing feedback information to an application using a deployment of a mobile communication network to achieve quality of service for the application, the method being implemented by at least one network function of the mobile communication network, the method comprising: obtaining an applicable request indicating a plurality of quality of service profiles; providing feedback information on the probability of variation in the quality of service expected to be achieved by the mobile communications network for the application function over a period of time between a plurality of levels corresponding to the quality of service profile indicated in the applicable request; Includes. The main advantage of the proposed method is that it provides flexibility for application integration and adaptation, as well as customization of 5GCN.

[0014] When the mobile communication network is a 5G communication network, the method may be specifically implemented by a Network Data Analysis Function (NWDAF) and / or a Network Publishing Function (NEF).

[0015] The multiple quality of service profiles indicated in the applicable request may include, for example, a reference QoS profile as the nominal requirement of the application and one or more alternative QoS profiles as degraded modes. Each of the multiple quality of service profiles indicated in the applicable request may be defined by one or more values ​​or ranges of QoS parameters, including reliability parameters such as speed / throughput, delay, network availability, block error rate, jitter, bandwidth, etc.

[0016] The expected variation in QoS achieved over time between levels is a discrete representation of the expected variation in the value of the QoS parameter over time within a period of time, in levels.

[0017] Optionally, the period is the network's operational time for providing a service to the application, or at least a portion of that operational time. The network's period and / or operational time may be set, for example, to a default or to a predetermined value, such as the lifetime of the application. The network's period and / or operational time may be estimated, for example, before, after, or together with estimating the statistical distribution of the QoS profile specified in the request.

[0018] The feedback information provided to the application function can be interpreted by the application function as a prediction of the statistical distribution of the QoS profile specified in the request, which prediction may be performed by and / or made available to the entity implementing the proposed method.

[0019] Optionally, the feedback information includes an expected time spent at each level of the plurality of levels. In one example, the expected time spent at each level of the plurality of levels may be relative as a percentage of the period. For example, the feedback information may include an indication that the QoS expected to be delivered for the application function should correspond to a reference QoS profile for at least x% of the time period and correspond to a particular alternative QoS profile for up to y% of the period. In one example, the expected time spent at each level of the plurality of levels may be absolute as a duration. For example, the feedback information may include, for each QoS profile, a minimum and / or maximum expected duration of a time interval during which the QoS expected to be delivered for the application function satisfies such QoS profile. Optionally, the feedback information includes a maximum expected duration of a time interval during which the quality of service expected to be achieved by the mobile communication network for the application corresponds to a given level of the plurality of levels. This maximum expected duration may be an absolute duration (expressed in units of time) or a relative duration (expressed as a percentage of the period). Optionally, the feedback information includes a percentage of time during which the quality of service expected to be achieved by the mobile communication network for the application corresponds to a given level of the plurality of levels.

[0020] Optionally, the method may further include, after obtaining the applicable request, estimating the feedback information based on measured quality of service parameters of communications in the mobile communication network classified according to the quality of service profiles indicated in the applicable request. Optionally, the method may further include obtaining a plurality of levels by classifying the measured quality of service parameters with respect to a plurality of quality of service profiles indicated in the applicable request. As an example, the measured quality of service parameters may be organized into time-stamped sets. The time-stamped sets may be classified to obtain several groups or clusters. Each such cluster may be assigned a level. Some or all of these levels may be associated with a respective one of the quality of service profiles indicated in the applicable request. It is then a simple matter to count the number of subsequent time-stamped sets classified as belonging to the same level, for example, to determine the duration of a time interval during which the quality of service expected to be achieved by the mobile communication network for the application over that period corresponds to that level.

[0021] Optionally, the method may further include obtaining from the application function an additional applicable message indicating an expected long-term satisfaction associated with the feedback information provided to the application function.

[0022] Optionally, the method may further comprise allocating resources of the mobile communication network based on the expected long-term satisfaction indicated in the applicable message.

[0023] Optionally, the method may further comprise providing new feedback information to the application function related to a new probability of variation in the quality of service output by the mobile communications network for the application function over a period of at least one of a plurality of levels.

[0024] Optionally, the method may further comprise inferring new feedback information based on the expected long-term satisfaction indicated in the applicable message and new measured quality of service parameters of communications in the mobile communications network classified according to the quality of service profile indicated in the applicable request.

[0025] 1. A device configured to implement at least part of the core functionality of a mobile communications network, the device comprising: obtaining an applicable request indicating a plurality of quality of service profiles; providing feedback information to an application that is using a deployment of a mobile communications network to achieve a quality of service for the application, the feedback information relating to a probability of variation in the quality of service expected to be achieved by the mobile communications network for the application over a period of time between a plurality of levels corresponding to a quality of service profile indicated in the applicable request; The present invention further proposes a device further configured to perform the following:

[0026] It is further proposed a computer program comprising instructions that, when the program is executed by a computer, cause the computer to carry out the above method.

[0027] Unless otherwise indicated, and as will be apparent from the description that follows, use of terms such as "computing," "calculating," "generating," and the like throughout this specification will be understood to refer to the acts and / or processes of a computer or computing system, or similar electronic computing device, that manipulate and / or transform data represented as physical quantities, such as electronic quantities, in the registers and / or memory of the computing system into other data that is similarly represented as physical quantities in the memory, registers, or other such information storage, transmission, or display device of the computing system.

[0028] Embodiments of the present invention may include an apparatus for performing the operations herein. This apparatus may be specially constructed for the desired purposes, or it may comprise a general-purpose computer or a digital signal processor ("DSP") selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored on a computer-readable storage medium such as any type of disk, including, but not limited to, a floppy disk, an optical disk, a CD-ROM, a magneto-optical disk, a read-only memory (ROM), a random access memory (RAM), an electrically programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic or optical card, or any other type of medium suitable for storing electronic instructions and which can be coupled to a computer system bus.

[0029] The processes presented herein are not inherently related to any particular computer or other apparatus. Various general-purpose systems may be used with programs in accordance with the teachings herein, or it may prove convenient to construct specialized apparatus to perform the desired method. The desired structure for a variety of these systems will be apparent from the description below. Further, embodiments of the present invention are not described with reference to any particular programming language. It will be understood that a variety of programming languages ​​can be used to implement the teachings of the present invention as described herein. Other features, details and advantages are set forth in the following detailed description and drawings. [Brief explanation of the drawings]

[0030] [Figure 1]

[0013] Figure 1 illustrates a state-of-the-art workflow for integrating applications in industrial deployments over private mobile communication networks. [Figure 2] FIG. 1 illustrates a basic workflow of a procedure for integrating applications in an industrial deployment across a private mobile communication network, according to one embodiment of the present disclosure. [Figure 3] FIG. 1 illustrates a general setup for a processing procedure implemented by a network function of a mobile communication network to derive statistical information representative of the expected variation in the quality of service output by the mobile communication network for an application function over a period of time, according to one embodiment of the present disclosure. [Figure 4] FIG. 2 illustrates a workflow of a signaling procedure between an application function and a network function of a mobile communication network according to one embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0031] Referring now to Figure 1, there is shown the basic signaling procedures corresponding to a state-of-the-art workflow for industrial deployment of applications over private 5G mobile communication networks. Further details regarding this state-of-the-art workflow can be found in section 4.15.6.6 of TS 23.502.

[0032] A 5G mobile communication network or 5G system includes a 5G core network (5GCN) and a 5G New Radio Access Network. User equipment (UE) may connect to the 5GCN via the 5G New Radio Access Network and further to a data network (DN) such as the Internet.

[0033] A key component of 5GCN is network functions separated by providing services to manage network behavior. These network functions are: -Access and Mobility Management Function (AMF) that acts as a single entry point for UE connectivity; -A respective Session Management Function (SMF) selected by the AMF to manage a user session for a service requested by the UE; A User Plane Function (UPF) that transports IP data traffic (user plane) between the user equipment (UE) and external networks; and - Policy Control Function (PCF) that provides a policy control framework; -Unified Data Management (UDM) function for managing network user data; - Application Function (AF) that provides application services; A Network Data Analysis Function (NWDAF) that collects data from other 5G network functions, processes this data, and outputs analyses, such as statistical analyses or machine learning predictions, in order to support other 5G network functions; A Network Exposure Function (NEF) that exposes 3GPP core network capabilities to third parties; - etc. Includes.

[0034] The 5G network stack includes an open wireless architecture layer, a network layer, an open transport layer, and an application layer (APP), which can be mapped to the presentation and application layers of the OSI stack.

[0035] This basic signaling procedure concerns communication between the application layer (APP) and the 5G core network (5GCN) and can be summarized as follows:

[0036] The application layer (APP) requests a 5G core network (5GCN) to establish a communication session between the core network and a user equipment (UE) with a specific application quality of service (QoS).

[0037] This request (102) can be, for example: - User equipment address (UE address), e.g., UE IP address; - Application Feature Identifier (AF Identifier), -flow description, - external application identifiers, - Reference QoS, one or more QoS parameters, - Alternative Service Requirements (AQP), -Data Network Name (DNN), such as Access Point Name (APN), - etc. The applicable request contains various information that may include:

[0038] The 5GCN responds to the request with a QoS Response message (104) that includes an authorization response (ACK / NACK) for the session and further includes the QoS implemented in the network, which can be the Reference QoS or one of several possible alternative QoS profiles.

[0039] This basic signaling procedure enables 5GCN to perform AQP dynamic adaptation of QoS by managing available resources according to UE mobility and changing radio conditions. However, this reactive approach does not specifically optimize the long-term quality of experience of applications, nor the resource management by 5GCN.

[0040] Certain embodiments of the present disclosure now present improved signaling procedures through application-level messaging that provide a proactive solution to this lack of optimization.

[0041] According to this improved signaling procedure, the 5GCN transmits additional information about the probability of QoS adaptation changes along with the QoS response message, including expected multi-level QoS information that can be calculated by the 5GCN from RAN level parameters, for example.

[0042] This expected multi-level QoS information: Helps apps proactively adapt to changing network conditions and minimize outages for inelastic applications; further enabling the APP to send in the QoS request additional satisfaction information representing a long-term expected (statistical) QoE obtained from statistics of the expected QoS sent by the APP; It further helps 5GCN provision the right amount of resources according to satisfaction feedback from applications.

[0043] Figure 2 illustrates the basic workflow for integrating an application into a 5G deployment according to this particular embodiment.

[0044] This basic workflow presents exemplary requests and message transmissions between an APP and a 5GCN.

[0045] The APP sends a QoS request (202) to the 5GCN, which includes a list of QoS profiles. The list of QoS profiles may include a reference QoS profile (refQoS) corresponding to a preferred QoS mode. The list of QoS profiles may also include one or more alternative QoS profiles (AQPs), each corresponding to a degraded QoS mode.

[0046] This network function then feeds back 204 the APP statistics related to future QoS output by the network for the application. This feedback may be provided periodically or through notifications.

[0047] The statistical information relates to the QoS profile included in the QoS request (202).

[0048] The statistical information transmitted in the QoS response message differs from the IQN scheme defined in 5GAA TR 23-700-81, since the Advance Quality of Service Notification (IQN) is used to transmit relative network coverage, data rate, or delay degradation, all of which are not related to the QoS profile requested by the application. Furthermore, the current state of the art does not provide any possibility for 5GCN to transmit QoS statistics to applications.

[0049] In an embodiment, the statistical information may include multiple sets of parameters, one set of parameters associated with a corresponding one of the QoS profile indicated in the QoS request (202).

[0050] In embodiments, the statistical information may include sets of parameters associated with multiple or all of the QoS profiles indicated in the QoS request. For example, the statistical information may include a first set of parameters associated with a reference QoS profile indicated in the QoS request and a second set of parameters associated with multiple alternative QoS profiles indicated in the QoS request.

[0051] Any intermediate action that may be provided between the reception of the QoS request (202) and the feedback of the statistical information (204) does not affect the signaling procedure between the APP and the network function.

[0052] Such intermediate effects include, but are not limited to: - processing QoS requests; - Forwarding the QoS request to another entity in the 5GCN responsible for processing the QoS request; - executing or triggering an inference procedure for inferring statistical information (204); - retrieving statistical information (204), especially if the statistical information has been predetermined; - etc. may include:

[0053] Additional implementation details regarding the inference procedure and the nature of the statistical information are provided in further sections of this document.

[0054] The 5GCN may further open a data flow corresponding to the session associated with the QoS request. The 5GCN may further allocate resources, such as radio resources or virtual resources, according to a QoS profile adaptation mechanism responsive to statistical information (204) feedback to the AF.

[0055] Optionally, the feedback from the 5GCN to the APP can be combined with subsequent feedback from the APP to the 5GCN to form a feedback sequence.

[0056] For example, statistical information (204) feedback to an APP may enable the determination of information (206) related to the expected long-term satisfaction with the application, may be considered readily available to the application, and may be provided by the APP to the network function of the 5GCN as feedback related to the application's behavior when presented along with statistical information (204) previously provided by that network function of the 5GCN.

[0057] The feedback sequence may be repeated several times.

[0058] In particular, when the 5GCN obtains information (206) related to the expected long-term satisfaction of the application, the 5GCN may perform a novel process to re-estimate the probability of the future QoS output by the network, based not only on the QoS requirements but also on the long-term satisfaction provided as feedback by the AF. These successive iterations allow finding a QoS profile adaptation mechanism that is optimized in terms of maximizing the long-term satisfaction of the application.

[0059] The 5GCN may further open a data flow corresponding to the session associated with the QoS request and may allocate resources according to the optimized QoS profile adaptation mechanism.

[0060] Here, further implementation details are provided for the possible use of statistical information (204) on the APP side. The statistical information relates to QoS profile conformance during a specific observation time, also called "network state information".

[0061] The application APP - Estimate application layer related satisfaction metrics such as Quality of Experience (QoE) distribution, also known as "long-term quality-of-experience"; Estimating application layer outage related metrics such as application lifetime, which is defined as the average time that the system QoS can remain in a degraded QoS mode, i.e., the minimum AQP of the application layer QoS profile; Estimate application layer resource allocation metrics, which are assumed to allow the application layer control plane, i.e., AF, to adapt application layer packet rates to address QoS degradation in the network. Statistical information may be used to

[0062] The statistical information may further be used to derive rate adaptation actions at the application layer.

[0063] Depending on the application type, i.e., critically inelastic or non-critically resilient, various application layer actions defined above may be crucial. For example, a critically inelastic application may use the AQP distribution to estimate the probability of an outage event, i.e., that the AQP degraded mode will be maintained for a sufficiently long time. APP outage is defined, for example, when time is greater than the application's lifetime. For non-critically inelastic applications, application layer rate control can be calculated to maximize the overall rate of network state information. One solution in this case may be to adapt the application rate to AQP statistics, for example, through some form of Transmission Control Protocol (TCP).

[0064] When the statistical information 204 includes information denoted f_refQoS related to a reference QoS profile specified in the QoS request 202 and information denoted f_AQP related to one or more alternative QoS profiles specified in the QoS request, the application layer APP may calculate a long-term satisfaction parameter or index for the application denoted IDX based on f_refQoS and f_AQP. The long-term satisfaction parameter may be related, for example, to the mean opinion score (MoS) of the application client or to a specific application layer-related metric such as application layer throughput, application E2E delay, application resilience, application lifetime, etc. The satisfaction metric may be a function of the various application layer metrics mentioned above. This long-term satisfaction index may then be used by the 5GCN to reorganize resources within the 5G Radio Access Network (RAN) and / or transport network resources to maximize application satisfaction.

[0065] Further implementation details applicable to embodiments in which the statistical information (204) is periodically reported by the 5GCN to the APP are now provided. These additional implementation details are shown in Figure 3, which describes a general setup of the processing that may be performed by the 5GCN to derive the statistical information.

[0066] The 5GCN receives an initial QoS request from the APP and sets up a PDU session with QoS parameters selected as corresponding to one of the QoS profiles specified in the QoS request. For example, the PDU session may be set up with QoS parameters closest to the reference QoS profile or closest to an alternative QoS profile, when applicable.

[0067] The GCN stores the QoS profiles specified in the initial QoS request. For each stored QoS profile, the GCN defines a corresponding class, with the classes ranked from the highest QoS, e.g., corresponding to a reference QoS profile indicating an optimal QoS mode for the APP, to the lowest QoS, e.g., corresponding to an alternative QoS profile indicating a most degraded QoS mode for the APP.

[0068] The 5GCN further measures the QoS parameters of an ongoing PDU Session periodically, i.e., every T second period during which the following is performed: Optionally, the 5GCN may periodically measure the QoS parameters of multiple ongoing PDU sessions of the same application.

[0069] A QoS classifier (302) may be provided to enable the 5GCN to perform classification of measurements into stored QoS profiles.

[0070] Specifically, the QoS classifier may be configured to enable the 5GCN to perform deterministic classification of measurements against stored QoS profiles. For example, the 5GCN may count occurrences of events classified as members of a reference QoS profile or as members of an alternative QoS profile. f_refQoS is taken as the total number of classes corresponding to the reference QoS, and f_AQP as the total number of classes corresponding to the application's alternative QoS profiles.

[0071] Alternatively, the QoS classifier may be configured to enable the 5GCN to perform probabilistic classification of measurements against the APP's stored QoS profiles. The 5GCN may, for example, use a Restricted Boltzmann Machine (RBM) to extract features from the network QoS measurements that are used to calculate a probability distribution of the occurrence of an event being classified as a member of the reference QoS profile or an alternative QoS profile. f_refQoS may be obtained as the distribution of features corresponding to the reference QoS profile, and f_AQP may be obtained as the distribution of features corresponding to the alternative QoS profile.

[0072] In addition to the QoS classifier, a model fitter (304) may be further provided so that the 5GCN can perform model fitting.

[0073] For example, the model fitter may be configured to enable the 5GCN to perform Markov model fitting by evaluating the average number of transition occurrences between the reference QoS profile and the alternative QoS class over multiple measurement periods (NT), where N is an integer. In other words, the 5GCN may count the number of events in which the output of the classification changes from the reference QoS profile to the alternative QoS profile. The 5GCN may also evaluate the duration that the output classification remains in the same class. These counts define a Markov transition matrix that can be used to calculate stationary probabilities of the reference and alternative QoS classes.

[0074] Alternatively, or in combination with Markov model fitting, the model fitter may be configured to enable the 5GCN to perform general Bayesian filtering, which selects from the statistical output of the classifier to reconstruct a multivariate distribution of reference and alternative QoS classes.

[0075] Here we provide examples of statistical information (204) that may be transmitted by 5GCN to the application layer.

[0076] The above classification and processing steps allow determining the average duration during the observation period T that the 5G system serves the APP with the reference QoS profile or with an alternative QoS profile. Then, as part of the statistical information, f_refQoS may include the average duration that the 5G system serves the APP with the reference QoS profile, while f_AQP may include, for one or more alternative QoS profiles, the average duration that the 5G system serves the APP with the alternative QoS profile.

[0077] The statistical information may include a probability distribution of the measured QoS provided by the 5G system during the time period T, represented by the combined distribution of the reference and alternative QoS profiles and by the probability not associated with a particular APP request. This information may be obtained through probabilistic classification of the measurements as described above.

[0078] The statistical information may include the probability that the reference QoS profile is not met during time period T, i.e., the probability that only one or more alternative QoS profiles are achieved during time period T. This information can be obtained by model fitting as described above.

[0079] The statistical information may include the probability that neither the reference nor the alternative QoS profile is satisfied during the period T, i.e., the probability that the network cannot realize any of the QoS profiles specified in the request from the APP. This information represents the outage probability of the APP in the current deployment of the APP in 5GCN.

[0080] The statistical information may include a cumulative distribution of the measured QoS provided by the 5G system during the period T, i.e., the probability that the QoS class obtained from the measurements is below a particular QoS class.

[0081] The statistics are as follows: the quantile of the cumulative distribution at which the reference QoS profile is not met during the period T, and / or the quantiles of the cumulative distribution at which the reference and alternative QoS profiles are not met during the period T, and / or a quantile of a probability distribution representing the proportion of UEs that experience a QoS that does not meet the reference QoS profile during the period T; and / or - quantile of the probability distribution representing the fraction of UEs that experience QoS that neither the reference QoS profile nor the alternative QoS profile meets during time period T It may also include quantiles such as

[0082] The above quantiles are useful for evaluating the performance of an APP during a period T, since the probability of achieving the reference QoS represents the average optimal behavior of the APP. The probability of not meeting the reference QoS represents the average period the APP is in a degraded mode, which may be important for inelastic applications. The quantiles of the probability distribution or cumulative distribution are related to the QoS performance of some specific UEs in the deployment, and can therefore represent the efficiency of the APP deployment.

[0083] The APP may use the above quantiles to predict APP tier outages during time period T. For example, critical applications may inherently have low tolerance for long periods of being served by 5GCN in a degraded mode.

[0084] Here we provide an example of information (206) related to expected long-term satisfaction with the application.

[0085] This information (206) is sent as feedback from the application layer APP in response to statistical information (204) previously sent by the 5GCN, such as f_refQoS and f_AQP.

[0086] This feedback describes the overall long-term behavior or application reaction of the application with respect to the statistical distribution of the requested QoS parameters predicted by the 5GCN, in other words, the feedback describes the overall long-term behavior or application reaction of the application to the expected QoS variations between the QoS profiles specified in the QoS request (202) according to the statistical information (204).

[0087] Information (206) is -Long-term APP Satisfaction Index (IDX), and / or - the maximum quality of experience (QoE) achieved in the APP by extrapolating the QoE distribution over any period of time, and / or -In an APP, the average QoE achieved by extrapolating the QoE distribution over any period of time, etc. The QoE statistics may include various types of QoE statistics, such as one or more of:

[0088] The index, maximum QoE, and average QoE may be expressed in the form of application classes, i.e., as a result of classification by the APP of QoE statistics associated with QoS statistics provided by the 5GCN in response to a QoS request.

[0089] The long-term APP satisfaction index, maximum QoE, and average QoE are examples of information (206) that can be used by the 5GCN to simplify the calculation of the QoS distribution in the case of multiple applications and to filter the transmission of the QoS distribution to the APPs.

[0090] The information sent as feedback from the application layer APP (206) may be used by the 5GCN as input for a QoE-based resource reorganization step.

[0091] In embodiments where statistical information (204) is periodically reported by the 5GCN to the APP, this statistical information takes into account such resource reorganization, thereby enabling joint optimization of the long-term QoE for the APP and resource allocation by the 5GCN over several iterations.

[0092] In the above description, the workflow for integrating an application into a 5G deployment is shown as a series of transmissions of data, such as request or feedback messages, from the APP to the network functions of the 5GCN and vice versa.

[0093] Figure 4 shows a possible implementation of the above workflow within a 5GCN as a series of communications using the N33 reference point between the AF handling the application and the NWDAF or NEF as the "5GCN Network Function" mentioned above.

[0094] The AF sends a QoS request (202) to the NWDAF or NEF (402).

[0095] The NWDAF or NEF then feeds back (404) the statistical information (204) to the AF.

[0096] Optionally, the AF then feeds back (406) information related to the expected long-term satisfaction (206) of the application.

[0097] Optionally, steps 404 to 406 may be repeated periodically or upon repetition of step 402 .

Claims

1. 1. A method for providing feedback information to an application using a deployment of a mobile communication network to achieve quality of service for the application, the method being implemented by at least one network function of the mobile communication network, the method comprising: - obtaining applicable requests indicating multiple quality of service profiles; providing feedback information regarding the probability of variation in the quality of service expected to be achieved by the mobile communication network for said application over a period of time between a plurality of levels corresponding to the quality of service profile indicated in said applicable request; A method comprising:

2. 2. The method of claim 1, wherein the feedback information comprises a maximum expected duration of a time interval during which the quality of service expected to be achieved by the mobile communication network for the application over the period corresponds to a given level of the plurality of levels.

3. 3. The method of claim 1, wherein the feedback information comprises a percentage of time over the period of time that the quality of service expected to be achieved by the mobile communication network for the application is expected to correspond to a given level of the plurality of levels.

4. The method of claim 1 , wherein the period of time is at least a portion of an operating time of the network for providing services to the application.

5. 5. The method according to claim 1, further comprising the step of, after obtaining the applicable requirements, estimating the feedback information based on measured quality of service parameters of communications in the mobile communication network.

6. The method of claim 5 , wherein estimating the feedback information comprises inferring the feedback information based on the plurality of levels.

7. 7. The method of claim 1, further comprising the step of obtaining the plurality of levels by classifying the measured quality of service parameters with respect to a plurality of quality of service profiles indicated in the applicable request.

8. 8. The method of claim 1, further comprising obtaining from the application function an additional applicable message indicating an expected long-term satisfaction associated with the feedback information provided to the application function.

9. 9. The method of claim 8, further comprising allocating resources of the mobile communications network based on the expected long-term satisfaction indicated in the applicable message.

10. 10. The method of claim 8 or 9, further comprising the step of providing new feedback information to the application function, the new feedback information relating to a new probability of variation in the quality of service output by the mobile communications network for the application function over the period of time corresponding to at least one of the plurality of levels.

11. 11. The method of claim 10, further comprising the step of inferring the new feedback information based on the expected long-term satisfaction indicated in the applicable message and new measured quality of service parameters of communications in the mobile communication network classified according to the quality of service profile indicated in the applicable request.

12. 1. A device configured to implement at least part of the core functionality of a mobile communication network, said device comprising: - obtaining applicable requests indicating multiple quality of service profiles; - providing feedback information to the application using the deployment of the mobile communication network to achieve a quality of service for the application, the feedback information relating to a probability of variation in the quality of service expected to be achieved by the mobile communication network for the application over a period of time between a plurality of levels corresponding to a quality of service profile indicated in the applicable request; and further configured to perform device.

13. A computer program comprising instructions that, when said program is executed by a computer, cause said computer to carry out the method of any one of claims 1 to 11.

Images