WO2023036440A1

WO2023036440A1 - Techniques for handling qos prediction parameters

Info

Publication number: WO2023036440A1
Application number: PCT/EP2021/075052
Authority: WO
Inventors: Mate BOBAN
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2021-09-13
Filing date: 2021-09-13
Publication date: 2023-03-16
Also published as: EP4344472A1

Abstract

The present disclosure relates to techniques for handling QoS prediction parameters, in particular a network device for determining QoS prediction parameters in a mobile communication network. The network device comprises a communication interface configured to send a request message towards a plurality of UEs, the request message requesting transmissions of measurement data from the plurality of UEs. The communication interface is configured to receive the measurement data. The network device comprises a processor configured to determine a QoS model of the mobile communication network based on the received measurement data. The processor is configured to determine QoS prediction parameters for at least one UE based on the QoS model, and to transmit a prediction message comprising the QoS prediction parameters to the UE. The processor is configured to generate the request message upon the basis of the QoS model.

Description

Techniques for handling QoS prediction parameters

TECHNICAL FIELD

The present disclosure relates to techniques for handling QoS (Quality of Service) prediction parameters, in particular a network device for determining QoS prediction parameters and a user equipment (UE) for receiving QoS prediction parameters and corresponding methods. The disclosure particularly relates to method for triggering data collection for generating predictions and for triggering the QoS prediction updates.

BACKGROUND

In the field of mobile networks, prediction of network performance behavior such as Quality of Service (QoS) prediction ahead of time is essential to improve the network performance. Predictive QoS frameworks are discussed for example in the White Paper: 5GAA, “Making 5G Proactive and Predictive for the Automotive Industry,” 2019 or by ITU-R Focus group on Machine Learning for Future Networks including 5G (FG-ML5G): “Unified architecture for machine learning in 5G and future networks”. However, these frameworks focus on data acquisition only without giving solutions on how to provide prediction updates, when to provide prediction updates or when to provide/acquire training data. Furthermore, in predicting which data to acquire, there is no assumption on the knowledge of any UE parameters beforehand (e.g., location, trajectory, SINR, or any other parameter).

SUMMARY

It is an objective of this disclosure to provide a concept for improving network performance by providing adequate QoS prediction data.

A particular objective of this disclosure is to provide a concept for efficiently providing and/or acquiring training data for predicting QoS and for efficiently providing QoS prediction updates to those users and other entities requiring the QoS prediction.

One or more of these objectives are achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the figures.

A basic idea of this disclosure is to introduce a novel method for a) triggering QoS prediction model updates from prediction service to users, which takes into account the prediction accuracy and latency requirements set forth by the users; and b) triggering measurement/training data updates from users to prediction service, which takes into account the accuracy of collected data, temporal properties of the data, density of users, and QoS evaluation metric and which can be either deterministic or stochastic. Additionally, the disclosure defines required messages and parameters for the method.

According to a first aspect, the disclosure relates to a network device for determining one or more of Quality-of-Service, QoS, prediction parameters in a mobile communication network, the network device comprising: a communication interface configured to send a request message towards a plurality of User Equipments, UEs, the request message requesting transmissions of measurement data from the plurality of User Equipments, UEs, and wherein the communication interface is configured to receive the measurement data; and a processor configured to determine a QoS model of the mobile communication network based on the received measurement data; wherein the processor is configured to determine one or more QoS prediction parameters for at least one UE of the plurality of UEs based on the QoS model, and to transmit a prediction message comprising the one or more QoS prediction parameters to the at least one UE, and wherein the processor is configured to generate the request message upon the basis of the QoS model.

The measurement data may be related to communications-related information in the mobile communication network, e.g., throughput, latency, jitter, etc., or to other non-communication related information, e.g. UE location information, etc.

Such a network device can advantageously implement a QoS prediction service (QPS). The network device can advantageously trigger QoS prediction model updates from QPS to UEs, taking into account the prediction accuracy and latency requirements set forth by the users and received from the users. The network device can advantageously trigger measurement/training data updates from users to QPS, which takes into account the accuracy of collected data, temporal properties of the data, density of users, and QoS evaluation metric and which can be either deterministic or stochastic. The network device is configured to provide all the required messages and parameters.

Alternatively, the QPS can be implemented on another network element and the network device can forward all messages to the QPS of the other network element. In an exemplary implementation of the network device, the request message comprises one or more parameters indicative of rules for the requesting of the transmissions of the measurement data.

This provides the advantage that the network device or the prediction service (QPS) is efficiently informed about the rules for the requesting of the transmissions of the measurement data.

In an exemplary implementation of the network device, the communication interface is configured to receive one or more parameters indicative of requirements of at least one of the plurality of UEs for determining the one or more QoS prediction parameters; and the processor is configured to determine the one or more QoS prediction parameters for the at least one UE based on the corresponding requirements of the at least one UE.

This provides the advantage that the network device or the prediction service (QPS) is efficiently informed about the requirements of the UEs for determining the QoS prediction parameters and can consider these requirements when determining the QoS prediction parameters.

In an exemplary implementation of the network device, the processor is configured to transmit the prediction message to at least one of the following: a specific UE, a specific group of UEs, all UEs within a certain geographic area.

This provides the advantage that the prediction message is transmitted only to those UEs that may need the prediction message, thereby saving communication resources.

In an exemplary implementation of the network device, the rules for the requesting of the transmissions of the measurement data are based on at least one of the following: a confidence of prediction, a latency requirement of the at least one UE for receiving the prediction message, a periodicity requirement of the at least one UE for receiving the prediction message, a frequency of updating the QoS model, a type of the QoS model, a type of the one or more QoS prediction parameters, a periodicity requirement of the network device for transmitting the prediction message.

This provides the advantage that the measurements can be focused on specific needs of the UEs in order to provide the correct measurement data. The rules for the requesting of the transmissions are not limited to the above-described items or parameters. The rules may also comprise further parameters not indicated above or derivatives of the above parameters.

In an exemplary implementation of the network device, the rules for the requesting of the transmissions of the measurement data are deterministic or probabilistic.

In deterministic case, if a set of prescribed rules are satisfied, UE will transmit measurement data, otherwise it will not. In probabilistic case, there is a function that describes measurement data transmission probability that may depend on the accuracy/quality of the collected data and its timeliness. Hence these rules provide flexibility for requesting the measurement data transmissions.

In an exemplary implementation of the network device, the QoS model comprises one of the following models: a table mapping, a table lookup, a Deep Neural Network, a Random Forest, a Kalman filter.

This provides the advantage that the QoS model can be flexibly selected depending on the specific requirements of the UE.

The QoS model is not limited to the above-described models. The QoS model may also comprise other models, e.g., other machine learning models or other mathematical models not listed above.

In an exemplary implementation of the network device, the one or more QoS prediction parameters comprise at least one of the following parameters of the mobile communication network: a predicted uplink, downlink, or sidelink throughput, a predicted uplink, downlink, or sidelink latency, a predicted jitter on uplink, downlink, or sidelink, a predicted service uptime or downtime, a predicted quality of experience.

This provides the advantage that the network device can provide a variety of different predictions for different parameters that may be required for the UE to improve its capabilities.

The QoS prediction parameters are not limited to the above-described parameters. A lot of further parameters can be used as well. According to a second aspect, the disclosure relates to a user equipment, UE, of a plurality of UEs of a mobile communication network, for receiving one or more Quality-of-Service, QoS, prediction parameters from a network device of the mobile communication network, the UE comprising: a communication interface configured to receive a request message from the network device of the mobile communication network, the request message requesting transmissions of measurement data; and a processor configured to transmit the measurement data via the communication interface to the network device upon reception of the request message, wherein the communication interface is configured to receive a prediction message from the network device, the prediction message comprising one or more QoS prediction parameters.

Such a User Equipment can advantageously receive QoS predictions from a QoS prediction service (QPS) implemented in the network. The UE can be advantageously triggered and updated by the network to receive QoS prediction model updates, taking into account the prediction accuracy and latency requirements of the UE. The UE can advantageously perform measurement/training data updates based on the request message from the network device, taking into account the accuracy of collected data, temporal properties of the data, density of users, and QoS evaluation metric. Triggering of measurement data update can be either deterministic or stochastic.

The one or more QoS prediction parameters may be based on a QoS model of the mobile communication network. This QoS model can be precisely determined by the network device based on measurement data of the plurality of UEs.

In an exemplary implementation of the UE, the processor is configured to execute an application using the one or more QoS prediction parameters.

This provides the advantage that the application can use the QoS prediction parameters for optimally performing the application.

In an exemplary implementation of the UE, the processor is configured to perform at least one of the following tasks based on the one or more QoS prediction parameters: estimate a future speed of the UE, select a route for the UE, select a maneuver for the UE, generate and/or update a radio map, estimate a future throughput on uplink, downlink, or sidelink, estimate a future latency on uplink, downlink, or sidelink, estimate a future jitter on uplink, downlink, or sidelink, estimate a future uptime or downtime of uplink, downlink, or sidelink, estimate a future quality of experience of uplink, downlink, or sidelink, generate and/or update a high-definition map.

This provides the advantage that these tasks can be optimally performed when prediction results from the QPS are available to improve the performance.

These tasks performed by the UE based on the QoS prediction parameters are not limiting. Many other tasks can be performed by the UE which may be advantageously performed when using the QoS prediction parameters.

In an exemplary implementation of the UE, the measurement data comprise at least one of the following: a signal to interference and noise ratio for a given location of the UE, a Reference Signal Received Power of an uplink, downlink, or sidelink, a Channel State Information of an uplink, downlink, or sidelink, an estimated location of the UE, an estimated speed of the UE, an estimated location of another object, an estimated speed of another object, a bearing of detected objects and the UE itself.

This provides the advantage that the UE can perform different measurements and provide results of these measurements to the QPS which can improve its QoS model based on these measurement data.

The measurement data are not limited to these examples. Any other measurement data may be provided by the UE as well.

In an exemplary implementation of the UE, the request message comprises one or more parameters indicative of rules for the requesting of the transmissions of the measurement data.

In an exemplary implementation of the UE, the processor is configured to transmit the measurement data to the network device based on the rules for the requesting of the transmissions of the measurement data.

This provides the advantage that custom-fit measurement data can be transmitted to the network device and processed by the QPS. In an exemplary implementation of the UE, the processor is configured to transmit one or more parameters to the network device, the one or more parameters being indicative of requirements of the UE for determining the one or more QoS prediction parameters by the network device.

This provides the advantage that the network device or the prediction service (QPS) is efficiently informed about the requirements of the UE for determining the QoS prediction parameters and can consider these requirements when determining the QoS prediction parameters.

In an exemplary implementation of the UE, the one or more parameters are indicative of requirements to periodicity and/or QoS model accuracy.

This improves the performance and accuracy of the prediction.

Requirements to periodicity specify how frequently the one or more QoS prediction parameters should be transmitted by the network device. Requirements to QoS model accuracy specify the precision or accuracy of the QoS model and/or the QoS prediction parameters as needed by the UE.

According to a third aspect, the disclosure relates to a method for determining one or more of Quality-of-Service, QoS, prediction parameters in a mobile communication network, the method comprising: sending a request message towards a plurality of User Equipments, UEs, the request message requesting transmissions of measurement data from the plurality of User Equipments, UEs; receiving the measurement data; determining a QoS model of the mobile communication network based on the received measurement data; determining one or more QoS prediction parameters for at least one UE of the plurality of UEs based on the QoS model; transmitting a prediction message comprising the one or more QoS prediction parameters to the at least one UE; and generating the request message upon the basis of the QoS model.

The method may be performed by or on a network device according to the first aspect and provides the same advantages as described above with respect to the network device of the first aspect.

According to a fourth aspect, the disclosure relates to a method for receiving one or more Quality-of-Service, QoS, prediction parameters from a network device of a mobile communication network, the method comprising: receiving a request message from the network device of the mobile communication network, the request message requesting transmissions of measurement data; transmitting the measurement data to the network device upon reception of the request message; and receiving a prediction message from the network device, the prediction message comprising one or more QoS prediction parameters.

The method may be performed by or on a User Equipment according to the second aspect and provides the same advantages as described above with respect to the UE of the second aspect.

According to a fifth aspect, the disclosure relates to a computer program product including computer executable code or computer executable instructions that, when executed, causes at least one computer to execute the method according to the third or fourth aspect.

The computer program product may run on any of the components of a communication system described below with respect to Figure 7. For example, the computer program product may run on a UE as shown in Figure 7. Such a UE may comprise a processing circuitry for instance, a processor, for processing and generating data, e.g., the program code described above, a communication interface including, for instance, a transmitter, a receiver and an antenna, for exchanging data with the other components of the communication system 700, and a non- transitory memory for storing data, e.g. the program code described above. The computer program product may also run on a network device as shown in Figure 7.

According to a sixth aspect, the disclosure relates to a computer-readable medium, storing instructions that, when executed by a computer, cause the computer to execute the method according to the third or fourth aspect. Such a computer readable medium may be a nontransient readable storage medium. The computer may be, for example, a user device, e.g., the user device according to the second aspect comprising a processor, a communication interface and a memory as shown in Figure 7 or a network device, e.g. as shown in Figure 7 comprising a processor, a communication interface and a memory as shown in Figure 7. The computer-readable medium may be stored in the memory of the user device. The instructions stored on the computer-readable medium may be executed by the processor of the user device or the network device. BRIEF DESCRIPTION OF THE DRAWINGS

Further embodiments of the invention will be described with respect to the following figures, in which:

Fig. 1 shows a schematic diagram illustrating a mobile network 100 according to the disclosure;

Fig. 2 shows a signaling flowchart 200 illustrating messages exchanged between a network device 110 according to the disclosure and a plurality of UEs 120 according to the disclosure;

Fig. 3 shows a schematic diagram illustrating a system 300 for triggering prediction updates based on prediction accuracy according to a first embodiment and a corresponding prediction accuracy diagram 301 ;

Fig. 4 shows a schematic diagram illustrating a system 400 for triggering prediction updates based on prediction accuracy and latency according to a second embodiment;

Fig. 5 shows a schematic diagram illustrating a mobile network 500 according to the disclosure for triggering measurement data updates;

Fig. 6 shows a schematic diagram illustrating an exemplary 5G network 600 including QoS prediction functionality according to the disclosure in different parts of the network 600;

Fig. 7 shows a schematic diagram illustrating a communication system 700 for triggering data collection and QoS prediction updates according to the disclosure;

Fig. 8 shows a schematic diagram illustrating a method 800 for determining one or more QoS prediction parameters in a mobile communication network; and

Fig. 9 shows a schematic diagram illustrating a method 900 for receiving one or more QoS prediction parameters from a network device of a mobile communication network. DETAILED DESCRIPTION OF EMBODIMENTS

In order to describe the invention in detail, the following terms, abbreviations and notations will be used:

QoS quality-of-service

QPS QoS prediction service

MD measurement data

QP QoS prediction

SINR signal to interference plus noise ratio

RMSE root mean squared error

HD high density

UE user equipment

LTE long term evolution

NR new radio

AMF access and mobility management function (entity)

In the following detailed description, reference is made to the accompanying drawings, which form a part thereof, and in which is shown by way of illustration specific aspects in which the disclosure may be practiced. It is understood that other aspects may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present disclosure is defined by the appended claims.

It is understood that comments made in connection with a described method may also hold true for a corresponding device or system configured to perform the method and vice versa. For example, if a specific method step is described, a corresponding device may include a unit to perform the described method step, even if such unit is not explicitly described or illustrated in the figures. Further, it is understood that the features of the various exemplary aspects described herein may be combined with each other, unless specifically noted otherwise.

The methods, devices and systems described herein may be implemented in mobile communication networks, in particular radio and/or core part of LTE, 5G, or 5G beyond. The described devices may include integrated circuits and/or passives and may be manufactured according to various technologies. For example, the circuits may be designed as logic integrated circuits, analog integrated circuits, mixed signal integrated circuits, optical circuits, memory circuits and/or integrated passives. The devices described herein may be configured to transmit and/or receive radio signals. Radio signals may be or may include radio frequency signals radiated by a radio transmitting device (or radio transmitter or sender). However, devices described herein are not limited to transmit and/or receive radio signals, also other signals designed for transmission in deterministic communication networks may be transmitted and/or received.

The devices and systems described herein may include processors or processing devices, memories and transceivers, i.e., transmitters and/or receivers. The term “processor” or “processing device” describes any device that can be utilized for processing specific tasks (or blocks or steps). A processor or processing device can be a single processor or a multi-core processor or can include a set of processors or can include means for processing. A processor or processing device can process software or firmware or applications etc.

The devices and systems described herein may include communication interfaces such as transceivers or transceiver devices. A transceiver is a device that is able to both transmit and receive information or signal through a transmission medium, e.g. a radio channel. It is a combination of a transmitter and a receiver, hence the name transceiver. Transmission is usually accomplished via radio waves. By combining a receiver and transmitter in one consolidated device, a transceiver allows for greater flexibility than what either of these could provide individually.

In the present disclosure, the user equipment (UE) may be, for example, a mobile phone, an intelligent terminal, a tablet computer (tablet), a notebook computer (laptop), a video game console, a multimedia player, vehicle, device to device (D2D) equipment, or any smart device which supports a positioning function.

Fig. 1 shows a schematic diagram illustrating a mobile network 100 according to the disclosure. The mobile network 100 shown in Figure 1 represents the system model used hereinafter. The mobile network 100 comprises one or more User Equipments (UEs) 120, e.g., cars, that are connected to a network device 110 via one or more base stations 130 or radio networks, respectively. A QoS Prediction Service (QPS) is implemented in the network device 110 which communicates via QP (QoS Prediction) 114 and MD (measurement data) 122 messages with the one or more UEs 120.

The communication can be described as follows. A) UEs 120 send measurement data (MD) 122 to QoS Prediction Service (QPS) of network device 110. QPS is an entity/function; QPS is under network control and it can reside in various locations/entities: e.g., as part of MEC, part of NWDAF in CN, Al entity/function in RAN, etc. Furthermore, it can be distributed across different network entities depending on the required functionality (e.g., the overall system parameters and rules can be stored in core network, whereas functionalities related to ensuring low latency can be placed in RAN).

B) QPS collects MDs 122 and builds a QoS Predictions (QP) based on a model, also called a QoS model hereinafter. Many types of model can be applied, e.g., table mapping/lookup, DNN, RF, Kalman filter, etc. QP can contain prediction for any number of QoS parameters. E.g., the output of the model can be predicted uplink or downlink throughput, predicted uplink or downlink latency, jitter, service uptime or downtime, quality of experience, etc.

C) QPS sends the QP 114 to UEs 120. UEs 120 can use the predictions contained in QP to, for example, enhance the performance of applications benefiting from such predictions. Some applications benefiting from QP are for example: estimating the possible UE speed in the future, selection of a route, selection of an appropriate maneuver, etc. Further application examples that benefit from predictions are: Radio Map generation, HD Map generation and updating. In terms of the examples of what might be contained in MD 122 and QP 114, MD 122 for Radio Map can contain SINR for given UE location; MD 122 for HD Map can contain estimated location, speed, bearing of the detected objects and of the UE itself. In terms of QP, for Radio Map it can contain a prediction for a part of Radio map relevant for UE or a group of UEs. For HD Map, QP can contain the predicted locations, speed, and bearing of a set of objects over a period of time in the future.

Depending on the specific embodiment, the following pieces of information, in the form of either standalone messages or collected as part of a set of already existing or new message exchanges, and related parameters may be needed. i) QP message containing the output of the model prediction (e.g., Radio Map, HD Map, etc.). The generated QP can be UE specific, UE-group specific (e.g., part of HD Map based on UE location), or applicable to all UEs within a certain area (e.g., base station coverage). ii) Furthermore, y_p the confidence of prediction (i.e., QP prediction accuracy) may be required to be defined. The range of the confidence can be defined as percentage/proportion (e.g., y_p e [0,1]). The value of y_p depends on both the type of model and the input data. As one example of the value of y_p: for a given location, the y_p for a radio map can be 95% for an SINR RMSE

< 3 dB. iii) Furthermore, T_UE is the UE latency requirement for QP. That is, if y_p is below a given target, UE may need to receive a new QP within T_UE seconds. iv) MD message contains measurement data message containing tuple {RM, LOG, y_d}, with y_d being optional. RM is defined as Radio Measurements (e.g., SINR, BER, RSRP,,...); LOG is location information (e.g., latitude, longitude, altitude) and y_d 6 [0,1] is the accuracy of the collected measurement data, defined in the same way as y_p. The reliability/accuracy of y_d is affected by, e.g., noisy measurements of radio parameters, errors in location estimates, etc. v) Furthermore, appropriate QoS evaluation metric may be needed to be defined. For example, in case of radio map prediction based on SINR, SINR Root Mean Squared Error (RMSE) can be a QoS evaluation metric. Another example is: in case of predicting throughput (uplink, downlink, or sidelink) based on parameters (e.g., SINR, BER, LOG), Mean Averaged Error (MAE) of throughput or RMSE of throughput can be the QoS evaluation metric. In case of HD Map prediction, RMSE of the location of detected objects can be the QoS evaluation metric. vi) Another useful parameter that may be defined is temporal reference discount for MD, 8 e [0,1], In particular, 8 can take the form of a function of time 8 = f(t - t'), where t is the current time and t' is the time when the measurement data tuple was collected; by definition, 8 = 1 if there is no temporal dependency. vii) QPS or any network entity involved in the prediction can at any time require further input from UEs 120 and consequently generate a request in addition to any existing rules for triggering MD 122. These additional requirements are denoted RQPS and contain any additional QPS requirements on MD, e.g., minimum or maximum periodicity of MD, additional information to be collected as part of MD, etc.

Fig. 2 shows a signaling flowchart 200 illustrating messages exchanged between a network device 110 according to the disclosure and a plurality of UEs 120 according to the disclosure. The network device 110 may correspond to the network device 110 described above with respect to Figure 1. Similarly, the UEs 120 may correspond to the UEs 120 described above with respect to Figure 1 . The network device 110 can be used for determining one or more of Quality-of-Service (QoS) prediction parameters 113 in a mobile communication network, e.g. a mobile communication network 100 as shown in Figure 1.

The network device 110 comprises a communication interface (not shown in Figure 2) configured to send a request message 112 towards a plurality of User Equipments 120, the request message 112 requesting transmissions of measurement data 122 from the plurality of User Equipments 120. The communication interface is configured to receive the measurement data 122. This measurement data 122 may correspond to the MD 122 described above with respect to Figure 1 .

The network device 110 comprises a processor (not shown in Figure 2) configured to determine a QoS model of the mobile communication network 100 based on the received measurement data 122. The processor is configured to determine one or more QoS prediction parameters 113 for at least one UE 120a of the plurality of UEs 120 based on the QoS model, and to transmit a prediction message 114 comprising the one or more QoS prediction parameters 113 to the at least one UE 120a. This prediction message 114 may correspond to the QP message 114 described above with respect to Figure 1. The processor is configured to generate the request message 112 upon the basis of the QoS model.

The measurement data 122 may be related to communications in the mobile communication network 100 or to other non-communication related information, e.g., UE location information, etc.

The request message 112 may comprise one or more parameters indicative of rules 111 for the requesting of the transmissions of the measurement data 122.

The communication interface may be configured to receive one or more parameters 124 indicative of requirements 123 of at least one of the plurality of UEs 120 for determining the one or more QoS prediction parameters 113. The processor may be configured to determine the one or more QoS prediction parameters 113 for the at least one UE 120a based on the corresponding requirements 123 of the at least one UE 120a.

The processor may be configured to transmit the prediction message 114 to at least one of the following: a specific UE 120a, a specific group of UEs 120a, 120b, all UEs 120a, 120b within a certain geographic area. The rules 111 for the requesting of the transmissions of the measurement data 122 may be based on at least one of the following: a confidence of prediction, a latency requirement of the at least one UE 120a for receiving the prediction message 114, a periodicity requirement of the at least one UE 120a for receiving the prediction message 114, a frequency of updating the QoS model, a type of the QoS model, a type of the one or more QoS prediction parameters 113, a periodicity requirement of the network device 110 for transmitting the prediction message 114.

The rules for the requesting of the transmissions are not limited to the above-described items or parameters. The rules may also comprise further parameters not indicated above or derivatives of the above parameters.

The rules 111 for the requesting of the transmissions of the measurement data 122 may be deterministic or probabilistic.

The QoS model may for example comprises one of the following models: a table mapping, a table lookup, a Deep Neural Network, a Random Forest decision tree, a Kalman filter.

The one or more QoS prediction parameters 113 may for example comprise at least one of the following parameters of the mobile communication network: a predicted uplink, downlink, or sidelink throughput, a predicted uplink, downlink, or sidelink latency, a predicted jitter on uplink, downlink, or sidelink, a predicted service uptime or downtime, a predicted quality of experience.

The QoS prediction parameters are not limited to the above-described parameters. A lot of further parameters can be used as well.

Figure 2 shows a plurality of UEs 120. One or more of these UEs are described in the following.

The user equipment 120a is one exemplary UE 120a of a plurality of UEs 120 of the mobile communication network 100 that may be used for receiving one or more Quality-of-Service (QoS) prediction parameters 113 from a network device 110 of the mobile communication network, e.g., a network device 110 as described above or with respect to Figure 1. The UE 120a comprises: a communication interface (not shown in Figure 2) configured to receive a request message 112 from the network device 110 of the mobile communication network. The request message 112 requests transmissions of measurement data 122, e.g., MD 122 as described above with respect to Figure 1.

The UE 120a comprises a processor (not shown in Figure 2) configured to transmit the measurement data 122 via the communication interface to the network device 110 upon reception of the request message 112. The communication interface is configured to receive a prediction message 114 from the network device 110, e.g., a prediction message QP 114 as described above with respect to Figure 1 . The prediction message 114 comprises one or more QoS prediction parameters 113.

The one or more QoS prediction parameters 113 may be based on a QoS model of the mobile communication network. This QoS model can be precisely determined by the network device 110 based on measurement data 122 of the plurality of UEs 120.

The processor may be configured to execute an application using the one or more QoS prediction parameters 113.

The processor may be configured to perform at least one of the following tasks based on the one or more QoS prediction parameters 113: estimate a future speed of the UE 120a, select a route for the UE 120a, select a maneuver for the UE 120a, generate and/or update a radio map, estimate a future throughput on uplink, downlink, or sidelink, estimate a future latency on uplink, downlink, or sidelink, estimate a future jitter on uplink, downlink, or sidelink, estimate a future uptime or downtime of uplink, downlink, or sidelink, estimate a future quality of experience of uplink, downlink, or sidelink, generate and/or update a high-definition map.

The measurement data 122 may comprise at least one of the following: a signal to interference and noise ratio for a given location of the UE 120a, a Reference Signal Received Power of an uplink, downlink, or sidelink, a Channel State Information of an uplink, downlink, or sidelink, an estimated location of the UE (120a), an estimated speed of the UE (120a), an estimated location of another object, an estimated speed of another object, a bearing of detected objects and the UE (120a) itself.

The processor may be configured to transmit the measurement data 122 to the network device 110 based on the rules 111 for the requesting of the transmissions of the measurement data 122.

The processor may be configured to transmit one or more parameters 124 to the network device 110 which are indicative of requirements 123 of the UE 120a for determining the one or more QoS prediction parameters 113 by the network device 110.

The one or more parameters can be indicative of requirements to periodicity and/or QoS model accuracy.

With respect to the system model shown in Figure 1 , Figure 2 shows the signaling flowchart for providing the rules for MD triggering, providing MDs 122, and providing QPs 114. Note that the message exchange process in the figure is iterative and not necessarily synchronous (i.e. , the frequency of update of MD rules is lower than the frequency of providing MD messages; similarly, UEs may need to share the requirements on QP only if those requirements are updated; this update may be not synchronized with updating the QPs).

The parameters needed for generating QP 114 can be set by (one of) the network entity 110 that is implementing QPS functionality. Furthermore, UEs 120 may provide their own requirements on QP generation, which are particularly relevant for UE-specific, UE-group- specific QPs. QPs 114 may be generated by the model, i.e., the QoS model as indicated above, running on QPS and trained using the data in provided MDs 122. QP triggering can be dependent on both the QPS parameters and the requirements by the applications running on UEs 120. In particular, it can be a function of any combination of the following: the confidence of prediction (i.e., QP prediction accuracy) y_p, the UE latency requirement for QP T_UE , the UE requirement on periodicity of QPs, the frequency of updating of the model generating QPs, the type of the model running on QPS, the type of QP generated (UE-specific, UE-group-specific, applicable to any UE), additional requirements by network or entity hosting the QPS (denoted NQPS) (e.g., minimum or maximum periodicity of QP, additional information to be shared as part of QP, etc.), among other.

The rules for MD triggering can be set by (one of) the network entity, i.e., network entity 110, that is implementing QPS functionality. In particular, MD triggering can be either deterministic or probabilistic. In deterministic case, if a set of prescribed rules are satisfied, UE 120 can transmit MD 122; otherwise it will not. In probabilistic case, there is a function that describes MD transmission probability that may depend on the accuracy/quality of the collected data and its timeliness, e.g., P[MD] = 8 - y_d, where P[MD] is the probability of sending an MD 122. Furthermore, MD triggering can also depend on how many UEs 120 are active in the area (i.e., the UE density). That is, if there is a larger number of UEs 120 in an area, each UE can send MDs 122 less frequently. The density can be applicable for both deterministic and probabilistic case. In deterministic case, density can be considered as part of the rules; e.g., if density is below a threshold, UEs 120 send more frequently; otherwise, less frequently. In probabilistic case, the probability of triggering MDs 122 can be affected by the density, so that P[MD] = 8 ■ Ya ■ f( ), where f(.) is some function of density (e.g., normalized density) and a e (0,1] is the UE density in the area (signaled to UE by network or estimated by UE).

Fig. 3 shows a schematic diagram illustrating a system 300 for triggering prediction updates based on prediction accuracy according to a first embodiment and a corresponding prediction accuracy diagram 301 .

The system 300 may correspond to the systems described above with respect to Figures 1 and 2. The system 300 comprises a network device 110 on which a QoS prediction service (QPS) is implemented which communicates via a base station 130 to an exemplary number of three UEs 120a, 120b, 120c.

This first embodiment for triggering prediction updates is based on prediction accuracy. It is assumed that latency requirement is relaxed or non-existent for this first embodiment. UEs 120a, 120b, 120c provide M Ds 122a, 122b, 122c to QPS of network device 110 according to MD triggering rules. For example, for first UE 120a MD 122a is triggered if prediction accuracy falls below 97.5 percent; for second UE 120b MD 122b is triggered if prediction accuracy falls below 95.0 percent; for third UE 120c MD 122c is triggered if prediction accuracy falls below 90.0 percent.

QPS builds a QoS model 301 , e.g., a Radio Map, and derives y_p (i.e., the prediction accuracy) estimate based on MDs and the specific QoS metric shown in Figure 3.

Figure 3 shows a radio map example. The QoS metric is SINR (signal to interference plus noise) RMSE (root mean squared error), for example. The relevant information from MD is SINR. The QoS prediction may contain a relevant part of the Radio Map.

Alternatively, a HD (high density) map can be used: QoS metric can be location (e.g. latitude, longitude, altitude) RMSE; as relevant information from MD the object location (e.g. latitude, longitude, altitude) can be estimated; The QoS prediction may contain a relevant part of the HD Map.

QPS may update the QoS model 301 and y_p as new MDs arrive from UEs 120a, 120b, 120.

For each UE, QPS keeps track if the prediction accuracy y_p between current model (e.g., Radio Map) and that provided to UE is above a target. For example if SINR RMSE of Radio Map at a specific location with respect to the updated map is less than what is needed to achieve the target y_p, as shown in Figure 3.

When the prediction accuracy y_p falls below a target required by the application running on UE, QPS triggers a QP (Radio Map) update as shown in Figure 3.

Fig. 4 shows a schematic diagram illustrating a system 400 for triggering prediction updates based on prediction accuracy and latency according to a second embodiment.

The system 400 may correspond to the systems described above with respect to Figures 1 and 2. The system 400 comprises a network device 110 on which a QoS prediction service (QPS) is implemented which communicates via a base station 130 to an exemplary number of three UEs 120a, 120b, 120c. This second embodiment for triggering prediction updates is based on both prediction accuracy and latency. Latency requirement as well as prediction accuracy is crucial for this second embodiment.

The following processing example may be implemented in the system 400.

QPS on network device 110 has built a QoS model, e.g. a HD map, and a prediction accuracy Yp-

UE A 120a provides another MD 122 to QPS.

Even without MD 122 from UE A 120a, current QoS model at QPS satisfies y_p for UE B 120b; with MD 122 from UE A 120a, y_p would be improved, but requires additional processing time, i.e. a time delay.

UE B 120b has an outdated model, i.e. y_p is below target, and needs QP 114b.

If T_UE for UE B 120b does not allow enough time for including MD 122 from UE A 120a into the QoS model, QPS sends QP 114b to UE B 120b without including MD 122 from UE A 120a; otherwise, QPS incorporates MD 122 from UE A 120a in QP 114b to UE B 120b.

UE C 120c has an outdated QoS model and needs QP 114c.

By the time UE C 120c arrives, QPS has sufficient time to include both MD 122 from UE A 120a and MD from UE B 120b, irrespective of T_UE for UE C 120C.

Fig. 5 shows a schematic diagram illustrating a mobile network 500 according to the disclosure for triggering measurement data updates.

The system 500 may correspond to the systems described above with respect to Figures 1 and 2. The system 500 comprises a network device 110 on which a QoS prediction service (QPS) is implemented which communicates via a base station 130 to one or more UEs 120.

Two embodiments for triggering measurement data updates can be implemented. In a first embodiment, triggering measurement data updates is performed deterministically while in a second embodiment triggering measurement data updates is performed probabilistically. The first embodiment for triggering measurement data MD 122 from UEs 120 to the prediction service (QPS) on network device 110 relies on a set of rules that can be set forth by a network entity. This network entity can be either the network device 110 with QPS or another entity. The rules can for example take into account the data accuracy y_d and temporal characteristics ( 8 ) of the data. If the rules are satisfied, UE 120 may transmit MD 122. For example, if 8 > 0.9 & y_d > 0.95, UE 120 may transmits MD 122; otherwise not.

The second embodiment for triggering measurement data MD relies on rules and on probability measure depending on data accuracy y_d and temporal characteristics ( 8 ) of the data. For example, MD transmission probability can be defined as the product of data accuracy and temporal characteristics according to: P[MD] = 8 • y_d.

Furthermore, if MD triggering depends on how many UEs 120 are active in the area (UE density), MD transmission probability can be for example defined as P[MD] = 8 - y_d

where a e (0,1] is the UE density in the area. This can be signaled to UE 120 by network or estimated by UE 120.

Fig. 6 shows a schematic diagram illustrating an exemplary 5G network 600 including QoS prediction functionality 601 , 602 according to the disclosure in different parts of the network 600.

The 5G network 600 may be an exemplary mobile communication network as described above with respect to Figures 1 to 5. The 5G network 600 comprises a radio access network (NG- RAN) 610, a core network 620 and a data network 630. The radio access network 610 may include one or more base stations 130 as described above for connecting the plurality of UEs 120a, 120b, 120c, 120d. The NG-RAN 610 may perform RAN prediction. The core network 620 may include a NWDAF (network data analytics function) network element 621 , a UDM (unified data management) element 622, a PCF (policy control function) element 623, a NEF (network exchange function) element 624, an AF (application function) element 625, an AMF (access and mobility management function) element 626, an SMF (session management function) element 627 and a UPF (user plane function) element 628. The data network 630 may include a V2X (vehicle-to-anything) application 631.

Input data is collected via paths 603 highlighted in dark. Prediction results are delivered via paths 604 highlighted in light color. In terms of the deployment of QPS in the network entities, depending on the target set of applications that need to be supported, QPS 601 , 602 can be deployed in many different entities. QPS can reside as part of MEC (multi access edge computing), part of NWDAF (network data analytics function) 621 in core network 620, as an Al (artificial intelligence) entity/function in RAN side 610 of the network. In particular, deploying some QPS functions in RAN 610 can be beneficial when relevance area that the generated QPs refer to is small (e.g., around a base station).

Furthermore, if applications running on UEs require very low latency, related QPS functions can be placed in RAN. Some examples of functionalities enabled by RAN QPS are: a) Generating and signaling QPs (especially UE-specific and UE-group-specific QPs); b) Signaling between QPS and UE (in particular, any QPS-LIE interaction and requests except for triggering rules); c) Receiving and processing MDs received from UEs.

QPS functions in core network 620 can be beneficial when network-wide functionalities and parameters are concerned. As shown in Figure 6, QPS functionalities 601 , 602 can be included as part of the existing NWDAF 621 . Some examples of functionalities enabled by core network QPS are: a) Sharing definition and updates of triggering rules; b) Generating network-wide model (e.g., city scale Radio Map).

Fig. 7 shows a schematic diagram illustrating a communication system 700 for triggering data collection and QoS prediction updates according to the disclosure.

The communication system 700 may include one or more user equipments 120a, 120b or UEs, respectively, according to an example, a base station 130 and a network device 110, e.g., as described above with respect to Figures 1 and 2.

In the example shown in Figure 7, the first UE 120a and the second UE 120b are, by way of example, portable devices, in particular smartphones 120a, 120b. However one or more of these user devices 120a, 120b may also be, by way of example, laptop computer 120a, 120b, mobile vehicle or machine-type device. The user devices 120a, 120b may be configured to communicate with the base station 130, for instance, via llu channel 704a, 704b. The Uu channel 704a, 704b or also referred to as E-UTRAN Uu interface, also known as LTE Uu or simply LTE radio interface, allows data transfer between the ENodeB (or base station 130) and the UEs. The base station 130 may be configured to communicate with the network device 110 via communication link 714.

As can be seen from Figure 7, the first and second user devices 120a, 120b may comprise a processing circuitry 123a, 123b for instance, a processor 123a, 123b, for processing and generating data, a communication interface or transceiver 125a, 125b, including, for instance, a transmitter, a receiver and an antenna, for exchanging data with the other components of the communication system 700, and a non-transitory memory 127a, 127b for storing data. The processor 123a, 123b of the user device 120a, 120b may be implemented in hardware and/or software.

The hardware may comprise digital circuitry, or both analog and digital circuitry. Digital circuitry may comprise components such as application-specific integrated circuits (ASICs), field- programmable arrays (FPGAs), digital signal processors (DSPs), or general-purpose processors. The non-transitory memory 127a, 127b may store data as well as executable program code which, when executed by the processor 123a, 123b, causes the respective user device 120a, 120b to perform the functions, operations and methods described in this disclosure.

In an example, the network device 110 may have a similar architecture as the user devices 120a, 120b, i.e., may comprise a processor 113 for processing and generating data, a communication interface 115 or transceiver for exchanging data with the other components of the communication system 700 as well as a memory 117 for storing data.

In the following, the functionalities of the entities in Figure 17 are described in detail.

The network device 110 that may correspond to the network entity 110 described above with respect to Figure 2 can be used for determining one or more of Quality-of-Service (QoS) prediction parameters 113 in a mobile communication network, e.g., a mobile communication network 100 as shown in Figures 1 and 2.

As described above, the network device 110 comprises a communication interface 115 configured to send a request message 112 towards a plurality of User Equipments 120a, 120b, the request message 112 requesting transmissions of measurement data 122 from the plurality of User Equipments 120a, 120b. The communication interface 115 is configured to receive the measurement data 122. This measurement data 122 may correspond to the MD 122 described above with respect to Figure 1.

As described above, the network device 110 comprises a processor 113 configured to determine a QoS model of the mobile communication network 100 based on the received measurement data 122. The processor 113 is configured to determine one or more QoS prediction parameters 113 for at least one UE 120a of the plurality of UEs 120 based on the QoS model, and to transmit a prediction message 114 comprising the one or more QoS prediction parameters 113 to the at least one UE 120a. This prediction message 114 may correspond to the QP message 114 described above with respect to Figure 1. The processor 113 is configured to generate the request message 112 upon the basis of the QoS model.

As described above, the UE 120a, 120b comprises a communication interface 125a, 125b configured to receive a request message 112 from the network device 110 of the mobile communication network. The request message 112 requests transmissions of measurement data 122, e.g., MD 122 as described above with respect to Figure 1.

The UE 120a, 120b comprises a processor 123a, 123b configured to transmit the measurement data 122 via the communication interface 125a, 125b to the network device 110 upon reception of the request message 112. The communication interface 125a, 125b is configured to receive a prediction message 114 from the network device 110, e.g., a prediction message QP 114 as described above with respect to Figure 1. The prediction message 114 comprises one or more QoS prediction parameters 113.

Fig. 8 shows a schematic diagram illustrating a method 800 for determining one or more QoS prediction parameters 113 in a mobile communication network. The method 800 may implement the functionality of the network device 110 described above with respect to Figures 1 to 7.

The method 800 comprises sending 801 a request message 112 towards a plurality of User Equipments (UEs) 120, the request message 112 requesting transmissions of measurement data 122 from a plurality of UEs 120, e.g., as described above with respect to Figure 2.

The method 800 comprises receiving 802 the measurement data 122, e.g., as described above with respect to Figure 2. The method 800 comprises determining 803 a QoS model of the mobile communication network based on the received measurement data 122, e.g., as described above with respect to Figure 2.

The method 800 comprises determining 804 one or more QoS prediction parameters 113 for at least one UE 120a of the plurality of UEs 120 based on the QoS model, e.g., as described above with respect to Figure 2.

The method 800 comprises transmitting 805 a prediction message 114 comprising the one or more QoS prediction parameters 113 to the at least one UE 120a, e.g., as described above with respect to Figure 2.

The method 800 comprises generating 806 the request message 112 upon the basis of the QoS model, e.g., as described above with respect to Figure 2.

Fig. 9 shows a schematic diagram illustrating a method 900 for receiving one or more QoS prediction parameters from a network device of a mobile communication network. The method 900 may implement the functionality of any of the UEs 120 described above with respect to Figures 1 to 7.

The method 900 comprises receiving 901 a request message 112 from the network device of the mobile communication network, the request message 112 requesting transmissions of measurement data 122, e.g., as described above with respect to Figure 2.

The method 900 comprises transmitting 902 the measurement data 122 to the network device upon reception of the request message 112, e.g., as described above with respect to Figure 2.

The method 900 comprises receiving 903 a prediction message 114 from the network device, the prediction message 114 comprising one or more QoS prediction parameters 113, e.g., as described above with respect to Figure 2.

The present disclosure also supports a computer program product including computer executable code or computer executable instructions that, when executed, causes at least one computer to execute the performing and computing steps described herein, in particular the methods and procedures described above. Such a computer program product may include a readable non-transitory storage medium storing program code thereon for use by a computer. The program code may perform the processing and computing steps described herein, in particular the methods and procedures described above.

While a particular feature or aspect of the disclosure may have been disclosed with respect to only one of several implementations, such feature or aspect may be combined with one or more other features or aspects of the other implementations as may be desired and advantageous for any given or particular application. Furthermore, to the extent that the terms "include", "have", "with", or other variants thereof are used in either the detailed description or the claims, such terms are intended to be inclusive in a manner similar to the term "comprise". Also, the terms "exemplary", "for example" and "e.g." are merely meant as an example, rather than the best or optimal. The terms “coupled” and “connected”, along with derivatives may have been used. It should be understood that these terms may have been used to indicate that two elements cooperate or interact with each other regardless whether they are in direct physical or electrical contact, or they are not in direct contact with each other.

Although specific aspects have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a variety of alternate and/or equivalent implementations may be substituted for the specific aspects shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific aspects discussed herein.

Although the elements in the following claims are recited in a particular sequence with corresponding labeling, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those elements, those elements are not necessarily intended to be limited to being implemented in that particular sequence.

Many alternatives, modifications, and variations will be apparent to those skilled in the art in light of the above teachings. Of course, those skilled in the art readily recognize that there are numerous applications of the invention beyond those described herein. While the present invention has been described with reference to one or more particular embodiments, those skilled in the art recognize that many changes may be made thereto without departing from the scope of the present invention. It is therefore to be understood that within the scope of the appended claims and their equivalents, the invention may be practiced otherwise than as specifically described herein.

Claims

CLAIMS:

1. A network device (110) for determining one or more of Quality-of-Service, QoS, prediction parameters (113) in a mobile communication network, the network device (110) comprising: a communication interface configured to send a request message (112) towards a plurality of User Equipments, UEs (120), the request message (112) requesting transmissions of measurement data (122) from the plurality of User Equipments, UEs (120), and wherein the communication interface is configured to receive the measurement data

(122); and a processor configured to determine a QoS model of the mobile communication network based on the received measurement data (122); wherein the processor is configured to determine one or more QoS prediction parameters (113) for at least one UE (120a) of the plurality of UEs (120) based on the QoS model, and to transmit a prediction message (114) comprising the one or more QoS prediction parameters (113) to the at least one UE (120a), and wherein the processor is configured to generate the request message (112) upon the basis of the QoS model.

2. The network device (110) of claim 1 , wherein the request message (112) comprises one or more parameters indicative of rules (111) for the requesting of the transmissions of the measurement data (122).

3. The network device (110) of claim 2, wherein the communication interface is configured to receive one or more parameters (124) indicative of requirements (123) of at least one of the plurality of UEs (120) for determining the one or more QoS prediction parameters (113); and wherein the processor is configured to determine the one or more QoS prediction parameters (113) for the at least one UE (120a) based on the corresponding requirements

(123) of the at least one UE (120a).

27

4. The network device (110) of claim 2 or 3, wherein the processor is configured to transmit the prediction message (114) to at least one of the following: a specific UE (120a), a specific group of UEs (120a, 120b), all UEs (120a, 120b) within a certain geographic area.

5. The network device (110) of any of claims 2 to 4, wherein the rules (111) for the requesting of the transmissions of the measurement data (122) are based on at least one of the following: a confidence of prediction, a latency requirement of the at least one UE (120a) for receiving the prediction message (114), a periodicity requirement of the at least one UE (120a) for receiving the prediction message (114), a frequency of updating the QoS model, a type of the QoS model, a type of the one or more QoS prediction parameters (113), a periodicity requirement of the network device (110) for transmitting the prediction message (114).

6. The network device (110) of any of claims 2 to 5, wherein the rules (111) for the requesting of the transmissions of the measurement data (122) are deterministic or probabilistic.

7. The network device (110) of any of the preceding claims, wherein the QoS model comprises one of the following models: a table mapping, a table lookup, a Deep Neural Network, a Random Forest decision tree, a Kalman filter.

8. The network device (110) of any of the preceding claims, wherein the one or more QoS prediction parameters (113) comprise at least one of the following parameters of the mobile communication network: a predicted uplink, downlink, or sidelink throughput, a predicted uplink, downlink, or sidelink latency, a predicted jitter on uplink, downlink, or sidelink, a predicted service uptime or downtime, a predicted quality of experience.

9. A user equipment, UE (120a), of a plurality of UEs (120) of a mobile communication network, for receiving one or more Quality-of-Service, QoS, prediction parameters (113) from a network device (110) of the mobile communication network, the UE (120a) comprising: a communication interface configured to receive a request message (112) from the network device of the mobile communication network, the request message (112) requesting transmissions of measurement data (122); and a processor configured to transmit the measurement data (122) via the communication interface to the network device upon reception of the request message (112), wherein the communication interface is configured to receive a prediction message (114) from the network device, the prediction message (114) comprising one or more QoS prediction parameters (113).

10. The UE (120a) of claim 9, wherein the processor is configured to execute an application using the one or more QoS prediction parameters (113).

11 . The UE (120a) of claim 9 or 10, wherein the processor is configured to perform at least one of the following tasks based on the one or more QoS prediction parameters (113): estimate a future speed of the UE (120a), select a route for the UE (120a), select a maneuver for the UE (120a), generate and/or update a radio map, estimate a future throughput on uplink, downlink, or sidelink, estimate a future latency on uplink, downlink, or sidelink, estimate a future jitter on uplink, downlink, or sidelink, estimate a future uptime or downtime of uplink, downlink, or sidelink, estimate a future quality of experience of uplink, downlink, or sidelink, generate and/or update a high-definition map.

12. The UE (120a) of any of claims 9 to 11 , wherein the measurement data (122) comprise at least one of the following: a signal to interference and noise ratio for a given location of the UE (120a), a Reference Signal Received Power of an uplink, downlink, or sidelink, a Channel State Information of an uplink, downlink, or sidelink, an estimated location of the UE (120a), an estimated speed of the UE (120a), an estimated location of another object, an estimated speed of another object, a bearing of detected objects and the UE (120a) itself.

13. The UE (120a) of any of claims 9 to 12, wherein the request message (112) comprises one or more parameters indicative of rules (111) for the requesting of the transmissions of the measurement data (122).

14. The UE (120a) of claim 13, wherein the processor is configured to transmit the measurement data (122) to the network device (110) based on the rules (111) for the requesting of the transmissions of the measurement data (122).

15. The UE (120a) of any of claims 9 to 14, wherein the processor is configured to transmit one or more parameters (124) to the network device (110), the one or more parameters being indicative of requirements (123) of the UE (120a) for determining the one or more QoS prediction parameters (113) by the network device (110).

16. The UE (120a) of claim 15, wherein the one or more parameters are indicative of requirements to periodicity and/or QoS model accuracy.

17. A method (800) for determining one or more of Quality-of-Service, QoS, prediction parameters (113) in a mobile communication network, the method (800) comprising: sending (801) a request message (112) towards a plurality of User Equipments, UEs (120), the request message (112) requesting transmissions of measurement data (122) from the plurality of User Equipments, UEs (120); receiving (802) the measurement data (122); determining (803) a QoS model of the mobile communication network based on the received measurement data (122); determining (804) one or more QoS prediction parameters (113) for at least one UE (120a) of the plurality of UEs (120) based on the QoS model; transmitting (805) a prediction message (114) comprising the one or more QoS prediction parameters (113) to the at least one UE (120a); and generating (806) the request message (112) upon the basis of the QoS model.

31

18. A method (900) for receiving one or more Quality-of-Service, QoS, prediction parameters (113) from a network device (110) of a mobile communication network, the method (900) comprising: receiving (901) a request message (112) from the network device of the mobile communication network, the request message (112) requesting transmissions of measurement data (122); transmitting (902) the measurement data (122) to the network device upon reception of the request message (112); and receiving (903) a prediction message (114) from the network device, the prediction message (114) comprising one or more QoS prediction parameters (113).

19. A computer program product including computer executable code or computer executable instructions that, when executed, causes at least one computer to execute the method (800, 900) according to claim 17 or 18.

32