Network node for QoS notification control

Abstract

The invention relates to enhanced quality of service (QoS) notification control, such that in case of a QoS change of a QoS flow, a client device or application function (AF) can be notified, whether the QoS change was predicted or not. A session management function (SMF) or policy control function (PCF) determines a prediction result for a QoS flow of a PDU session based on a comparison of a potential QoS change of the QoS flow with an actual change of the QoS flow. The prediction result can be a Boolean flag set to "true", if the potential QoS change of the QoS flow is the same as the actual change of the QoS flow, and set to "false" in the opposite case. The SMF or PCF sends the determined prediction result for the QoS flow to a client device or AF associated with the PDU session. Thus, if an application knows in advance about a QoS change, the application can mitigate the impact, e.g. by taking measures. Furthermore, the invention also relates to corresponding methods and computer programs.

Description

Technical Field

[0001] This invention relates to a first network node and a second network node for enhanced QoS notification control and for enhanced QoS sustainability analysis. Furthermore, the invention also relates to corresponding methods and computer programs. Background Technology

[0002] Industry applications rely on fulfilling Service Level Specifications (SLS). The GSMA generic network slice template (GST) serves as the service level agreement (SLA) information for communication between vertical industries and communication service providers. SLA requirements need to be fulfilled in a coordinated manner from both the management and control planes. The SLS includes the ServiceProfile data type. The ServiceProfile data type represents the attributes that a network slice instance (NSI) in a fifth-generation (5G) network should support regarding network slice requirements. Network slices can be tailored based on specific requirements and must comply with the SLA agreed upon between the network slice customer (NSC) and the network slice provider (NSP).

[0003] Deviations in 5G system (5GS) performance from any parameter described in the SLS can have serious consequences for applications and financial losses for network operators as a result of any penalties applied under the SLA. SLS monitoring and measurement are critical components of ensuring and validating the agreed SLS delivery. Therefore, 5G systems should provide monitoring and event management throughout the infrastructure, effectively and efficiently managing the infrastructure 24 / 7 to ensure minimal downtime and immediate notification of performance degradation or service unavailability.

[0004] SLS implementation is based on a set of end-to-end key performance indicators (KPIs) that should be monitored by 5GS. If any deviations are observed on those parameters, the application should be notified, and a slice-level modification process should be triggered to compensate for the deviations. Summary of the Invention

[0005] The purpose of embodiments of the present invention is to provide a solution that mitigates or solves the drawbacks and problems of conventional solutions.

[0006] The above and other objectives are achieved by the subject matter claimed in the independent claims. Further advantageous embodiments of the invention can be found in the dependent claims.

[0007] According to a first aspect of the invention, the above and other objectives are achieved by a first network node for a communication system, the first network node being configured to:

[0008] A first message is transmitted to the second network node, the first message indicating a request for a potential QoS change of the quality of service (QoS) stream of the protocol data unit (PDU) session established by the first network node;

[0009] Receive a second message from the second network node, the second message indicating the potential QoS change of the QoS flow;

[0010] The prediction result of the QoS flow is determined by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow;

[0011] A third message is transmitted to the client device or application function (AF) associated with the PDU session, the third message indicating the prediction result of the QoS flow.

[0012] Potential QoS changes can be understood, for example, as notifications provided by network nodes to network consumers when a network consumer requests QoS sustainability analysis for the target analysis period.

[0013] The advantage of the first network node according to the first aspect is that it can compare the notification of a QoS change that has already occurred with its previously received predictions of potential QoS changes that may have already occurred. In this case, the first network node knows whether the QoS change that has occurred is something that was predicted in the past (in such cases, it also knows the prediction reference), thus distinguishing between systemic (or recurring) problems and transient problems.

[0014] Proper operation of an application requires it to fulfill the necessary QoS (Quality of Service) requirements within the 3GPP framework. Typically, any QoS change not triggered by the application can cause problems. Furthermore, anticipating QoS changes in advance reduces the potential harm to the application. In fact, if an application knows about QoS changes in advance, it can mitigate their impact by taking countermeasures. Other benefits of being a first network node include:

[0015] ● Supports 5GS performance predictability KPIs as part of the network slicing SLS and SLS guarantee process, which can be calculated based on prediction results and relevant time periods when QoS changes occur.

[0016] ●Supports 5GS performance predictability KPIs as part of the analysis of observed slice performance and application performance.

[0017] ● The application is able to meet the requirements for 5GS performance predictability in sliced ​​SLS.

[0018] ● The application is able to receive information disclosures regarding the predictability of current 5GS performance, which is supported as part of SLS.

[0019] In one implementation of the first network node according to the first aspect, the first network node is used for:

[0020] Upon receiving a fourth message from the radio access network (RAN), the predicted result of the QoS flow is determined, the fourth message indicating the actual change in the QoS flow.

[0021] The advantage of this implementation is that by calculating the prediction result upon receiving the fourth message from the RAN, the first network node can assess whether 5GS performs in a predictable manner for a specific communication service, despite QoS non-fulfillment related to actual QoS changes. As mentioned above, the predictability of 5GS is a desired feature because higher predictability determines higher application availability. Since the first network node participates in session management for a specific communication service, the advantage of assessing the prediction result is that such information can be shared with the application as an indication of the predictability of 5GS. The application can use the information to verify whether the 5GS's performance in terms of predictability complies with the SLA. Another advantage is that the first network node can share such information with both ends of the application, namely, sharing the information with the UE side of the application (e.g., via SMF, access and mobility management function (AMF), RAN, and user equipment (UE)) or with the application side connected to the AF (e.g., via PCF and AF).

[0022] In one implementation of the first network node according to the first aspect, the first network node is used for:

[0023] Upon receiving a fifth message from the client device or the AF, the first message is transmitted to the second network node, the fifth message indicating a request for a prediction result of the QoS flow.

[0024] The advantage of this implementation is that the first network node participating in session management for a specific communication service can combine real-time information about the session in question with applicable predictions for the same session by transmitting a first message to the second network node upon receiving a fifth message from the client device or AF. An example of such a combination is calculating predictions in the event of QoS changes. The first network node can also use the predictions for the session in question for other purposes, such as resource management, policy decisions, or network configuration changes.

[0025] In one implementation of the first network node according to the first aspect, the prediction result of the QoS flow is a Boolean flag.

[0026] The advantage of this implementation is that it provides a low-overhead indication of the prediction results.

[0027] In one implementation of the first network node according to the first aspect, the second message further indicates an identifier of the potential QoS change of the QoS flow, wherein the first network node is used for:

[0028] The potential QoS changes of the QoS flow are identified based on the identifier.

[0029] The advantage of this implementation is that the consumer of NWDAF QoS sustainability analysis can use a novel, introduced identifier to refer to each prediction included in the potential QoS change notification when the analysis notification is correlated with the actual QoS change. In fact, a notification in Section 6.9 of TS23.288 (regarding potential QoS changes or QoS sustainability analysis) may refer to a relatively long analysis target period or a relatively large area (cell or tracking area), thus containing many predictions. The identifier can be used as a prediction reference to identify each predicted change in the KPI in question relative to a pre-configured threshold defined by the analysis consumer.

[0030] In one implementation of the first network node according to the first aspect, the first network node is used for:

[0031] A sixth message is transmitted to the second network node, the sixth message indicating the prediction result of the QoS flow.

[0032] The advantage of this implementation is that the first network node participates in the session management of a specific communication service, and the second network node can receive such information in real time whenever the QoS of the communication service in question changes. The second network node can use this information, such as the communication service ID, time interval, service node, and service network slice, to track and build reports on which communication services are behaving as predicted or not as predicted.

[0033] In one implementation of the first network node according to the first aspect, the first network node is a session management function (SMF) or a policy control function (PCF), wherein at least one of the following:

[0034] The first message is either Nnwdaf_AnalyticsSubscription_Subscribe or Nnwdaf_AnalyticsInfo_Request;

[0035] The second message is either Nnwdaf_AnalyticsSubscription_Notify or Nnwdaf_AnalyticsInfo_Response;

[0036] When the first network node is an SMF, the third message is an Npcf_SMPolicyControl_Update request or a PDU SESSION MODIFICATION COMMAND; when the first network node is a PCF, the third message is an Npcf_PolicyAuthorization_Notify or an Npcf_SMPolicyControl_Update response.

[0037] The fourth message is an Nsmf_PDUSession_UpdateSMContext request when the first network node is an SMF, or an Npcf_SMPolicyControl_Update request when the first network node is a PCF.

[0038] When the first network node is an SMF, the fifth message is an Nsmf_PDUSession_UpdateSMContext request, an Nsmf_PDUSession_CreateSMContext request, or an Npcf_PolicyAuthorization_Subscribe request. When the first network node is a PCF, the fifth message is an SM policy association establishment or modification or an Npcf_PolicyAuthorization_Subscribe request.

[0039] The sixth message is the Nsmf_EventExposure_Subscribe notification.

[0040] The advantage of this implementation is that the first network node can be deployed in the SMF or PCF node portion of the communication service session management and policy control.

[0041] According to a second aspect of the invention, the above and other objectives are achieved by a second network node for a communication system, the second network node being used for:

[0042] Receive a first message from the first network node, the first message indicating a request for a potential QoS change of the QoS flow of the PDU session established on the first network node;

[0043] A second message is transmitted to the first network node, the second message indicating the potential QoS change of the QoS flow and an identifier of the potential QoS change of the QoS flow.

[0044] The advantage of the second network node according to the second aspect is that the consumer of NWDAF QoS sustainability analysis can refer to each prediction included in the potential QoS change notification when relating the analysis notification to the actual QoS change. In fact, the notification in Section 6.9 of TS 23.288 (notification regarding potential QoS changes or QoS sustainability analysis) may refer to a relatively long analysis target period or a relatively large area (cell or tracking area), and therefore includes many predictions. The prediction reference can identify each predicted change in the KPI in question relative to a pre-configured threshold defined by the analysis consumer.

[0045] In one implementation of the second network node according to the second aspect, the second network node is used for:

[0046] A seventh message is transmitted to the first network node, the seventh message indicating a request for the prediction result of the QoS flow;

[0047] A sixth message is received from the first network node, the sixth message indicating the prediction result of the QoS flow, the prediction result being a comparison between the potential QoS change of the QoS flow and the actual change of the QoS flow.

[0048] The advantage of this implementation is that the prediction results can be used to compare the actual behavior of the communication service with its predicted behavior.

[0049] In one implementation of the second network node according to the second aspect, the second network node is used for:

[0050] Based on the prediction results of the QoS flow, the key performance indicators (KPIs) of the QoS flow and the observed time intervals are determined.

[0051] The advantage of this implementation is that KPIs provide an indication of the predictability of 5GS performance for the specific communication service or application in question. KPIs can be used as part of the SLA / SLS requested by network slicing customers from network slicing providers. 5GS services with higher values ​​for these KPIs may be charged higher prices by network slicing providers because they may determine greater application availability.

[0052] In one implementation of the second network node according to the second aspect, the second network node is used for:

[0053] The KPI of the QoS flow is determined based on the ratio between the sum of time intervals in which the QoS flow changes and the prediction result of the QoS flow is true, and the sum of time intervals in which the QoS change occurs.

[0054] The advantage of this implementation is that this type of KPI can be used to assess potential gains or losses in application availability. In fact, for a time interval where a QoS change occurs and the prediction is true, application availability is expected to be higher compared to a time interval where a QoS change occurs and the prediction is false. Indeed, if the prediction is true, 5GS performs predictably despite the QoS change and can notify the application in advance of the impending QoS change. Applications can use this advance notification to respond to impending QoS changes and continue operating (increased availability).

[0055] In one implementation of the second network node according to the second aspect, the second network node is used for:

[0056] When a request for the KPI is received from an NF consumer, the KPI is transmitted to the NF consumer.

[0057] The advantage of this implementation is that if the NF consumer is a network slice management system, such KPIs can be used to trigger a potential slice adaptation process.

[0058] According to a third aspect of the invention, the above and other objectives are achieved by a method for a first network node, the method comprising:

[0059] A first message is transmitted to the second network node, the first message indicating a request for a potential QoS change of the QoS flow of the PDU session established on the first network node;

[0060] Receive a second message from the second network node, the second message indicating the potential QoS change of the QoS flow;

[0061] The prediction result of the QoS flow is determined by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow;

[0062] A third message is transmitted to the client device or AF associated with the PDU session, the third message indicating the prediction result of the QoS flow.

[0063] The method according to the third aspect can be extended to an implementation corresponding to the implementation of the first network node according to the first aspect. Therefore, the implementation of the method includes the features of the corresponding implementation of the first network node.

[0064] The advantages of the method according to the third aspect are the same as the advantages of the corresponding implementation of the first network node according to the first aspect.

[0065] According to a fourth aspect of the invention, the above and other objectives are achieved by a method for a second network node, the method comprising:

[0066] Receive a first message from the first network node, the first message indicating a request for a potential QoS change of the QoS flow of the PDU session established on the first network node;

[0067] A second message is transmitted to the first network node, the second message indicating the potential QoS change of the QoS flow and an identifier of the potential QoS change of the QoS flow.

[0068] The method according to the fourth aspect can be extended to an implementation corresponding to the implementation of the second network node according to the second aspect. Therefore, the implementation of the method includes the features of the corresponding implementation of the second network node.

[0069] The advantages of the method according to the fourth aspect are the same as the advantages of the corresponding implementation of the second network node according to the second aspect.

[0070] The present invention also relates to a computer program characterized by program code that, when run by at least one processor, causes the at least one processor to perform any method according to embodiments of the present invention. Furthermore, the present invention relates to a computer program product comprising a computer-readable medium and the aforementioned computer program, wherein the computer program is contained in the computer-readable medium and includes one or more from the group consisting of: read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), flash memory, electrically EPROM (EEPROM), and hard disk drives.

[0071] Other applications and advantages of embodiments of the present invention will become apparent from the following detailed description. Attached Figure Description

[0072] The accompanying drawings are intended to illustrate and explain different embodiments of the invention, in which:

[0073] – Figure 1 This illustrates a first network node according to an embodiment of the present invention;

[0074] – Figure 2 This illustrates a method for a first network node according to an embodiment of the present invention;

[0075] – Figure 3 The second network node is shown in an embodiment of the present invention;

[0076] – Figure 4 This invention illustrates a method for a second network node according to an embodiment of the present invention;

[0077] – Figure 5 This illustrates the QoS notification process of an embodiment of the present invention;

[0078] – Figure 6 This illustrates the slice QoE analysis process of an embodiment of the present invention;

[0079] – Figure 7 This illustrates the QoS notification process of an embodiment of the present invention;

[0080] – Figure 8 This illustrates the QoS sustainability analysis process of an embodiment of the present invention;

[0081] – Figure 9 This illustrates the SLS guarantee process of an embodiment of the present invention;

[0082] – Figure 10 This illustrates the slice QoE analysis process of an embodiment of the present invention;

[0083] – Figure 11 SMF event notification, illustrating an embodiment of the present invention;

[0084] – Figure 12 A flowchart illustrating a QoS sustainability analysis subscription triggered by a PDU session event according to an embodiment of the present invention is shown.

[0085] – Figure 13 A flowchart illustrating a QoS sustainability analysis subscription triggered by QoS notification control according to an embodiment of the present invention is shown.

[0086] – Figure 14 The calculation of the "asPredicted" Boolean flag is shown in an embodiment of the present invention. Detailed Implementation

[0087] Some applications are inherently adaptive. This means they can have several operating modes (also known as application modes, such as normal mode, security mode, and adaptation mode) that are triggered to run based on the QoS available from the network. Switching between application modes may not be immediate: the switch may take hundreds of milliseconds to several seconds to reconfigure the process control network based on a new set of parameters, or even several minutes to change the relevant application-level parameters. The transition between application modes must be completed before the network's QoS changes. Therefore, it is important to anticipate upcoming QoS changes and notify applications in advance so that they can switch to the appropriate mode, allowing them to continue operating under the upcoming QoS conditions.

[0088] In this context, it is important that SLS be supplemented with additional KPIs, as stated in the GSMA NEST: Performance Prediction - Availability; Performance Prediction - Predicted Frequency. GSMA GST v4.0 has already included end-to-end (E2E) KPIs related to performance prediction in the SLS template. These parameters are crucial for industrial applications because performance prediction triggers a switch from normal mode to adaptive mode (either way is possible), which in turn helps reduce application unavailability. Since reducing application unavailability depends on the reliability of performance prediction (i.e., predictions are accurate enough to predict achievable 5GS performance), it is also important to introduce new KPIs to measure the quality of performance prediction.

[0089] This paper defines this type of KPI as follows: Performance Prediction - 5GS performance predictability is an indicator of 5GS performance prediction service quality, and therefore also an indicator of how it helps reduce application unavailability. This invention proposes introducing this new E2E KPI and its calculation method in 5GS. It also introduces an additional information element (IE) in the notification generated by 5GS in the event of performance degradation. The IE describes whether such performance degradation was previously predicted. This information is relevant to the application because if the performance degradation was previously predicted, the application is still likely to handle the degradation, while unpredicted performance degradation could be more detrimental to the application. Some aspects of 5G performance prediction were introduced in 3GPPRel-16, as specified in Section 6.9, as a new set of analyses. This type of analysis, called QoS sustainability analysis, allows applications or service consumers such as 5GS network functions (NFs) to request predictions for target periods (time intervals). Therefore, this invention uses terms such as "performance prediction" to refer to the analytical services provided by such functions. However, if new performance prediction services are introduced in the future to predict 5GS performance, this new KPI could also refer to such new prediction services.

[0090] Furthermore, this invention focuses on the analysis of “posterior” computational predictability based on system observations and related data collection. “Posterior” computational predictability means first observing the system for a period of time, and then, after collecting sufficient information, calculating the E2E KPI according to a mathematical formula. 5GS performance predictability can be defined by borrowing some concepts from weather forecasting science. In weather forecasting science, given the current state of the atmosphere, a simulation means the case most similar to a repository of other past atmospheric states. The temporal divergence of the observed simulation (i.e., similar observed atmospheric states) provides an estimate of the forecast divergence. In 5GS, predictability can be defined by determining the time intervals in which the system behaves as predicted (as in the “simulation”) and the time intervals in which the system behaves differently from the prediction.

[0091] Because different applications rely on the prediction of different 5GS KPIs, 5GS predictability is relevant to the specific application being considered. This means that KPIs should measure how well 5GS predicts these KPIs, which are important to the specific application because the application may be sensitive to them. For example, motion control may depend on latency prediction, as failure to meet latency requirements can lead to application unavailability, while remote control actuation depends on latency prediction as well as the data rate of the uplink path.

[0092] Therefore, the proposed definition of 5GS performance predictability is as follows: For a specific application, 5GS performance predictability is the ratio of performance degradation or service unavailability events that occur and have been notified in advance to the total number of performance degradation or service unavailability events that occur.

[0093] 5GS performance predictability is an end-to-end KPI that measures the quality of the performance prediction service provided to an application by 5GS within an observed time window. A 5GS performance predictability value close to 1 is an indicator of a good prediction service. 5GS performance predictability is directly related to application availability: higher predictability means lower application unavailability, because higher performance degradation can be compensated for by triggering adaptive modes rather than switching the system to safe mode.

[0094] An embodiment of the present invention defines 5GS performance predictability as a novel 5GS E2E KPI, with the aim of supporting such a new KPI as part of the SLS of network slicing in 3GPP systems. Embodiments of the present invention introduce support for the new KPI through a solution based on the following existing mechanisms:

[0095] ● Notification control mechanism, triggered by the RAN when a GBR QoS flow is fulfilled / not fulfilled.

[0096] ●QoS sustainability analysis supports the prediction of potential QoS changes.

[0097] ● Service experience analysis observed by network slicing, supporting the analysis of services related to network slices.

[0098] Because GBR-type QoS flows support notification control mechanisms, the embodiments described below in the 3GPP context cover, for example, the predictability of 5G system (5GS) performance for GBR-type QoS flows. The advantage of these embodiments of the invention is that minimal changes to 5GS are required since all services are already in place.

[0099] Therefore, embodiments of the present invention may address at least two main issues.

[0100] Question 1: Definition of the reporting and assurance process for this type of 5GS E2E KPI. Therefore, the same parameters should be supported according to the following aspects of the 3GPP system:

[0101] ○ In the defined end-to-end 5GS KPI.

[0102] ○ In the relevant NWDAF analysis reported as part of the slice QoE and in the NSI and NSSI performance assurance process.

[0103] ○ During the service experience of the application.

[0104] Question 2: Define the mechanism for disclosing such KPIs to the application, including a mechanism for notifying the application in the event of a performance degradation event:

[0105] For applications that act as consumers of predictive analytics services, which performance degradation / service unavailability events were actually predicted or not?

[0106] ○Which events may have been handled as adaptation mode or not?

[0107] ○ Which events cause application unavailability and which do not, because application unavailability can be mitigated through adaptation patterns that allow continued operation under specific circumstances.

[0108] Solutions to Problem 1 could involve modifications to the QoS notification control process, for example:

[0109] ● The modified process informs the UE that the QoS of a specific QoS flow has been modified, and includes the "asPredicted" Boolean flag as an additional information element (IE) upon UE request. The "asPredicted" Boolean flag corresponds to the Boolean flag previously described for the prediction outcome (PO).

[0110] ● The modified process, used to inform the AF that the QoS of a specific QoS flow has been modified, includes the "asPredicted" boolean flag as an additional IE, depending on the request. In this notification, 5GS can also add a predicted reference as another additional IE, which can inform the AF of the predicted QoS change in terms of timing and method. The predicted reference can be a new IE, as described in the last bullet point.

[0111] ● A new procedure for SMF-NWDAF interaction allows the SMF to become a consumer of QoS sustainability analysis in order to determine the "asPredicted" boolean flag. The SMF can perform this new function by comparing the currently implemented QoS with the previously predicted QoS (for each QoS flow of GBR type).

[0112] ● Additional information element (optional) for QoS sustainability analysis on NWDAF: A unique event ID for each threshold reached or exceeded, which can be used as a "predictive reference". The unique event ID corresponds to the identifier previously described for potential QoS changes.

[0113] Solutions to problem 2 could involve supporting 5GS performance predictability guarantees, for example:

[0114] ●SMF will include the "asPredicted" boolean flag in the events reported to NWDAF.

[0115] ● A new trigger associated with the RAN fulfillment / non-fulfillment of a GBR QoS flow: event id = "QoS flow identifier (QFI) assignment".

[0116] ● The revised slice QoE analysis also includes 5GS performance predictability to be reported as part of the slice QoE, to support appropriate assurance processes.

[0117] Figure 1 A first network node 100 according to an embodiment of the present invention is shown. Figure 1 In the illustrated embodiment, the first network node 100 includes a processor 102, a transceiver 104, and a memory 106. The processor 102 is coupled to the transceiver 104 and the memory 106 via a communication component 108 known in the art. The first network node 100 can be used for wireless and wired communication in wireless and wired communication systems, respectively. Wireless communication capability can be provided via an antenna or antenna array 110 coupled to the transceiver 104, while wired communication capability can be provided via a wired communication interface 112 coupled to the transceiver 104.

[0118] Processor 102 may be referred to as one or more general-purpose CPUs, one or more digital signal processors (DSPs), one or more application-specific integrated circuits (ASICs), one or more field-programmable gate arrays (FPGAs), one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, or one or more chipsets. Memory 106 may be read-only memory, random access memory, or non-volatile random access memory (NVRAM). Transceiver 104 may be transceiver circuitry, a power controller, an antenna, or an interface for communicating with other modules or devices. In embodiments, transceiver 104 may be a separate chipset, or it may be integrated with a processor within a chipset. In some implementations, transceiver 104, memory 106, and processor 102 are integrated into a single chipset.

[0119] In this invention, the first network node 100 performing certain actions can be understood to mean that the first network node 100 includes appropriate components for performing said actions, such as processor 102 and transceiver 104.

[0120] According to an embodiment of the present invention, a first network node 100 is configured to transmit a first message 510 to a second network node 300, the first message 510 indicating a request for a potential QoS change of the QoS flow of a PDU session established by the first network node 100. The first network node 100 is configured to receive a second message 520 from the second network node 300, the second message 520 indicating a potential QoS change of the QoS flow. The first network node 100 is configured to determine a QoS flow prediction result (PO) based on a comparison between the potential QoS change of the QoS flow and the actual change of the QoS flow. The first network node 100 is further configured to transmit a third message 530 to a client device 610 or AF 620 associated with the PDU session, the third message 530 indicating the QoS flow prediction result (PO).

[0121] Figure 2 This can be shown at the first network node 100 (e.g. Figure 1 The flowchart illustrates the corresponding method 200 executed in the first network node (shown). Method 200 includes transmitting 202 a first message 510 to the second network node 300, the first message 510 indicating a request for a potential QoS change of the QoS flow for a PDU session established on the first network node 100. Method 200 includes receiving 204 a second message 520 from the second network node 300, the second message 520 indicating a potential QoS change of the QoS flow. Method 200 includes determining 206 a predicted QoS flow (PO) based on a comparison between the potential QoS change of the QoS flow and the actual change of the QoS flow. Method 200 includes transmitting 208 a third message 530 to the client device 610 or AF 620 associated with the PDU session, the third message 530 indicating the predicted QoS flow (PO).

[0122] Figure 3 A second network node 300 according to an embodiment of the present invention is shown. Figure 3 In the illustrated embodiment, the second network node 300 includes a processor 302, a transceiver 304, and a memory 306. The processor 302 is coupled to the transceiver 304 and the memory 306 via a communication component 308 known in the art. The second network node 300 can be used for wireless and wired communication in wireless and wired communication systems, respectively. Wireless communication capability can be provided via an antenna or antenna array 310 coupled to the transceiver 304, while wired communication capability can be provided via a wired communication interface 312 coupled to the transceiver 304.

[0123] Processor 302 may be referred to as one or more general-purpose CPUs, one or more digital signal processors (DSPs), one or more application-specific integrated circuits (ASICs), one or more field-programmable gate arrays (FPGAs), one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, or one or more chipsets. Memory 306 may be read-only memory, random access memory, or non-volatile random access memory (NVRAM). Transceiver 304 may be transceiver circuitry, a power controller, an antenna, or an interface for communicating with other modules or devices. In embodiments, transceiver 304 may be a separate chipset, or it may be integrated with a processor within a chipset. In some implementations, transceiver 304, memory 306, and processor 302 are integrated into a single chipset. In this invention, the second network node 300 performing certain actions can be understood to mean that the second network node 300 includes appropriate components for performing said actions, such as processor 302 and transceiver 304.

[0124] According to an embodiment of the present invention, a second network node 300 is configured to receive a first message 510 from a first network node 100, the first message 510 indicating a request for a potential QoS change of the QoS flow of a PDU session established with the first network node 100. The second network node 300 is further configured to transmit a second message 520 to the first network node 100, the second message 520 indicating a potential QoS change of the QoS flow and an identifier of the potential QoS change of the QoS flow.

[0125] Figure 4 This shows that it can be done at the second network node 300 (e.g. Figure 3 The flowchart illustrates the corresponding method 400 executed in the second network node shown. Method 400 includes receiving 402 a first message 510 from the first network node 100, the first message 510 indicating a request for a potential QoS change of a QoS flow for a PDU session established on the first network node 100. Method 400 also includes transmitting 404 a second message 520 to the first network node 100, the second message 520 indicating a potential QoS change of the QoS flow and an identifier of the potential QoS change of the QoS flow.

[0126] Figure 5 The signaling for a QoS notification control process according to an embodiment of the present invention is illustrated. The QoS notification control process is performed between a first network node 100, a second network node 300, and a client device 610 or AF 620. In the embodiment, the first network node 100 is a Session Management Function (SMF) or Policy Control Function (PCF), and the second network node 300 is a Network Data Analytics Function (NWDAF).

[0127] exist Figure 5 In step 1, the first network node 100 transmits a first message 510 to the second network node 300. The first message 510 indicates a request for a potential QoS change of the QoS flow of the PDU session established by the first network node 100. The PDU session can be a PDU session between client device 610 and AF 620.

[0128] First network node 100 transmits first message 510 to subscribe to notification from second network node 300 regarding potential QoS changes related to the QoS flow of the PDU session established with first network node 100. When first network node 100 is an SMF or PCF, first message 510 may be Nnwdaf_AnalyticsSubscription_Subscribe or Nnwdaf_AnalyticsInfo_Request.

[0129] In the 3GPP implementation, the first message 510 can be considered as a request to receive a prediction analysis applicable to the QoS sustainability analysis of a QoS flow, such as a prediction analysis that is as specific as possible to the QoS flow in question. The first network node 100 can issue a subscription request for the analysis, set the "analysis ID" to "QoS sustainability", and set "analysis filter information" containing QoS requirements.

[0130] In one embodiment, the first network node 100 transmits the first message 510 to the second network node 300 upon receiving the fifth message 550 from the client device 610 or the AF 620. The fifth message 550 indicates a request for a prediction result (PO) of the QoS flow. Therefore, in this embodiment, the first network node 100 can subscribe to notifications of the QoS flow when triggered by the client device 610 or the AF 620. Figure 5 In the illustrated embodiment, a fifth message 550 from client device 610 or AF 620 is indicated in step 1a.

[0131] Typically, client device 610 or AF 620 requests prediction results to provide such information to the application, allowing the application to know whether a particular QoS change had been previously predicted. In this regard, the second message 520 may also include an additional prediction reference IE containing an identifier of a potential QoS change for the QoS flow. Predicted QoS changes reduce potential harm compared to unpredictable ones because the application can implement countermeasures. Predictability can also be disclosed in the SLA and is related to the billing of 5GS services. More predictable services are better than less predictable ones.

[0132] The fifth message 550 can be transmitted as part of a PDU session establishment or PDU modification process. For example, when client device 610 establishes a PDU session for at least one QoS flow with a resource type guaranteed bit rate (GBR) or delay critical GBR, when client device 610 executes a PDU session modification request, or when AF 620 executes an AF session request for QoS of the QoS flow in question. By transmitting the fifth message 550 and thus a request for a prediction result (PO) of the QoS flow, client device 610 or AF 620 indicates that for the QoS flow, and in the event that the QoS is not fulfilled, client device 610 or AF 620 wishes to be notified whether the QoS change is based on a previous prediction.

[0133] When the first network node 100 is an SMF, the fifth message 550 can be an Nsmf_PDUSession_UpdateSMContext request, an Nsmf_PDUSession_CreateSMContext request, or an Npcf_PolicyAuthorization_Subscribe request. When the first network node 100 is a PCF, the fifth message 550 can be an SM policy association establishment or modification request, or an Npcf_PolicyAuthorization_Subscribe request.

[0134] The second network node 300 receives the data. Figure 5 In step 1, the first message 510 is transmitted from the first network node 100. Therefore, the second network node 300 receives the request indicated in the first message 510, namely, a request for a potential QoS change of the QoS flow of the PDU session established by the first network node 100. The second network node 300 is thus informed that the first network node 100 wishes to receive notification related to the potential QoS change, which is relevant to the QoS flow in question.

[0135] exist Figure 5 In step 2, the second network node 300 transmits a second message 520 to the first network node 100. The second message 520 indicates a potential QoS change in the QoS flow, and in embodiments may also indicate an identifier (ID) of the potential QoS change event for the QoS flow. The second network node 300 may transmit the second message 520 in response to an application request. For example, the AF may send Npcf_PolicyAuthorization_Subscribe because it wants to know about any events related to a specific AF session bound to the PDU session, since the AF acts as a control plane function connected to the application. Applications are typically interested in knowing whether certain changes have occurred in the AF session.

[0136] The second message 520 can indicate more than one potential QoS change event, for example, different potential QoS change events referencing different time intervals and / or different locations. Each potential QoS change event refers to a time interval and / or a location with a different identifier. For example, a potential QoS change message referencing different time intervals can contain potential QoS change event IDs. For example: "Potential QoS Change Event ID 1", for the interval from 10.00 to 10.10 and QoS parameter X = 10; "Potential QoS Change Event ID 2", for the time interval from 10.10 to 10.20 and QoS parameter X = 9; and so on.

[0137] The first network node 100 receives a second message 520 from the second network node 300, indicating a potential QoS change in the QoS flow. When the first network node 100 is an SMF or PCF, the second message 520 can be Nnwdaf_AnalyticsSubscription_Notify or Nnwdaf_AnalyticsInfo_Response.

[0138] In an embodiment where the second message 520 also indicates an identifier for a potential QoS change in the QoS flow, the first network node 100 can identify a potential QoS change in the QoS flow based on the identifier. The identifier can serve as evidence that a QoS prediction has been made in the past for the relevant prediction result. In this way, whoever receives the prediction result knows the prediction result.

[0139] exist Figure 5 In step 3, the first network node 100 determines the predicted QoS flow (PO) based on a comparison between the potential QoS changes of the QoS flow and the actual changes of the QoS flow. The first network node 100 can obtain the actual changes of the QoS flow from the radio access network (RAN) 810 in the fourth message 540, such as... Figure 5As indicated in optional step 3a. Therefore, in this embodiment, the first network node 100 can determine the predicted outcome (PO) of the QoS flow upon receiving the fourth message 540 from RAN 810, where the fourth message 540 indicates the actual change in the QoS flow. Thus, in this case, the fourth message 540 acts as a trigger. When the QoS of the QoS flow changes, for example because RAN 810 can no longer fulfill the previous QoS of the QoS flow, RAN 810 can transmit the fourth message 540. If the first network node 100 is an SMF, the fourth message 540 can be transmitted from RAN 810 to the first network node 100 via AMF; if the first network node 100 is a PCF, the fourth message can be transmitted via both AMF and SMF. When the first network node 100 is an SMF, the fourth message 540 can be an Nsmf_PDUSession_UpdateSMContext request. When the first network node 100 is a PCF, the fourth message 540 can be an Npcf_SMPolicyControl_Update request.

[0140] In an embodiment, the predicted outcome (PO) of a QoS flow can be a Boolean flag indicating whether a potential QoS change is the same as the actual change, i.e., whether the QoS change is as predicted. For example, if the potential QoS change is the same as the actual change of the QoS flow, the predicted outcome (PO) can be set to a first value, such as "true" / "1". If the potential QoS change is different from the actual change of the QoS flow, the predicted outcome can be set to a second value, such as "false" / "0".

[0141] exist Figure 5 In step 4, the first network node 100 transmits a third message 530 to the client device 610 or AF 620 associated with the PDU session. The third message 530 indicates the determination prediction result (PO) of the QoS flow. When the first network node 100 is an SMF (Small, Medium, and Large), the third message 530 can be an Npcf_SMPolicyControl_Update request or a PDU SESSIONMODIFICATION COMMAND. When the first network node 100 is a PCF (Policy Authorization), the third message 530 can be an Npcf_PolicyAuthorization_Notify or an Npcf_SMPolicyControl_Update response.

[0142] When the predicted outcome (PO) of the QoS flow is a Boolean flag, the Boolean flag can be set to "true" / "1" if the potential QoS change is the same as the actual change in the QoS flow, and to "false" / "0" if the potential QoS change is different from the actual change in the QoS flow. When the Boolean flag is set to "true" / "1", the third message 530 can also indicate an identifier of the potential QoS change in the QoS flow, which can be used by the client device 610 or AF 620 to identify the potential QoS change.

[0143] Figure 6 The signaling for a slice quality of experience (QoE) analysis process according to an embodiment of the present invention is shown. The slice QoE analysis process is performed between a first network node 100, a second network node 300, and a network function (NF) consumer 710. In this embodiment, the first network node 100 is an SMF or PCF, and the second network node 300 is an NWDAF.

[0144] exist Figure 6 In step 1, the second network node 300 transmits a seventh message 570 to the first network node 100, indicating a request for a QoS prediction result (PO). In this way, the second network node 300 can indicate to the first network node 100 that it wants to know whether a potential QoS change is the same as the actual change in the event of a QoS change. The seventh message 560 can be Nsmf_EventExposure_Subscribe.

[0145] exist Figure 6 In step 2, the first network node 100 transmits a sixth message 560 to the second network node 300. The sixth message 560 indicates the prediction result (PO) of the QoS flow. When the first network node 100 is an SMF or PCF, the sixth message 560 can be an Nsmf_EventExposure_Subscribe notification. The first network node 100 may transmit the sixth message 560 when determining the prediction result (PO) of the QoS flow based on an indication that the QoS of the QoS flow from RAN 810 has changed, as shown in the reference. Figure 5 As stated above.

[0146] The second network node 300 receives a sixth message 560 from the first network node 100, and therefore receives a QoS flow indication prediction result (PO). The prediction result (PO) can be a comparison between a potential QoS change in the QoS flow and an actual change in the QoS flow. When the QoS flow prediction result (PO) is a Boolean flag, the Boolean flag can be set to "true" if the potential QoS change is the same as the actual change in the QoS flow, and set to "false" if the potential QoS change is different from the actual change in the QoS flow. When the Boolean flag is set to "true", the sixth message can also indicate an identifier of the potential QoS change in the QoS flow, which can be used to identify the potential QoS change.

[0147] exist Figure 6 In step 3, the second network node 300 determines the key performance indicators (KPIs) and observed time intervals for the QoS flow based on the prediction results (PO) of the QoS flow. The KPIs indicate the QoE associated with the QoS flow. The second network node 300 can determine the KPIs for the QoS flow based on the ratio between the sum of time intervals where QoS changes occur and the prediction results (PO) of the QoS flow are true, and the sum of time intervals where QoS changes occur. In other words, the determined KPIs for the QoS flow indicate the ratio of performance degradation or service unavailability events that occur and are predicted / notified in advance to the total number of performance degradation or service unavailability events that occur. Therefore, the KPIs for the QoS flow can be viewed as a measure of how well the performance prediction service performs within the observed time window, with values ​​close to 1 indicating good prediction service. The KPI values ​​for the QoS flow are directly related to application unavailability. The higher the KPI value for the QoS flow, the lower the application unavailability, because the higher portion of the performance degradation can be compensated for by triggering an adaptation mode rather than switching the system to a safe mode.

[0148] The second network node 300 can share defined KPIs with one or more NF consumers. Figure 6 In the illustrated embodiment, in step 4, the second network node 300 transmits the determined KPI to the NF consumer 710. In this embodiment, the second network node 300 transmits the KPI to the NF consumer 710 upon receiving a request for the KPI from the NF consumer 710, such as... Figure 6 As indicated in optional step 1a. NF consumer 710 can request KPIs from the second network node 300 using Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Subscribe.

[0149] In the following invention, another embodiment of the invention will be presented to provide a deeper understanding of the disclosed solution. In this regard, the following embodiments are also set in a 3GPP context, and therefore the terminology, expressions, protocols, interfaces, and architectures used are also set therein. Thus, in such implementation examples, the client device corresponds to the UE. However, embodiments of the invention are not limited thereto and can be implemented in any suitable communication system. Generally, implementations of embodiments of the invention may include, but are not limited to, modifications to the following 3GPP system procedures:

[0150] 1. QoS notification control process, including UE and AF notifications, and cases of alternative QoS templates.

[0151] 2. NWDAF QoS Sustainability Process.

[0152] 3. SLS guarantee process.

[0153] 4. QoE analysis process of NWDAF slices.

[0154] 5. SMF event notification for the event "QFI allocation".

[0155] The following sections describe the details of the proposed modifications, as well as implementation examples of embodiments of the present invention.

[0156] Changes to the QoS notification control process, including UE and AF notifications.

[0157] Figure 7 The signaling related to proposed modifications to the existing QoS notification control process is illustrated. The purpose of introducing novel aspects of embodiments of the present invention is as follows:

[0158] ● To enable the 5GS to notify the AF and UE whether such QoS changes were previously predicted (in a previous analysis notification) when any QoS flow of GBR or Delay-Critical GBR type occurs, i.e., a prediction result, novel information can be added to the "asPredicted" Boolean flag IE. This Boolean flag IE can assume a value of "true" if the QoS change was previously predicted, or vice versa. "Previously predicted" means that the application has already received a QoS sustainability analysis notification applicable to the specific QoS change.

[0159] ●When QoS changes have been predicted, relevant references to such predictions are also included in the notification, i.e., prediction references.

[0160] Additional information can be added as novel information to notifications sent to the UE and / or AF for GBR and delay-critical GBR QoS flows. This information can be sent when the AF or UE requests additional novel information. When making a request to 5GS, the AF or UE can set a Boolean flag for a specific "asPredicted" request to indicate a request for such additional novel information.

[0161] Now refer to Figure 7 The process is described. The process assumes the following steps as prerequisites:

[0162] exist Figure 7 In step 1, the UE has established a PDU session for at least one QoS flow with resource type GBR or delay key GBR. As a novel feature, the UE has set the Boolean flag of the "asPredicted" request to the value "true" for GBR or delay key GBR type QoS flows to indicate that for the QoS flow, and in the event that the QoS is not fulfilled, the UE expects to be notified whether the QoS change is based on a previously predicted change. The Boolean flag of the "asPredicted" request can also be set in the context of a PDU session modification request performed by the UE, or as part of an AF session request with QoS issued by the AF for the QoS flow in question. The Boolean flag of the "asPredicted" request can be set when the SMF knows that the relevant QoS flow exists and has an assigned QoS template. The establishment of a QoS flow may be caused by several operations. Therefore, the Boolean flag of the "asPredicted" request can be set in any operation that causes the establishment of a QoS flow.

[0163] exist Figure 7 In step 1bis, after receiving a request to establish or modify a PDU session with a relevant QoS flow (where the Boolean flag of the "asPredicted" request is switched to "true" / "1"), the SMF subscribes to the NWDAF to receive a predicted analysis applicable to the QoS sustainability analysis of the QoS flow. The SMF may issue a subscription request to receive the predicted analysis, which is as specific as possible to the QoS flow in question. The SMF may also issue a subscription request for the analysis, for example, by setting the "Analysis ID" to "QoS Sustainability" and setting "Analysis Filter Information" containing at least one of the following QoS requirements:

[0164] ● In the “5QI” filter, it sets the 5QI of the QoS flow under discussion, including applicable additional QoS parameters and corresponding values, such as guaranteed flow bit rate (GFBR), maximum flow bit rate (MFBR), etc.

[0165] ● Set QoS characteristic attributes based on the relevant values ​​of the QoS flow under discussion, including resource type, packet delay budget (PDB), packet error rate (PER), and their values.

[0166] ● The "Analysis Target Period" time window is a window setting that starts from the current time and continues for an appropriate future time interval until the QoS flow is expected to run or continue. This interval depends on the application and is based on static configuration or previously collected analysis settings, such as 5 minutes, 10 minutes, 1 hour, etc.

[0167] ●Based on current location settings such as those provided by the AMF, area expansion is configured based on the locations where QoS flows are expected to operate in the near future. This area expansion depends on the application and can be determined based on static configuration or previously collected analysis, such as one or more cells.

[0168] ● Set a reporting threshold for the QoS flow persistence KPI so that an NWDAF notification is triggered whenever the predicted QoS differs from the QoS currently implemented for the QoS flow in question.

[0169] ● Set the S-NSSAI based on the single network slice selection assistance information (S-NSSAI) of the PDU session. If the PDU session references more than one S-NSSAI, the S-NSSAI that the QoS flows are expected to run in the near future must be considered.

[0170] exist Figure 7 In step 2, the AF has already subscribed to the PCF for QoS fulfillment and / or non-fulfillment events. This is done when the AF issues an NpcfPolicyAuthorization_Subscribe for the QoS flow in question. This request may also include a Boolean flag for an "asPredicted" request set to "True" to indicate that the AF has requested additional information if the QoS change is based on or differs from a previously predicted event.

[0171] The following steps describe the events and actions that occur in the network when a QoS fulfillment or non-fulfillment event of a critical GBR type QoS flow is triggered or delayed in a Next Generation RAN (NG-RAN). It is assumed that notification control is active for this type of QoS flow, and the Boolean flag for the "asPredicted" request is set to "true", as shown below. Figure 7 As described in steps 1-2.

[0172] exist Figure 7 In step 3, a QoS change occurs because NG-RAN is unable to fulfill the GBR or delay critical GBR of the QoS flow. Therefore, NG-RAN sends a notification control message PDU SESSIONRESOURCE NOTIFY (i.e., PDU session ID and N2 SM information) to the core network (CN), containing a PDU session resource notification delivery IE (i.e., QoS flow notification list IE, notification reason IE, and current QoS parameter setting index IE) for the QoS flow in question, indicating the new QoS that NG-RAN can fulfill. AMF receives this message.

[0173] exist Figure 7 In step 4, when the AMF receives a message from the NG-RAN, it sends an Nsmf_PDUSession_UpdateSMContext request (i.e., SM context ID and N2 SM information) to the SMF responsible for managing the PDU session in question. This message can be used by the AMF to inform the SMF of QoS notification control notifications and contains information about the relevant QoS flow, including new QoS that the QoS flow can fulfill. If the QoS flow has a set of alternative QoS templates, it also contains IE currentQosProfileIndex = N or nullQoSProfileIndex = true, where N is the QoS template that the NG-RAN can fulfill, or nullQoSProfileIndex = true if alternative QoS templates cannot be fulfilled.

[0174] Triggered by the message received in step 4 Figure 7 In step 5, the SMF informs the PCF of the new QoS applied to the QoS flow in question by issuing an Npcf_SMPolicyControl_Update request with additional novel information (i.e., asPredicted). The SMF can determine the additional novel information based on the notification received in step 1bis:

[0175] ● The "asPredicted" Boolean flag is set to "true" if the QoS is based on a previously predicted value, and to "false" if the QoS differs from the prediction, or if there is no prediction that can be associated with the relevant QoS change event reported in the message received in step 4.

[0176] ● "predictionReference" information containing a unique ID for the predicted event.

[0177] Add novel information for each unfulfilled event. SMF can add this information within a qncReport array containing one or more elements of QosNotificationControlInfo. Within such elements, the notifType field of QosNotifType can take the value GUARANTEED or NOT_GUARANTEED. Because novel information is added for unfulfilled events, this novel information is added for notifType fields with the value "NOT_GUARANTEED".

[0178] exist Figure 7 In step 6, if the AF has already requested the notification as described in step 2, the PCF informs the AF to issue an Npcf_PolicyAuthorization_Notify, which includes the new QoS for the QoS stream in question and as follows: Figure 7 The novel predictability information specified in step 5 of the process.

[0179] exist Figure 7 In step 7, the PCF sends an Npcf_SMPolicyControl_Update response to the SMF following the request in step 5.

[0180] exist Figure 7 In step 8, as one of the events triggered by the message received in step 4, the SMF sends a PDU SESSION MODIFICATION COMMAND with authorized QoS rule information to the UE. If the Boolean flag of the "asPredicted" request is set to "true" for the PDU session, the SMF also includes novelty information with an optional Boolean flag of "asPredicted". Such novelty IE is set to "true" if the QoS changes included in the authorized QoS rules of the PDU SESSION MODIFICATION COMMAND are based on previously delivered predictions, or set to "false" if there is no reference to a prediction related to the QoS changes in question.

[0181] exist Figure 7 In step 9, the UE responds to the PDU SESSION MODIFICATION COMMAND message received in step 8 by sending a PDU SESSION MODIFICATION COMPLETE message to the SMF and indicating acceptance of the PDU SESSION MODIFICATION COMMAND message. Alternatively, if the UE rejects the PDU SESSION MODIFICATION COMMAND, the UE can send a PDU SESSION MODIFICATION REJECT message. The rejection message does not contain any novelty or modification.

[0182] exist Figure 7 In step 10, the SMF responds to the message received by the AMF in step 4 by issuing an Nsmf_PDUSession_UpdateSMContext response, which confirms that the update of the PDU session is related to the change in the QoS flow in question.

[0183] Changes to the NWDAF QoS Sustainability Process

[0184] According to embodiments of the invention, this document also provides changes to QoS sustainability analysis. QoS sustainability analysis is used to calculate predictions of QoS within a defined location area and time window. These changes introduce a unique ID for each predicted QoS change event, intended to be used as a reference each time an AF or UE requests such information, when a notification of a QoS change has occurred.

[0185] Figure 8 The QoS sustainability analysis process of an embodiment of the present invention is illustrated.

[0186] exist Figure 8 In step 1, NF is used as an NF consumer, for example... Figure 7 The SMF subscribes to the NWDAF for QoS sustainability analysis, providing relevant analysis filter information and target analysis periods. Therefore, the NF consumer transmits either the Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Subscribe message to the NWDAF.

[0187] exist Figure 8In step 2, NWDAF receives the subscription request and begins calculating and analyzing the response. To do this, NWDAF receives user plane function (UPF) information, and in step 3, it receives information through data collection from operations, administration, and maintenance (OAM).

[0188] exist Figure 8 In step 4, NWDAF derives the requested analysis based on the data collected from OAM. As part of the analysis information provided to NF consumers, NWDAF introduces a novel unique ID for each event as part of the prediction. This unique ID is called the "prediction ID". NWDAF outputs multiple QoS sustainability analysis entries, the number of which is limited to the maximum number of objects provided as part of the analysis report information. Each entry should have a unique prediction ID. NWDAF may detect the need for notifications (potential QoS change notifications) based on comparing the analysis of the requested 5QI with the reporting threshold provided by consumers in any cell within the requested analysis target period, or add entries to the notification in such cases. Table 2 below describes the IE of QoS sustainability analysis, including the novel "prediction ID".

[0189] QoS Sustainability Analysis List (1...Maximum) Predict Id A unique ID used to identify the prediction Applicable Areas A list of TAIs or cell IDs within the location information analyzed by the application. Applicable Time Period The time period within the target analysis period of the application analysis. Cross-reporting threshold The statistical or expected values ​​of QoS KPIs meet or exceed the reporting thresholds. Confidence level Confidence level of the prediction.

[0190] Table 2

[0191] For each analysis and each cross-reporting threshold included in the notification, a unique ID is introduced. This unique ID can be used later to refer to the prediction, for example, in a later QoS change notification. Information about the timing and manner of the predicted QoS changes is provided.

[0192] Finally, Figure 8 In step 5, NWDAF responds to the NF consumer by transmitting an Nnwdaf_AnalyticsInfo_Response or Nnwdaf_AnalyticsSubscription_Notify message to the NF consumer.

[0193] Changes to the SLS assurance process

[0194] According to embodiments of the present invention, the SLS guarantee process is enhanced to include 5GS performance predictability. The SLS guarantee process is used for performance guarantees of network slice instances (NSI) or network slice subnet instances (NSSI), such as... Figure 9 As shown.

[0195] exist Figure 9 In step 1, the NSMS_Consumer is authorized to request the NSMS_Producer to allocate a new NSI. The NSMS_Producer uses the deployment services provided by the NSSMS_Producer to create the NSI. After the NSI is allocated, the NSMS_Producer and NSSMS_Producer perform NSI and NSSI performance monitoring.

[0196] After step 1, a loop is executed, including steps 2 and 3 and optional steps 4a and 4b, which are described in the following invention.

[0197] exist Figure 9 In step 2, either NSMS_Producer or NSSMS_Producer can obtain the slice QoE analysis provided by NWDAF. Additionally, 5GS performance predictability is added as a relevant metric related to the performance of QoS sustainability analysis (as a service of NWDAF), because optimizing such analysis and using predictions to implement appropriate application countermeasures can help reduce application unavailability.

[0198] exist Figure 9 In step 3, NSMS_Producer or NSSMS_Producer checks whether NSI or NSSI meets the performance requirements by utilizing the end-to-end KPIs, performance measurements, and slice QoE analysis provided by NWDAF.

[0199] exist Figure 9 In step 4a, if the NSI performance requirements cannot be met, NSMS_Producer triggers the NSI modification process. NSMS_Producer modifies the network slice capacity or network slice configuration to ensure performance requirements are met.

[0200] exist Figure 9 In step 4b, if the NSSI performance requirements cannot be met, the CN's NSSMS_Producer modifies the configuration of virtualized resources and the 5G core network (5GC) NF to ensure performance requirements. The access network's NSSMS_Producer reconfigures the RRMPolicy to optimize performance.

[0201] This SLS assurance process covers slice-level performance parameters. NWDAF reports the fulfillment of measured slice-level KPIs based on these parameters. See details below. Figure 10Furthermore, the Operations Support System (OSS) adjusts the NSI parameters whenever any observed parameters deviate from the promised SLS. For example, parameters such as allocated radio resources, virtual network function (VNF) instances, central processor unit (CPU) power, and memory can be adjusted whenever the service level is not "as needed."

[0202] Changes to the QoE analysis procedure for NWDAF slices

[0203] Figure 10 The NWDAF slice QoE analysis process according to an embodiment of the present invention is illustrated. The NWDAF slice QoE analysis process has been modified, introducing the following main changes:

[0204] ● When requesting event notification for event ID "QFI Allocation" from the SMF, the NWDAF can switch the "asPredictedRequested" boolean flag to "true" to indicate whether the NWDAF is also requesting information on whether the change is based on predictions in the event of a QoS change. This functionality requires modification of the trigger for event ID "QFI Allocation" as described in the next bullet point.

[0205] ● You can also trigger the event ID "QFI allocation" for an additional event, which is an event that changes the QoS flow template due to RAN fulfillment or non-fulfillment of an event.

[0206] ●The information included in Nsmf_EventExposure_Notify may include, along with information about QFI allocation, whether the QoS change event was based on a previous prediction, and a reference to such predicted events, i.e., the relevant unique ID.

[0207] ● When responding to NF consumers, NWDAF can calculate the end-to-end KPI "5GS Performance Predictability" and include the new analytics in the Nnwdaf_AnalyticsInfo_Request response or Nnwdaf_AnalyticsSubscription_Notify.

[0208] exist Figure 10In step 1, the NF consumer sends an analytics request / subscription message to the NWDAF, which sends the analytics ID (service experience), the target of the analytics report (any UE), the analytics filter information (application ID), the analytics target period (S-NSSAI), the DNN, and the region of interest to the NWDAF by calling Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Subscribe. The NF consumer can be an NSSMS_Producer, such as... Figure 9 The description.

[0209] exist Figure 10 In step 2a, the NWDAF collects input data from the relevant NF by invoking Nnef_EventExposure_Subscribe (if the NF is an AF) or Naf_EventExposure_Subscribe (if it is a generic 5GC NF). In the proposed solution, one applicable NF is the SMF, and the service invoked is an Nsmf_EventExposure_Subscribe request with event ID = QFI allocation. Assuming that each time the SMF receives QoS notification control for a GBR-type QoS flow (a novel modification introduced in this IPR), the SMF also triggers an event ID = QFI allocation for an NG-RAN fulfilled / unfulfilled event, the NWDAF can add a specific novel element to this Nsmf_EventExposure_Subscribe request to signal that the SMF also requests prediction results (e.g., an asPredicted flag).

[0210] exist Figure 10 In step 2b, the SMF acknowledges the request by sending an Nsmf_EventExposure_Subscribe response to the NWDAF. When the SMF receives the notification control message and the event ID = QFI allocation is triggered, the SMF can send an Nsmf_EventExposure_Subscribe notification as in step 2c, but with the addition of an "asPredicted" boolean flag. This flag switches to "true" if the QoS change is as predicted or if an applicable prediction exists; otherwise, it switches to "false" if the QoS change differs from the prediction in Nsmf_EventExposure_Notify in step 2c. If "asPredicted" is set to "true", the SMF can also provide a reference to the prediction by incorporating a unique prediction ID into the notification / response.

[0211] Compared to the current process in Clause 6.4.5 of TS23.288, Figure 10Steps 3a to 3d in the present invention remain unchanged.

[0212] exist Figure 10 In step 4, according to the proposed implementation, NWDAF also performs a predictability analysis of requests to derive network slices by computing end-to-end 5GS.

[0213] exist Figure 10 In step 5, NWDAF provides data analytics to the NF consumer, i.e., the observed service experience, which can be a series of values. NWDAF provides data analytics to the NF consumer via an Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Notify response. The appropriate method can be selected based on the service used in step 1, indicating the extent to which the QoS parameters used satisfy the mean opinion score (MoS) agreed upon between the mobile network operator (MNO) and the end user, or between the MNO and the external application service provider (ASP).

[0214] If the NF consumer is a PCF and it is determined that the application SLA is not met, the NF consumer can consider the observed service experience and the operator's policies, which include the SLA and the required service experience, and these policies can be used to determine the range of new QoS parameters to be applied to the service.

[0215] Changes to the procedures related to SMF event notifications assigned by the event QFI

[0216] According to an embodiment of the invention, the possibility of triggering the event ID "QFI allocation" also for NG-RAN fulfillment / non-fulfillment events is introduced. In this case, the following changes are introduced to the process related to the SMF event notification of the event "QFI allocation", such as... Figure 11 As shown.

[0217] exist Figure 11 In step 1, the SMF sends a QFI allocation event notification to the NF service consumer, including the following information:

[0218] a) The QFI for the QoS flow ID assigned by the application, as the "qfi" attribute;

[0219] b) The data network name (DNN) of the assigned PDU session, as the "dnn" attribute;

[0220] c) A slice of the assigned PDU session, as the "snssai" attribute;

[0221] d) Description of application traffic, as an attribute of “appId”, “fDescs”, or “ethfDescs”.

[0222] The following novel information is proposed to be added to messages provided by SMF to NF service consumers (e.g., NWDAF):

[0223] e) If a QFI allocation event is triggered under conditions of QoS fulfillment / non-fulfillment and the QoS is as predicted, the "asPredicted" boolean flag is set to "1", or if a QFI allocation event is triggered under conditions of QoS fulfillment / non-fulfillment and the QoS is different from the prediction, the "asPredicted" boolean flag is set to "0".

[0224] f) Refer to “Prediction ID”, which refers to the relevant prediction ID set by NWDAF in QoS sustainability analysis.

[0225] Adding this new information, along with the event ID "QFI Assignment" triggered for each QoS fulfillment / non-fulfillment notification, introduces the possibility of notifying the NWDAF of every QoS change starting from the initial QoS and reverting to the initial QoS of the QoS flow after each QoS change. This notification provides information on whether a QoS change was predicted. Based on this information, the NWDAF can calculate the 5GS performance predictability KPI for a specific QoS flow according to the formula:

[0226]

[0227] In other words, the KPI for a QoS flow is determined based on the ratio between the sum of time intervals in which QoS changes occur and the prediction result of the QoS flow is true, and the sum of time intervals in which QoS changes occur. NWDAF can calculate the KPI globally for the entire 5GS, a specific 5QI, or a set of 5QIs.

[0228] Other implementation methods

[0229] In other implementations of this invention, most of the novel logic can be added to the SMF, as referenced in [reference]. Figure 12 The diagram illustrates a flowchart of a QoS sustainability analysis subscription from SMF triggered by a PDU session establishment, modification, or release event.

[0230] In step 1, the SMF detects a PDU session establishment, modification, or release event that causes the allocation or modification of a GBR or delay of a critical GBR type QoS flow. This triggers the SMF to check the following conditions in steps 2-4 to determine whether to subscribe to QoS sustainability analytics:

[0231] ●Whether NotifControl was set to True in step 2 for the QoS flow.

[0232] ● In step 3, whether the UE or AF involved sets the Boolean flag of the "asPredicted" request to "true".

[0233] ● As a result of PDU session establishment or modification, is the QoS flow allocated in step 4, i.e., is the check result in step 4 "yes"?

[0234] If all three conditions are true, the SMF node subscribes to the QoS sustainability analysis in step 5, thereby subscribing to the analysis that receives predictions of the QoS flows in question from the NWDAF. The SMF can subscribe to the QoS sustainability analysis with the following parameters: 5QI = QFI, QoS characteristic attributes including resource type, PDB, PER and their values, location = characteristics of the QoS flow, S-NSSAI = S-NSSAI of the PDU session, [optional maximum number of objects], analysis target period = 10-15 minutes from now to the future, reporting threshold = based on an alternative QoS template (if available) or based on characteristics of the QoS flow (e.g., MFBR, GFBR), notification relevance ID, notification target address, etc.

[0235] If the SMF detects a PDU session establishment, modification, or release event in step 1, and this event results in the release or deallocation of a GBR or a delay of a critical GBR type QoS flow, then the SMF checks the following conditions in steps 2-4 to determine whether to unsubscribe from the QoS sustainability analysis:

[0236] ●Whether NotifControl was set to True in step 2 for the QoS flow.

[0237] ● In step 3, whether the UE or AF involved sets the Boolean flag of the "asPredicted" request to "true".

[0238] ● As a result of PDU session establishment or modification, whether the QoS flow is released or deallocated in step 4, i.e., whether the check result in step 4 is "No".

[0239] If all three conditions are true, the SMF node unsubscribes from the QoS sustainability analysis in step 6. In this way, the SMF can unsubscribe each time a QoS flow is deallocated.

[0240] NWDAF detects the need for notifications regarding potential QoS changes by comparing the expected value of the target 5QI KPI with the reporting thresholds provided by consumers in any cell within the requested area during the requested analysis target period. The expected KPI value is derived from 5QI statistics obtained from OAM. OAM information may also include planned or unplanned outage detections and other information not covered in detail by 3GPP. The analysis feedback contains information about where and when potential QoS changes may occur and which reporting thresholds may be exceeded. The actions taken by SMF upon receiving an NWDAF notification will be detailed in the next paragraph.

[0241] According to embodiments of the present invention, novel logic is introduced into the SMF, allowing the NF to calculate the value of the "asPredicted" Boolean flag each time the NG-RAN receives a QoS notification control message, and notify the UE and / or AF accordingly. If such novel information is requested, a notification may be sent to the UE and / or AF, i.e., the "asPredicted" Boolean flag is set to "true" for either the UE or the AF.

[0242] Figure 13 This illustration shows a QoS notification control triggered by a QoS sustainability analysis subscription from the SMF in an embodiment of the present invention. When a QoS notification control event is received from the NG-RAN via the AMF in step 1, the SMF checks whether each QoSFlowNotifyItem contained in the QoSFlowsNotifyList relates to an unfulfilled notification reason in step 2. If they relate to an unfulfilled event, i.e., the check result in step 2 is "yes", the SMF checks whether the notification contains a reference to an alternative QoS template that fulfills the original QoS template in step 3, i.e., nullQoSProfileIndex = false. If the check result in step 3 is "yes", the SMF compares the QoS contained in the new mandatory QoS template with the QoS predicted for the QoS in question in step 4, based on location, time interval, and other information included in the subscription. If the check result in step 3 is "no", the "asPredicted" Boolean flag is set to "false" in step 5.

[0243] If the new forced QoS template matches the predicted QoS template, i.e., the check result in step 4 is "yes", then the "asPredicted" boolean flag is set to "true" in step 6. If no applicable prediction is found, i.e., the check result in step 4 is "no", then the asPredicted boolean flag is set to "false" in step 5.

[0244] Whenever a new prediction for a managed QoS flow is received, the SMF can store the prediction internally as part of the local PDU context and save it for later use in the event that a QoS non-fulfillment event is received for a QoS flow requesting "asPredicted".

[0245] Therefore, according to Figure 14 In the embodiments of the present invention shown, the SMF can use information received from the NWDAF regarding QoS sustainability analysis notifications to internally store predictions for specific QoS flows applicable to GBR or delay-critical GBR types within a future analysis target period. Each time the SMF receives a QoS sustainability analysis notification, it can determine whether it applies to the QoS flow in question. If the prediction report forecasts a potential QoS change, and if a QoS non-fulfillment notification control message is also received, the SMF can determine whether a QoS change has been predicted. If the QoS change has been predicted, the SMF can set the "asPredicted" boolean flag to "true," or if the QoS change was not predicted in an event ID "QoS Allocation" notification, a non-access stratum (NAS) notification to the UE, or a notification to the PCF and AF, the SMF can set the "asPredicted" boolean flag to "false." If "asPredicted" is set to "true," the prediction reference IE can be set according to the value of the prediction ID contained in the applicable analysis entry in the QoS sustainability analysis notification.

[0246] For SMF events, the SMF can trigger an event ID "QFI allocation" when a QoS notification control occurs, adding an "asPredicted" boolean flag and a "predicted reference" (if applicable to event notifications provided to NF consumers). In this implementation, the NF consumer can be an NWDAF.

[0247] Alternative implementations can be based on the definition of a new SMF event that can be triggered when QoS notification control occurs, including the same IE as the above implementations, namely the "asPredicted" boolean flag and "predicted reference" (if applicable). In such cases, NWDAF is configured as an NF consumer.

[0248] The first network node 100 here can be identified as an SMF (Mobile Management Entity (MME) in 5G Long Term Evolution (LTE) and Long Term Evolution Advanced (LTE-A)) or a PCF (PCRF in LTE and LTE-A). The second network node 300 here can be identified as an NWDAF in 5G New Radio (NR). The NWDAF can be a general-purpose function / node used for communication in 3GPP-related LTE and LTE-Advanced.

[0249] Furthermore, any method provided according to embodiments of the present invention can be implemented in a computer program having code components, which, when run by a processing component, causes the processing component to perform method steps. The computer program is included in a computer-readable medium of the computer program product. The computer-readable medium can substantially include any memory, such as read-only memory (ROM), programmable read-only memory (PROM), erasable PROM (EPROM), flash memory, electrically erasable EPROM (EEPROM), or hard disk drive.

[0250] Furthermore, those skilled in the art will recognize that embodiments of the first network node 100 and the second network node 300 include the necessary communication capabilities for performing the solution, in the form of, for example, functions, components, units, elements, etc. Examples of other such components, units, elements, and functions include: processors, memories, buffers, control logic, encoders, decoders, rate matchers, derate matchers, mapping units, multipliers, decision units, selection units, switches, interleavers, deinterleavers, modulators, demodulators, inputs, outputs, antennas, amplifiers, receiver units, transmitter units, DSPs, MSDs, TCM encoders, TCM decoders, power supply units, power feeders, communication interfaces, communication protocols, etc., which are suitably arranged together to perform the solution.

[0251] Specifically, the processors of the first network node 100 and the second network node 300 may include, for example, a central processing unit (CPU), a processing unit, processing circuitry, a processor, an application-specific integrated circuit (ASIC), a microprocessor, or one or more instances of other processing logic capable of interpreting and executing instructions. Therefore, the expression "processor" can refer to a processing circuitry system that includes multiple processing circuits, such as any, some, or all of the aforementioned processing circuits. The processing circuitry system can also perform data processing functions for inputting, outputting, and processing data, including data buffering and device control functions such as call processing control, user interface control, etc.

[0252] Finally, it should be understood that the present invention is not limited to the above embodiments, but also relates to and incorporates all embodiments within the scope of the appended independent claims.

Claims

1. A first network node (100) for a communication system (500), characterized in that, The first network node (100) is used for: A first message (510) is transmitted to the second network node (300), the first message (510) indicating a request for a potential QoS change of the Quality of Service (QoS) stream of the Protocol Data Unit (PDU) session established by the first network node (100); Receive a second message (520) from the second network node (300), the second message (520) indicating the potential QoS change of the QoS flow; The prediction result PO of the QoS flow is determined by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow; A third message (530) is transmitted to the client device (610) or application function AF (620) associated with the PDU session, the third message (530) indicating the prediction result PO of the QoS flow; Upon receiving a fourth message (540) from the radio access network (RAN) (810), the predicted result PO of the QoS flow is determined, the fourth message (540) indicating the actual change of the QoS flow.

2. The first network node (100) according to claim 1, characterized in that, The first network node (100) is used for: Upon receiving a fifth message (550) from the client device (610) or the AF (620), the first message (510) is transmitted to the second network node (300), the fifth message (550) indicating a request for the prediction result PO of the QoS flow.

3. The first network node (100) according to claim 1, characterized in that, The prediction result PO for the QoS flow is a Boolean flag.

4. The first network node (100) according to claim 1, characterized in that, The second message (520) also indicates the identifier ID of the potential QoS change of the QoS flow, wherein the first network node (100) is used for: The potential QoS changes of the QoS flow are identified based on the identifier ID.

5. The first network node (100) according to claim 2, characterized in that, The first network node (100) is used for: A sixth message (560) is transmitted to the second network node (300), the sixth message (560) indicating the prediction result PO of the QoS flow.

6. The first network node (100) according to claim 5, characterized in that, The first network node (100) is a Session Management Function (SMF) or a Policy Control Function (PCF), wherein at least one of the following is true: The first message (510) is either Nnwdaf_AnalyticsSubscription_Subscribe or Nnwdaf_AnalyticsInfo_Request; The second message (520) is either Nnwdaf_AnalyticsSubscription_Notify or Nnwdaf_AnalyticsInfo_Response; When the first network node (100) is an SMF, the third message (530) is an Npcf_SMPolicyControl_Update request or a PDU SESSION MODIFICATION COMMAND; when the first network node (100) is a PCF, the third message (530) is an Npcf_PolicyAuthorization_Notify or an Npcf_SMPolicyControl_Update response. The fourth message (540) is an Nsmf_PDUSession_UpdateSMContext request when the first network node (100) is an SMF, or an Npcf_SMPolicyControl_Update request when the first network node (100) is a PCF; When the first network node (100) is an SMF, the fifth message (550) is an Nsmf_PDUSession_UpdateSMContext request, an Nsmf_PDUSession_CreateSMContext request, or an Npcf_PolicyAuthorization_Subscribe request; when the first network node (100) is a PCF, the fifth message (550) is an SM policy association establishment or modification or an Npcf_PolicyAuthorization_Subscribe request. The sixth message (560) is the Nsmf_EventExposure_Subscribe notification.

7. A second network node (300) for a communication system (500), characterized in that, The second network node (300) is used for: Receive a first message (510) from the first network node (100), the first message (510) indicating a request for a potential QoS change of the Quality of Service (QoS) stream of the Protocol Data Unit (PDU) session established by the first network node (100); A second message (520) is transmitted to the first network node (100), the second message (520) indicating the potential QoS change of the QoS flow and the identifier ID of the potential QoS change of the QoS flow; The first network node (100) is used for: The prediction result PO of the QoS flow is determined by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow; A third message (530) is transmitted to the client device (610) or application function AF (620) associated with the PDU session, the third message (530) indicating the prediction result PO of the QoS flow; Upon receiving a fourth message (540) from the radio access network (RAN) (810), the predicted result PO of the QoS flow is determined, the fourth message (540) indicating the actual change of the QoS flow.

8. The second network node (300) according to claim 7, characterized in that, The second network node (300) is used for: A seventh message (570) is transmitted to the first network node, the seventh message (570) indicating a request for the prediction result PO of the QoS flow; A sixth message (560) is received from the first network node (100), the sixth message (560) indicating the prediction result PO of the QoS flow.

9. The second network node (300) according to claim 8, characterized in that, The second network node (300) is used for: Based on the prediction results PO of the QoS flow, the key performance indicators (KPIs) and observed time intervals of the QoS flow are determined.

10. The second network node (300) according to claim 9, characterized in that, The second network node (300) is used for: The KPI of the QoS flow is determined based on the ratio between the sum of the time intervals in which the QoS flow changes and the predicted PO of the QoS flow is true, and the sum of the time intervals in which the QoS change occurs.

11. The second network node (300) according to claim 9 or 10, characterized in that, The second network node (300) is used for: When a request for the KPI is received from a network function NF consumer (710), the KPI is transmitted to the NF consumer (710).

12. A method (200) for a first network node (100), characterized in that, The method (200) includes: Transmit (202) a first message (510) to the second network node (300), the first message (510) indicating a request for a potential QoS change of the Quality of Service (QoS) stream of the Protocol Data Unit (PDU) session established by the first network node (100); Receive (204) a second message (520) from the second network node (300), the second message (520) indicating the potential QoS change of the QoS flow; The prediction result PO of the QoS flow is determined (206) by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow; Transmit (208) a third message (530) to the client device (610) or application function AF (620) associated with the PDU session, the third message (530) indicating the prediction result PO of the QoS flow; Upon receiving a fourth message (540) from the radio access network (RAN) (810), the predicted result PO of the QoS flow is determined, the fourth message (540) indicating the actual change of the QoS flow.

13. A method (400) for a second network node (300), characterized in that, The method (400) includes: Receive (402) a first message (510) from the first network node (100), the first message (510) indicating a request for a potential QoS change of the Quality of Service (QoS) stream of the Protocol Data Unit (PDU) session established by the first network node (100); Transmit (404) a second message (520) to the first network node (100), the second message (520) indicating the potential QoS change of the QoS flow and the identifier ID of the potential QoS change of the QoS flow; The first network node (100) is used for: The prediction result PO of the QoS flow is determined by comparing the potential QoS changes of the QoS flow with the actual changes of the QoS flow; A third message (530) is transmitted to the client device (610) or application function AF (620) associated with the PDU session, the third message (530) indicating the prediction result PO of the QoS flow; Upon receiving a fourth message (540) from the radio access network (RAN) (810), the predicted result PO of the QoS flow is determined, the fourth message (540) indicating the actual change of the QoS flow.

14. A computer program product comprising computer-executable instructions, characterized in that, When the computer-executable instructions are run on the processor, the method according to claim 12 or 13 is performed.

Images