Apparatus, methods and computer programs

By introducing prompt classification and priority sorting functions on the user device and network sides, the problems of improper handling of prompt modalities and criticality in the existing technology are solved, and the GenAI agent achieves efficient resource scheduling and task response optimization in mobile networks.

CN123075596APending Publication Date: 2026-11-10NOKIA TECHNOLOGIES OY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610600801.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2025-05-09
Filing Date
2026-04-30
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing cue creation and manipulation mechanisms fail to effectively handle different modalities and criticalities of cues in mobile networks, resulting in poor applicability and resource scheduling efficiency of the GenAI agent, especially in critical tasks.

Method used

By introducing Prompt Classification and Prioritization Function (PCPF) on the user equipment (UE) and network sides, generative AI capabilities are used to classify prompts modally and critically. Service routing strategies are determined based on classification identifiers, enabling effective classification and prioritization of GenAI proxy prompts, and resource scheduling is performed through PDU sessions or network slicing.

Benefits of technology

It improves the processing efficiency and resource utilization of the GenAI agent in mobile networks, ensures timely response to critical tasks and reasonable allocation of non-critical tasks, and optimizes the use of network resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123075596A_ABST
    Figure CN123075596A_ABST
Patent Text Reader

Abstract

Various embodiments of this disclosure relate to apparatuses, methods, and computer programs. An apparatus is provided comprising: at least one processor; and at least one memory storing instructions, which, when executed by the at least one processor, cause the apparatus to at least: obtain a prompt from an application of a user equipment; determine at least one characteristic associated with the prompt; determine a classification identifier associated with the prompt based on the at least one characteristic; determine, for the prompt, a description of a service route for providing the prompt to a network based on the classification identifier and from a service routing policy; and provide the prompt to the network according to the description of the service route.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Various embodiments of this disclosure generally relate to methods, apparatus, and computer programs, and specifically, but not exclusively, to LLM-aware UE prompt classification, prioritization, and processing performed by a 6G system. Background Technology

[0002] A communication system can be viewed as a facility that enables communication sessions between two or more communication devices or provides communication devices with access to a network. Mobile or wireless communication networks are an example of communication networks. Application servers can provide services to communication devices.

[0003] Mobile or wireless communication networks can operate according to standards, such as those provided by 3GPP (3rd Generation Partnership Project) or ETSI (European Telecommunications Standards Institute). Examples of mobile or wireless communication networks operating according to 3GPP standards are often referred to as 4G (fourth generation) networks, 5G (fifth generation) networks, 5G Advanced networks, and 6G networks. Summary of the Invention

[0004] Some embodiments of this disclosure will be described with respect to certain aspects. These aspects are not intended to indicate key or essential features of the various exemplary embodiments of this disclosure, nor are they intended to limit its scope. Other features, aspects, and elements will be apparent to those skilled in the art in light of this disclosure. For example, it should be understood that additional aspects may be provided by any combination of any two or more of the aspects described herein.

[0005] In a first aspect, a method for a user equipment is provided, the method comprising: obtaining a prompt from an application of the user equipment; determining at least one characteristic associated with the prompt; determining a classification identifier associated with the prompt based on the at least one characteristic; determining, for the prompt, a description of a service route for providing the prompt to a network based on the classification identifier and from a service routing policy; and providing a prompt to the network based on the description of the service route.

[0006] Applications may include generative artificial intelligence (AI), applications, or generative AI agents.

[0007] The description of the service route can identify at least one of the Packet Data Unit (PDU) sessions or network slices used to provide prompts to the network.

[0008] At least one characteristic may include at least one of the following: the input modality of the prompt, the output modality of the prompt, the number of prompts including the prompt obtained by the user device over a period of time, or the criticality of the prompt.

[0009] Providing hints to the network based on the description of the service route may include placing the hints in a queue associated with a classification identifier, which is linked to the hint.

[0010] Providing information to the network based on a description of a service route may include providing information to the network via a Packet Data Unit (PDU) session, which is identified by the description of the service route.

[0011] Providing hints to the network based on the description of the service route may include providing hints to the network via network slices, which are identified by the description of the service route.

[0012] In a second aspect, an apparatus is provided, comprising components for performing the method according to the first aspect.

[0013] In a third aspect, an apparatus is provided, comprising at least one processor and at least one memory storing instructions, wherein the instructions, when executed by the at least one processor, cause the apparatus to at least perform the method according to the first aspect.

[0014] In a fourth aspect, a non-transitory computer-readable medium comprising instructions is provided, wherein the instructions, when executed by at least one processor of the device, cause the device to perform the method according to the first aspect.

[0015] In a fifth aspect, a computer program including instructions is provided, which, when executed by a device, cause the device to perform at least the method according to the first aspect.

[0016] Some embodiments of the present invention are defined in the dependent claims.

[0017] Many different aspects have been described above. As previously stated, it should be understood that additional aspects may be provided by combination of any two or more of the above-described aspects (or otherwise in this disclosure).

[0018] Various other aspects are also described in the following detailed description and claims. Attached Figure Description

[0019] Some embodiments will be described by way of non-limiting and illustrative example only with reference to the accompanying drawings, wherein: Figure 1 A schematic diagram of an example communication network is shown; Figure 2 A schematic diagram of an example core network of a communication network is shown; Figure 3 A schematic diagram of the example device is shown; Figure 4 A flowchart based on the example method is shown; Figure 5An example schematic architecture of UE-side PCPF and network-side PCPF is shown; Figure 6 A flowchart based on the example method is shown; Figure 7 A flowchart based on the example method is shown; Figure 8 A flowchart based on the example method is shown; Figure 9 An example call flow is shown; Figure 10 A schematic diagram of an example device is shown. Detailed Implementation

[0020] The following embodiments are provided by way of non-limiting and illustrative example. Although the specification may refer to "a," "an," or "some" embodiments in several places in the text, this does not necessarily mean that the same embodiment is referred to every time, or that a particular feature applies only to a single embodiment. Individual features of different embodiments may also be combined to provide other embodiments. Furthermore, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is intended that such a feature, structure, or characteristic may be applied in conjunction with other embodiments, whether explicitly described or not.

[0021] It should be understood that although the terms "first," "second," etc., may be used in this document to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another.

[0022] For the purposes of this disclosure, the phrases "at least one of A or B", "at least one of A and B", and "A and / or B" mean (A), (B), or (A and B). For the purposes of this disclosure, the phrases "A, B, and / or C" mean (A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C).

[0023] As used herein, the term “or” means non-exclusive “or” unless otherwise stated (e.g., “otherwise” or “or in alternatives”).

[0024] As used herein, unless explicitly stated otherwise, performing a feature, step, or function in response to A does not indicate that the feature, step, or function is performed immediately after A occurs, because one or more intermediate features, steps, or functions may be performed (at least partially) between the occurrence of the feature, step, or function and A. Similarly, performing a feature, step, or function based on A does not indicate that the feature, step, or function is performed solely based on A, because the feature, step, or function may also be based on one or more other features, steps, or functions besides A.

[0025] The embodiments described herein can be implemented in communication networks such as any of the following radio access technologies (RATs): Global Microwave Access Interoperability (WiMAX), Global System for Mobile Communications (GSM, 2G), GSM EDGE Radio Access Network (GERAN), General Packet Radio Service (GRPS), Universal Mobile Telecommunications System based on Basic Wideband Code Division Multiple Access (W-CDMA) (UMTS, 3G), High-Speed ​​Packet Access (HSPA), Long Term Evolution (LTE), Advanced LTE and Enhanced LTE (eLTE), 5G (also known as NR), or any future RAT (such as 6G). Furthermore, communication within the communication network can utilize any suitable wireless communication technology, including but not limited to: Code Division Multiple Access (CDMA), Frequency Division Multiple Access (FDMA), Time Division Multiple Access (TDMA), Frequency Division Duplex (FDD), Time Division Duplex (TDD), Multiple-Input Multiple-Output (MIMO), Orthogonal Frequency Division Multiple Access (OFDM), and / or Discrete Fourier Transform Extended OFDM (DFT-s-OFDM).

[0026] As used herein, the term "network device" or "network node" refers to a node in a communications network through which user equipment can access the network and / or be configured to control radio communications and manage radio resources within a cell. A network node or network device may be referred to as a base station (BS), access point (AP), or access node. Depending on the technology applied, a network device may be, for example, a Node B (Node B or NB), an evolved Node B (eNodeB or eNB), an NR NB (also known as a gNB), a Remote Radio Unit (RRU), a Radio Head (RH), a Remote Radio Head (RRH), a relay, an Integrated Access and Backhaul (IAB) node, a low-power node, a non-terrestrial network (NTN) or non-land network device (such as satellite network equipment, low Earth orbit (LEO) satellites, and geostationary Earth orbit (GEO) satellites), or an aircraft network device.

[0027] Furthermore, in the context of split radio access networks (RAN), network equipment can refer to a centralized unit (CU) and / or a distributed unit (DU) of a base station. The interface between the CU and DU may be referred to as the F1 interface in NR. In a split RAN architecture, node operations can be performed at least partially operatively coupled to the central / centralized unit (CU) (e.g., server, host, or node) of the DU (e.g., radio head / node). A CU can control one or more DUs, at least acting as a transmit / receive (Tx / Rx) node. In some embodiments, a DU may include, for example, a radio link control (RLC), a media access control (MAC) layer, and a physical (PHY) layer, while a CU may include layers above the RLC layer, such as the packet data convergence protocol (PDCP) layer, radio resource control (RRC), and internet protocol (IP) layer. Other functional splitting is also possible. In practice, any processing task can be performed in a CU or DU, and the boundaries of responsibility transfer between the CU and DU may depend on the applied implementation.

[0028] The term "terminal device" refers to any terminal device that can be configured to perform wireless communication. For example, a terminal device can be referred to as a communication device, user equipment (UE), subscriber station (SS), or mobile station (MS). Terminal devices can include mobile phones, cellular phones, smartphones, VoIP phones, wireless local loop phones, tablets, wearable terminal devices, personal digital assistants (PDAs), portable computers, desktop computers, image capture terminal devices such as digital cameras, gaming terminal devices, music storage and playback devices, in-vehicle wireless terminal devices, USB dongles, Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMDs), vehicles, drones, medical devices and applications (e.g., remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in industrial and / or automated processing chain environments), consumer electronics devices, devices operating on commercial and / or industrial wireless networks, etc.

[0029] As used herein, the term "resource" can refer to radio resources in the time domain, frequency domain, spatial domain, and / or code domain. Some examples of resources may include, for example, physical resource blocks (PRBs), radio frames, subframes, time slots, subbands, frequency regions, subcarriers, beams, etc. The terms "transmit" and / or "receive" can refer to wireless transmission and / or reception over radio resources via a wireless propagation channel.

[0030] refer to Figure 1An example of a user equipment (UE) 100 configured to communicate with a data network 102 via a communication network 104 is shown. The communication network 104 includes an access network 106 and a core network 108 including one or more network functions 110.

[0031] Access network 106 provides wireless connectivity (e.g., radio connectivity) to UE 100 and connects UE 100 to core network 108 via a backhaul network. Access network 106 may include a radio access network (RAN), a non-terrestrial network (e.g., a satellite network), a wireless local area network (WLAN), or any other type of network that provides wireless connectivity to UE 100 and connects UE 100 to core network 108. Access network 106 provides wireless connectivity (e.g., radio connectivity) to UE 100 via at least one access node of access network 106. For example, the access node of access network 106 may be a radio access node and may provide radio connectivity to UE 102. Radio access nodes may include gNodeB (gNB), ng-eNodeB (ng-eNB), and eNodeB (eNB). The access node of access network 106 may be an access point, such as a wireless local area network (WLAN) access point, and may provide wireless connectivity to UE 100. In some implementations, the access node of access network 106 may be an access point that provides a wired connection (typically referred to as fixed access) to UE 100. The backhaul network between access network 106 and core network 108 may include a fixed network, a satellite network, or a combination thereof. Core network 108 connects access network 106 to data network (DN) 102 via its N6 interface (not shown). As described above, the core network includes network functions (NFs) 110 (typically referred to as network functions 110 and collectively as network functions 110). Each corresponding network function 110 may be implemented as software running on dedicated hardware (e.g., one or more physical computing devices, such as servers), a virtualized network function (VNF) instantiated on a physical or virtual machine, or a container provided by the infrastructure of a cloud computing system. Data network 102 may be an external public network, a private data network, or an intranet data network (e.g., for providing IP Multimedia Subsystem (IMS) services). UE 100 (also referred to herein as a mobile terminal) can be configured to access access network 106, register with core network 108, establish one or more data sessions with core network 108, and access services provided by core network 108 and / or application functions hosted on application server (not shown) on data network 102.

[0032] Core network 108 has a service-based architecture. The network functions 110 of core network 108 include Access and Mobility Function (AMF), Authentication Server Function (AUSF), Network Openness Function (NEF), Network Repository Function (NRF), Network Slice Selection Function (NSSF), Policy Control Function (PCF), Session Management Function (SMF), User Plane Function (UPF), Unionized Data Repository (UDM), and Network Data Analysis Function (NWDAF). For ease of illustration, in Figure 2 Other network functions of core network 108, such as binding support function (BSF) and billing function (CHF), are not shown.

[0033] When user equipment 100 moves between different radio access networks, cells or locations, the AMF handles access, authorization, authentication of user equipment including UE 100, and manages the mobility of user equipment 100.

[0034] The SMF is responsible for establishing, maintaining, and terminating Protocol Data Unit sessions in core network 108. The SMF manages user plane resources and interacts with the UPF in core network 108 to ensure that data packets are correctly routed and forwarded.

[0035] The UDM performs the authentication process, stores and manages user data, including subscriber profiles, authentication credentials, and authorization policies; implements security mechanisms to protect user data and resources of the communication network (e.g., core network 108) from unauthorized access attacks and vulnerabilities; and interacts with other network functions of the core network 108 (such as PCF) to enforce access control policies, Quality of Service (QoS) parameters, and service restrictions based on user profiles and subscription plans. The UDM is also responsible for managing the registration of network functions 110 serving user equipment 100.

[0036] Each NF 110 of core network 108 can provide one or more services to other network functions 110 of core network 108 via an application programming interface (API). Each NF 110 can also register itself and the services it supports (e.g., services it provides to other network functions) with the NRF of core network 108. The NRF can be used by any network function 110 to discover other network functions 110 (or instances of NF 110) and services supported by other NF 110 (e.g., services provided by other NFs). Any NF 110 can consume (e.g., use) services provided and exposed by another NF 110. An NF 110 that consumes services from another network function is generally referred to as a network function service consumer (NFc). A network function 110 that provides and exposes one or more services is referred to as a network function service producer (NFp).

[0037] Figure 3The physical and logical components of an exemplary device 300 including one or more network functions 110 of a communication network 104 are shown. Although an example implementation of device 300 is shown and discussed below, examples of other devices that can implement the NF can be included, which may include... Figure 3 The logical and physical components shown are different logical and physical components. Furthermore, although... Figure 3 A single instance of each logical and / or physical component of device 300 is shown, but there may be... Figure 3 Multiple instances of each logical and / or physical component shown.

[0038] Device 300 includes one or more processors 302, such as a central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), application-specific logic circuit, graphics processing unit (GPU), tensor processing unit, neural processing unit, dedicated artificial intelligence processing unit, hardware accelerator, quantum processor, or any combination thereof. One or more processors 302 may generally be referred to as processor 302 and are collectively referred to as processor 302.

[0039] Device 300 also includes one or more memories 304 (generally referred to as memory 304 and collectively referred to herein as "memory 304"), which may include volatile or non-volatile memory (e.g., flash memory, random access memory (RAM), and / or read-only memory (ROM)). Memory 304 may store machine-executable instructions for execution by at least one of one or more processors 302. For example, machine-executable instruction 306 may be shown as stored in memory 304, which may be executed by at least one of one or more processors 302, causing device 300 to perform the example operations described herein. Memory 304 may store machine-executable instructions for execution by processor 302, such as those referenced above. Figure 1 The machine-executable instructions describe other network functions of the core network 108.

[0040] In addition to machine-executable instructions 306, memory 304 may also store data, information, rules and / or policies.

[0041] In some examples, device 300 may also include one or more electronic storage units (not shown), such as solid-state drives, hard disk drives, disk drives, and / or optical disk drives. In some examples, one or more datasets and / or modules may be provided by external memory (e.g., an external drive that communicates with the computing system via wired or wireless communication), or may be provided by transient or non-transitory computer-readable media. Examples of non-transitory computer-readable media include RAM, ROM, erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, CD ROM, or other portable storage. Storage units and / or external memory may be used in conjunction with memory 304 to implement data storage, retrieval, and caching functions of device 300.

[0042] For example, the processor and memory 304 of device 300 can communicate with each other via a communication bus. In some embodiments, device 300 is a distributed computing system that includes multiple computing devices (e.g., servers) communicating with each other via a data network, and optionally one or more additional components. In some embodiments, the various operations described herein can be performed by different computing devices (e.g., servers) of the distributed computing system. In some embodiments, device 100 is a virtual machine provided by the infrastructure (e.g., hypervisor, processor, and memory) of a cloud computing system.

[0043] Generative AI is a method of generating content based on the knowledge gained from task-specific training and fine-tuning using LLMs. One approach to generative AI is the Large Language Model (LLM), which can be used to generate plausible text / language based on input queries. Currently, many examples of proprietary and open-source models are available and used in various applications. A general-purpose LLM (base model) is typically obtained through extensive training on large amounts of data to capture relationships between words and acquire general capabilities for text understanding, processing, and generation. This process is called pre-training. Fine-tuning is the process of adapting a general-purpose model to a domain-specific task, such as understanding technical text and recommending administrative actions, as in the telecommunications domain. Fine-tuning can be accomplished by selectively adjusting / training a subset of model parameters or a newly added set of parameters.

[0044] LLMs use a single input and output data modality: (natural) language. Large multimodal models (LMMs) combine various data modalities, such as text, audio, visual, and sensor data, to capture the correlations between different modalities. This approach can be applied to any kind of data, including web data.

[0045] Small language models (SLMs) are less computationally intensive than LLMs in both training and inference, and can provide good performance, for example, if trained and used for a specific problem.

[0046] Models including LLM / LMM and / or SLM can be used to authorize AI agents, referred to as GenAI agents in this paper. Such GenAI agents can leverage LLM / LMM / SLM capabilities to reason, use external tools, and maintain action and / or environmental awareness memories to complete tasks.

[0047] GenAI agents can be specifically designed to perform different types of tasks. These tasks can include consumer tasks (e.g., sightseeing planning based on timing, traffic planning based on weather forecasts, or personalized fitness planning and guidance) or network-related tasks (e.g., optimizing network parameters for energy conservation in a network).

[0048] Compared to other network entities or network functions, GenAI agents can act autonomously. For example, leveraging the capabilities of LLM or LMM (Large Multimodal Models), GenAI agents can understand complex tasks and create execution plans with lists of actions (e.g., CRUD operations, using tools, services, collaborating with other GenAI agents), and act according to those plans.

[0049] Gen AI agents can share their capabilities and collaborate with other Gen AI agents based on orchestration.

[0050] In 6G systems, GenAI agents can be deployed in the network and UE devices, for example, as part of an application deployed at the UE or as an integral part of the UE's operating system. Given the expected number of agents, the number of requests, and the processing demands associated with those requests, requests (or prompts) from GenAI agents at the UE toward the network can place a significant burden on the network as it processes such requests.

[0051] Processing all UE agent requests (also known as prompts) at once may be impractical and unnecessary, as not all prompts will have the same importance. For example, in the case of a single UE and a single GenAI agent at that UE, a prompt in one context may be entertainment-related and have very low criticality, while in another context, a prompt related to a health issue that leads to an emergency call may be identified as a highly critical prompt.

[0052] Furthermore, even prompts from the same UE or the same application may be associated with different data modalities, and to implement such a prompt, different resources from the network may be required, such as text-based question-and-answer prompts, rather than high-quality video generation and delivery.

[0053] Existing techniques for creating and manipulating cues within the context of LLM focus on modulating the content of the cues while adhering to responsible AI practices. Such practices concentrate on filtering the user's input text (cues) and the LLM's output to remove all inappropriate content and provide unbiased manipulation. Such mechanisms are available in commercially viable tools, such as those in the public cloud, and can include the ability to set guardrails that define the operational constraints of the chosen LLM model.

[0054] Further management of cues involves applying cue engineering techniques to guide the LLM to produce desired outputs. These techniques include zero-shot learning, few-shot learning, and mind chain cues. The goals of cue engineering can be to format the LLM's output, set rules for the output, and assign "roles" to the LLM, among others.

[0055] However, existing prompt creation and manipulation mechanisms do not include prompt classification and processing in mobile networks.

[0056] In order to handle cues of different modalities and criticalities and adapt them to network resources, the network needs to be able to understand the cue modality and criticality and classify the cues accordingly. The lack of such processing can hinder the applicability and adoption of GenAI agents, especially in highly critical tasks or tasks requiring specific / additional network resources.

[0057] Figure 4 A flowchart of a method according to an example embodiment is shown. The method can be performed at a device. The device may include a user device. The user device may include a prompting categorization and prioritization function (PCPF).

[0058] At 401, the method includes obtaining a prompt from an application on the user device.

[0059] At 402, the method includes determining at least one characteristic associated with the prompt.

[0060] At 403, the method includes determining a classification identifier associated with the prompt based on at least one characteristic.

[0061] At 404, the method includes: for the prompt, determining a description of the service route used to provide the prompt to the network, based on a classification identifier and from a service routing policy.

[0062] At 405, the method includes providing prompts to the network based on a description of the service route.

[0063] At least one application may include a generative AI application or a generative AI agent (GenAI agent). (See reference) Figure 4The described method can provide GenAI agent cue classification and priority sorting functionality (PCPF) at the UE.

[0064] PCPF can leverage generative AI capabilities related to cue understanding to identify whether a particular application / agent is a cue-based application / agent for interacting with users, and how critical and demanding the implementation of a particular cue is for the network.

[0065] At least one characteristic may include at least one of the following: the input modality of the prompt, the output modality of the prompt, the number of prompts including the prompt received by the user equipment within a time period, or the criticality of the prompt. For example, PCPF may classify UE GenAI agent prompts into low, medium, or high volume categories, where volume is the number of prompts including the prompt received by the user equipment within a time period.

[0066] Alternatively or additionally, cues can be categorized based on modality category (i.e., input or output modality, such as voice, video, text, or augmented reality modality).

[0067] Alternatively or additionally, prompts may be categorized or prioritized based on criticality categories (e.g., low, medium, high, or mission-critical; V2X collision avoidance; emergency call criticality). Prompt priority or criticality may, for example, be based on the semantics of the prompt.

[0068] The category identifier can be a parameter associated with the cue category, such as PromptTypeID. The category identifier can indicate information about at least one characteristic (e.g., modality and criticality).

[0069] Based on this classification and priority ordering, the network performs corresponding actions to satisfy the UE GenAI agent prompts based on their classification and priority ordering, and incorporates them as part of the URSP rule service description.

[0070] Based on the classification identifier, customized URSP rules for prompts can be derived, and the UE can appropriately handle services from the same application but caused by different prompt types. This is an example of determining the description of the service route used to provide the prompt to the network based on the classification identifier and the service routing policy, and providing the prompt to the network according to the description of the service route.

[0071] The description of a service route can identify at least one of the Packet Data Unit (PDU) sessions or network slices used to provide cues to the network. Since the same UE GenAI application / agent may be subject to different user cues, and therefore may require different processing for different cues, relying on the public AppID as part of the service descriptor within the URSP rules may be insufficient to differentiate between different service requirements of the network. The use of classification identifiers can provide more efficient processing of network cues based on the classification and prioritization of cues.

[0072] Figure 5 The example LLM prompting priority ranking architecture is shown. This example consists of prompting classification and prioritization functions (PCPF) installed at the UE and the network (e.g., co-located with the AMF function, or co-located with a similar function to be defined for 6G). The prompting classification and prioritization function (PCPF) installed at the UE is referred to as the UE-side PCPF, and the PCPF co-located with the NF (e.g., the AMF function, etc.) is referred to as the network-side PCPF function.

[0073] exist Figure 5 In the example shown, a prompt received from a prompt-based application is first queued in a joint FIFO queue. The UE-side PCPF function processes the prompt to classify it (determine its modality) and determine its priority (criticality) level at the UE. At the end of the UE-PCPF processing, each prompt is classified based on its modality and priority and assigned a prompt priority category. During this process, each prompt receives a corresponding PromptTypeID. This is an example of determining the classification identifier associated with a prompt based on at least one characteristic associated with the prompt.

[0074] Providing hints to the network using service routing can include placing the hints in a queue associated with a category identifier, which is linked to the hint. For example, such as Figure 5 As shown in the diagram, each prompt is queued into its corresponding prompt priority category queue (one prompt per PromptTypeID), as illustrated in the previous figure. The prompt can then be provided to the network based on the description of the service route. Providing the prompt to the network based on the service route description can include providing the prompt to the network via network slices or via Packet Data Unit (PDU) sessions, where network slices are identified by the service route description and PDU sessions are identified by the service route description.

[0075] Figure 6 A flowchart of a method according to an example embodiment is shown. The method can be performed at a device. This device may include, is, or be included in a user equipment. The user equipment may include a prompting categorization and prioritization function (PCPF).

[0076] At 601, the method includes providing information to the network, wherein the information relates to a classification identifier of multiple prompts from at least one application of a user device, and wherein the information includes an identifier of the at least one application.

[0077] At 602, the method includes receiving an updated service routing policy from the network in response to information provided to the network.

[0078] At 603, the method includes implementing an updated service routing policy in response to subsequent prompts received from at least one application.

[0079] Applications can include generative artificial intelligence (AI) applications or generative AI agents. (See reference...) Figure 6 The described method may include: obtaining multiple prompts from at least one application on a user device; determining at least one corresponding characteristic associated with a corresponding prompt among the multiple prompts; and determining a classification identifier associated with the corresponding prompt based on the at least one corresponding characteristic.

[0080] In the example, the PCPF at the UE can identify and notify the PCF of the prompt-driven nature of the GenAI agent, i.e., a specific application / agent at the UE interacts with the user via natural language prompts of different categories and priorities. The PCPF can provide primary prompt categories based on the type and volume of content the network needs to support, and prompt priority / criticality based on the semantics of the prompts. This is an example of providing information to the network, where the information involves classification identifiers of multiple prompts from at least one application of the user equipment, and where the information includes the identifier of at least one application. This information can be referred to as a UEPrompt report. The classification identifier can be referred to as PromptTypeID.

[0081] If at least one of the classification identifier or the identifier of at least one application is new, the UE can determine to provide information to the network. In the example, after discovering the PromptTypeID for each prompt at the UE side, the PCPF can provide a UEPrompt report from the prompt-based application to the network-side PCPF. The UEPrompt report includes a list of applications with prompt-based interactions with the user, and for each AppID in the list, a list associated with the PromptTypeID is provided whenever a prompt-based application is first added to the UE or whenever the PromptTypeID is changed.

[0082] Once a new type of alert or application has been identified, such as an emergency alert, which indicates a change in UE state, a policy update may be needed to provide the network with the appropriate routing for the alert. For example, if the currently available URSP policy does not allow the use of PDU sessions or slices corresponding to an emergency situation, the URSP policy should be updated.

[0083] In the example, implementing the updated service routing policy includes: determining a description of the service route to the network based on a classification identifier and from the updated service routing policy, in response to subsequent prompts, and providing a prompt to the network according to the description of the service route. The description of the service route may identify at least one of a Packet Data Unit (PDU) session or a network slice for providing a prompt to the network.

[0084] refer to Figure 6 The described example allows the UE-side PCPF to immediately report changes to the network-side PCPF. This information from different UEs in the network can be collected at the network-side PCPF, which can identify the necessity of policy updates as a response to changes in the type of alerts detected at single or multiple UEs, such as an emergency (an emergency alert is detected). The network-side PCPF can then transmit this information to the PCF.

[0085] Figure 7 A flowchart illustrating a method according to an example is shown. This method can be performed at a device. This device may include, is, or be included in a network function. The network function may include a prompting classification and prioritization function (PCPF).

[0086] At 701, the method includes receiving information from a user device, wherein the information relates to a classification identifier of multiple prompts from at least one application of the user device, wherein the information includes an identifier of at least one application.

[0087] At 702, the method includes determining, based on the received information, a service routing policy to be provided to the user equipment.

[0088] At 703, the method includes providing the received information to the policy control function based on the determination.

[0089] At 704, the method includes receiving an updated service routing policy from the policy control function in response.

[0090] At point 705, the method includes providing updated service routing policies to the user equipment.

[0091] Determining the service routing policy to be updated for user equipment based on the received information may include: determining that at least one of the classification identifier or the identifier of at least one application is new.

[0092] In the example, the UE-side PCPF notifies the network-side PCPF of changes to the AppID and PromptTypeID (e.g., at least one of the classification identifier or at least one application identifier is new). The network-side PCPF can be a separate entity or part of an existing NF, such as the AMF. The network-side PCPF can also notify the Access and Mobility Function (AMF) of AppID and PromptTypeID changes.

[0093] Then, the AMF can signal the PCF (e.g., using a signal like...) Figure 9 The Npcf_UEPolicyControl call flow shown here changes the UE status and transmits information related to the classification identifier and the identifier of at least one application.

[0094] For reference Figure 7 The described method may include updating received information based on subscription information related to the user equipment. For example, the network-side PCPF may update the prompt classification based on information from the UDM regarding the UE's subscription, i.e., adapting the prompt classification to the available UE services.

[0095] Figure 8 A flowchart of a method according to an example embodiment is shown. The method can be performed at a device. This device may include, is, or is included in a Policy Control Function (PCF).

[0096] At 801, the method includes receiving information from a user device, wherein the information relates to a classification identifier of multiple prompts from at least one application of the user device, wherein the information includes an identifier of at least one application.

[0097] At 802, the method includes obtaining subscription information related to the user's device.

[0098] At 803, the method includes determining a business routing policy for updates to prompts from at least one application based on the received information and subscription information.

[0099] At 804, the method includes providing the user equipment with an updated service routing policy.

[0100] In one example, based on information related to the classification identifiers (e.g., PromptTypeID) of multiple prompts from at least one application of the user equipment and the identifier of at least one application (e.g., AppID), the PCF can update the URSP policy to use the appropriate PDU session / slice corresponding to the prompt type (e.g., the urgency of the situation). This is an example of determining an updated service routing policy for prompts from at least one application. Policy decisions can also be based on UDM values ​​(e.g., subscription information or subscription data associated with the UE) and information provided by the UE.

[0101] Then, the AMF and UE are updated to reflect the changes in the URSP policy. This is an example of providing updated service routing policies to user equipment.

[0102] The example process can be summarized as follows.

[0103] In the example UE-side PCPF prompt classification and reporting, the UE-side PCPF detects changes in prompt types and sends a UE prompt report to the network-side PCPF. This may involve new message types, such as updated messages including AppID and PromptTypeID. This is an example of providing information to the network that pertains to the classification identifiers of multiple prompts from at least one application of the user equipment, and that this information includes the identifier of at least one application.

[0104] In the example UE prompt report forwarding, the network-side PCPF forwards the UE prompt report to the AMF. Alternatively, the AMF forwards the UE report to the PCF. The message type can be a modified message Npcf_UEPolicyControl_Update (including information related to the classification identifier (e.g., PromptTypeID) of multiple prompts from at least one application of the user equipment and the identifier (e.g., AppID) of at least one application).

[0105] In the example UE policy update, the PCF processes the AppID and PromptTypeID changes and determines whether any policy update is required (for URSP policies). If so, the PCF sends a policy update request to the AMF. (The message type can be Npcf_UEPolicyControl_Update, as per Table 4.2.1 (which includes information on classification identifiers (e.g., PromptTypeID) for multiple prompts from at least one application from the user equipment and identifiers (e.g., AppID) for at least one application)).

[0106] In the example policy update implementation, the AMF acknowledges the policy update request and sends information about the new URSP policy to the UE. This is an example of receiving the updated service routing policy from the policy control function and providing the updated service routing policy to the user equipment.

[0107] The call flow for exchanges between PCPFs on the UE side can be based on the Npcf_UEPolicyControl API call flow, which is used by the AMF to notify the PCF of changes in UE status. Figure 9 The example call flow is shown below. The example call flow is built on top of the standard process used for UE configuration update procedures.

[0108] exist Figure 9 In step 1 of the example call flow, the UE-side PCPF classifies incoming prompts based on modality, volume, and criticality / priority. The UE-side PCPF can leverage built-in generative AI capabilities to understand and classify prompts. This is an example of obtaining prompts from the user equipment application, determining at least one characteristic associated with the prompt, and determining a classification identifier associated with the prompt based on at least one characteristic. The prompt category is then transmitted to the network-side PCPF (in... Figure 2 The message is described as part of the AMF, but this function could be part of a gNB or any other NF, in which case another message would be needed to forward the prompt category to the AMF. This is an example of providing information to the network, where the information pertains to the classification identifiers of multiple prompts from at least one application on a user device, and where the information includes the identifier of at least one application.

[0109] exist Figure 9 In step 2 of the example call flow, the network-side PCPF can also update the prompt classification based, for example, on the prompt type from different applications or different UEs and on the UE subscription information from the UDM. In the simplest scenario, the network-side PCPF can use the prompt classification received from the UE-side PCPF and send a request to the PCF to update the UE policy according to the prompt type. Such a request is only sent to the PCF if a new prompt type not previously seen at the UE is about to occur. This is an example of receiving information from a user equipment, where the information involves classification identifiers of multiple prompts from at least one application of the user equipment, including the identifier of at least one application, determining an updated service routing policy to be provided to the user equipment based on the received information, and providing the received information to the policy control function based on this determination.

[0110] Optionally, the PCF can notify the network-side PCPF of relevant UE subscription data that helps in better cue classification. This input can be used by the network-side PCPF for future cue classification from the UE. Alternatively, the network-side PCPF can forward subscription data related to cue classification to the UE-side PCPF, allowing this input to be considered for cue classification already performed at the UE-side PCPF.

[0111] exist Figure 9 In step 3 of the example call flow, upon receiving information about the new prompt type for the UE and the UE application, the PCF can obtain information about the UE's subscription in order to gain a complete picture of the UE. This is an example of obtaining subscription information related to the user equipment.

[0112] exist Figure 9 In step 4 of the example call flow, based on information about the new prompt type at the UE and the UE application, as well as information about the UE's subscription, the PCF decides to update the URSP rules at the UE to better handle services from applications that generate new prompt types that have not been seen before. This is an example of determining an updated service routing policy for prompts from at least one application based on received information and subscription information.

[0113] In step 5 of the example call flow, the PCF notifies the AMF of the new URSP rules that need to be configured at the UE. This is an example of providing updated service routing policies to user equipment.

[0114] In step 6 of the example call flow, if the PCF has not yet subscribed to the UE response notified by the AMF of the update on the UE policy information, the PCF subscribes to the AMF to be notified of the UE response of the update on the UE policy information.

[0115] In step 7 of the example call flow, the PCF invokes the Namf_Communication_N1N2MessageTransfer service operation provided by the AMF. This message includes SUPI and the UE policy container.

[0116] In step 8 of the example call flow, if the UE is CM-CONNECTED on 3GPP access or non-3GPP access, the AMF transparently transmits the UE policy container (UE policy information) received from the PCF to the UE. The UE policy container can include policies on how to handle different categories of prompts. For example, in addition to the application descriptor OSAppID and PromptTypeID, the service descriptor of the URSP rule can also have a corresponding routing description on how to handle services from a specific application with a specific prompt.

[0117] In step 9 of the example call flow, the UE updates the UE policy provided by the PCF and sends the result to the AMF.

[0118] In step 10 of the example call flow, the AMF uses Namf_Communication_N1MessageNotify to forward the UE's response to the PCF.

[0119] Figure 10 A block diagram of apparatus 10 is shown by way of example. Apparatus 10 includes, for example, at least one processor 12 and at least one memory 14 storing instructions 15, which, when executed by the at least one processor, cause apparatus 10 to perform at least one or more methods (or portions thereof) as disclosed herein, and any embodiments (or corresponding portions thereof). In the example, at least one memory and instructions (e.g., computer program code, software) are configured, together with at least one processor, to cause apparatus 10 to perform one or more methods (or portions thereof) as disclosed herein, and any embodiments (or corresponding portions thereof).

[0120] The processor 12 may include circuitry, or be configured as one or more circuits, which are configured to perform phases of the method according to the embodiments described herein.

[0121] As used herein, the term "circuit" may refer to one or more of the following: (a) a hardware circuit implementation only, such as an implementation in analog and / or digital circuits only; and (b) a combination of hardware circuits and software, such as, if applicable: (i) a combination of (multiple) analog and / or digital hardware circuits with software / firmware; and (ii) any part of a hardware processor having software (including (multiple) digital signal processors, software, and (multiple) memories, which work together to enable a device such as a user equipment to perform various functions); and (c) hardware circuits and / or processors, such as a microprocessor or a portion thereof, which require software (e.g., firmware) to operate, but may be absent when the software is not required to operate. This definition of circuit applies to all uses of the term herein, including in any claim. As another example, as used herein, the term circuit also covers implementations of hardware circuits or processors (or multiple processors) only, or a portion thereof and its accompanying software and / or firmware. The term circuit also covers (for example and if applicable to a particular claim element) baseband integrated circuits or processor integrated circuits for mobile devices or similar integrated circuits in servers, cellular network devices or other computing or networking devices.

[0122] The memory 14 can be implemented using any suitable data storage technology. The memory may include a database for storing data. The memory 14 may be at least partially external to the device 10, but may be accessible by the device 10.

[0123] Instruction 15 may be included in a computer-readable medium or a non-transitory computer-readable medium. As used herein, the term non-transitory is a limitation on the medium itself (i.e., tangible rather than tactile), rather than a limitation on the persistence of data storage (e.g., random access memory (RAM) and read-only memory (ROM)).

[0124] Device 10 includes a radio interface 16. The radio interface 16 can provide communication capabilities to device 10. The radio interface 16 may include a receiver configured to receive information according to at least one cellular or non-cellular standard. The radio interface 16 may include a transmitter configured to transmit information according to at least one cellular or non-cellular standard. The receiver may include more than one receiver. The transmitter may include more than one transmitter. The radio interface 16 may include a transceiver configured to receive and transmit information according to at least one cellular or non-cellular standard. The transceiver may include more than one transceiver.

[0125] Device 10 may include a user interface 18, which includes at least one of, for example, a keyboard, microphone, touch display, monitor, speaker, etc. User interface 18 can be used to control the device by a user. User interface 18 may be external to device 10. For example, device 10 may be connected to another device, such as a computer, via a wireless or wired connection, and device 10 may be controlled by a user via the computer.

[0126] refer to Figure 10 The described apparatus may be caused or configured to perform, or may include means for at least performing Figure 4 or Figure 6 The methods and / or components of any one or more embodiments described herein.

[0127] refer to Figure 3 The described apparatus may be caused or configured to perform, or may include means for at least performing Figure 7 or Figure 8 The methods and / or components of any one or more embodiments described herein.

[0128] The core network functions described herein can be implemented as core network entities comprising a combination of hardware processing circuitry and software and / or firmware including machine-readable instructions, or software including machine-readable instructions executable by at least one processor of the hardware processing circuitry. The hardware processing circuitry includes at least one processor and at least one memory storing machine-readable instructions executable by at least one processor of the hardware processing circuitry. The processor includes any one or a combination of accelerators, microprocessors, cores of multi-core microprocessors, microcontrollers, programmable integrated circuits, programmable gate arrays, digital signal processors, central processing units, graphics processing units, and tensor processing units. The memory includes any or a combination of volatile or non-volatile memory (e.g., flash memory, cache, random access memory (RAM), and / or read-only memory (ROM)). The memory stores machine-readable instructions for the software and / or firmware executed by at least one processor of the hardware processing circuitry. The machine-readable instructions are executable by at least one processor of the hardware processing circuitry, causing the hardware processing circuitry to perform the actions or operations of the methods described herein.

[0129] In embodiments, at least some of the processes described herein may be performed by means including components for performing at least some of the processes described herein. Means for performing the method steps disclosed herein may include software and / or hardware components of device 300. For example, at least one processor 302, memory 304, and computer program code form means for performing one or more methods (or portions thereof) disclosed herein, and any embodiments (or corresponding portions thereof). As used herein, the term “component” should be interpreted in the singular form, referring to a single element, or in the plural form, referring to a combination of single elements. Therefore, the term “component for [performing A, B, C]” should be interpreted to encompass means in which only one component for performing A, B, and C is present, or in which separate components for performing A, B, and C are present, or in which partially or completely overlapping components for performing A, B, and C are present. Furthermore, the terms "component for performing A, component for performing B, component for performing C" should be interpreted as covering an apparatus in which only one component exists for performing A, B, and C, or in which a single component exists for performing A, B, and C, or in which components for performing A, B, and C partially or completely overlap.

[0130] Although this disclosure has been described above with reference to non-limiting and illustrative examples based on the accompanying drawings, it is clear that the scope of this disclosure is not limited thereto, and modifications can be made in many different ways. As technology advances, it will become apparent to those skilled in the art how to further implement and / or modify this disclosure in various ways. Furthermore, it will be apparent to those skilled in the art that the embodiments described herein can, but are not required to, be combined in various ways with other embodiments described herein.

Claims

1. A method for a user equipment, the method comprising: Receive prompts from the application on the user device; Determine at least one characteristic associated with the prompt; A classification identifier associated with the prompt is determined based on at least one of the characteristics; In response to the prompt, based on the classification identifier and from the service routing policy, a description of the service route used to provide the prompt to the network is determined; as well as The prompt is provided to the network according to the description of the service route.

2. The method according to claim 1, wherein the application includes generative artificial intelligence (AI) applications or generative AI agents.

3. The method according to claim 1 or claim 2, wherein the description identifier of the service route is at least one of a Packet Data Unit (PDU) session or a network slice, for providing the prompt to the network.

4. The method according to claim 1 or claim 2, wherein the at least one characteristic comprises at least one of the following: The input modality of the prompt; The output mode of the prompt; The number of prompts, including the prompt, obtained by the user equipment within a time period; or The importance of the prompt.

5. The method of claim 1 or claim 2, wherein providing the prompt to the network according to the description of the service route includes placing the prompt in a queue associated with the classification identifier, the classification identifier being associated with the prompt.

6. The method of claim 1 or claim 2, wherein providing the prompt to the network according to the description of the service route includes providing the prompt to the network via a Packet Data Unit (PDU) session, the PDU session being identified by the description of the service route.

7. The method of claim 1 or claim 2, wherein providing the prompt to the network according to the description of the service route includes providing the prompt to the network via a network slice, the network slice being identified by the description of the service route.

8. An apparatus for communication, comprising at least one processor and at least one memory storing instructions, the instructions, when executed by the at least one processor, causing the apparatus to perform at least the following: Receive prompts from the application on the user device; Determine at least one characteristic associated with the prompt; A classification identifier associated with the prompt is determined based on at least one of the characteristics; In response to the prompt, based on the classification identifier and from the service routing policy, a description of the service route used to provide the prompt to the network is determined; as well as The prompt is provided to the network according to the description of the service route.

9. A communication apparatus comprising components for: Receive prompts from the application on the user device; Determine at least one characteristic associated with the prompt; A classification identifier associated with the prompt is determined based on at least one of the characteristics; In response to the prompt, based on the classification identifier and from the service routing policy, a description of the service route used to provide the prompt to the network is determined; as well as The prompt is provided to the network according to the description of the service route.

10. A computer program comprising instructions that, when executed by a device, cause the device to perform: Receive prompts from the application on the user device; Determine at least one characteristic associated with the prompt; A classification identifier associated with the prompt is determined based on at least one of the characteristics; In response to the prompt, based on the classification identifier and from the service routing policy, a description of the service route used to provide the prompt to the network is determined; as well as The prompt is provided to the network according to the description of the service route.