Method and apparatus for configuring artificial intelligence and machine learning traffic transmissions in wireless communication network
By introducing policy control functions in 5G networks, dynamically configuring and managing AI/ML traffic transmission, the problem of inefficient AI/ML traffic transmission in the existing technology has been solved, and the optimization of AI/ML traffic transmission and service quality improvement has been achieved.
Patent Information
- Application Number
- CN202380030199.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-03-07
- Filing Date
- 2023-03-29
- Publication Date
- 2024-11-12
AI Technical Summary
The lack of support for artificial intelligence (AI) and machine learning (ML) traffic transmission in existing 5G networks, and the inability to effectively configure and manage AI/ML transmission strategies, resulting in inefficiency and unstable service quality.
Determine whether the AI/ML protocol data unit (PDU) session or transmission policy is affected by introducing policy control function (PCF) in a wireless communication network, receiving update notifications for AI/ML transmission configuration information from the unified data repository (UDR). , and notify the session management function (SMF) for reconfiguration to ensure the optimized transmission of AI/ML traffic.
It realizes dynamic configuration and optimized transmission of AI/ML traffic, improves network efficiency and service quality, adapts to the needs of different AI/ML applications, and enhances the flexibility and scalability of 5G networks.
Smart Images

Figure CN118947209A_ABST
Abstract
Description
Technical Field
[0001] Certain examples of the present disclosure provide techniques related to configuring artificial intelligence (AI) and / or machine learning (ML) traffic transmission. Background Art
[0002] 5G mobile communication technology defines a wide frequency band, making high transmission rates and new services possible, and can be implemented not only in "below 6 GHz" frequency bands such as 3.5 GHz, but also in "above 6 GHz" frequency bands called mmWave including 28 GHz and 39 GHz. In addition, 6G mobile communication technology (called super 5G system) has been considered to be implemented in terahertz frequency bands (e.g., 95 GHz to 3 THz bands) to achieve a transmission rate 50 times faster than 5G mobile communication technology and an ultra-low latency of one-tenth of 5G mobile communication technology.
[0003] In the early stage of the development of 5G mobile communication technology, in order to support services and meet performance requirements related to enhanced mobile broadband (eMBB), ultra-reliable low latency communication (URLLC) and massive machine type communication (mMTC), beamforming and massive MIMO for alleviating radio wave path loss and increasing radio wave transmission distance in mmWave, dynamic operation of parameter sets (e.g., operating multiple subcarrier spacings) and time slot formats supporting efficient utilization of mmWave resources, initial access technology for supporting multi-beam transmission and broadband, definition and operation of bandwidth parts (BWPs), new channel coding and decoding methods (such as low-density parity-check (LDPC) codes for large amounts of data transmission and polar codes for highly reliable control information transmission), L2 preprocessing, and standardization of network slicing for providing dedicated networks dedicated to specific services have been underway.
[0004] Currently, discussions are ongoing on improvements and performance enhancements of initial 5G mobile communication technologies in consideration of the services that they will support, and there is already physical layer standardization on technologies such as Vehicle-to-Everything (V2X) for assisting driving determination of autonomous vehicles based on information about the location and status of the vehicle sent by the vehicle and for enhancing user convenience, NR-U (New Radio Unlicensed) designed to enable system operation to comply with various regulatory-related requirements in unlicensed bands, NR UE power saving, a non-terrestrial network (NTN) as UE-satellite direct communication for providing coverage in areas where communication with terrestrial networks is unavailable, and positioning.
[0005] In addition, standardization is being conducted in air interface architecture / protocols on technologies such as Industrial Internet of Things (IIoT) for supporting new services through interworking and integration with other industries, Integrated Access and Backhaul (IAB) for providing nodes for network service area expansion by supporting wireless backhaul links and access links in an integrated manner, mobility enhancement including conditional handover and dual active protocol stack (DAPS) handover, and two-step random access (2-step RACH for NR) for simplifying the random access procedure. Standardization is also being conducted in system architecture / services on 5G baseline architecture (e.g., service-based architecture or service-based interface) for combining network function virtualization (NFV) and software defined network (SDN) technologies, and mobile edge computing (MEC) for receiving services based on UE location.
[0006] As 5G mobile communication systems are commercialized, an exponentially increasing number of connected devices will be connected to the communication network, and therefore, it is expected that enhanced functions and performance of 5G mobile communication systems and integrated operations of connected devices will be necessary. To this end, new research is planned in relation to extended reality (XR) for efficiently supporting augmented reality (AR), virtual reality (VR), mixed reality (MR), etc., 5G performance improvement and complexity reduction by utilizing artificial intelligence (AI) and machine learning (ML), AI service support, metaverse service support, and drone communication.
[0007] Furthermore, such development of 5G mobile communication systems will serve as a foundation for developing not only new waveforms for providing coverage in the terahertz band for 6G mobile communication technology, multi-antenna transmission technologies (such as full-dimensional MIMO (FD-MIMO), array antennas, and large antennas), metamaterial-based lenses and antennas for improving coverage of terahertz band signals, high-dimensional spatial multiplexing technologies using orbital angular momentum (OAM), and reconfigurable smart surfaces (RIS), but also full-duplex technologies for improving the frequency efficiency of 6G mobile communication technology and improving system networks, AI-based communication technologies for implementing system optimization by utilizing satellites and artificial intelligence (AI) from the design stage and internalizing end-to-end AI support functions, and next-generation distributed computing technologies for implementing services with a complexity level that exceeds the limits of UE operating capabilities by utilizing ultra-high performance communication and computing resources.
[0008] AI / ML is used across a range of application areas across industry sectors. In mobile communication systems, conventional algorithms (e.g., speech recognition, image recognition, video processing) in mobile devices (e.g., smartphones, cars, robots) are increasingly being replaced by AI / ML models to enable a variety of applications.
[0009] The above information is presented as background information only to assist with an understanding of the present disclosure. No determination has been made, and no assertion is made, as to whether any of the above might be applicable as prior art to the present invention. Summary of the invention
[0010] Technical issues
[0011] Certain examples of the present disclosure provide methods, apparatus, and systems for configuring AI and / or ML traffic transmission in a 3rd Generation Partnership Project (3GPP) fifth generation (5G) network.
[0012] Solution to the problem
[0013] The purpose of some examples of the present disclosure is to at least partially address, solve and / or alleviate at least one of the problems and / or disadvantages associated with the related art, such as at least one of the problems and / or disadvantages described herein. The purpose of some examples of the present disclosure is to provide at least one advantage over the related art, such as at least one of the advantages described herein.
[0014] According to an embodiment of the present disclosure, a method and apparatus for configuring artificial intelligence / machine learning (AI / ML) traffic transmission in a wireless communication network.
[0015] According to an embodiment of the present disclosure, a method and apparatus for transmitting AI / ML transmission configuration information in a wireless communication network.
[0016] According to an embodiment of the present disclosure, a method for configuring artificial intelligence / machine learning (AI / ML) traffic transmission in a wireless communication network is provided. The method includes receiving, by a policy control function (PCF), a notification of an update of AI / ML configuration information from a unified data repository (UDR) storing AI / ML transmission configuration information; determining, by the PCF, whether an AI / ML protocol data unit (PDU) session or transmission policy is affected by the update; and based on determining that the AI / ML session or transmission policy is affected by the update, notifying a session management function (SMF) by the PCF that a session management (SM) policy is updated. The SMF can be configured to reconfigure a PDU session for transmitting AI / ML traffic based on an updated SM policy.
[0017] In an example, the method further comprises sending, by the SMF, information about the reconfigured PDU session to an access and mobility management function (AMF).
[0018] In an example, the method further comprises updating, by the SMF, the user equipment (UE) based on the AI / ML transmission configuration information.
[0019] In an example, updating the UE based on the AI / ML transmission configuration information includes updating one or more of an AI / ML application function (AF) address, an AI / ML domain name system (DNS) server address, an AI / ML traffic type, and AI / ML authentication information.
[0020] In an example, the method further includes indicating, by the UE, a capability for receiving AI / ML transmission configuration information during a PDU session establishment or PDU session modification procedure.
[0021] In an example, the AI / ML transmission configuration information is received by the UDR as part of an AI / ML application function (AF) request.
[0022] In an example, the AI / ML AF request is received directly from the AI / ML AF or via a Network Exposure Function (NEF).
[0023] In an example, the AI / ML AF request is part of a PDU session establishment procedure or a PDU session modification procedure for updating AI / ML transmission configuration information or associated validity parameters.
[0024] In an example, the AI / ML AF request also includes a traffic description including one or more of a data network name (DNN), single network slice selection assistance information (S-NSSAI), an application identifier, an application ID, and traffic filtering information.
[0025] In an example, the AI / ML AF request also includes one or more of potential location information of the AI / ML application, a target UE identifier, spatial validity information, temporal validity information, user plane latency requirements, quality of experience requirements, and an indication associated with the AI / ML traffic type.
[0026] In an example, the PDU session transports AI / ML traffic, and reconfiguring the PDU session includes reconfiguring a user plane of the PDU session.
[0027] In an example, reconfiguring the user plane of the PDU session includes one or more of allocating a new prefix to the UE, updating a user plane function (UPF) with new traffic steering rules, and determining whether to relocate the UPF.
[0028] In an example, AI / ML transport configuration information is received by the UDR from an AI / ML application function (AF) or a network exposure function (NEF).
[0029] In an example, the AI / ML transmission configuration information is pre-configured by the AI / ML service provider on the AI / ML AF and / or the AI / ML application client on the UE.
[0030] In an example, the AI / ML transmission configuration information includes one or more of an AI / ML application function (AF) address, an AI / ML domain name system (DNS) server address, an AI / ML traffic type, and AI / ML authentication information.
[0031] In an example, the AI / ML transmission configuration information is determined by a service level agreement (SLA) between a mobile network operator (MNO) and an AI / ML application service provider associated with the AI / ML AF.
[0032] In the example, the AI / ML transmission configuration information is for each AI / ML application ID.
[0033] According to an embodiment of the present disclosure, a policy control function (PCF) for configuring artificial intelligence / machine learning (AI / ML) traffic transmission in a wireless communication network is disclosed. The PCF may include a transceiver and a processor, which is coupled to the transceiver and configured to control the transceiver to receive a notification of an update of AI / ML transmission configuration information from a unified data repository (UDR) storing the AI / ML transmission configuration information, determine whether an AI / ML protocol data unit (PDU) session or transmission policy is affected by the update, and based on determining that the AI / ML PDU session or transmission policy is affected by the update, notify a session management function (SMF) that the updated session management (SM) policy is updated. The SMF may be configured to reconfigure the PDU session for transmitting AI / ML traffic based on the updated SM policy.
[0034] According to an embodiment of the present disclosure, a wireless communication network including a plurality of network entities is provided, wherein the plurality of network entities include a unified data repository (UDR), a policy control function (PCF), and a session management function (SMF). The UDR may be configured to receive AI / ML transmission configuration information, update the AI / ML transmission configuration information based on the received AI / ML configuration information, and notify the policy control function (PCF) of the update of the AI / ML configuration information. The PCF may be configured to receive a notification of an update of the AI / ML configuration information from the UDR, and determine whether an AI / ML protocol data unit (PDU) session or transmission policy is affected by the update, and based on determining that the AI / ML PDU session or transmission policy is affected by the update, notify the session management function (SMF) that the session management (SM) policy is updated. The SMF may be configured to reconfigure the PDU session for transmitting AI / ML traffic based on the updated SM policy.
[0035] In an example, the wireless communication network further includes an access and mobility management function (AMF), wherein the SMF is configured to send information about the reconfigured PDU session to the AMF.
[0036] In an example, the wireless communication network further includes a user equipment (UE), wherein the SMF is configured to update the UE based on the AI / ML transmission configuration information.
[0037] In an example, the wireless communication network is a 3GPP 5G network.
[0038] Embodiments or examples disclosed in the specification and / or drawings that fall outside the scope of claims should be understood as examples useful for understanding the present invention.
[0039] Other aspects, advantages and salient features of the present invention will become clear to those skilled in the art through the following detailed description in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] The present disclosure is illustrated in the accompanying drawings, and the same reference numerals indicate corresponding parts in the various figures throughout the accompanying drawings. The embodiments of the present invention will be better understood according to the following description with reference to the accompanying drawings, in which:
[0041] Figure 1 It is a sample architecture of AI / ML transfer model;
[0042] Figure 2 is an example call flow diagram illustrating the impact of AI / ML AF on traffic routing and / or reconfiguration of AI / ML traffic; and
[0043] Figure 3 is a block diagram of exemplary network entities that may be used in certain examples of the present disclosure. DETAILED DESCRIPTION
[0044] The following description of examples of the present disclosure is provided with reference to the accompanying drawings to assist in a comprehensive understanding of the invention as defined by the claims. The description includes various specific details to assist in understanding, but these specific details should be regarded as merely exemplary. Therefore, it will be appreciated by those of ordinary skill in the art that various changes and modifications may be made to the examples described herein without departing from the scope of the present invention.
[0045] The same or similar components may be denoted by the same or similar reference numerals even though they may be shown in different drawings.
[0046] For the sake of clarity and conciseness, and to avoid obscuring the subject matter of the present invention, detailed descriptions of techniques, structures, configurations, functions, or processes known in the art may be omitted.
[0047] The terms and words used herein are not limited to the bibliographical or standard meanings, but, are merely used to enable a clear and consistent understanding of the invention.
[0048] Throughout the description and claims of this specification, the words “includes,” “comprising,” and “containing,” and variations of the words mean “including but not limited to,” and are not intended to (and do not) exclude other features, elements, components, integers, steps, processes, operations, functions, characteristics, properties, and / or groups thereof.
[0049] Throughout the description and claims of this specification, singular forms such as "a", "an" and "the" include plural referents unless the context requires otherwise. For example, reference to "an object" includes reference to one or more of such objects.
[0050] Throughout the description and claims of this specification, language of the general form "X for Y" (where Y is some action, process, operation, function, activity or step, and X is some component for performing the action, process, operation, function, activity or step) includes that component X is suitable for, configured to or arranged to specifically but not necessarily exclusively perform Y.
[0051] Features, elements, components, integers, steps, processes, operations, functions, characteristics, properties and / or groups thereof described or disclosed in conjunction with a particular aspect, embodiment, example or claim are to be understood to be applicable to any other aspect, embodiment, example or claim described herein unless incompatible therewith.
[0052] Certain examples of the present disclosure provide techniques related to artificial intelligence (AI) and / or machine learning (ML) traffic transmission. For example, certain examples of the present disclosure provide methods, devices, and systems for AI and / or ML traffic transmission in a third generation partnership project (3GPP) fifth generation (5G) network. However, those skilled in the art will appreciate that the present invention is not limited to these examples and may be applied in any suitable system or standard (e.g., one or more existing and / or next generation wireless communication systems or standards, including any existing or future versions of the same standard specification, such as 3GPP 5G).
[0053] The following examples apply to 3GPP 5G and use terms associated with 3GPP 5G. However, those skilled in the art will appreciate that the technology disclosed herein is not limited to 3GPP 5G. For example, the functions and other features of the various network entities disclosed herein may be applied to corresponding or equivalent entities or features in other communication systems or standards. Corresponding or equivalent entities or features may be considered to be entities or features that play the same or similar role, perform the same or similar functions or purposes within the network.
[0054] Those skilled in the art will also appreciate that the transmission of information between network entities is not limited to the specific form, type, or order of messages described with respect to the examples disclosed herein.
[0055] Specific network entities may be implemented as network elements on dedicated hardware, as software instances running on dedicated hardware, and / or as virtualized functions instantiated on an appropriate platform (e.g., on a cloud infrastructure).
[0056] Those skilled in the art will appreciate that the present invention is not limited to the specific examples disclosed herein. For example:
[0057] The technology disclosed herein is not limited to 3GPP 5G;
[0058] One or more entities in the examples disclosed herein may be replaced with one or more alternative entities that perform equivalent or corresponding functions, procedures, or operations;
[0059] One or more of the messages in the examples disclosed herein may be replaced with one or more alternative messages, signals, or other types of information carriers conveying equivalent or corresponding information;
[0060] One or more additional entities and / or messages may be added to the examples disclosed herein;
[0061] In some examples, one or more non-essential entities and / or messages may be omitted;
[0062] The functionality, processes, or operations of a particular entity in one example may be divided between two or more separate entities in an alternative example;
[0063] The functions, processes, or operations of two or more separate entities in one example may be performed by a single entity in an alternative example;
[0064] Information carried by a particular message in one example may be carried by two or more separate messages in an alternative example;
[0065] Information carried by two or more separate messages in one example may be carried by a single message in an alternative example; and / or
[0066] In alternative examples, the order in which operations are performed and / or the order in which messages are sent may be modified, if possible.
[0067] Certain examples of the present disclosure may be provided in the form of an apparatus / device / network entity configured to perform one or more defined network functions and / or methods thereof. Certain examples of the present disclosure may be provided in the form of a system (e.g., a network or wireless communication system) including one or more such apparatus / device / network entities and / or methods thereof.
[0068] In the present disclosure, UE may refer to one or both of a mobile terminal (MT) and a terminal equipment (TE). MT may provide common mobile network functions, such as one or more of radio transmission and switching, speech coding and decoding, error detection and correction, signaling transmission, and access to a SIM. An International Mobile Equipment Identity (IMEI) code or any other suitable type of identification may be attached to the MT. TE may provide any suitable service to a user via the MT functions. However, it may not contain any network functions itself.
[0069] In this article, the following references are cited:
[0070] [1] 3GPP TS22.261V18.5.0
[0071] [2] 3GPP TS23.501V17.3.0
[0072] [3] 3GPP TS23.502V17.3.0
[0073] 5G systems can support various types of AI / ML operations, including the following three defined in [1]:
[0074] Splitting of AI / ML operations between AI / ML endpoints;
[0075] AI / ML model / data distribution and sharing on 5G systems; and / or
[0076] Distributed / federated learning on 5G systems.
[0077] AI / ML operations / models can be split into multiple parts, for example, based on the current task and environment. The intention is to offload the computationally intensive, energy-intensive parts to the network endpoints and leave the privacy-sensitive and latency-sensitive parts at the end devices. The device executes the operation / model up to a specific part / layer and then sends the intermediate data to the network endpoint. The network endpoint executes the remaining parts / layers and feeds the inference results back to the device.
[0078] Multifunctional mobile terminals may need to switch AI / ML models, for example in response to task and environment changes. The assumption of adaptive model selection is that the model to be selected is available for the mobile device. However, as AI / ML models become more and more diverse, and with limited storage resources in the UE, not all candidate AI / ML models can be preloaded built-in. Online model distribution (i.e., new model download) may be required, where the AI / ML model can be distributed from the network (NW) endpoint to the device when the device needs it to adapt to the changing AI / ML tasks and environment. For this purpose, it may be necessary to continuously monitor the model performance at the UE.
[0079] The cloud server can train a global model by aggregating local models trained by each of multiple end devices (e.g., UE). In each training iteration, the UE performs training based on the model downloaded from the AI server using local training data. The UE then reports the temporary training results to the cloud server, for example, via a 5G UL channel. The server aggregates the temporary training results from the UE and updates the global model. The updated global model is then distributed to the UE, and the UE can perform training for the next iteration.
[0080] For example, based on [1], different levels of interaction are expected between the UE and AF as AI / ML endpoints to exchange AI / ML models, intermediate data, local training data, inference results and / or model performance as application AI / ML traffic.
[0081] However, support for transmitting application AI / ML traffic between AI / ML endpoints (e.g., UE and AF) (e.g., via 5GS) as described above is currently not defined in the existing 5GC data delivery / traffic routing mechanism.
[0082] What is desired is the process and signaling that defines how AI / ML AF affects AI / ML data delivery and / or traffic routing (e.g., through 5GS).
[0083] The AI / ML application may be part of the TE that uses services provided by the MT to support AI / ML operations, and the AI / ML application client may be part of the MT. Alternatively, part of the AI / ML application client may be in the TE, and part of the AI / ML application client may be in the MT.
[0084] The processes disclosed herein may refer to various network functions / entities. The functions and definitions of certain network functions / entities (such as those indicated below) are known to those skilled in the art and are defined, for example, in at least [2] and [3]:
[0085] Application Function (AF);
[0086] Network Exposure Function (NEF);
[0087] Unified Data Repository (UDR);
[0088] Policy Control Function (PCF);
[0089] Session Management Function (SMF);
[0090] User Plane Function (UPF);
[0091] Access and mobility management function (AMF);
[0092] User Equipment (UE); and / or
[0093] Network Repository Function (NRF)
[0094] However, as mentioned above, the skilled person will understand that the present disclosure is not limited to the definitions given in [2] and [3] and that equivalent functions / entities may be used.
[0095] Architecture
[0096] Figure 1 A representation of an architecture according to an exemplary embodiment of the present disclosure is shown.
[0097] like Figure 1 As shown, various entities are connected via multiple interfaces or reference points S11-S17. However, those skilled in the art will appreciate that the present disclosure is not limited to Figure 1 Examples shown. For example, in alternative examples, more or fewer interfaces may be provided, and / or interfaces between different entities may be provided. In addition, any suitable terminology may be used to refer to an interface.
[0098] exist Figure 1 In the example of , reference points S11 and S15 manage the interaction between different logical functions expected from the application functions (AF). These may be implemented, for example, centrally or in a distributed manner as part of separate network entities.
[0099] The AI / ML AF 102 may be a network-side endpoint for AI / ML operations, which may be responsible for AI / ML operations, for example, in the case of federated learning, such as splitting model training, distributing models to UE 104, or collecting and aggregating local models, inference feedback, etc. from multiple UEs. The latter responsibility may be similar to a data collection application function (DCAF). Unlike the DCAF, the processed model or data may not only be exposed to the network data analysis function (NWDAF), but may also be used by other 5GC NFs (e.g., via the provisioning AF as described below) or by other consumer AFs (described below). In addition, the AI / ML AF 102 may play other roles, for example, providing assistance in UPF (re)selection and / or in provisioning transport configuration information in collaboration with the AI / ML application server, and / or providing assistance in generating AI / ML UE policies or routing selection policies (URSP) policies for UEs by the PCF (described below).
[0100] The Provisioning AF 106 may be responsible for provisioning external parameters and models (e.g., collected via the S11 reference point) and / or exposing corresponding events defined, for example, for each AI / ML operation to the 5GC NF through a service-based interface.
[0101] The consumer AF 108 (eg, client consumer AF) may represent AF logic that may act as an external consumer of AI / ML AF models and / or AI / ML operations, for example, via an S15 reference point.
[0102] An AF (AI / ML AF 102, Supply AF 106 or Consumer AF 108) (e.g., when in a trusted domain) may register in a Network Repository Function (NRF), including, for example, DNN, S-NSSAI, supported application IDs, supported event IDs, and any relevant group IDs. The AF may be discovered by other 5GC NFs via the NRF service.
[0103] Reference points S12, S13, S16, and S17 may govern how AI / ML traffic types are collected or distributed between the UE 104 and the network. For example, the S16 interface may be used to collect local trained models, inference results, and / or model performance from the AI / ML application to the direct AI / ML application client 110 on the UE 104. It may also be used to distribute (global) AI / ML models to the AI / ML application 112 on the UE 104 via the direct AI / ML application client 110.
[0104] The S12 reference point may be used, for example, in the case of direct reporting between the UE 104 and the network. In some examples, S12 may be implemented via a user plane PDU session established between the UE and an anchor user plane function (UPF) within the 5GC user plane. In the case of direct reporting between the UE and the network via S12, the AI / ML AF 102 may also assist in UPF (re)selection in collaboration with one or more AI / ML application servers (AI / ML AS) 114 via the S14 reference point.
[0105] For the case of indirect reporting between the UE and the network (e.g., using an indirect AI / ML application client 118), a combination of S17 and S13 may be used. In some examples, S17 may be implemented outside the 3GPP domain.
[0106] In various examples, including one or more or all of the above, NEF 120 may be utilized to expose services when interactions are expected to occur between untrusted entities outside of 3GPP and trusted entities within the 3GPP domain.
[0107] Transfer configuration information
[0108] In various examples of the present disclosure, various examples of AI / ML transmission configuration information may be used. For example, the transmission configuration information may include one or more of address information, traffic type information, auxiliary data or metadata, authentication or security information, and other configuration information. A number of non-limiting examples will now be described.
[0109] For AI / ML AF in a trusted domain and AI / ML AF in an untrusted domain, a service level agreement (SLA) between a mobile network operator (MNO) and an AI / ML application service provider (e.g., ASP) 116 may determine AI / ML transmission configuration information (e.g., for each AI / ML application ID) using any combination of one or more of the following:
[0110] AI / ML AF address;
[0111] AI / ML DNS server address;
[0112] AI / ML traffic type;
[0113] AI / ML metadata information;
[0114] Authentication information; and / or
[0115] Report mode.
[0116] For the AI / ML / AF address, any suitable type of address may be used. For example, the AI / ML AF address may be a fully qualified domain name (FQDN) and / or an IP address and / or a non-IP address that a UE or an AI / ML application client on the UE may communicate with the AI / ML AF or any associated AI / ML application server.
[0117] For the AI / ML DNS server address, any suitable type of address may be used. The AI / ML DNS server address may optionally be used by the UE or an AI / ML application client on the UE to resolve the AI / ML AF address from the FQDN of the AI / ML AF or any associated AI / ML application server to the IP address of the AI / ML AF or any associated AI / ML application server.
[0118] The AI / ML traffic type may indicate the traffic type that the UE and / or the AI / ML application client on the UE may support, for example when interacting with the AI / ML AF or any associated AI / ML application server, or vice versa (e.g., subject to user consent). Non-limiting examples of traffic types include any combination of one or more of AI / ML models, intermediate data, local training data, inference results, and model performance as application AI / ML traffic. In some examples, a unified AI / ML traffic type may be used for all traffic between the UE (or the AI / ML application client on the UE) and the AI / ML AF (or any associated AI / ML application server).
[0119] For AI / ML metadata information, any suitable type of metadata may be used, such as possible AI / ML processing algorithms and associated parameters supported by the AI / ML AF or any associated AI / ML application server, such as for anonymization, aggregation, normalization, federated learning, etc.
[0120] For example, the authentication information may include information that enables the AI / ML AF (or any associated AI / ML application server) and / or the UE (or an AI / ML application client on the UE) to verify the authenticity of the exchanged AI / ML traffic.
[0121] For example, the reporting mode may include direct reporting through 3GPP or indirect reporting via non-3GPP.
[0122] Sharing / application of transmission configuration information
[0123] In some examples, the AI / ML transmission configuration information may be (pre)configured on the AI / ML AF and / or the AI / ML application client on the UE, for example, by the AI / ML application service provider. In some examples, the AI / ML transmission configuration information may be dynamically configured.
[0124] In some examples, the UE may indicate the possibility and / or capability of receiving AI / ML transmission configuration information (or associated policies). For example, such an indication may be made as part of a protocol configuration option (PCO) during a PDU session establishment and / or PDU session modification procedure. The UE may receive at least a portion of the AI / ML transmission configuration information (or associated policies) via any suitable entity (e.g., SMF or AMF) (e.g., via non-access stratum (NAS) messages and commands). This may also be shared as part of the AI / ML UE policy or routing policy (URSP) from the PCF.
[0125] To this end, the AI / ML service provider may use AF requests to influence traffic routing directly (e.g., for AI / ML AF in a trusted domain) or indirectly via NEF (e.g., for AI / ML AF in an untrusted domain), for example as part of a PDU session establishment and / or modification process to update AI / ML transport configuration information and / or associated validity parameters.
[0126] The AI / ML AF request may include any suitable type of information as a traffic description, such as any combination of one or more of a DNN, S-NSSAI, application identifier, application ID, and traffic filtering information addressing the AI / ML AF or any associated AI / ML application server. If the request is via the NEF, the AF request may use the (external) AF service identifier as a traffic description, which the NEF may then convert to any combination of one or more of a DNN, S-NSSAI, application identifier, application ID, and traffic filtering information.
[0127] In some examples, in addition to the AI / ML transmission configuration information, the request may also include one or more other parameters. For example, the parameters may include one or more parameters for enabling the 5GC (e.g., UDR, PCF, or SMF) to compile / generate a transmission policy and associated validity parameters.
[0128] For example, an AF request may include one or more of the following:
[0129] - Potential location information of AI / ML applications, for example in the form of DNAI (e.g., for AI / ML AF in a trusted domain).
[0130] - A target UE identifier, for example, if the transmission configuration information applies to a single UE (e.g., for AI / ML operation splitting or AI / ML model distribution), a group of UEs (e.g., for AI / ML model distribution or federated learning), or any UE (e.g., for supporting any type of AI / ML operation). Any suitable type of identifier may be used. For example, for an AI / ML AF in a trusted domain, the identifier may include a subscription permanent identifier (SUPI), an internal UE identifier, and / or an internal group ID. For an AI / ML AF in an untrusted domain, the identifier may include, for example, a general public subscription identifier (GPSI), an external UE identifier, and / or an external group ID that will be converted by the NEF to a SUPI, an internal UE identifier, and / or an internal group ID.
[0131] - Spatial validity information, such as whether there are any geographic boundaries for transmitting the configuration information. Any suitable type of spatial validity information may be used. For example, for an AI / ML AF in a trusted domain, this information may include a Tracking Area Identifier (TAI) or other suitable resolution of location data. For an AI / ML AF in an untrusted domain, this information may include, for example, a geographic area to be converted from a NEF to a TAI or other resolution of location data.
[0132] - Time validity information, such as whether there is any expiry time for the transmission configuration information.
[0133] - User plane latency requirements, e.g., whether AI / ML traffic types are associated with certain latency requirements to support AI / ML operations.
[0134] - Any other service or quality of experience requirements, for example, whether the AI / ML operation is associated with certain service requirements or quality of experience requirements.
[0135] -Indications associated with certain AI / ML traffic types, types of AI / ML operations, or AI / ML traffic in general.
[0136] Examples of the impact of AI / MLAF on traffic routing and / or reconfiguration of AI / ML traffic
[0137] In various examples of the present disclosure, transmission configuration information may be shared / transmitted between various network entities. Each network entity may store the received transmission configuration information and / or perform update and / or (re) configuration according to the received / stored transmission configuration information.
[0138] The transmission configuration information may include one or more information items as disclosed above and / or any other suitable information. The transmission configuration information may be shared, for example, using an AF request as described above or any other suitable technique. In some examples, the architecture disclosed above or any other suitable architecture may be used to share the transmission configuration information, and to send any other message for performing an update and / or (re) configuration according to the transmission configuration information.
[0139] Figure 2 A representation of a call flow according to an embodiment of the present invention is shown. However, those skilled in the art will appreciate that the present disclosure is not limited to Figure 2 For example, the transmission configuration information may be shared / transmitted between any suitable network entities. In addition, any suitable network entity may store the received transmission configuration information and / or perform an update and / or (re) configuration based on the received / stored transmission configuration information.
[0140] In operations S21 and S22, the AI / ML AF 216 (e.g., AI / ML AF 102) (or NEF 214 (e.g., NEF 120)) may create, update (or delete from) the UDR 212 using AI / ML transmission configuration information and other relevant parameters (e.g., via a unified data management (UDM) service).
[0141] In operation S23a, UDR 212 may store or update the new AI / ML transmission configuration information and other related parameters (or remove the old transmission configuration information, if any). In operation 23b, NEF 214 may respond to the message of operation S22 (eg, create, update, or delete).
[0142] In operation S24, UDR 212 may notify PCF 210. This operation may be based on an earlier subscription of the PCF 210 to the modification of the AF request. For example, any combination of one or more of DNN, S-NSSAI, AI / ML application identifier, SUPI, and internal group identifier may be used as a data key to address PCF 210.
[0143] In operation S25, the PCF 210 may determine whether the AI / ML PDU session or transmission policy is affected, and may update the SM policy, and may notify the SMF 208 based on the SM policy control update.
[0144] In operation S26, SMF 208 may take appropriate actions to reconfigure the user plane of the PDU session that transports the AI / ML traffic. Non-limiting examples of such actions include one or more of the following:
[0145] Allocating a new prefix to UE 202 (e.g., UE 104);
[0146] Updating UPF 206 (e.g., in the target DNAI) with new traffic steering rules; and / or
[0147] Determine whether to relocate UPF 206 (e.g., in coordination with AI / ML AS) by considering requirements provided by AI / ML AF 216, such as regarding location information, target UE ID, spatial validity, temporal validity, UP waiting time and / or service requirements and / or any other suitable indication associated with AI / ML operations.
[0148] In operation S27, SMF 208 may send the target DNAI to AMF 204 to trigger SMF / I-SMF (re)selection, and then notify AMF 204 of the target DNAI information of the current PDU session or the next PDU session, for example via the Nsmf_PDUSession_SMContextStatusNotify service operation.
[0149] In operation S28, the SMF 208 may also update the UE 102 with new or revised AI / ML transmission policies and other session management (SM) subscription information (e.g., via a non-access stratum (NAS) message). Non-limiting examples include one or more of the following:
[0150] Update AI / ML AF address;
[0151] Update AI / ML DNS server address;
[0152] Update AI / ML metadata information;
[0153] Update AI / ML traffic types; and / or
[0154] Updated AI / ML certification information.
[0155] Those skilled in the art will understand that Figure 2 The call flows are merely examples, and various alternatives are within the scope of this disclosure.
[0156] For example, as an alternative process, the PCF (e.g., PCF 210) may use a user configuration update process to update the UE AI / ML policy or the URSP on the UE for the AI / ML transmission policy (e.g., via AMF 204). If so, the traffic descriptor in the AI / ML policy or URSP may be interpreted as the AI / ML transmission policy. For example, the application descriptor may match the AI / ML application OS Id and the OSAPP Id on the UE. The IP descriptor and the domain descriptor (or non-IP descriptor) may match the AI / ML AF address. The connection capability may match the AI / ML traffic type. The routing descriptor (RSD) may match the session and service continuity (SSC), S-NSSAI, DNN, PDU session type, time window, and location criteria set for each AI / ML traffic type or for each unified traffic type. This may be based on the AI / ML transmission configuration information in step S21. In some examples, an access type preference and / or a non-seamless offload indication (or similar information elements) may be used to indicate the use of direct reporting via 3GPP (i.e., S12 reference point) versus indirect reporting via non-3GPP (i.e., a combination of S17 and S13).
[0157] In the above example, the AI / ML application client (e.g., direct AI / ML application client 110) on the UE side (e.g., UE 104) can pass a portion of the AI / ML transmission configuration information to the AI / ML application 112 on the UE 104, for example, based on the S16 interface or based on another logic outside the scope of 3GPP.
[0158] In the above example, the UE (e.g., UE 104) or the AI / ML application client on the UE 104 (e.g., the direct AI / ML application client 110) can correctly convert the FQDN of the AI / ML AF (e.g., AI / ML AF 102) or any associated AI / ML application server (e.g., AI / ML AS 114) to the IP address of the AI / ML AF or any associated AI / ML application server. This can be done, for example, by accessing a local, private, or global DNS server. As described above, the DNS server address or related configuration for the UE can also be optionally shared as part of the transport configuration information if needed (e.g., for private DNS).
[0159] In some examples, the AI / ML AF (e.g., AI / ML A 102) (or NEF (e.g., NEF 120)) may find the PDU session serving SUPI, DNN, S-NSSAI from UDM and the assigned IPv4 address or IPv6 prefix or both from SMF. The AI / ML AF (or NEF) may store the UE IP address or any other external UE ID during PDU session establishment to the UE (or AI / ML application client on the UE). The AI / ML AF (or NEF) may associate and store the mapping of the UE IP address (or any other external UE ID) and the retrieved SUPI (e.g., via UDM / SMF) using the IPv4 address or IPv6 prefix assigned by the SMF.
[0160] The skilled person will understand that Figure 2 One or more of the operations may be optional in some examples, such as the operations indicated by dashed lines.
[0161] Figure 3 is an example that can be used in the present disclosure (such as Figure 1 and / or Figure 2For example, a UE (e.g., UE 104 or UE 202), an AI / ML AF (e.g., AI / ML AF 102 or AI / MLAF 216), an NEF (e.g., NEF 120 or NEF 214), an UDR (e.g., UDR 212), a PCF (e.g., PCF 210), an SMF (e.g., SMF 208), an UPF (e.g., UPF 206), an AMF (e.g., AMF 204), and / or other NFs may be configured to communicate with each other in a manner that is consistent with the present disclosure. Figure 3 The network entities are provided in the form of the network entities shown. The skilled person will understand that the network entities can be implemented as network elements on dedicated hardware, software instances running on dedicated hardware, and / or virtualized functions instantiated on a suitable platform (e.g., on a cloud infrastructure), for example.
[0162] The entity 300 may include a processor (or controller) 301, a transmitter 303, and a receiver 305. The receiver 305 may be configured to receive one or more messages from one or more other network entities by wire or wirelessly, for example, as described above. The transmitter 303 may be configured to send one or more messages to one or more other network entities by wire or wirelessly, for example, as described above. The processor 301 may be configured to perform one or more operations, for example, according to the operations described above.
[0163] The technology described herein can be implemented using any appropriately configured device and / or system. Such a device and / or system can be configured to perform a method according to any aspect, embodiment, example or claim disclosed herein. Such a device may include one or more elements, such as one or more of a receiver, a transmitter, a transceiver, a processor, a controller, a module, a unit, etc., each of which is configured to perform one or more corresponding processes, operations and / or method steps for implementing the technology described herein. For example, the operation / function of X can be performed by a module (or X module) configured to perform X. One or more elements can be implemented in the form of hardware, software, or any combination of hardware and software.
[0164] It should be understood that examples of the present disclosure may be implemented in the form of hardware, software, or any combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage (e.g., a storage device such as ROM), whether erasable or rewritable, or in the form of memory (e.g., RAM, memory chip, device, or integrated circuit), or stored on an optical or magnetically readable medium (e.g., CD, DVD, disk, or tape, etc.).
[0165] It should be understood that storage devices and storage media are embodiments of machine-readable storage that are suitable for storing one or more programs including instructions that, when executed, implement certain examples of the present disclosure. Therefore, certain examples provide a program that includes code for implementing a method, apparatus, or system according to any example, embodiment, aspect, and / or claim disclosed herein, and / or a machine-readable storage storing such a program. In addition, such a program may be transmitted electronically via any medium (e.g., a communication signal carried by a wired or wireless connection).
[0166] While the invention has been shown and described with reference to certain embodiments, it will be understood by those skilled in the art that various changes in form and details may be made without departing from the scope of the invention as defined by the following claims.
[0167] Acronyms and Definitions
[0168] 3GPP: Third Generation Partnership Project
[0169] 5G: 5th Generation
[0170] 5GC: 5G Core
[0171] 5GS: 5G System
[0172] AF: Application Function
[0173] AI: Artificial Intelligence
[0174] AMF: Access and Mobility Management Function
[0175] AS: Application Server
[0176] ASP: Application Service Provider
[0177] DCAF: Data Collection Application Function
[0178] DNAI: Data Network Access Identifier
[0179] DNN: Data Network Name
[0180] DNS: Domain Name Server
[0181] FQDN: Fully Qualified Domain Name
[0182] GPSI: General Public Subscription Identifier
[0183] ID: ID / Identifier
[0184] IMEI: International Mobile Equipment Identity
[0185] IP: Internet Protocol
[0186] I-SMF: Intermediate SMF
[0187] ML: Machine Learning
[0188] MNO: Mobile Network Operator
[0189] MT: Mobile Terminal
[0190] NAS: Non-Access Stratum
[0191] NEF: Network Exposure Function
[0192] NRF: Network Repository Function
[0193] NW: Network
[0194] NWDAF: Network Data Analysis Function
[0195] OS: Operating System
[0196] OSAPP: OS application
[0197] PCF: Policy Control Function
[0198] PCO: Protocol Configuration Options
[0199] PDU: Protocol Data Unit
[0200] RSD: Routing Descriptor
[0201] SIM: Subscriber Identity Module
[0202] SLA: Service Level Agreement
[0203] SM: Session Management
[0204] SMF: Session Management Function
[0205] S-NSSAI: Single Network Slice Selection Auxiliary Information
[0206] SSC: Session and Service Continuity
[0207] SUPI: Subscription Persistent Identifier
[0208] TAI: Tracking Area Identifier
[0209] TE: terminal equipment
[0210] TS: Technical Specification
[0211] UDM: Unified Data Manager
[0212] UDR: Unified Data Repository
[0213] UE: User Equipment
[0214] UL: Uplink
[0215] UP: User Plane
[0216] UPF: User Plane Function
[0217] URSP: UE Routing Selection Policy
Claims
1. A method for configuring artificial intelligence / machine learning (AI / ML) traffic transmission in a wireless communication network, the method include: receiving ( S24 ) a notification of update of the AI / ML configuration information by the policy control function (PCF) from a unified data repository (UDR) storing the AI / ML transmission configuration information; The PCF determines whether an AI / ML protocol data unit (PDU) session or transport policy is affected by the update; as well as Based on determining that the AI / ML PDU session or transmission policy is affected by the update, the PCF notifies (S25) a session management function (SMF) that the session management (SM) policy is updated, The SMF is configured to reconfigure ( S26 ) a PDU session for transmitting AI / ML traffic based on the updated SM policy.
2. The method according to claim 1, further comprising: include: Information about the reconfigured PDU session is sent by the SMF to the Access and Mobility Management Function (AMF).
3. The method according to claim 1, further comprising: include: The SMF updates the user equipment (UE) based on AI / ML transmission configuration information, Among them, updating the UE based on the AI / ML transmission configuration information includes: updating one or more of the AI / ML application function (AF) address, the AI / ML domain name system (DNS) server address and the AI / ML authentication information.
4. The method according to claim 3, further comprising: include: Indicates by the UE the capability to receive AI / ML transmission configuration information during the PDU session establishment or PDU session modification procedure.
5. The method according to claim 1, in, The AI / ML transmission configuration information is received by the UDR as part of an AI / ML application function (AF) request, wherein the AI / ML AF request is received directly from the AI / ML AF or via a network exposure function (NEF), The AI / ML AF request is part of a PDU session establishment procedure or a PDU session modification procedure for updating AI / ML transmission configuration information or associated validity parameters.
6. The method according to claim 5, in, The AI / ML AF request also includes a traffic description, wherein the traffic description includes one or more of a data network name (DNN), single network slice selection assistance information (S-NSSAI), an application identifier, an application ID, and traffic filtering information.
7. The method according to claim 5, in, The AI / ML AF request also includes one or more of potential location information of the AI / ML application, a target UE identifier, spatial validity information, temporal validity information, user plane latency requirements, quality of experience requirements, and an indication associated with the AI / ML traffic type.
8. The method according to claim 1, in, The PDU session transmits AI / ML traffic, and reconfiguring the PDU session includes reconfiguring a user plane of the PDU session, Among them, reconfiguring the user plane of the PDU session includes allocating a new prefix to the UE, updating the user plane function (UPF) with new traffic steering rules, and determining whether to relocate the UPF.
9. The method according to claim 1, in, The AI / ML transmission configuration information is received by the UDR from the AI / ML application function (AF) or the network exposure function (NEF).
10. The method according to claim 5, in, The AI / ML transmission configuration information is pre-configured by the AI / ML service provider on the AI / ML AF and / or the AI / ML application client on the UE.
11. The method according to claim 5, in, The AI / ML transmission configuration information includes one or more of an AI / ML application function (AF) address, an AI / ML domain name system (DNS) server address, an AI / ML traffic type, and AI / ML authentication information.
12. The method according to claim 5, in, The AI / ML transmission configuration information is determined by a service level agreement (SLA) between a mobile network operator (MNO) and an AI / ML application service provider associated with the AI / ML AF.
13. The method according to claim 1, in, The AI / ML transmission configuration information is for each AI / ML application ID.
14. A policy control function (PCF) for configuring artificial intelligence / machine learning (AI / ML) traffic transmission in a wireless communication network, the PCF include: Transceiver; as well as a processor coupled to the transceiver and configured to control the transceiver to: receiving notification of updates to the AI / ML transmission configuration information from a unified data repository (UDR) storing the AI / ML transmission configuration information, Determine if AI / ML protocol data unit (PDU) sessions or transport policies are affected by the update, and Based on determining that the AI / ML PDU session or transmission policy is affected by the update, notifying a session management function (SMF) that a session management (SM) policy is updated, The SMF is configured to reconfigure the PDU session for transmitting AI / ML traffic based on the updated SM policy.
15. The PCF according to claim 14, in, The SMF is configured to send information about the reconfigured PDU session to an Access and Mobility Management Function (AMF), and The SMF is configured to update the user equipment (UE) based on the AI / ML transmission configuration information.
Citation Information
Cited By
Service execution method and device, computer equipment, storage medium and product
CN120547620A
Service execution method and apparatus, computer device, storage medium, and product
CN120547620B
