System and method for AIML services lifecycle operations control and management

Abstract

Disclosed is a method for execution by a VAL (Vertical Application Layer) server. The method involves sending, to an AIML (Artificial Intelligence / Machine Learning) enablement server, a request concerning AIML service lifecycle management. In accordance with an embodiment of the disclosure, the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode. In this way, the AIML service operation can be conditionally managed according to the AIML service operation mode configuration. A corresponding method for the VAL server is also disclosed. Corresponding apparatuses are also disclosed.

Description

[0001] SYSTEM AND METHOD FOR

[0002] AIML SERVICES LIFECYCLE OPERATIONS CONTROL AND MANAGEMENT

[0003] Related Applications

[0004] [1] This patent application claims priority from US provisional patent application no. 63 / 681 ,563 entitled “System and Method for AIML Services Lifecycle Operations Control and Management” and filed on August 9, 2024, the entire disclosure of which being incorporated herein by reference.

[0005] Field of the Disclosure

[0006] [2] This disclosure relates to mobile communication systems, and more particularly to managing AIML (Artificial Intelligence I Machine Learning) service operations.

[0007] Background

[0008] [3] Example details of an application enabling layer architecture, capabilities, and services to support AI / ML (Artificial Intelligence I Machine Learning) services at an application layer are provided in 3GPP (3rd Generation Partnership Project) TR (Technical Report) 23.700-82, entitled Study on application layer support for AI / ML services, version 19.0.0 (2024-06-25), hereinafter “3GPP TR 23.700-82”.

[0009] [4] In AI / ML, there are various AIML (Artificial Intelligence I Machine Learning) services or AIML operations / tasks like model training, inference, data collection, client discovery, selection, etc. for any AIML process or application. AIML services are to be controlled for their lifecycle to manage them like start the AIML service or stop.

[0010] [5] Unfortunately, AIML services are subject to various conditions like data unavailability, device run of power or low on compute, poor network coverage etc., which can lead to interruptions of the AIML services. Also, such information about the current status of the AIML services may not be available and therefore there may be need for mechanisms to handle the interruptions.

[0011] [6] The lifecycle of an AIML service or application can be dependent on an AIML application library or SDK (Software Development Kit) used by the AIML application, which is generally specific to the application. Unfortunately, this creates a vendor-specific lock- in problem to manage the lifecycle of the AIML service.

[0012] [7] Some embodiments disclosed herein set out to solve, address, or mitigate one or more of the foregoing deficiencies.

[0013] Summary of the Disclosure

[0014] [8] Disclosed is a method for execution by a VAL (Vertical Application Layer) server. The method involves sending, to an AIML (Artificial Intelligence I Machine Learning) enablement server, a request concerning AIML service lifecycle management. In accordance with an embodiment of the disclosure, the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode. In this way, the AIML service operation can be conditionally managed according to the AIML service operation mode configuration.

[0015] [9] In some implementations, the AIML service operation mode includes start, continue, pause, or finish. In some implementations, the AIML service operation mode configuration is compute-based, network utilization-based, time-limit-based and / or performance-based. Permutations or combinations are possible. For example, in some implementations, the AIML service operation mode configuration is network utilizationbased, time-limit-based and performance-based.

[0016]

[0010] In some implementations, the AIML service operation mode configuration comprises a network utilization-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when a latency is greater than x milliseconds. In some implementations, the AIML service operation mode configuration comprises a time-limit-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service after a time limit threshold value. In some implementations, the AIML service operation mode configuration comprises a performance-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when model accuracy is 99% achieved. In some implementations, the AIML service operation mode configuration comprises threshold values for compute utilization or power utilization, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service based on the threshold values.

[0017]

[0011] In some implementations, the request includes an AIML service operation ID, an AIML client ID, an AIML clients group ID, and an AIML service operation mode.

[0018]

[0012] In some implementations, the method also involves receiving, from the AIML enablement server, a response to the request, wherein the response conveys a status of the AIML service operation.

[0019]

[0013] In some implementations, the request is a first request, and the method also includes sending, to the AIML enablement server, a second request to update the AIML service lifecycle management. Furthermore, the method includes receiving, from the AIML enablement server, a response to the second request, wherein the response conveys a result of the update.

[0020]

[0014] Also disclosed is a non-transitory CRM (computer readable medium) having recorded thereon statements and instructions that, when executed by a processor of a network node, configure the network node to implement a method as summarized above.

[0021]

[0015] Also disclosed is a network node implementing a VAL (Vertical Application Layer) server. The network node has a network interface configured to communicate with other network nodes, and control circuitry coupled to the network interface. The control circuitry is configured to send, to an AIML (Artificial Intelligence I Machine Learning) enablement server via the network interface, a request concerning AIML service lifecycle management. In accordance with an embodiment of the disclosure, the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode. In this way, the AIML service operation can be conditionally managed according to the AIML service operation mode configuration.

[0022]

[0016] In some implementations, the control circuitry is further configured to implement a method as summarized above.

[0023]

[0017] Also disclosed is a method for execution by an AIML (Artificial Intelligence / Machine Learning) enablement server. The method involves receiving, from a VAL (Vertical Application Layer) server, a request concerning AIML service lifecycle management. In accordance with an embodiment of the disclosure, the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode. Thus, the method further involves implementing the AIML service operation mode for the AIML service operation, in accordance with the request, wherein implementing the AIML service operation mode is conditionally triggered in accordance with the AIML service operation mode configuration. In this way, the AIML service operation can be conditionally managed according to the AIML service operation mode configuration.

[0024]

[0018] In some implementations, the AIML service operation mode includes start, continue, pause, or finish. In some implementations, the AIML service operation mode configuration is compute-based, network utilization-based, time-limit-based and / or performance-based. Permutations or combinations are possible. For example, in some implementations, the AIML service operation mode configuration is network utilizationbased, time-limit-based and performance-based.

[0025]

[0019] In some implementations, the AIML service operation mode configuration comprises a network utilization-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when a latency is greater than x milliseconds. In some implementations, the AIML service operation mode configuration comprises a time-limit-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service after a time limit threshold value. In some implementations, the AIML service operation mode configuration comprises a performance-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when model accuracy is 99% achieved. In some implementations, the AIML service operation mode configuration comprises threshold values for compute utilization or power utilization, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service based on the threshold values.

[0026]

[0020] In some implementations, the request includes an AIML service operation ID, an AIML client ID, an AIML clients group ID, and an AIML service operation mode.

[0027]

[0021] In some implementations, the method also involves sending, to the VAL server, a response to the request, wherein the response conveys a status of the AIML service operation.

[0028]

[0022] In some implementations, the request is a first request, and the method also includes receiving, from the VAL server, a second request to update the AIML service lifecycle management. The method also includes updating the AIML service operation mode for the AIML service operation, in accordance with the second request. The method also includes sending, to the VAL server, a response to the second request, wherein the response conveys a result of the update.

[0029]

[0023] Also disclosed is a non-transitory CRM (computer readable medium) having recorded thereon statements and instructions that, when executed by a processor of a network node, configure the network node to implement a method as summarized above.

[0030]

[0024] Also disclosed is a network node implementing an AIML (Artificial Intelligence / Machine Learning) enablement server. The network node has a network interface configured to communicate with other network nodes, and control circuitry coupled to the network interface. The control circuitry is configured to receive, from a VAL (Vertical Application Layer) server, a request concerning AIML service lifecycle management, wherein the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode. The control circuitry is also configured to implement the AIML service operation mode for the AIML service operation, in accordance with the request, wherein implementing the AIML service operation mode is conditionally triggered in accordance with the AIML service operation mode configuration. In this way, the AIML service operation can be conditionally managed according to the AIML service operation mode configuration.

[0031]

[0025] In some implementations, the control circuitry is further configured to implement a method as summarized above.

[0032]

[0026] Other aspects and features of the present disclosure will become apparent, to those ordinarily skilled in the art, upon review of the following description of the various embodiments of the disclosure.

[0033] Brief Description of the Drawings

[0034]

[0027] Embodiments will now be described with reference to the attached drawings in which:

[0035] Figure 1 is a block diagram of a communication system, in accordance with an embodiment of the disclosure;

[0036] Figure 2 is a sequence drawing of a method of AIML service lifecycle management;

[0037] Figure 3 is a schematic of an example cellular communications system in which some embodiments of the present disclosure may be implemented;

[0038] Figures 4A and 4B are block diagrams of a wireless communication system represented as a 5G network architecture in which some embodiments of the present disclosure may be implemented;

[0039] Figures 5 and 7 are block diagrams of a radio access node according to some embodiments of the present disclosure; Figure 6 is a block diagram that illustrates a virtualized embodiment of a radio access node according to some embodiments of the present disclosure;

[0040] Figures 8 and 9 are block diagrams of a wireless communication device; and

[0041] Figure 10 is a schematic of an example communication system according to some embodiments of the present disclosure.

[0042] Detailed Description of Embodiments

[0043]

[0028] It should be understood at the outset that although illustrative implementations of one or more embodiments of the present disclosure are provided below, the disclosed systems and / or methods may be implemented using any number of techniques. The disclosure should in no way be limited to the illustrative implementations, drawings, and techniques illustrated below, including the exemplary designs and implementations illustrated and described herein, but may be modified within the scope of the appended embodiment along with their full scope of equivalents.

[0044] Introduction

[0045]

[0029] Referring first to Figure 1 , shown is a block diagram of a communication system 100, in accordance with an embodiment of the disclosure. The communication system 100 has a VAL server 110 and an AIML enablement server 120 operatively coupled to one or more AIML enablement clients 140a-140c via at least one network 102. Details of the AIML enablement clients 140a-140c are omitted for simplicity. The communication system 100 may have other components that are not shown for simplicity. For example, the communication system 100 may include components of a core network and components of a radio access network.

[0046]

[0030] The VAL server 110 has a network interface 115 configured to communicate with other nodes of the communication system 100, a CRM 119, and control circuitry 116 coupled to the network interface 115 and the CRM 119. In some implementations, the control circuitry 116 includes a processor 117 that executes software, which can stem from a memory 118. However, other implementations are possible and are within the scope of this disclosure. The VAL server 110 can have additional components, but these are not shown for simplicity.

[0047]

[0031] The AIML enablement server 120 has a network interface 125 configured to communicate with other nodes of the communication system 100, a CRM 129, and control circuitry 126 coupled to the network interface 125 and the CRM 129. In some implementations, the control circuitry 126 includes a processor 127 that executes software, which can stem from a memory 128. However, other implementations are possible and are within the scope of this disclosure. The AIML enablement server 120 can have additional components, but these are not shown for simplicity.

[0048]

[0032] The control circuitry 116 of the VAL server 110 and the control circuitry 126 of the AIML enablement server 120 operates to implement a method of AIML service lifecycle management. The operation by the VAL server 110 and the AIML enablement server 120 will be described below with reference to Figure 2. Although the method of Figure 2 is described below with reference to the communication system 100 shown in Figure 1 , it is to be understood that the method of Figure 2 is applicable to other communication systems. In general, the method of Figure 2 is applicable to any appropriately configured communication system.

[0049]

[0033] At step 2-1 , the VAL server 110 sends an AIML service lifecycle management request to the AIML enablement server 120. In some implementations, the request contains one ore more of an AIML service operation ID, an AIML client ID, an AIML clients group ID, and an AIML service operation mode like start, pause, continue, and finish. The AIML service operation ID can represent both the process ID and service operation (e.g. model training) to identify the AIML service. In some implementations, the VAL server 110 can also configure the AIML service operation mode status reporting configuration to receive the AIML service operation mode notifications from the AIML clients. In some implementations, the VAL server 110 can provide the AIML enablement server 120 with an AIML service operation mode configuration like automatic AIML service operation mode configuration, compute, network utilization-based, time-limit or performance-based to configure conditional triggering of the AIML service operation modes for the AIML enablement clients 140.

[0050]

[0034] At step 2-2, upon receiving the AIML service lifecycle management request, the AIML enablement server 120 performs steps 2-2 and step 2-3. At step 2-2, the AIML enablement server 120 sends an AIML enablement client service operation request to the AIML enablement client(s) 140. In some implementations, the request contains an AIML client ID, an AIML service operation mode, an AIML service operation mode status reporting configuration, and an AIML service operation mode configuration. The AIML enablement client 140 on receiving the AIML service operation mode performs the service operation mode for the AIML service operation.

[0051]

[0035] The start operation mode defines the initiation of the AIML service. The pause operation mode defines the temporary stop of the AIML service. During the pause operation mode, the AIML enablement client 140 can save the current state of the AIML service operation like the current state of the model training. The continue operation mode defines the resuming of the paused operation mode. The AIML enablement client 140 can restore the saved state and resume the AIML operation without any interruptions. The finish mode is defined to stop the AIML operation.

[0052]

[0036] If the automatic AIML service operation mode configuration is ‘Yes’ then the AIML enablement server 120 monitors the AIML enablement clients 140 as per the AIML service operation mode configuration and sends the AIML service operation mode in the AIML enablement client service operation request. If automatic AIML service operation mode configuration is ‘No’, then the AIML enablement server 120 includes the AIML service operation mode configuration in the AIML enablement client service operation request for the AIML enablement clients 140, which is applied by the AIML enablement client 140 after receiving the step 2-2 request message.

[0053]

[0037] At step 2-3, the AIML enablement client 140 sends a response indicating the success or failure of the AIML enablement client service operation response. Based on the AIML service operation mode status reporting configuration, the AIML client reports the service operation mode status to the AIML enablement server 120 or to the configured endpoint. The reporting can be configured as periodic or event-based. For event-based the AIML enablement client 140 monitors the state transition of the AIML service operation and reports to the AIML enablement server 120. The AIML enablement client 140 uses the AIML service operation mode configuration for the conditional triggering of the requested AIML service operation mode for the AIML service operation.

[0054]

[0038] There are many ways that the conditional triggering can occur based on the AIML service operation mode configuration. In some implementations, the AIML service operation mode configuration includes a network utilization-based configuration, and wherein the conditional triggering of the AIML service operation mode includes stopping an AIML service when a latency is greater than x milliseconds. In some implementations, the AIML service operation mode configuration includes a time-limit-based configuration, and wherein the conditional triggering of the AIML service operation mode includes stopping an AIML service after a time limit threshold value. In some implementations, the AIML service operation mode configuration includes a performance-based configuration, and wherein the conditional triggering of the AIML service operation mode includes stopping an AIML service when model accuracy is 99% achieved. In some implementations, the AIML service operation mode configuration comprises threshold values for compute utilization or power utilization, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service based on the threshold values.

[0055]

[0039] At step 2-4, the AIML enablement server 120 provides the AIML service lifecycle management response to the VAL server 110. The message includes AIML service operation ID and the status of the AIML service operation.

[0056]

[0040] At step 2-5, the VAL server 110 sends AIML service lifecycle management update request to the AIML enablement server 120 with the AIML service operation ID and update information.

[0057]

[0041] At step 2-6, the AIML enablement server 120 provides the AIML service lifecycle management update response to the VAL server 110. The message includes the result of the update and the AIML service operation ID.

[0042] According to another embodiment of the disclosure, there is provided a non- transitory CRM having recorded thereon statements and instructions that, when executed by the processor 117 of the network node 114, implement a method as described herein. The non-transitory computer readable medium can be the memory 118 and / or the CRM 119 of the network node 114 shown in Figure 2, or some other non-transitory CRM.

[0058]

[0043] According to another embodiment of the disclosure, there is provided a non- transitory CRM having recorded thereon statements and instructions that, when executed by the processor 107 of the UE 104a, implement a method as described herein. The non- transitory computer readable medium can be the memory 108 and / or the CRM 109 of the UE 104a shown in Figure 2, or some other non-transitory CRM.

[0059]

[0044] Examples of a non-transitory CRM include memory, an SSD (Solid State Drive), a hard disk drive, a CD (Compact Disc), a DVD (Digital Video Disc), a BD (Blu-ray Disc), a memory stick, etc. Other non-transitory CRMs are also possible.

[0060]

[0045] The illustrated examples described herein focus on software implementations. However, other implementations are possible and are within the scope of this disclosure. Other implementations can include additional or alternative hardware components, such as any appropriately configured FPGA (Field-Programmable Gate Array), ASIC (Application-Specific Integrated Circuit), and / or microcontroller, for example. Thus, the data acquisition circuitry 116 of the network node 114 and the data acquisition circuitry 106 of the UE 104a can instead be implemented with any suitable combination of hardware, software and / or firmware.

[0061]

[0046] Further example details are provided in the following sections. It is to be understood that the following sections are very specific and are provided merely for exemplary purposes, such that other implementations are possible and within the scope of the disclosure. Further Example Details

[0062]

[0047] Some embodiments provide a network-controlled Al ML service lifecycle management procedure for a VAL server.

[0063]

[0048] Some embodiments provide AIML service operation modes like start, continue, pause, finish to manage the AIML service operation. The AIML service operation is identified with the AIML service process and / or an AIML service operation ID.

[0064]

[0049] Some embodiments provide an AIML service operation mode configuration to control the AIML service operation and enable conditional triggering of the AIML service operation mode for the requested AIML service operation. The AIML service operation mode configuration is as follows in the table.

[0065]

[0050] In some embodiments, a request is sent to AIML enablement clients to configure the AIML enablement clients with the AIML service operation mode configuration and to enable the conditional triggering of the AIML service operation mode for the requested AIML service operation. It can also configure the AIML enablement clients with AIML service operation mode status reporting configuration as follows:

[0051] Based on the status received from the AIML enablement clients, the AIML enablement server can provide the status of the AIML service operation to the VAL server.

[0066]

[0052] The VAL server can request the network (AIML enablement server) to manage the lifecycle of the AIML service or AIML operation / tasks like model training / inference, device selection, etc.

[0053] It defines the AIML service operation modes like start, pause, continue, finish to manage the lifecycle of the AIML service.

[0067]

[0054] These modes are additionally supported with AIML service operation mode configurations to enable the conditional triggering and / or scheduled triggering of the AIML service operation modes.

[0055] The AIML service is dynamic. For example, the participating device in model training can go offline anytime (like the user shutdowns the application) or the device in poor network coverage. Therefore the solution provides the configuration of event-based state transition reporting of the current status of the AIML service operation. It can also be configured for periodic status reporting of the AIML service operation.

[0068]

[0056] AIML, ML lifecycle management, AIML services, instantiation, termination.

[0069]

[0057] Some or all of the following potential advantages may be realized by one of more embodiments:

[0070] • It provides the vendor-neutral interoperable network-controlled mechanism for the AIML service or AIML operation / tasks lifecycle management using AIML service operation modes.

[0071] • It provides the AIML service operation mode configuration to enable conditional triggering and / or scheduled mechanisms of the AIML lifecycle management. The conditional control operation mode provides cost and resource efficiency.

[0072] • It provides the AIML service operation mode status reporting of the state transition of the AIML service operation during the AIML lifecycle management.

[0073]

[0058] This section relates to how embodiments disclosed herein can be implemented in one or more standards, especially 3GPP TS 23.482. It is note that this section is very specific and that other implementations are possible.

[0074]

[0059] Reason for change:

[0075] • 3GPP TR 23.700-82 Solution#11 is recommended for normative work for the AIML enablement service lifecycle management. The procedure enables to manage the service lifecycle with service operation modes like pause, etc. The service operation mode provides better resource optimization like model training service can be configured to stop automatically if it is using maximum memory or cpu. By using service operation mode, VAL server can pause the AIML service and perform some action related to the service and again resume or continue the AIML service. For example, the VAL server can pause the model training, and load new data set, and again resume the model training service to train on new dataset. There are cases where the model saturates or it does not improve the model performance, in such situations, the VAL server can stop the AIML service, which saves compute and associated costs. Sometimes it is important to pause the model training and do some hyperparameter optimization and then resume the model training. Similarly in FL model training, the VAL server may pause the training of devices which are in poor network coverage. Above explained examples describe managing the AIML service lifecycle is critical to AIML services.

[0076] • Since the AIML service can be compute heavy and due to resource constraints, it is important to consider the efficient usage of the resources. Therefore the solution provides the AIML service operation mode configurations to enable the conditional triggering and / or scheduled triggering of the AIML service operation modes. Also, the heterogeneity of AIML clints like tiny loT devices which are low in resources(like network, compute, power) may require the network controlled conditional triggering for efficient usage of client resources.

[0077] • Also the AIML service is dynamic. For example, the participating device in model training can go offline anytime(like the user shutdowns the application) or the device in poor network coverage. Therefore the solution provides the configuration of event-based state transition reporting of the current status of the AIML service operation. It can also be configured for periodic status reporting of the AIML service operation.

[0078]

[0060] Summary of change:

[0079] Addition of AIML service lifecycle management procedure.

[0061] Consequences if not approved:

[0080] • Fail to control and manage the Al ML service lifecycle

[0081] 8.x AIML service lifecycle management procedure

[0082] 8.x. 1 General

[0083]

[0062] The lifecycle management of the AIML services is an essential requirement for the applications to manage the AIML services like model training, inference, discovery etc. An AIML service (as defined in TR) is equivalent to an AIMLE service and assists in performing or enabling one or more AIML operations.

[0084] 8.x.2 AIML service lifecycle management procedure

[0085]

[0063] Figure 2 shows a AIML service lifecycle management procedure. Example steps are listed below.

[0086]

[0064] Step 2-1 : The VAL server sends AIML service lifecycle management request to the AIML Enablement server. The request contains AIML service operation ID, AIML client ID, AIML clients group ID, AIML service operation mode like start, pause, continue, and finish. AIML service operation ID can represent both the process ID and service operation (e.g. model training) to identify the AIML service. The VAL server can also configure the AIML service operation mode status reporting configuration to receive the AIML service operation mode notifications from the AIML clients. Optionally, the VAL server can provide the AIML server with AIML service operation mode configuration like automatic AIML service operation mode configuration, compute, network utilization-based, time-limit or performance-based to configure conditional triggering of the AIML service operation modes for the AIML clients.

[0087]

[0065] Step 2-2: The AIML Enablement server sends the AIML Enablement client service operation request to the AIML Enablement client(s). The request contains AIML client ID, AIML service operation mode, AIML service operation mode status reporting configuration, AIML service operation mode configuration. The AIML Enablement client on receiving the Al ML service operation mode performs the service operation mode for the Al ML service operation.

[0088]

[0066] The Start operation mode defines the initiation of the Al ML service. The pause operation mode defines the temporary stop of the Al ML service. During the pause operation mode, the Al ML client can save the current state of the Al ML service operation like the current state of the model training. The continue operation mode defines the resuming of the paused operation mode. The AIML client can restore the saved state and resume the AIML operation without any interruptions. The finish mode is defined to stop the AIML operation.

[0089]

[0067] If the Automatic AIML service operation mode configuration is Yes then the AIML Enablement server monitors the AIML enablement clients as per the AIML service operation mode configuration and sends the required AIML service operation mode in the in the AIML Enablement client service operation request. If Automatic AIML service operation mode configuration is No, then the AIML Enablement server includes the AIML service operation mode configuration in the AIML Enablement client service operation request for the AIML enablement clients, which is applied by the AIML enablement client after receiving the step 2 request message.

[0090]

[0068] Step 2-3: The AIML Enablement client sends a response indicating the success or failure of the AIML Enablement client service operation response. Based on the AIML service operation mode status reporting configuration, the AIML client reports the service operation mode status to the AIML Enablement server or to the configured endpoint. The reporting can be configured as periodic or event-based. For event-based the AIML client monitors the state transition of the AIML service operation and reports to the AIML Enablement server. The AIML Enablement client uses the AIML service operation mode configuration for the conditional triggering of the requested AIML service operation mode for the AIML service operation.

[0091]

[0069] Step 2-4: The AIML Enablement server provides the AIML service lifecycle management response to the VAL server. The message includes AIML service operation ID and the status of the AIML service operation.

[0070] Step 2-5: The VAL server sends Al ML service lifecycle management update request to the AIML Enablement server with the AIML service operation ID and update information.

[0092]

[0071] Step 2-6: The AIML Enablement server receives the request and performs steps 2 and step 3 The AIML Enablement server provides the AIML service lifecycle management update response to the VAL server. The message includes the result of the update and the AIML service operation ID.

[0093] 8.x.3 Information flows

[0094] 8.x.3. 1 AIML service lifecycle management request / update request

[0072] Table 8.X.3.1-1 shows the request sent by an AIML service consumer(e.g. VAL server) to an AIML Enablement server for the AIML service lifecycle management request / update request.

[0095] Table 8.X.3.1 -1 : AIML service lifecycle management request / update request

[0096] 8.x.3.2 AIML service lifecycle management response / update response

[0097]

[0073] Table 8.X.3.2-1 shows the request sent by an AIML Enablement server to an AIML service consumer(e.g. VAL server) for the AIML service lifecycle management response or update response.

[0098] Table 8.X.3.1-1 : AIML service lifecycle management response / update response

[0099] 8.x.3.3 AIML Enablement client service operation request

[0074] Table 8.X.3.3-1 shows the request sent by an AIML Enablement server to an AIML Enablement client for the AIML Enablement client service operation request.

[0100] Table 8.X.3.3-1 : AIML Enablement client service operation request

[0101] 8.x.3.4 AIML Enablement client service operation response

[0102]

[0075] Table 8.X.3.4-1 shows the request sent by an AIML Enablement client to an

[0103] AIML Enablement server for the AIML Enablement client service operation response or update response.

[0104] Table 8.X.3.4-1 : AIML Enablement client service operation response

[0105] Additional Details

[0106]

[0076] Additional details are provided below with reference to Figures 3 through 10. It is to be understood that these details are very specific for exemplary purposes only.

[0077] Figure 3 illustrates one example of a cellular communications system 500 in which embodiments of the present disclosure may be implemented. In the embodiments described herein, the cellular communications system 500 is a 5GS (5G system) including a NG-RAN (Next Generation RAN) and a 5GC (5G Core). In this example, the RAN includes base stations 102-1 and 502-2, which in the 5GS include NR base stations (gNBs) and optionally next generation eNBs (ng-eNBs) (e.g., LTE RAN nodes connected to the 5GC), controlling corresponding (macro) cells 504-1 and 504-2. The base stations 502-1 and 502-2 are generally referred to herein collectively as base stations 502 and individually as base station 502. Likewise, the (macro) cells 504-1 and 504-2 are generally referred to herein collectively as (macro) cells 504 and individually as (macro) cell 504. The RAN may also include a number of low power nodes 506-1 through 506-4 controlling corresponding small cells 508-1 through 508-4. The low power nodes 506-1 through 506-4 can be small base stations (such as pico or femto base stations) or RRHs (Remote Radio Heads), or the like. Notably, while not illustrated, one or more of the small cells 508-1 through 508-4 may alternatively be provided by the base stations 502. The low power nodes 506-1 through 506-4 are generally referred to herein collectively as low power nodes 506 and individually as low power node 506. Likewise, the small cells 508-1 through 508-4 are generally referred to herein collectively as small cells 508 and individually as small cell 508. The cellular communications system 500 also includes a core network 510, which in the 5G System (5GS) is referred to as the 5GC. The base stations 502 (and optionally the low power nodes 506) are connected to the core network 510.

[0107]

[0078] The base stations 502 and the low power nodes 506 provide service to wireless communication devices 512-1 through 512-5 in the corresponding cells 504 and 508. The wireless communication devices 512-1 through 512-5 are generally referred to herein collectively as wireless communication devices 512 and individually as wireless communication device 512. In the following description, the wireless communication devices 512 are oftentimes UEs, but the present disclosure is not limited thereto.

[0108]

[0079] Referring now to Figure 4A, shown is a block diagram of a wireless communication system represented as a 5G network architecture composed of core NFs (Network Functions), where interaction between any two NFs is represented by a point-to- point reference point / interface. Figure 4A can be viewed as one particular implementation of the system 500 of Figure 3.

[0109]

[0080] Seen from the access side the 5G network architecture shown in Figure 4A includes a plurality of UEs 613 connected to either a RAN 607 or an (Access Network) as well as an AMF 600. Typically, the R(AN) 607 comprises base stations, e.g. such as eNBs or gNBs or similar. Seen from the core network side, the 5GC NFs shown in Figure 4A include a NSSF 602, an AUSF 604, a UDM 606, the AMF 600, a SMF 608, a PCF 610, and an AF (Application Function) 612.

[0110]

[0081] Reference point representations of the 5G network architecture are used to develop detailed call flows in the normative standardization. The N1 reference point is defined to carry signaling between the UE 613 and AMF 600. The reference points for connecting between the AN 607 and AMF 600 and between the AN 607 and UPF 614 are defined as N2 and N3, respectively. There is a reference point, N11 , between the AMF 600 and SMF 608, which implies that the SMF 608 is at least partly controlled by the AMF 600. N4 is used by the SMF 608 and UPF 614 so that the UPF 614 can be set using the control signal generated by the SMF 608, and the UPF 614 can report its state to the SMF 608. N9 is the reference point for the connection between different UPFs 614, and N14 is the reference point connecting between different AMFs 600, respectively. N15 and N7 are defined since the PCF 610 applies policy to the AMF 600 and SMF 608, respectively. N12 is utilized for the AMF 600 to perform authentication of the UE 613. N8 and N10 are defined because the subscription data of the UE 613 is utilized for the AMF 600 and SMF 608.

[0111]

[0082] The 5GC network aims at separating UP and CP. The UP carries user traffic while the CP carries signaling in the network. In Figure 4A, the UPF 614 is in the UP and all other NFs, i.e., the AMF 600, SMF 608, PCF 610, AF 612, NSSF 602, AUSF 604, and UDM 606, are in the CP. Separating the UP and CP guarantees each plane resource to be scaled independently. It also allows UPFs to be deployed separately from CP functions in a distributed fashion. In this architecture, UPFs may be deployed very close to UEs to shorten the RTT (Round Trip Time) between UEs and data network for some applications involving low latency.

[0083] The core 5G network architecture is composed of modularized functions. For example, the AMF 600 and SMF 608 are independent functions in the CP. Separated AMF 600 and SMF 608 allow independent evolution and scaling. Other CP functions like the PCF 610 and AUSF 604 can be separated as shown in Figure 4A. Modularized function design enables the 5GC network to support various services flexibly.

[0112]

[0084] Each NF interacts with another NF directly. It is possible to use intermediate functions to route messages from one NF to another NF. In the CP, a set of interactions between two NFs is defined as service so that its reuse is possible. This service enables support for modularity. The UP supports interactions such as forwarding operations between different UPFs.

[0113]

[0085] Referring now to Figure 4B, shown is a block diagram of a 5G network architecture using service-based interfaces between the NFs in the CP, instead of the point-to-point reference points / interfaces used in the 5G network architecture of Figure 4A. However, the NFs described above with reference to Figure 4B correspond to the NFs shown in Figure 4A. The service(s) etc. that a NF provides to other authorized NFs can be exposed to the authorized NFs through the service-based interface. In Figure 4B, the service based interfaces are indicated by the letter “N” followed by the name of the NF, e.g. Namf for the service based interface of the AMF 600 and Nsmf for the service based interface of the SMF 608, etc. The NEF 603 and the NRF 601 in Figure 4B are not shown in Figure 4A discussed above. However, it should be clarified that all NFs depicted in Figure 4A can interact with the NEF 603 and the NRF 601 of Figure 4B as necessary, though not explicitly indicated in Figure 4A.

[0114]

[0086] Some properties of the NFs shown in Figures 4A and 4B may be described in the following manner. The AMF 600 provides UE-based authentication, authorization, mobility management, etc. A UE 613 even using multiple access technologies is basically connected to a single AMF 600 because the AMF 600 is independent of the access technologies. The SMF 608 is responsible for session management and allocates IP (Internet Protocol) addresses to UEs. It also selects and controls the UPF 614 for data transfer. If a UE 613 has multiple sessions, different SMFs 608 may be allocated to each session to manage them individually and possibly provide different functionalities per session. The AF 612 provides information on the packet flow to the PCF 610 responsible for policy control in order to support QoS. Based on the information, the PCF 610 determines policies about mobility and session management to make the AMF 600 and SMF 608 operate properly. The AUSF 604 supports authentication function for UEs or similar and thus stores data for authentication of UEs or similar while the UDM 606 stores subscription data of the UE 613. The DN (Data Network), not part of the 5GC network, provides Internet access or operator services and similar.

[0115]

[0087] An NF may be implemented either as a network element on a dedicated hardware, as a software instance running on a dedicated hardware, or as a virtualized function instantiated on an appropriate platform, e.g., a cloud infrastructure.

[0116]

[0088] Figure 5 is a schematic block diagram of a radio access node 700 according to some embodiments of the present disclosure. Optional features are represented by dashed boxes. The radio access node 700 may be, for example, a base station 102 or 106 or a network node that implements all or part of the functionality of the base station 102 or gNB described herein. As illustrated, the radio access node 700 includes a control system 702 that includes one or more processors 704 (e.g., CPUs (Central Processing Units), ASICs (Application Specific Integrated Circuits), FPGAs (Field Programmable Gate Arrays), and / or the like), memory 706, and a network interface 708. The one or more processors 704 are also referred to herein as processing circuitry. In addition, the radio access node 700 may include one or more radio units 710 that each includes one or more transmitters 712 and one or more receivers 714 coupled to one or more antennas 716. The radio units 710 may be referred to or be part of radio interface circuitry. In some embodiments, the radio unit(s) 710 is external to the control system 702 and connected to the control system 702 via, e.g., a wired connection (e.g., an optical cable). However, in some other embodiments, the radio unit(s) 710 and potentially the antenna(s) 716 are integrated together with the control system 702. The one or more processors 704 operate to provide one or more functions of a radio access node 700 as described herein. In some embodiments, the function(s) are implemented in software that is stored, e.g., in the memory 706 and executed by the one or more processors 704.

[0089] Figure 6 is a schematic block diagram that illustrates a virtualized embodiment of the radio access node 700 according to some embodiments of the present disclosure. This discussion is equally applicable to other types of network nodes. Further, other types of network nodes may have similar virtualized architectures. Again, optional features are represented by dashed boxes.

[0117]

[0090] As used herein, a “virtualized” radio access node is an implementation of the radio access node 700 in which at least a portion of the functionality of the radio access node 700 is implemented as a virtual component(s) (e.g., via a virtual machine(s) executing on a physical processing node(s) in a network(s)). As illustrated, in this example, the radio access node 700 may include the control system 702 and / or the one or more radio units 710, as described above. The control system 702 may be connected to the radio unit(s) 710 via, for example, an optical cable or the like. The radio access node 700 includes one or more processing nodes 800 coupled to or included as part of a network(s) 802. If present, the control system 702 or the radio unit(s) 710 are connected to the processing node(s) 800 via the network 802. Each processing node 800 includes one or more processors 804 (e.g., CPUs, ASICs, FPGAs, and / or the like), memory 806, and a network interface 808.

[0118]

[0091] In this example, functions 810 of the radio access node 700 described herein are implemented at the one or more processing nodes 800 or distributed across the one or more processing nodes 800 and the control system 702 and / or the radio unit(s) 810 in any desired manner. In some particular embodiments, some or all of the functions 810 of the radio access node 700 described herein are implemented as virtual components executed by one or more virtual machines implemented in a virtual environment(s) hosted by the processing node(s) 800. As will be appreciated by one of ordinary skill in the art, additional signaling or communication between the processing node(s) 800 and the control system 702 is used in order to carry out at least some of the desired functions 810. Notably, in some embodiments, the control system 702 may not be included, in which case the radio unit(s) 810 communicates directly with the processing node(s) 800 via an appropriate network interface(s).

[0092] In some embodiments, a computer program including instructions which, when executed by at least one processor, causes the at least one processor to carry out the functionality of radio access node 700 or a node (e.g., a processing node 800) implementing one or more of the functions 810 of the radio access node 700 in a virtual environment according to any of the embodiments described herein is provided. In some embodiments, a carrier comprising the aforementioned computer program product is provided. The carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium (e.g., a non-transitory computer readable medium such as memory).

[0119]

[0093] Figure 7 is a schematic block diagram of the radio access node 700 according to some other embodiments of the present disclosure. The radio access node 700 includes one or more modules 800, each of which is implemented in software. The module(s) 800 provide the functionality of the radio access node 700 described herein. This discussion is equally applicable to the processing node 800 of Figure 6 where the modules 800 may be implemented at one of the processing nodes 800 or distributed across multiple processing nodes 800 and / or distributed across the processing node(s) 800 and the control system 702.

[0120]

[0094] Figure 8 is a schematic block diagram of a wireless communication device 900 according to some embodiments of the present disclosure. As illustrated, the wireless communication device 900 includes one or more processors 902 (e.g., CPUs, ASICs, FPGAs, and / or the like), memory 904, and one or more transceivers 906 each including one or more transmitters 908 and one or more receivers 910 coupled to one or more antennas 912. The transceiver(s) 906 includes radio-front end circuitry connected to the antenna(s) 912 that is configured to condition signals communicated between the antenna(s) 912 and the processor(s) 902, as will be appreciated by on of ordinary skill in the art. The processors 902 are also referred to herein as processing circuitry. The transceivers 906 are also referred to herein as radio circuitry. In some embodiments, the functionality of the wireless communication device 900 described above may be fully or partially implemented in software that is, e.g., stored in the memory 904 and executed by the processor(s) 902. Note that the wireless communication device 900 may include additional components not illustrated in Figure 8 such as, e.g., one or more user interface components (e.g., an input / output interface including a display, buttons, a touch screen, a microphone, a speaker(s), and / or the like and / or any other components for allowing input of information into the wireless communication device 900 and / or allowing output of information from the wireless communication device 900), a power supply (e.g., a battery and associated power circuitry), etc.

[0121]

[0095] In some embodiments, a computer program including instructions which, when executed by at least one processor, causes the at least one processor to carry out the functionality of the wireless communication device 900 according to any of the embodiments described herein is provided. In some embodiments, a carrier comprising the aforementioned computer program product is provided. The carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium (e.g., a non-transitory computer readable medium such as memory).

[0122]

[0096] Figure 9 is a schematic block diagram of the wireless communication device 900 according to some other embodiments of the present disclosure. The wireless communication device 900 includes one or more modules 1000, each of which is implemented in software. The module(s) 1000 provide the functionality of the wireless communication device 900 described herein.

[0123]

[0097] Each station 1106A, 1106B, 1106C is connectable to the core network 1104 over a wired or wireless connection 1110. A first UE 1112 located in coverage area 1108C is configured to wirelessly connect to, or be paged by, the corresponding base station 1106C. A second UE 1114 in coverage area 1108A is wirelessly connectable to the corresponding base station 1106A. While a plurality of UEs 1112, 1114 are illustrated in this example, the disclosed embodiments are equally applicable to a situation where a sole UE is in the coverage area or where a sole UE is connecting to the corresponding base station 1106.

[0124]

[0098] The telecommunication network 1100 is itself connected to a host computer 1116, which may be embodied in the hardware and / or software of a standalone server, a cloud-implemented server, a distributed server, or as processing resources in a server farm. The host computer 1116 may be under the ownership or control of a service provider, or may be operated by the service provider or on behalf of the service provider. Connections 1118 and 1120 between the telecommunication network 1100 and the host computer 1116 may extend directly from the core network 1104 to the host computer 1116 or may go via an optional intermediate network 1122. The intermediate network 1122 may be one of, or a combination of more than one of, a public, private, or hosted network; the intermediate network 1122, if any, may be a backbone network or the Internet; in particular, the intermediate network 1122 may comprise two or more sub-networks (not shown).

[0125]

[0099] The communication system of Figure 10 as a whole enables connectivity between the connected UEs 1112, 1114 and the host computer 1116. The connectivity may be described as an OTT (Over-the-Top) connection 1124. The host computer 1116 and the connected UEs 1112, 1114 are configured to communicate data and / or signaling via the OTT connection 1124, using the access network 1102, the core network 1104, any intermediate network 1122, and possible further infrastructure (not shown) as intermediaries. The OTT connection 1124 may be transparent in the sense that the participating communication devices through which the OTT connection 1124 passes are unaware of routing of uplink and downlink communications. For example, the base station 1106 may not or need not be informed about the past routing of an incoming downlink communication with data originating from the host computer 1116 to be forwarded (e.g., handed over) to a connected UE 1112. Similarly, the base station 1106 need not be aware of the future routing of an outgoing uplink communication originating from the UE 1112 towards the host computer 1116.

[0126]

[0100] Any appropriate steps, methods, features, functions, or benefits disclosed herein may be performed through one or more functional units or modules of one or more virtual apparatuses. Each virtual apparatus may comprise a number of these functional units. These functional units may be implemented via processing circuitry, which may include one or more microprocessor or microcontrollers, as well as other digital hardware, which may include DSPs (Digital Signal Processor), special-purpose digital logic, and the like. The processing circuitry may be configured to execute program code stored in memory, which may include one or several types of memory such as ROM (Read Only Memory), RAM (Random Access Memory), cache memory, flash memory devices, optical storage devices, etc. Program code stored in memory includes program instructions for executing one or more telecommunications and / or data communications protocols as well as instructions for carrying out one or more of the techniques described herein. In some implementations, the processing circuitry may be used to cause the respective functional unit to perform corresponding functions according one or more embodiments of the present disclosure.

[0127]

[0101] While processes in the figures may show a particular order of operations performed by certain embodiments of the present disclosure, it should be understood that such order is exemplary (e.g., alternative embodiments may perform the operations in a different order, combine certain operations, overlap certain operations, etc.).

[0128]

[0102] Numerous modifications and variations of the present disclosure are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended embodiments, the disclosure may be practised otherwise than as specifically described herein.

Claims

Claims:

1. A method for execution by a VAL (Vertical Application Layer) server, comprising: sending, to an AIML (Artificial Intelligence I Machine Learning) enablement server, a request concerning AIML service lifecycle management; wherein the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode.

2. The method of claim 1 , wherein the AIML service operation mode comprises at least one of: start, continue, pause, and finish.

3. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration is compute-based, network utilization-based, time-limit-based and / or performance-based.

4. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration is network utilization-based, time-limit-based and performance-based.

5. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration comprises a network utilization-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when a latency is greater than x milliseconds.

6. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration comprises a time-limit-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service after a time limit threshold value.

7. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration comprises a performance-based configuration, and wherein the conditionaltriggering of the AIML service operation mode comprises stopping an AIML service when model accuracy is 99% achieved.

8. The method of claim 1 or claim 2, wherein the AIML service operation mode configuration comprises threshold values for compute utilization or power utilization, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service based on the threshold values.

9. The method of any one of claims 1 to 8, wherein the request comprises: an AIML service operation ID, an AIML client ID, an AIML clients group ID, and an AIML service operation mode.

10. The method of any one of claims 1 to 9, further comprising: receiving, from the AIML enablement server, a response to the request, wherein the response conveys a status of the AIML service operation.

11. The method of any one of claims 1 to 10, wherein the request is a first request, and the method further comprises: sending, to the AIML enablement server, a second request to update the AIML service lifecycle management; and receiving, from the AIML enablement server, a response to the second request, wherein the response conveys a result of the update.

12. A non-transitory CRM (computer readable medium) having recorded thereon statements and instructions that, when executed by a processor of a network node, configure the network node to implement a method according to any one of claims 1 to 11 .

13. A network node implementing a VAL (Vertical Application Layer) server, comprising: a network interface configured to communicate with other network nodes; control circuitry coupled to the network interface and configured to:send, to an AIML (Artificial Intelligence I Machine Learning) enablement server via the network interface, a request concerning AIML service lifecycle management; wherein the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode.

14. The network node of claim 13, wherein the control circuitry is further configured to implement a method according to any one of claims 2 to 11 .

15. A method for execution by an AIML (Artificial Intelligence / Machine Learning) enablement server, comprising: receiving, from a VAL (Vertical Application Layer) server, a request concerning AIML service lifecycle management, wherein the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode; and implementing the AIML service operation mode for the AIML service operation, in accordance with the request, wherein implementing the AIML service operation mode is conditionally triggered in accordance with the AIML service operation mode configuration.

16. The method of claim 15, wherein the AIML service operation mode comprises at least one of: start, continue, pause, and finish.

17. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration is compute-based, network utilization-based, time-limit-based and / or performance-based.

18. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration is network utilization-based, time-limit-based and performance-based.

19. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration comprises a network utilization-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when a latency is greater than x milliseconds.

20. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration comprises a time-limit-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service after a time limit threshold value.

21. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration comprises a performance-based configuration, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service when model accuracy is 99% achieved.

22. The method of claim 15 or claim 16, wherein the AIML service operation mode configuration comprises threshold values for compute utilization or power utilization, and wherein the conditional triggering of the AIML service operation mode comprises stopping an AIML service based on the threshold values.

23. The method of any one of claims 15 to 22, wherein the request comprises: an AIML service operation ID, an AIML client ID, an AIML clients group ID, and an AIML service operation mode.

24. The method of any one of claims 15 to 23, further comprising: sending, to the VAL server, a response to the request, wherein the response conveys a status of the AIML service operation.

25. The method of any one of claims 15 to 24, wherein the request is a first request, and the method further comprises: receiving, from the VAL server, a second request to update the AIML service lifecycle management; andupdating the AIML service operation mode for the AIML service operation, in accordance with the second request; sending, to the VAL server, a response to the second request, wherein the response conveys a result of the update.

26. A non-transitory CRM (computer readable medium) having recorded thereon statements and instructions that, when executed by a processor of a network node, configure the network node to implement a method according to any one of claims 15 to 25.

27. A network node implementing an AIML (Artificial Intelligence I ) enablement server, comprising: a network interface configured to communicate with other network nodes; control circuitry coupled to the network interface and configured to: receive, from a VAL (Vertical Application Layer) server, a request concerning AIML service lifecycle management, wherein the request conveys both an AIML service operation mode to manage an AIML service operation, and an AIML service operation mode configuration to configure conditional triggering of the AIML service operation mode; and implement the AIML service operation mode for the AIML service operation, in accordance with the request, wherein implementing the AIML service operation mode is conditionally triggered in accordance with the AIML service operation mode configuration.

28. The network node of claims 28, wherein the control circuitry is further configured to implement a method according to any one of claims 16 to 25.

Images