Network management operations using an artificial intelligence virtual representation

Abstract

A computer-implemented method performed by a node is provided for network management operations for a telecommunication network using an AI virtual representation of at least a portion of the telecommunication network. The method includes receiving (802) inputs from ML models 5 based on candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The method further includes initiating (810) the AI virtual representation based on the inputs to obtain a recommended action from the candidate actions; and determining (812) (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action, and (ii) a second change of a KPI of the 0 telecommunication network caused by executing the respective candidate action. The method further includes providing (814) an output including the recommended action to take in the telecommunication network based on the first change and the second change.

Description

[0001] NETWORK MANAGEMENT OPERATIONS USING AN ARTIFICIAL INTELLIGENCE VIRTUAL REPRESENTATION

[0002] TECHNICAL FIELD

[0003] The present disclosure is related to wireless communication systems and more particularly to a computer-implemented method performed by a node for network management operations for a telecommunication network using an artificial intelligence (Al) virtual representation of at least a portion of the telecommunication network.

[0004] BACKGROUND

[0005] A network virtual representation (e.g., a network digital twin) refers to a virtual or digital representation / copy of an actual network asset(s) or process(es) that is periodically synchronized with the network and its environment. Virtual representations may be a critical element in network design, operations, and maintenance with, for example, an ability to provide accurate up-to-date virtual representations of a network and its surrounding environment in-line with forecasted impact of changes to the network as a result of either internal or external influencers.

[0006] SUMMARY

[0007] Some approaches may use a virtual representation with respect to manufactured products and operational health and lifespan of manufactured products, or with respect to manufacturing operations involving a virtual factory or shop-floor, for example. However, such approaches do not include a virtual representation of a network (e.g., a telecommunication network) or a portion of a network. Additionally, they also do not address managing operations in the network, particularly based on multiple inputs; and may not address building, synchronization, and / or self-learning of a network virtual representation, particularly from a network management perspective.

[0008] Operational management of a network, such as a radio access network (RAN) can be complex. Some approaches for operational management of a network may include features to try to improve performance or optimize usage in some way, for example, radio sleep or hibernation features to reduce energy usage. Such features, however, may have a singular purpose, such as to reduce energy costs, and may not consider the impact on the network function and quality, as this is typically controlled with limits that aim to restrict such features from having a detrimental effect on network quality. Moreover, activation of such features and adjustment of their limits may be very difficult for network engineers to forecast and visualize the impact. Moreover, network functions and management systems may lack the ability to achieve zero touch control on the network as there may not be ability to forecast an actual effect of changes.

[0009] Certain aspects of the disclosure and their embodiments may provide solutions to these or other challenges. In some embodiments, a computer-implemented method is provided that is performed by a node for network management operations for a telecommunication network using an artificial intelligence (Al) virtual representation of at least a portion of the telecommunication network. The method includes receiving a plurality of inputs from a plurality of machine learning (ML) models. The plurality of inputs are based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The method further includes initiating the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions. The method further includes determining, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a key performance indicator, KPI, of the telecommunication network caused by executing the respective candidate action on the telecommunication network. The method further includes providing, from the Al virtual representation, an output including the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

[0010] According to some embodiments, a node is provided that is configured to perform network management operations for a telecommunication network using an Al virtual representation of at least a portion of the telecommunication network. The node includes processing circuitry and memory coupled with the processing circuitry. The memory includes instructions that when executed by the processing circuitry cause the node to perform operations. The operations include to receive a plurality of inputs from a plurality of ML models. The plurality of inputs are based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The operations further include to initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions. The operations further include to determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a KPI of the telecommunication network caused by executing the respective candidate action on the telecommunication network. The operations further include to provide, from the Al virtual representation, an output including the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

[0011] According to other embodiments, a computer program is provided that includes program code to be executed by processing circuitry of a node configured to perform network management operations for a telecommunication network using an Al virtual representation of at least a portion of the telecommunication. Execution of the program code causes the node to perform operations. The operations include to receive a plurality of inputs from a plurality of ML models. The plurality of inputs are based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The operations further include to initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions. The operations further include to determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a KPI of the telecommunication network caused by executing the respective candidate action on the telecommunication network. The operations further include to provide, from the Al virtual representation, an output including the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

[0012] According to other embodiments, a computer program product is provided that includes a non-transitory storage medium including program code to be executed by processing circuitry of a node. Execution of the program code causes the node to perform operations. The operations include to receive a plurality of inputs from a plurality of ML models. The plurality of inputs are based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The operations further include to initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions. The operations further include to determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a KPI of the telecommunication network caused by executing the respective candidate action on the telecommunication network. The operations further include to provide, from the Al virtual representation, an output including the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

[0013] Certain embodiments may provide one or more of the following technical advantages. Based on inclusion of operations of an Al virtual representation, network operations may be improved with a decreased risk of impacting quality metrics. Additionally, operations of the Al virtual representation may provide a zero touch approach that may achieve advances in targeted improvements, such as energy efficiency, increased bandwidth, reduced latency, lowered call drop rates, etc.

[0014] BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this application, illustrate certain non-limiting embodiments of inventive concepts. In the drawings:

[0016] Figure 1 is schematic diagram of a component view of an Al driven virtual representation for network operations in accordance with some embodiments;

[0017] Figure 2 is a flowchart of an overview of logic of an Al driven virtual representation in accordance with some embodiments;

[0018] Figure 3 is a flowchart for fetching dependency data and modelling dependencies in accordance with some embodiments;

[0019] Figure 4 is a flowchart for data validation and ML model deployment in accordance with some embodiments;

[0020] Figure 5 is a flowchart for ML model training in accordance with some embodiments;

[0021] Figure 6 is a flowchart for evaluation of the target output in accordance with some embodiments;

[0022] Figure 7 is a flowchart for data augmentation in accordance with some embodiments;

[0023] Figure 8 is a flow chart illustrating an example of operations performed by a node in accordance with some embodiments;

[0024] Figure 9 is a block diagram of a node in accordance with some embodiments;

[0025] Figure 10 is a block diagram of a communication system in accordance with some embodiments;

[0026] Figure 11 is a block diagram of a virtualization environment in accordance with some embodiments;

[0027] Figure 12 is schematic diagram of an example open- radio access network (O-RAN) architecture in accordance with some embodiments; and Figure 13 is a schematic diagram of an example implementation of a virtual representation distributed in a cloud native manner across different hosts / nodes.

[0028] DETAILED DESCRIPTION

[0029] Some of the embodiments contemplated herein will now be described more fully with reference to the accompanying drawings. Embodiments are provided by way of example to convey the scope of the subject matter to those skilled in the art, in which examples of embodiments of the present disclosure are shown. The present disclosure may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of present inventive concepts to those skilled in the art. It should also be noted that these embodiments are not mutually exclusive. Components from one embodiment may be tacitly assumed to be present / used in another embodiment.

[0030] ML may play a key role in virtual representations as it may model a system in a very detailed way to represent its real-time behaviours and such a ML model may be used to further understand how a network will respond with changes.

[0031] As discussed above, there currently exist certain challenges. Certain aspects of the disclosure and their embodiments may provide solutions to these or other challenges. Some embodiments include a five-dimension Al driven virtual representation to virtualize physical and functional aspects of a telecommunication network, e.g., a RAN. The virtual representation can be used for various use cases and targets. Non-limiting examples of network management use cases include (1) a root cause analysis (RCA) virtual representation to analyze specific issues in a network and explore root causes; (2) using a network virtual representation to identify energy usage patterns and forecasting for trial-and-error type interactions; (3) using a network virtual representation for managing data speed and capacity in the network; or (4) using a network virtual representation to support business to business (B2B) use cases and real-time applications in making decisions, where changes to the network are validated before they are applied.

[0032] Figure 1 is schematic diagram of a component view of an Al driven virtual representation for network operations in accordance with some embodiments. As shown in Figure 1, a node 100 includes a five-dimension Al virtual representation 104. Al virtual representation 104 in this example is an interactive virtualized representation of a network and functions of the network that is non-invasive but facilitates improvement of network operational states. Each of the five dimensions and how the Al virtual representation 104 is used are discussed further herein and may provide a fully automated interactive virtual representation for network adjustment with a reduced risk of network degradation.

[0033] Still referring to Figure 1, the five-dimensions include: (1) dependency modelling of a first ML model 106, (2) predictive modelling of a second ML model 108, (3) interactive augmentation with a third ML model 110, (4) a three-part targeted output 118 from a fourth ML model 112 showing impact on a target and surrounding effects; and (5) near real-time network 102 synchronization with an error feedback 120.

[0034] As shown, the virtual representation 104 in Figure 1 uses four ML models to facilitate its operation: (1) a predictive ML model shown as second ML model 108 that can provide near real-time forecasting; (2) a reinforcement learning (RL) ML model shown as fourth ML model 112 that can provide state action recommendations; (3) a graph based learning ML model shown as first ML model 106 that can leam dependencies between functions, states, procedures, and assets in the network 102; and (4) a data augmentation ML model shown as third ML model 110 that can leam how to augment dependency data based on actions. It is noted, however, that the present disclosure is not limited to this number or type of ML models, and other appropriate ML models may be used as deemed appropriate for a use case(s). The ML models can internally feed one another and can adaptively assist each other in their learning process particularly, for example, with the prediction ML model and the dependency ML model feeding into the state machine ML model. Such interactions are included in Figure 2 discussed further herein.

[0035] Some examples include an Al virtual representation that self organizes, based on a targeted use case and localized network conditions, as an O-RAN based rApp, which is software application designed to run on a non-real time RAN Intelligent Controller, as discussed further herein.

[0036] Additionally or alternatively, the Al virtual representation may include the following characteristics built-in and utilizing O-RAN non-real-time (non-RT) RAN intelligent controller (RIC) framework components: (1) Fully automated zero touch network operations and improvements (e.g., rApp communications via a non-RT RIC framework); and / or (2) a human assisted discovery and learning process (e.g., human input in a loop) involving an rApp - service management and orchestration (SMO) external exposure interface communication.

[0037] Moreover, some examples include an input 116 that includes a variable action mode that can provide flexibility to a user to choose actions on the virtual representation 104, and the virtual representation 104 can demonstrate the change (e.g., impact), if any, of the chosen actions towards a target output. Other examples include an input 116 that includes a fixed action mode that is autonomous, and intelligence for exploring actions that are either pre-defined or selfdiscovered by the virtual representation 104.

[0038] Still other examples include automatic important feature impact discovery, where network performance and configuration is analyzed using a combination of explainable Al (xAI) processes and graph based approaches to discover the impact of given input actions.

[0039] Yet other examples include a prediction ML model(s) combined with input augmentations to forecast near real-time changes in a network towards a virtual representation target output and overall network impact.

[0040] Each virtual representation 104 may be considered to be a unique entity with a specific objective as its output 118 (also referred to here as a target 118 or target output 118). The output 118 can include three parts (1) the proposed change in input state; (2) a result(s) that includes a change on at least a portion of the network 102 and / or an change on a KPI(s) in the network 102. The discovery of KPIs can be pre-defined and / or automatically identified using xAI processes, such as feature dependency methods; and / or (3) a representation of the result(s) in an intuitive manner, such as initiating a display of advanced visualizations.

[0041] As shown in Figure 1, the virtual representation 104 is synchronized (e.g., continuously synchronized) with the actual network 102 to try to ensure both accuracy and continuous learning through a feedback 120 loop that indicates error levels between projected views of a change(s) in the network 102 and actual states in real-time.

[0042] When first activated (e.g., spinning up the virtual representation to address a “what- if’ scenario from a network management perspective), the virtual representation 104 can sequentially undergo training loops where it consumes data 114 about the network 102 and uses this for establishing ground truths and more accurate prediction results. When the predicted target(s) falls with an acceptable defined error threshold, the virtual representation 104 can have further interactions as illustrated in the example shown in Figure 2.

[0043] Figure 2 is a flowchart of an overview of logic of an Al driven virtual representation 104 in accordance with some examples.

[0044] Figure 2 starts 200 with a determination regarding whether external data 202 should be fetched 206 related to an observation in the network 102. External data 202 may include, for example, data from an external platform, such as a weather source, that may have an association with an observation in network 102. Based on the fetched external data 202, first ML model 106 can be defined 208 (e.g., hyperparameters of the first ML model 106 can be defined). First ML model 106 can be a ML model such as a graph neural network (GNN) for dependency graph modelling relationships between datapoints.

[0045] Additionally, a further determination can be made regarding whether network data 204 should be fetched and / or reduced 210 that is related to the observation in the network 102. Network data 204 may include, for example, data from a network management component(s), a RAN component(s), a protocol(s) with respect to RAN (e.g., LI, L2, L3 protocol(s), a network management alarm(s), a network management alert(s), configuration management data, etc.. Based on the fetched network data 204, a prediction ML model(s) (e.g., second ML model 108) can be automatically selected.

[0046] Second ML model 108 can be a regression type ML model that learns from historical data to forecast / estimate future trends.

[0047] Once the first ML model 106 and second ML model 108 are defined / selected, states and profiles are built 214.

[0048] In operation 218, the second ML model 108 (e.g., a prediction ML model) is trained on training data 216; and, in operation 222, the second ML model 108 is validated using validation data 220.

[0049] In operation 224, a determination is made regarding whether the virtual representation 104 is ready. If no, operations return to operations 218 and 222 to re-train and validate the second ML model 108. If yes, operations proceed to use the virtual representation 104.

[0050] Input 116 can be a candidate action(s) to take in the virtual representation 104 of at least a portion of network 102. In one example, an action can be to tilt an antenna of a base station by a specified angle(s) in response to an observation in the network 102 that a call of a user equipment (UE) dropped. As previously discussed, the input 116 can be from a variable action mode that can provide flexibility for a user to choose an action(s) on the virtual representation 104, and the virtual representation 104 can demonstrate the change, if any, of the chosen action(s) towards a target output 118. For a variable mode, a user (e.g. human or a UE) can decide from among several / all possible actions, which action(s) to take. However, with variable inputs 116, this may result in little or no change in the target output 118 as the change may be unrelated or substantially unrelated. Continuing with the example of a dropped call, another example of a variable mode input 116 may include action such as to collect certain measurement data from network 102.

[0051] Alternatively or additionally, input 116 can be an input from a fixed action mode (also referred to herein as “fixed mode” or “machine to machine mode” (M2M mode)) that is autonomous, and intelligence for exploring actions (e.g., different antenna tilt angles) that is either pre-defined or self-discovered by the virtual representation 104. Fixed mode inputs 116 can be a set of valid and related inputs that are initially provided but also may be learned over time by the virtual representation 104 through its own internal process of using either or both graph theory and xAI. A fixed action input 116 may be received from a manual / human actor or an automated machine process, which can allow for the virtual representation 104 to be either a learning and discovery tool or a rapid and intelligent solution finder against a specified target outputll8.

[0052] As shown in the example in Figure 2, in operation 226, a determination is made regarding whether variable mode is being used. If yes, possible actions are retrieved 228; and a request is received 230 from input 116 selecting an action(s) from the retrieved possible actions.

[0053] Alternatively, a request for an action(s) is received in operation 230 based on input 116; and, in operation 232, a determination is made regarding whether M2M mode is being used. If no, user interaction is needed in operation 234 to specify the action(s) and operations proceed to operation 236. If yes, operations proceed to operation 236.

[0054] In operation 236, a result(s) is calculated and included in output 118. In operation 238, as part of feedback 120, the result(s) is reviewed to determine an error rate for the result(s); and in operation 230, a determination is made regarding whether the error rate is acceptable or not.

[0055] If the error rate is acceptable, the operations end and the virtual representation 104 is ready to be used again. If the error rate is not acceptable, operations return to operation 218 to re-train the fourth ML model 112 and proceed through operations 226-230.

[0056] As shown in the example in Figure 1, the virtual representation 104 uses both a historical (time = t (now and historical) and up-to date real-time (time = t+1 (actual) synchronization process to maintain an accurate state to predict future sates in the network 102.

[0057] Some operations of Figures 1 and 2 are now discussed in more detail.

[0058] Figure 3 is a flowchart for fetching dependency data and modelling dependencies in relation to at least operations 206, 208, and / or 214. These operations include initial steps to identify dependencies in data in terms of data ingestion, categories of data sets, and availability of the data. Operations start 300 and proceed to operation 302 to determine whether dependencies were identified in external data 202 and / or network data 204. If no, operations proceed to fixed action mode 304 which includes predefined dependencies / rules for self-learning data dependencies.

[0059] If yes, operations proceed to operation 306 to determine whether the data is available. If no, in operations 206 / 210 data retrieval is initiated. If data is available, in operation 308, a data broker(s) is established for each data set and dependency graphs are built using the first ML model 106 (e.g., a GNN or large language model (LLM)). In operation 214, the first ML model 106 is then processed further to build states and profiles, as identified in operation 214 in Figure 2.

[0060] Figure 4 is a flowchart for data validation and ML model deployment in relation to at least operations 218, 222, 224, and / or 230 of Figure 2. In the flowchart in Figure 4, the training data set(s) 216 are validated against the set target output 118 and deployed for further operations including deployment of the fourth ML model 112 for inference. Operations start 400 and proceed to operation 402 to determine whether a trained fourth ML model (s) 112 is available. If no, operations proceed to operation 402 to initiate retrieval of a fourth ML model 112.

[0061] If yes, operations proceed to operation 406, to determine whether a training data set 216 is validated for the target output 118. If yes, in operation 408, the fourth ML model 112 is deployed for inferencing.

[0062] If no, operations proceed to operation 410 to initiate data augmentation. For example, when there is a significant deviation in the training data set 216 to achieve a target 118, additional data is planned to be augmented and is done in a dynamic manner. The operations of Figure 4 can be important in order to arrive closer to a target 118 without significant deviance.

[0063] Figure 5 is a flowchart for ML model training in relation to at least operations 218, 222, and / or 230 in Figure 2. The flowchart in Figure 5 includes operations for training of the fourth ML model 112 including validating the training parameters (e.g., hyper parameters) of the fourth ML model 112 and training the fourth ML model 112 according to various ML training processes understood by one of skill in the art.

[0064] Operations start 500 and proceed to operation 502 to determine whether the version of fourth ML model 112 is a latest version. If no, in operation 504, a latest version of ML model 112 is imported and operations proceed to operation 502.

[0065] If yes, operations proceed to operation 505 to determine whether training data 216 is available. If no, in operation 508, fetching dependency data 214 is initiated.

[0066] If training data 216 is available, operations proceed to operation 510 to validate an infrastructure based on the infrastructure for the training data 216 on which the fourth ML model 112 was trained. As shown in block 512, the infrastructure may be an on-site infrastructure for the network 102 or may be an external infrastructure (e.g. , a weather source). From the validation operation 510, operations proceed to operation 514 to set a hyper parameter(s) of the fourth ML model 112. As shown in block 516, the hyper parameters are based on the selected ML model and training process. In operation 518, training of the fourth ML model 112 is initiated.

[0067] Figure 6 is a flowchart for evaluation of the target output 118 in relation to at least operations 226, 228, 232, 234, 236, and / or 238 in Figure 2. Analysis included in the target output 118 based on fixed or variable-action mode is used to determine whether the target output 118 was achieved. The analysis can include an implicit functionality that involves KPI analysis for analyzing impact(s) on the network 102. However, backward propagation also can be used where the target output 118 is set and backward propagated to understand and leam the KPIs, network logs, etc. involved to build and arrive at a target output 118.

[0068] Operations start 600 and proceed to operation 602 to determine whether a variable mode was used. If no, in operation 604, a determination is made whether the target output 118 is achieved based on a fixed action mode being set.

[0069] If a variable action mode was used, in operation 606, a determination is made whether the target output 118 is available. If yes, in operation 608, a display of an analysis of the target output 118 is initiated. The display of the analysis can include, for example, a visual and / or graphical analysis of how the action included in the target output 118 influences / changes the network 102.

[0070] Additionally, in operation 610, an analysis is performed of KPI(s) degradation and network changes. As shown in block 612, this analysis can be based on observed versus expected KPI degradation and network change(s). For example, an observed rate of dropped calls in network 102 (e.g., < 40%) versus an expected change from the analysis (e.g., < 60% dropped calls).

[0071] In operation 614, the virtual representation 104 and the target output 118 is saved, e.g., for future inferencing.

[0072] Figure 7 is a flowchart for data augmentation in relation to at least operations 228, 230, and / or 218 in Figure 2. Figure 7 shows a flow for the input data set(s) 116 . The data augmentation process builds upon the dynamicity of the virtual representation 104 to augment input data 116 as part of its self-learning process to establish new data brokers for additional data sets in order to try to achieve the target output 118.

[0073] Operations start 700 and proceed to operation 702 to determine whether augmentation of input data 116 is needed. If no, in operation 704, the process of the virtual representation 104 is continued. If yes, operations proceed to operation 706 to determine whether dynamic (e.g., on-the-fly or in real -time / near-real -time) data ingestion is needed. If no, in operation 708, an existing data broker is fetched for the data. If yes, operations proceed to operation 710 to identify a data source broker; and in operation 712, the data source broker is established. In operation 714, the process for prediction from the virtual representation 104 continues.

[0074] Implementation of the virtual representation 104 can be executed on a single host / node 100, as shown in Figure 1 for example. Alternatively, as shown in the example in Figure 13, implementation of the virtual representation 104 can be distributed in a cloud native manner across different hosts / nodes. Figure 13 differs from Figure 2 by including four example cloud hosts 1300, 1302, 1304, and 1306 for portions of the virtual representation 104. As such for brevity, portions of Figure 13 that correspond to Figure 2 are not repeated and are incorporated by reference with respect to Figure 13. In the non-limiting example shown in Figure 13, the predictor ML model (e.g., the second ML model 108) is deployed on cloud host 1304; a graph ML model (e.g., first ML model 106) is deployed on cloud host 1300; a data augmentation ML model (e.g., third ML model 110) is deployed on cloud host 1302; and a state ML model (e.g., fourth ML model 112) and target output 118 are deployed on cloud host 1306.

[0075] In another example implementation, virtual representation 104 can be an rApp or a service in a service management and orchestration (SMO) platform having a non-real-time (non-RT) RIC framework in an open-RAN (O-RAN) architecture, as shown in the example in Figure 12. This example O-RAN architecture includes:

[0076] (1) SMO framework 1200 includes a component that oversees orchestration, management, and automation of RAN elements, and supports the 01, Al, and 02 interfaces;

[0077] (2) Near-RT RIC 1204 includes a logical function that enables near-real-time control and optimization of O-RAN elements and resources via fine-grained data collection and actions over an E2 interface;

[0078] (3) Non-RT RIC 1202 includes a logical function that enables non-real-time control and optimization of RAN elements and resources, AI / ML workflow including model training and updates, and policy-based guidance of applications / features in near-RT RIC;

[0079] (4) O-RAN Central Unit (O-CU) 1214 includes a logical node hosting radio resource controls (RRC), service data adaptation protocol (SDAP) and packet data convergence protocol (PDCP) protocols;

[0080] (5) O-RAN Central Unit - Control Plane (O-CU-CP) 1208 includes a logical node hosting the RRC and the control plane part of the PDCP protocol; (6) O-RAN Central Unit - User Plane (O-CU-UP) 1210 includes a logical node hosting the user plane part of the PDCP protocol and the SDAP protocol;

[0081] (7) O-RAN Distributed Unit (O-DU) 1212 includes a logical node hosting radio link control (RLC) / medium access control (MAC)ZHigh-physical (PHY) layers based on a lower layer functional split;

[0082] (8) O-RAN Radio Unit (O-RU) 1214 includes a logical node hosting Low-PHY layer and radio frequency (RF) processing based on a lower layer functional split. This is similar, for example, to the Third Generation Partnership Project (3GPP’s) transmission / reception points (“TRP”) or remote radio head (“RRH”) but more specific in including the Low-PHY layer (fast Fourier transform (FFT)Zinverse FFT (iFFT), physical random access channel (PRACH) extraction);

[0083] (9) xAPP (not depicted) includes a network function which, for example, can be an independent software plug-in to the Near-RT RIC 1204 platform to provide functional extensibility to a RAN by third parties;

[0084] (10) rAPP (not depicted) includes a network function which, for example, can be an independent software plug-in to the Non-RT RIC 1202 platform to provide functional extensibility to the RAN by third parties;

[0085] (11) Node 1206 includes, for example, an open evolved nodeB (O-eNB); and

[0086] (12) O-Cloud 1216 includes hardware andZor software components that can provide cloud computing capabilities to execute RAN network functions.

[0087] Operations of a node 900 (implemented using the structure of Figure 9) will now be discussed with reference to the flow chart of Figure 8 according to some embodiments of the present disclosure. For example, modules may be stored in memory 906 or ML models 908 of Figure 9, and these modules may provide instructions so that when the instructions of a module are executed by respective node processing circuitry 904, node 900 performs respective operations of the flow chart.

[0088] In some embodiments, a computer-implemented method is provided performed by a node for network management operations for a telecommunication network using an Al virtual representation of at least a portion of the telecommunication network. The method includes receiving (operation 802 in Figure 8) a plurality of inputs from a plurality of ML models. The plurality of inputs are based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network. The method further includes initiating (operation 810 in Figure 8) the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions. The method further includes determining (operation 812 in Figure 8), with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a KPI of the telecommunication network caused by executing the respective candidate action on the telecommunication network. The method further includes providing (operation 814 in Figure 8), from the Al virtual representation, an output including the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

[0089] In some embodiments, the plurality of inputs from the plurality of ML models include one or more of (i) a first input from a first ML model including a relationship among data related to the observation from at least one the telecommunication network and a source external to the telecommunication network, (ii) a second input from a second ML model including an estimate of the first change and the second change of the respective candidate action, and (iii) a third input including feedback related to an accuracy of the output of the Al virtual representation.

[0090] The plurality of inputs, in some embodiments, are based on at least one of (i) a first input initiated from a user interface regarding the plurality of candidate actions, and (ii) a second input regarding the plurality of candidate actions that is specified for, or discovered by, the Al virtual representation.

[0091] In some embodiments, the method further includes synchronizing (operation 800 in Figure 8) in real-time or near real-time at least the portion of the telecommunication network and the Al virtual representation.

[0092] The first ML model, in some embodiments, includes a graph based ML model that learns dependencies in the data related to the observation between a function, a state, a procedure, and / or an asset in at least the potion of the telecommunication network.

[0093] The second ML model, in some embodiments includes a predictive model that provides the estimate in near real-time.

[0094] The output, in some embodiments, is from a fourth ML model including a reinforcement learning model.

[0095] The relationship from the first ML model can be identified based on (i) identifying the data from the observation that is related to the plurality of candidate actions, (ii) determining that the data is available, (iii) determining dependencies among the data, and (iv) building a plurality of states and profiles for the dependencies.

[0096] In some embodiments, the method further includes determining (operation 804 in Figure 8) whether data related to the plurality of inputs is available for the respective ML models in the plurality of ML models; and performing (operation 806 in Figure 8) one of (i) deploying the plurality of ML models when the data is available, and (ii) initiate obtaining the data when the data is not available.

[0097] In some embodiments, the method further includes training (operation 808 in Figure 8) at least one of the plurality of ML models.

[0098] The training (operation 808 in Figure 8) can include (i) determining whether the at least one of the plurality of ML models is a latest version, and (ii) when the at least one of the plurality of ML models is a latest version, determine whether training data is available, or (iii) when the at least one of the plurality of ML models is not a latest version, obtain the latest version.

[0099] The training (operation 808) can further include to (i) validate an infrastructure for the training, (ii) set at least one hyperparameter for the at least one of the plurality of ML models, and (iii) initiate the training.

[0100] Providing (operation 814 in Figure 8) the output can be based on setting the respective candidate actions and backward propagating the respective candidate actions to arrive at the recommended action based on the respective first change and the respective second change.

[0101] Obtaining the data can include using a third ML model to fetch the data via a data broker. Obtaining the data can also include (i) identifying a data source for the data, and (ii) establishing a data broker to obtain the data.

[0102] In some embodiments, the method further includes initiating (operation 816 in Figure 8) a display of at least one of the first impact and the second impact.

[0103] In other embodiments, the method further includes saving (operation 818 in Figure 8) the Al virtual representation and the output.

[0104] The Al virtual representation can be located in a cloud-based node.

[0105] In other embodiments, the Al virtual representation is distributed across a plurality of cloud-based nodes.

[0106] In yet other embodiments, the node includes a component comprising an rApp in a SMO platform.

[0107] In still other embodiments, the node includes a component for a non-real-time RAN (RIC) in an O-RAN.

[0108] In some embodiments, the Al virtual representation includes a digital twin.

[0109] Various operations from the flow chart of Figure 8 may be optional with respect to some embodiments of nodes configured to configured to perform network management operations for a telecommunication network using an Al virtual representation of at least a portion of the telecommunication network and related methods. For example, the operations of blocks 800, 804- 808, and / or 816-818 may be optional in some embodiments. Figure 9 shows a node 900 in accordance with some embodiments. As used herein, node refers to equipment capable, configured, arranged and / or operable to configured to perform network management operations for a telecommunication network using an Al virtual representation of at least a portion of the telecommunication network and related methods. Examples of nodes include, but are not limited to, cloud-based nodes, access points (APs) (e.g., radio access points), base stations (BSs) (e.g., radio base stations, Node Bs, eNBs, NR NodeBs (gNBs)), O-RAN nodes, or components of an O-RAN node (e.g., intelligent controller, O-RU, O- DU, O-CU).

[0110] Base stations may be categorized based on the amount of coverage they provide (or, stated differently, their transmit power level) and so, depending on the provided amount of coverage, may be referred to as femto base stations, pico base stations, micro base stations, or macro base stations. A base station may be a relay node or a relay donor node controlling a relay. A network node may also include one or more (or all) parts of a distributed radio base station such as centralized digital units and / or remote radio units (RRUs), sometimes referred to as Remote Radio Heads (RRHs). Such remote radio units may or may not be integrated with an antenna as an antenna integrated radio. Parts of a distributed radio base station may also be referred to as nodes in a distributed antenna system (DAS).

[0111] Other examples of network nodes include multiple transmission point (multi-TRP) 5G access nodes, multi-standard radio (MSR) equipment such as MSR BSs, network controllers such as radio network controllers (RNCs) or base station controllers (BSCs), base transceiver stations (BTSs), transmission points, transmission nodes, multi-cell / multicast coordination entities (MCEs), Operation and Maintenance (O&M) nodes, Operations Support System (OSS) nodes, Self-Organizing Network (SON) nodes, positioning nodes (e.g., Evolved Serving Mobile Location Centers (E-SMLCs)), and / or Minimization of Drive Tests (MDTs).

[0112] The node 900 includes a processing circuitry 904, a memory 906, a communication interface 902, ML models 908, and a power source (not shown). The node 900 may be composed of multiple physically separate components (e.g., a NodeB component and a RNC component, or a BTS component and a BSC component, etc.), which may each have their own respective components. In certain scenarios in which the node 900 comprises multiple separate components (e.g., BTS and BSC components), one or more of the separate components may be shared among several network nodes. For example, a single RNC may control multiple NodeBs. In such a scenario, each unique NodeB and RNC pair, may in some instances be considered a single separate network node. In some embodiments, the node 900 may be configured to support multiple radio access technologies (RATs). In such embodiments, some components may be duplicated (e.g., separate memory 906 for different RATs) and some components may be reused (e.g., a same antenna may be shared by different RATs). The node 900 may also include multiple sets of the various illustrated components for different wireless technologies integrated into node 900, for example GSM, WCDMA, LTE, NR, WiFi, Zigbee, Z-wave, LoRaWAN, Radio Frequency Identification (RFID) or Bluetooth wireless technologies. These wireless technologies may be integrated into the same or different chip or set of chips and other components within node 900.

[0113] The processing circuitry 904 may comprise a combination of one or more of a microprocessor, controller, microcontroller, central processing unit, digital signal processor, application-specific integrated circuit, field programmable gate array, or any other suitable computing device, resource, or combination of hardware, software and / or encoded logic operable to provide, either alone or in conjunction with other node 900 components, such as the memory 906 and / or ML models 908, to provide node 900 functionality.

[0114] In some embodiments, the processing circuitry 904 includes a system on a chip (SOC). In some embodiments, the processing circuitry 904 includes one or more of radio frequency (RF) transceiver circuitry and baseband processing circuitry. In some embodiments, the radio frequency (RF) transceiver circuitry and the baseband processing circuitry may be on separate chips (or sets of chips), boards, or units, such as radio units and digital units. In alternative embodiments, part or all of RF transceiver circuitry and baseband processing circuitry may be on the same chip or set of chips, boards, or units.

[0115] The memory 906 and / or ML models 908 may comprise any form of volatile or nonvolatile computer-readable memory including, without limitation, persistent storage, solid-state memory, remotely mounted memory, magnetic media, optical media, random access memory (RAM), read-only memory (ROM), mass storage media (for example, a hard disk), removable storage media (for example, a flash drive, a Compact Disk (CD) or a Digital Video Disk (DVD)), and / or any other volatile or non-volatile, non-transitory device-readable and / or computerexecutable memory devices that store information, data, and / or instructions that may be used by the processing circuitry 904. The memory 906 and / or ML models 908 may store any suitable instructions, data, or information, including a computer program, software, an application including one or more of logic, rules, code, tables, and / or other instructions capable of being executed by the processing circuitry 904 and utilized by the node 900. The memory 906 and / or ML models 908 may be used to store any calculations made by the processing circuitry 904 and / or any data received via the communication interface 902. In some embodiments, the processing circuitry 904, memory 906 and / or ML models 908 are integrated.

[0116] The communication interface 902 is used in wired or wireless communication of signaling and / or data between a node and other devices. The communication interface 902 can comprise port(s) / terminal(s) to send and receive data, for example to and from a network over a wired connection. The communication interface 902 also can include radio front-end circuitry that may be coupled to, or in certain embodiments a part of, an antenna. A radio signal may then be transmitted via the antenna. Similarly, when receiving data, the antenna may collect radio signals which are then converted into digital data by the radio front-end circuitry. The digital data may be passed to the processing circuitry 904. In other embodiments, the communication interface may comprise different components and / or different combinations of components.

[0117] In certain alternative embodiments, the node 900 does not include separate radio frontend circuitry, instead, the processing circuitry 904 includes radio front-end circuitry and is connected to an antenna. Similarly, in some embodiments, all or some of the RF transceiver circuitry is part of the communication interface 902. In still other embodiments, the communication interface 902 includes one or more ports or terminals, the radio front-end circuitry, and the RF transceiver circuitry, as part of a radio unit (not shown), and the communication interface 902 communicates with baseband processing circuitry, which is part of a digital unit (not shown).

[0118] Embodiments of the node 900 may include additional components beyond those shown in Figure 9 for providing certain aspects of the node’s functionality, including any of the functionality described herein and / or any functionality necessary to support the subject matter described herein. For example, the node 900 may include user interface equipment to allow input of information into the node 900 and to allow output of information from the node 900. This may allow a user to perform diagnostic, maintenance, repair, and other administrative functions for the node 900.

[0119] Figure 10 shows an example of a communication system 1000 in accordance with some embodiments.

[0120] In the example, the communication system 1000 includes a telecommunication network 1002 that includes an access network 1004, such as a RAN, and a core network 1006, which includes one or more core network nodes 1008. The access network 1004 includes one or more access network nodes, such as network nodes 1010a and 1010b (one or more of which may be generally referred to as network nodes 1010), or any other similar 3GPP access node or non- 3GPP access point. Moreover, as will be appreciated by those of skill in the art, the network nodes 1010 are not necessarily limited to an implementation in which a radio portion and a baseband portion are supplied and integrated by a single vendor. Thus, it will be understood that the network nodes 1010 may include disaggregated implementations or portions thereof. For example, in some embodiments, the telecommunication network 1002 includes one or more ORAN network nodes. An ORAN network node is a node in the telecommunication network 1002 that supports an ORAN specification (e.g., a specification published by the O-RAN Alliance, or any similar organization) and may operate alone or together with other nodes to implement one or more functionalities of any node in the telecommunication network 1002, including one or more network nodes 1010 and / or core network nodes 1008.

[0121] Examples of an ORAN network node include an O-RU, an O-DU, an O-CU, including an O-CU-CP or an O-CU-UP, a RAN intelligent controller (near-real time or non-real time) hosting software or software plug-ins, such as a near-real time RAN control application (e.g., xApp) or a non-real time RAN automation application (e.g., rApp), or any combination thereof (the adjective “open” designating support of an ORAN specification). The network node may support a specification by, for example, supporting an interface defined by the ORAN specification, such as an Al, Fl, Wl, El, E2, X2, Xn interface, an open fronthaul user plane interface, or an open fronthaul management plane interface. Intents and content-aware notifications described herein may be communicated from a 3GPP network node or an ORAN network node over 3GPP-defmed interfaces (e.g., N2, N3) and / or ORAN Alliance-defined interfaces (e.g., Al, 01). Moreover, an ORAN network node may be a logical node in a physical node. Furthermore, an ORAN network node may be implemented in a virtualization environment (described further herein) in which one or more network functions are virtualized. For example, as discussed with respect to Figures 11 and 12 for example, the virtualization environment may include an O-Cloud computing platform orchestrated by a SMO Framework 1200 via an O-2 interface defined by the O-RAN Alliance. The network nodes 1010 facilitate direct or indirect connection of user equipment (UE), such as by connecting wireless devices 1012a, 1012b, 1012c, and 1012d (one or more of which may be generally referred to as UEs 1012) to the core network 1006 over one or more wireless connections.

[0122] Example wireless communications over a wireless connection include transmitting and / or receiving wireless signals using electromagnetic waves, radio waves, infrared waves, and / or other types of signals suitable for conveying information without the use of wires, cables, or other material conductors. Moreover, in different embodiments, the communication system 9100 may include any number of wired or wireless networks, network nodes, UEs, and / or any other components or systems that may facilitate or participate in the communication of data and / or signals whether via wired or wireless connections. The communication system 1000 may include and / or interface with any type of communication, telecommunication, data, cellular, radio network, and / or other similar type of system.

[0123] The UEs 1012 may be any of a wide variety of communication devices, including wireless devices arranged, configured, and / or operable to communicate wirelessly with the network nodes 1010 and other communication devices. Similarly, the network nodes 1010 are arranged, capable, configured, and / or operable to communicate directly or indirectly with the UEs 1012 and / or with other network nodes or equipment in the telecommunication network 1002 to enable and / or provide network access, such as wireless network access, and / or to perform other functions, such as administration in the telecommunication network 1002.

[0124] In the depicted example, the core network 1006 connects the network nodeslOlO to one or more hosts, such as host 1016 or cloud-based node 1018. These connections may be direct or indirect via one or more intermediary networks or devices. In other examples, network nodes may be directly coupled to hosts. The core network 1006 includes one more core network nodes (e.g., core network node 1008) that are structured with hardware and software components. Features of these components may be substantially similar to those described with respect to the UEs, network nodes, nodes, and / or hosts, such that the descriptions thereof are generally applicable to the corresponding components of the core network node 1008. Example core network nodes include functions of one or more of a Mobile Switching Center (MSC), Mobility Management Entity (MME), Home Subscriber Server (HSS), Access and Mobility Management Function (AMF), Session Management Function (SMF), Authentication Server Function (AUSF), Subscription Identifier De-concealing function (SIDF), Unified Data Management (UDM), Security Edge Protection Proxy (SEPP), Network Exposure Function (NEF), and / or a User Plane Function (UPF).

[0125] The host 1016and / or node 1018 may be under the ownership or control of a service provider other than an operator or provider of the access network 1004 and / or the telecommunication network 1002, and may be operated by the service provider or on behalf of the service provider. The host 1016 and / or node 1018 may host a variety of applications to provide one or more service. Examples of such applications include rApps as discussed herein, network management services, live and pre-recorded audio / video content, data collection services such as retrieving and compiling data on various ambient conditions detected by a plurality of UEs, analytics functionality, social media, functions for controlling or otherwise interacting with remote devices, functions for an alarm and surveillance center, or any other such function performed by a server.

[0126] As a whole, the communication system 1000 of Figure 10 enables connectivity between the UEs, network nodes, nodes, and hosts. In that sense, the communication system may be configured to operate according to predefined rules or procedures, such as specific standards that include, but are not limited to: Global System for Mobile Communications (GSM); Universal Mobile Telecommunications System (UMTS); Long Term Evolution (LTE), and / or other suitable 2G, 3G, 4G, 5G standards, or any applicable future generation standard (e.g., 6G); wireless local area network (WLAN) standards, such as the Institute of Electrical and Electronics Engineers (IEEE) 802.11 standards (WiFi); and / or any other appropriate wireless communication standard, such as the Worldwide Interoperability for Microwave Access (WiMax), Bluetooth, Z-Wave, Near Field Communication (NFC) ZigBee, LiFi, and / or any low-power wide-area network (LPWAN) standards such as LoRa and Sigfox.

[0127] In some examples, the telecommunication network 1002 is a cellular network that implements 3GPP standardized features. Accordingly, the telecommunications network 1002 may support network slicing to provide different logical networks to different devices that are connected to the telecommunication network 1002. For example, the telecommunications network 1002 may provide Ultra Reliable Low Latency Communication (URLLC) services to some UEs, while providing Enhanced Mobile Broadband (eMBB) services to other UEs, and / or Massive Machine Type Communication (mMTC)ZMassive loT services to yet further UEs.

[0128] In some examples, the UEs 1012 are configured to transmit and / or receive information without direct human interaction. For instance, a UE may be designed to transmit information to the access network 1004 on a predetermined schedule, when triggered by an internal or external event, or in response to requests from the access network 1004. Additionally, a UE may be configured for operating in single- or multi-RAT or multi-standard mode. For example, a UE may operate with any one or combination of Wi-Fi, NR (New Radio) and LTE, i.e. being configured for multi -radio dual connectivity (MR-DC), such as E-UTRAN (Evolved-UMTS Terrestrial Radio Access Network) New Radio - Dual Connectivity (EN-DC).

[0129] In the example, the hub 1014 communicates with the access network 1004 to facilitate indirect communication between one or more UEs (e.g., UE 1012c and / or 1012d) and network nodes (e.g., network node 1010b). In some examples, the hub 1014 may be a controller, router, content source and analytics, or any of the other communication devices described herein regarding UEs. For example, the hub 1014 may be a broadband router enabling access to the core network 1006 for the UEs. As another example, the hub 1014 may be a controller that sends commands or instructions to one or more actuators in the UEs. Commands or instructions may be received from the UEs, network nodes 1010, or by executable code, script, process, or other instructions in the hub 1014. As another example, the hub 1014 may be a data collector that acts as temporary storage for UE data and, in some embodiments, may perform analysis or other processing of the data. As another example, the hub 1014 may be a content source. For example, for a UE that is a VR headset, display, loudspeaker or other media delivery device, the hub 1014 may retrieve VR assets, video, audio, or other media or data related to sensory information via a network node, which the hub 1014 then provides to the UE either directly, after performing local processing, and / or after adding additional local content. In still another example, the hub 1014 acts as a proxy server or orchestrator for the UEs, in particular in if one or more of the UEs are low energy loT devices. The hub 1014 may have a constant / persistent or intermittent connection to the network node 1010b. The hub 1014 may also allow for a different communication scheme and / or schedule between the hub 1014 and UEs (e.g., UE 1012c and / or 1012d), and between the hub 1014 and the core network 1006. In other examples, the hub 1014 is connected to the core network 1006 and / or one or more UEs via a wired connection. Moreover, the hub 1014 may be configured to connect to an M2M service provider over the access network 1004 and / or to another UE over a direct connection. In some scenarios, UEs may establish a wireless connection with the network nodes 1010 while still connected via the hub 1014 via a wired or wireless connection. In some embodiments, the hub 1014 may be a dedicated hub - that is, a hub whose primary function is to route communications to / from the UEs from / to the network node 1010b. In other embodiments, the hub 1014 may be a non-dedicated hub - that is, a device which is capable of operating to route communications between the UEs and network node 1010b, but which is additionally capable of operating as a communication start and / or end point for certain data channels.

[0130] Figure 11 is a block diagram illustrating a virtualization environment 1100 in which functions implemented by some embodiments may be virtualized. In the present context, virtualizing means creating virtual versions of apparatuses or devices which may include virtualizing hardware platforms, storage devices and networking resources. As used herein, virtualization can be applied to any device described herein, or components thereof, and relates to an implementation in which at least a portion of the functionality is implemented as one or more virtual components. Some or all of the functions described herein may be implemented as virtual components executed by one or more virtual machines (VMs) implemented in one or more virtual environments 1100 hosted by one or more of hardware nodes, such as a hardware computing device that operates as a network node, UE, core network node, or host. Further, in embodiments in which the virtual node does not require radio connectivity (e.g., a core network node or host), then the node may be entirely virtualized. In some embodiments, the virtualization environment 1100 includes components defined by the O-RAN Alliance, such as an O-Cloud environment orchestrated by a Service Management and Orchestration Framework 1200 via an O-2 interface.

[0131] Applications 1102 (which may alternatively be called software instances, virtual appliances, network functions, virtual nodes, virtual network functions, etc.) are run in the virtualization environment 1100 to implement some of the features, functions, and / or benefits of some of the embodiments disclosed herein.

[0132] Hardware 1104 includes processing circuitry, memory that stores software and / or instructions executable by hardware processing circuitry, and / or other hardware devices as described herein, such as a network interface, input / output interface, and so forth. Software may be executed by the processing circuitry to instantiate one or more virtualization layers 1106 (also referred to as hypervisors or virtual machine monitors (VMMs)), provide VMs 1108a and 1108b (one or more of which may be generally referred to as VMs 1108), and / or perform any of the functions, features and / or benefits described in relation with some embodiments described herein. The virtualization layer 1106 may present a virtual operating platform that appears like networking hardware to the VMs 1108.

[0133] The VMs 1108 comprise virtual processing, virtual memory, virtual networking or interface and virtual storage, and may be run by a corresponding virtualization layer 1106. Different embodiments of the instance of a virtual appliance 1102 may be implemented on one or more of VMs 1108, and the implementations may be made in different ways. Virtualization of the hardware is in some contexts referred to as network function virtualization (NFV). NFV may be used to consolidate many network equipment types onto industry standard high volume server hardware, physical switches, and physical storage, which can be located in data centers, and customer premise equipment.

[0134] In the context of NFV, a VM 1108 may be a software implementation of a physical machine that runs programs as if they were executing on a physical, non-virtualized machine. Each of the VMs 1108, and that part of hardware 1104 that executes that VM, be it hardware dedicated to that VM and / or hardware shared by that VM with others of the VMs, forms separate virtual network elements. Still in the context of NFV, a virtual network function is responsible for handling specific network functions that run in one or more VMs 1108 on top of the hardware 1104 and corresponds to the application 1102.

[0135] Hardware 1104 may be implemented in a standalone network node with generic or specific components. Hardware 1104 may implement some functions via virtualization. Alternatively, hardware 1104 may be part of a larger cluster of hardware (e.g. such as in a data center or CPE) where many hardware nodes work together and are managed via management and orchestration 1110, which, among others, oversees lifecycle management of applications 1102. In some embodiments, hardware 1104 is coupled to one or more radio units that each include one or more transmitters and one or more receivers that may be coupled to one or more antennas. Radio units may communicate directly with other hardware nodes via one or more appropriate network interfaces and may be used in combination with the virtual components to provide a virtual node with radio capabilities, such as a radio access node or a base station. In some embodiments, some signaling can be provided with the use of a control system 1112 which may alternatively be used for communication between hardware nodes and radio units.

[0136] Although computing devices described herein (e.g., nodes, UEs, network nodes, hosts) may include the illustrated combination of hardware components, other embodiments may comprise computing devices with different combinations of components. It is to be understood that these computing devices may comprise any suitable combination of hardware and / or software needed to perform the tasks, features, functions and methods disclosed herein. Determining, calculating, obtaining or similar operations described herein may be performed by processing circuitry, which may process information by, for example, converting the obtained information into other information, comparing the obtained information or converted information to information stored in the network node, and / or performing one or more operations based on the obtained information or converted information, and as a result of said processing making a determination.

[0137] Moreover, while components are depicted as single boxes located within a larger box, or nested within multiple boxes, in practice, computing devices may comprise multiple different physical components that make up a single illustrated component, and functionality may be partitioned between separate components. For example, a communication interface may be configured to include any of the components described herein, and / or the functionality of the components may be partitioned between the processing circuitry and the communication interface. In another example, non-computationally intensive functions of any of such components may be implemented in software or firmware and computationally intensive functions may be implemented in hardware.

[0138] In certain embodiments, some or all of the functionality described herein may be provided by processing circuitry executing instructions stored on in memory, which in certain embodiments may be a computer program product in the form of a non-transitory computer- readable storage medium. In alternative embodiments, some or all of the functionality may be provided by the processing circuitry without executing instructions stored on a separate or discrete device-readable storage medium, such as in a hard-wired manner. In any of those particular embodiments, whether executing instructions stored on a non-transitory computer-readable storage medium or not, the processing circuitry can be configured to perform the described functionality. The benefits provided by such functionality are not limited to the processing circuitry alone or to other components of the computing device, but are enjoyed by the computing device as a whole, and / or by end users and a wireless network generally.

Claims

CLAIMS1. A computer-implemented method performed by a node for network management operations for a telecommunication network using an artificial intelligence (Al) virtual representation of at least a portion of the telecommunication network, the method comprising: receiving (802) a plurality of inputs from a plurality of machine learning, ML, models, the plurality of inputs based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network; initiating (810) the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions; determining (812), with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a key performance indicator, KPI, of the telecommunication network caused by executing the respective candidate action on the telecommunication network; and providing (814), from the Al virtual representation, an output comprising the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

2. The method of Claim 1, wherein the plurality of inputs from the plurality of ML models comprise one or more of:(i) a first input from a first machine learning, ML, model comprising a relationship among data related to the observation from at least one the telecommunication network and a source external to the telecommunication network,(ii) a second input from a second ML model comprising an estimate of the first change and the second change of the respective candidate action, and(iii) a third input comprising feedback related to an accuracy of the output of the Al virtual representation.

3. The method of any one of Claims 1 to 2, wherein the plurality of inputs are based on at least one of (i) a first input initiated from a user interface regarding the plurality of candidate actions, and (ii) a second input regarding the plurality of candidate actions that is specified for, or discovered by, the Al virtual representation.

4. The method of any one of Claims 1 to 3, further comprising:synchronizing (800) in real-time or near real-time at least the portion of the telecommunication network and the Al virtual representation.

5. The method of any one of Claims 2 to 4, wherein the first ML model comprises a graph based ML model that leams dependencies in the data related to the observation between a function, a state, a procedure, and / or an asset in at least the potion of the telecommunication network.

6. The method of any one of Claims 2 to 5, wherein the second ML model comprises a predictive model that provides the estimate in near real-time.

7. The method of any one of Claims 1 to 6, wherein the output is from a fourth ML model comprising a reinforcement learning model.

8. The method of any one of Claims 2 to 7, wherein the relationship from the first ML model is identified based on (i) identifying the data from the observation that is related to the plurality of candidate actions, (ii) determining that the data is available, (iii) determining dependencies among the data, and (iv) building a plurality of states and profiles for the dependencies.

9. The method of any one of Claims 1 to 8, further comprising: determining (804) whether data related to the plurality of inputs is available for the respective ML models in the plurality of ML models; and performing (806) one of (i) deploying the plurality of ML models when the data is available, and (ii) initiate obtaining the data when the data is not available.

10. The method of any one of Claims 1 to 9, further comprising: training (808) at least one of the plurality of ML models.

11. The method of Claim 10, wherein the training (808) comprises (i) determining whether the at least one of the plurality of ML models is a latest version, and (ii) when the at least one of the plurality of ML models is a latest version, determine whether training data is available, or (iii) when the at least one of the plurality of ML models is not a latest version, obtain the latest version.

12. The method of Claim 11, wherein the training (808) further comprises (i) validating an infrastructure for the training, (ii) setting at least one hyperparameter for the at least one of the plurality of ML models, and (iii) initiating the training.

13. The method of any one of Claims 1 to 12, wherein the providing (814) the output is based on setting the respective candidate actions and backward propagating the respective candidate actions to arrive at the recommended action based on the respective first change and the respective second change.

14. The method of any one of Claims 9 to 13, wherein obtaining the data comprises using a third ML model to fetch the data via a data broker.

15. The method of Claim 14, wherein obtaining the data comprises (i) identifying a data source for the data, and (ii) establishing a data broker to obtain the data.

16. The method of any one of Claims 1 to 15, further comprising: initiating (816) a display of at least one of the first change and the second change.

17. The method of any one of Claims 1 to 16, further comprising: saving (818) the Al virtual representation and the output.

18. The method of any one of Claims 1 to 17, wherein the Al virtual representation is located in a cloud-based node.

19. The method of any one of Claims 1 to 18, wherein the Al virtual representation is distributed across a plurality of cloud-based nodes.

20. The method of any one of Claims 1 to 19, wherein the node comprises a component comprising an rApp in a service and management orchestration, SMO, platform.

21. The method of any one of Claims 1 to 20, wherein the node comprises a component for a non-real-time radio access network (RAN) intelligent controller (RIC) in an open-radio access network, O-RAN.

22. The method of any one of Claims 1 to 21, wherein the Al virtual representationcomprises a digital twin.

23. A node (100, 900, 1018, 1206, 1300-1306) configured to perform network management operations for a telecommunication network using an artificial intelligence (Al) virtual representation of at least a portion of the telecommunication network, the node comprising: processing circuitry (904); memory (906, 908) coupled with the processing circuitry, wherein the memory includes instructions that when executed by the processing circuitry cause the node to perform operations comprising: receive a plurality of inputs from a plurality of machine learning, ML, models, the plurality of inputs based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network; initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions; determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a key performance indicator, KPI, of the telecommunication network caused by executing the respective candidate action on the telecommunication network; and provide, from the Al virtual representation, an output comprising the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

24. The node of Claim 23, wherein the memory includes instructions that when executed by the processing circuitry causes the node to perform further operations comprising any of the operations of any one of Claims 2 to 22.

25. A computer program comprising program code (906, 908) to be executed by processing circuitry (904) of a node (100, 900, 1018, 1206, 1300-1306) configured to perform network management operations for a telecommunication network using an artificial intelligence (Al) virtual representation of at least a portion of the telecommunication, whereby execution of the program code causes the node to perform operations comprising: receive a plurality of inputs from a plurality of machine learning, ML, models, the plurality of inputs based on a plurality of candidate actions to take in the telecommunicationnetwork regarding an observation in the telecommunication network; initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions; determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a key performance indicator, KPI, of the telecommunication network caused by executing the respective candidate action on the telecommunication network; and provide, from the Al virtual representation, an output comprising the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

26. The computer program of Claim 25, whereby execution of the program code causes the node to perform operations according to any one of Claims 2 to 22.

27. A computer program product comprising a non-transitory storage medium (906, 908) including program code to be executed by processing circuitry (904) of anode (100, 900, 1018, 1206, 1300-1306), whereby execution of the program code causes the node to perform operations comprising: receive a plurality of inputs from a plurality of machine learning, ML, models, the plurality of inputs based on a plurality of candidate actions to take in the telecommunication network regarding an observation in the telecommunication network; initiate the Al virtual representation based on the plurality of inputs to obtain a recommended action from the plurality of candidate actions; determine, with the Al virtual representation, (i) a first change to at least the portion of the telecommunication network caused by executing a respective candidate action on at least the portion of the telecommunication network, and (ii) a second change of a key performance indicator, KPI, of the telecommunication network caused by executing the respective candidate action on the telecommunication network; and provide, from the Al virtual representation, an output comprising the recommended action from the plurality of candidate actions to take in the telecommunication network based on the first change and the second change.

28. The computer program product of Claim 27, whereby execution of the program code causes the node to perform operations according to any one of Claims 2 to 22.

Images