Communication control system and communication control method

Abstract

This communication control system comprises: a first control unit that, in an O-RAN specification radio access network, performs learning for inferring a predetermined communication control item on a model provided to correspond to each of the classifications of radio environments around a user terminal; and a second control unit that uses the model learned by the first control unit for the user terminal associated with the classification of the radio environment corresponding to the learning model. The second control unit comprises a data control unit that selects, from pieces of inference data using a model corresponding to the classification of one radio environment, inference data to be used for learning the model corresponding to the classification of the one radio environment in accordance with a predetermined data selection condition. The first control unit performs learning on the model corresponding to classification of the one radio environment using the inference data of the classification of the one radio environment selected by the data control unit.

Description

Communication control system and communication control method

[0001] This application claims priority to Japanese Patent Application No. 2024-037455, filed on March 11, 2024, the contents of which are incorporated herein by reference.

[0002] In recent years, in "Beyond 5G systems" beyond the fifth generation mobile communication system (5G), various services such as robot control, connected cars, augmented reality (AR), and virtual reality (VR) have been considered for providing by further improving communication performance, such as throughput, communication latency, and the number of terminal connections. AI (artificial intelligence) and machine learning (ML) have attracted attention as key technologies for this purpose. For example, in radio access networks (RANs), the application of reinforcement learning to various applications, such as beamforming control, radio resource allocation, and base station function allocation, is being considered in order to maximize network performance using limited network resources.

[0003] Additionally, the O-RAN (Open Radio Access Network) Alliance is studying the opening and intelligence of next-generation radio access networks such as 5G, and is studying interfaces and workflows necessary for control using AI and ML. O-RAN defines RICs (RAN Intelligent controllers), which are controllers that control the RAN. There are two types of RICs: "Non-RT RICs (Non-Real Time RAN Intelligent Controllers)," which are intended for long-term control, and "Near-RT RICs (Near-Real Time RAN Intelligent Controllers)," which are intended for short-term control.

[0004] It is being considered to realize low-latency inference by placing the learning unit having AI and ML functions in the "Non-RT RIC" where abundant computational resources are available, while placing the inference unit that performs inference using a trained model in the "Near-RT RIC." For example, Non-Patent Document 1 proposes placing the "Data Pipeline," which processes training data, in the "Near-RT RIC." Furthermore, Non-Patent Document 2 proposes a procedure in the A1 interface, which is an interface between the "Non-RT RIC" and the "Near-RT RIC," for collecting training data via the "Data Pipeline."

[0005] Furthermore, for example, Non-Patent Document 3 describes a distributed reinforcement learning technology for achieving high learning efficiency in online learning. This distributed reinforcement learning technology is characterized by collecting large amounts of learning data through parallel execution as parallel processing using a distributed architecture, having multiple distributed "Actors" collect learning data (experience) through trial and error with the environment, which is learned by a single "Learner," and the single "Learner" periodically supplies the learned unique model to each "Actor" to update the model common to each "Actor."

[0006] “Discussion on Data Collection for Training in Non-RT RIC”, KDDI, CMCC, O-RAN WG2, July 25 2023 “Study of AI / ML-Related Procedures for RAN Slice SLA Assurance Use Case”, KDDI, CMCC, O-RAN WG2, June 21, 2023D. Horgan, et al. “Distributed Prioritized Experience Replay”, ICLR, 2018

[0007] When applying distributed reinforcement learning technology to RAN control, in order to reduce the processing load of online learning, it is possible to consider, for example, reducing the size of models such as neural networks and reducing the amount of training data. However, for large-scale RANs, reducing the model size or the amount of training data makes it impossible to sufficiently learn diverse radio environments that may differ for each user equipment (UE), which may result in a decrease in the quality of the trained model and a decrease in the accuracy of inference using the trained model.

[0008] The present invention has been made in consideration of these circumstances, and its purpose is to improve the learning efficiency of models and the inference accuracy of trained models in the control of radio access networks (RANs).

[0009] One aspect of the present invention is a communication control system in an O-RAN (Open Radio Access Network) radio access network, comprising: a first control unit that performs learning to infer predetermined communication control items for models provided corresponding to each classification of radio environments around a user terminal; and a second control unit that uses the model learned by the first control unit for a user terminal associated with the classification of the radio environment corresponding to the learned model, wherein the second control unit comprises a data control unit that selects inference data to be used for learning the model corresponding to the classification of a radio environment from inference data based on a model corresponding to the classification of the radio environment in accordance with predetermined data selection conditions, and the first control unit performs the learning for the model corresponding to the classification of the radio environment using the inference data of the classification of the radio environment selected by the data control unit. One aspect of the present invention is a communication control system, wherein in the above-described communication control system, the first control unit is realized using a "Non-RT RIC (Non-Real Time RAN Intelligent Controller)" and the second control unit is realized using a "Near-RT RIC (Near-Real Time RAN Intelligent Controller)."

[0010] One aspect of the present invention is a communication control method in a radio access network conforming to O-RAN (Open Radio Access Network) specifications, comprising: a first control unit that performs learning to infer predetermined communication control items for models provided corresponding to each classification of radio environments around a user terminal; and a second control unit that uses the model learned by the first control unit for a user terminal associated with the classification of the radio environment corresponding to the learned model, wherein the second control unit selects inference data to be used for learning the model corresponding to the classification of the one radio environment from inference data based on a model corresponding to the classification of the one radio environment in accordance with predetermined data selection conditions; and the first control unit performs the learning for the model corresponding to the classification of the one radio environment using the inference data of the classification of the one radio environment selected by the second control unit.

[0011] According to the present invention, it is possible to obtain the effect of improving the learning efficiency of models and the inference accuracy of trained models in the control of radio access networks (RANs).

[0012] FIG. 1 is a block diagram showing an example of the configuration of a communication control system according to a first embodiment. FIG. 2 is a flowchart showing an example of the procedure of a communication control method according to the first embodiment. FIG. 3 is a block diagram showing an example of the configuration of a communication control system according to a second embodiment. FIG. 4 is a block diagram showing an example of the configuration of a communication control system according to a third embodiment. FIG. 5 is a sequence diagram showing an example of the procedure of a communication control method according to the third embodiment. FIG. 6 is a block diagram showing an example of the configuration of a communication control system according to a fourth embodiment. FIG. 7 is a block diagram showing an example of the configuration of a "Data Pipeline" according to the fourth embodiment. FIG. 8 is a block diagram showing an example of the configuration of a communication control system according to a fifth embodiment.

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0014] [First embodiment] Fig. 1 is a block diagram showing an example of the configuration of a communication control system according to Embodiment 1. In Fig. 1, the communication control system 1 includes a plurality of model learning units 10 (Learner #1 to #N, where N is an integer of 2 or more), storage units 20 (Repository #1 to #N) provided corresponding to each of the model learning units 10, and model execution units 30 (Actor for UE1 to UEk) provided corresponding to each of the user terminals (UE1 to UEk).

[0015] UE1-UEk are associated with classifications of the radio environment around the UE. The classification of the radio environment around the UE is, for example, whether the signal strength of the AP (access point) around the UE is "high" or "low." In this case, a signal strength threshold is set for determining whether the signal strength of the AP around the UE is high or low. Among UE1-UEk, UEs whose signal strength of the AP around themselves is equal to or greater than the signal strength threshold are associated with a classification in which the signal strength of the AP around the UE is "high." On the other hand, among UE1-UEk, UEs whose signal strength of the AP around themselves is less than the signal strength threshold are associated with a classification in which the signal strength of the AP around the UE is "low." This results in the organization of a group of UEs associated with a classification in which the signal strength of the AP around the UE is "high" and a group of UEs associated with a classification in which the signal strength of the AP around the UE is "low."

[0016] Examples of classification of the radio environment around the UE include classification by the RSRP (Reference Signal Received Power) value (average, total, etc.) of the APs around the UE, classification by the SINR (Signal to Interference Noise power Ratio) value (average, total, etc.) of the APs around the UE, classification by the moving speed of the UE, classification by geographical information of the AP to which the UE is connected, etc. Examples of geographical information of the AP include urban structure, UE-related information, and images (still images, video) taken by a camera located near the AP.

[0017] Examples of classification based on urban structure include classification by clustering the wireless environment based on building height data, classification by clustering the wireless environment based on satellite image data, classification by building area ratio, classification by building occupancy rate around the UE based on point cloud data, etc. The building occupancy rate around the UE is an index of whether the visibility around the UE is good.

[0018] Examples of classification based on UE-related factors include classification based on the density of UE distribution, classification based on the density of AP distribution, and classification based on traffic volume.

[0019] The model learning unit 10 is provided corresponding to each classification of the wireless environment around the UE. For example, if the wireless environment around the UE is classified into two classifications, namely, "high" and "low" signal strength of APs around the UE, a model learning unit 10 (Learner #1) corresponding to the classification of "high" signal strength of APs around the UE and a model learning unit 10 (Learner #2) corresponding to the classification of "low" signal strength of APs around the UE are provided.

[0020] The storage units 20 are provided corresponding to each of the model learning units 10. For example, if the wireless environment around the UE is classified into two categories, namely, "high" and "low" signal strength of APs around the UE, a storage unit 20 (Repository #1) corresponding to the category of "high" signal strength of APs around the UE and a storage unit 20 (Repository #2) corresponding to the category of "low" signal strength of APs around the UE are provided.

[0021] The model learning unit 10 learns a model using the storage unit 20 corresponding to itself. For example, if the wireless environment around the UE is classified into two categories, "high" and "low," the signal strength of the AP around the UE, the model learning unit 10 (Learner #1) corresponding to the category of "high" signal strength of the AP around the UE learns a model using the storage unit 20 (Repository #1) corresponding to the category of "high" signal strength of the AP around the UE. On the other hand, the model learning unit 10 (Learner #2) corresponding to the category of "low" signal strength of the AP around the UE learns a model using the storage unit 20 (Repository #2) corresponding to the category of "low" signal strength of the AP around the UE. In this embodiment, reinforcement learning is used as the model learning method.

[0022] A model execution unit 30 (Actor for UE1-UEk) is provided corresponding to each of the user terminals (UE1-UEk). A model learned by one model learning unit 10 is supplied to the model execution unit 30 corresponding to the user terminal associated with the classification of the wireless environment corresponding to that one model learning unit 10.

[0023] For example, if the radio environment around a UE is classified into two categories, "high" and "low," the model (Model #1) learned by the model learning unit 10 (Learner #1) corresponding to the "high" category of signal strength of APs around the UE is supplied to the model execution units 30 (Actor for UE1, Actor for UEk) corresponding to UE1 and UEk, respectively, associated with the "high" category of signal strength of APs around the UE. Meanwhile, the model (Model #2) learned by the model learning unit 10 (Learner #2) corresponding to the "low" category of signal strength of APs around the UE is supplied to the model execution unit 30 (Actor for UE2) corresponding to UE2, associated with the "low" category of signal strength of APs around the UE.

[0024] The model execution unit 30 executes a model (trained model) that has been trained by the model learning unit 10. By executing the trained model, one model execution unit 30 takes an action on a UE (target UE) corresponding to that one model execution unit 30, and obtains states and a reward from the target UE as a result of the action.

[0025] For example, if the wireless environment around a UE is classified into two categories, namely, "high" and "low" signal strength of APs around the UE, the model execution unit 30 (Actor for UE1, Actor for UEk) corresponding to the "high" signal strength category of APs around the UE takes action against the target UE (UE1, UEk) by executing the learned model (Model #1) of the model learning unit 10 (Learner #1) corresponding to the "high" signal strength category of APs around the UE, and obtains a state and a reward from the target UE (UE1, UEk) as a result of the action. On the other hand, the model execution unit 30 (Actor for UE2) corresponding to the "low" signal strength category of APs around the UE takes action against the target UE (UE2) by executing the learned model (Model #2) of the model learning unit 10 (Learner #2) corresponding to the "low" signal strength category of APs around the UE, and obtains a state and a reward from the target UE (UE2) as a result of the action.

[0026] In the model execution unit 30, the trained model performs inference based on the state and reward resulting from the action taken against the target UE. Inference data (Experience) indicating the result of inference by the trained model of one model execution unit 30 is stored in the storage unit 20 corresponding to that one model execution unit 30.

[0027] For example, if the wireless environment around a UE is classified into two categories, namely, "high" and "low" signal strength of APs around the UE, the inference data (Experience) indicating the result of inference by the trained model (Model #1) of the model execution unit 30 (Actor for UE1, Actor for UEk) corresponding to the category of "high" signal strength of APs around the UE is stored in the storage unit 20 (Repository #1) corresponding to the category of "high" signal strength of APs around the UE. On the other hand, the inference data (Experience) indicating the result of inference by the trained model (Model #2) of the model execution unit 30 (Actor for UE2) corresponding to the category of "low" signal strength of APs around the UE is stored in the storage unit 20 (Repository #2) corresponding to the category of "low" signal strength of APs around the UE.

[0028] Each model learning unit 10 uses inference data of the classification of the wireless environment corresponding to that model learning unit 10 to learn (reinforcement learning) a model that infers a predetermined communication control item.

[0029] For example, if the wireless environment around a UE is classified into two categories, namely, "high" and "low" signal strength of APs around the UE, the model learning unit 10 (Learner #1) corresponding to the category of "high" signal strength of APs around the UE learns a model that infers a predetermined communication control item using inference data (Experience) stored in the storage unit 20 (Repository #1) corresponding to the category of "high" signal strength of APs around the UE. On the other hand, the model learning unit 10 (Learner #2) corresponding to the category of "low" signal strength of APs around the UE learns a model that infers a predetermined communication control item using inference data (Experience) stored in the storage unit 20 (Repository #2) corresponding to the category of "low" signal strength of APs around the UE.

[0030] An example of a communication control item according to this embodiment is a cluster (AP cluster) of APs (access points) around a UE. For the communication control item "AP cluster," the state used by the model may be, for example, the current AP cluster, an SLA (Service Level Agreement), the RSRP value of an AP around the UE, or the SINR value of an AP around the UE. For the communication control item "AP cluster," the model infers (what action should be taken), for example, the action to be taken when the SLA is insufficient or the action to be taken when the SLA is excessive. For example, when the SLA is insufficient, the number of APs belonging to the AP cluster may be increased, while when the SLA is excessive, the number of APs belonging to the AP cluster may be decreased.

[0031] An example of a communication control item according to this embodiment is the selection of an AP to be used for interference suppression. In the case of the communication control item "selection of an AP to be used for interference suppression," the state used by the model may include, for example, the current AP cluster, a list of APs to be used for interference suppression, the SLA, the RSRP values ​​of APs around the UE, and the SINR values ​​of APs around the UE. In the case of the communication control item "selection of an AP to be used for interference suppression," the model infers (what action should be taken), for example, the action to be taken when the SLA is insufficient or the action to be taken when the SLA is excessive. For example, when the SLA is insufficient, the number of APs to be used for interference suppression may be increased, while when the SLA is excessive, the number of APs to be used for interference suppression may be decreased.

[0032] An example of a communication control item according to this embodiment is the placement of a vDU (virtual distributed unit). In the case of the communication control item "vDU placement," the state used by the model may include, for example, the current placement of the vDU on the virtual infrastructure (server identifier (server ID) and base ID), SLA, RSRP values ​​of APs around the UE, SINR values ​​of APs around the UE, and various server performances. In the case of the communication control item "vDU placement," the item inferred by the model (what action should be taken) may include, for example, a change in the vDU selected by the UE.

[0033] An example of a reward in reinforcement learning according to this embodiment is the difference between the communication speed (throughput) and the SLA. A specific example is "reward = (performance throughput) - (throughput determined by the SLA)." Another specific example is a reward of "1" when the performance throughput is equal to or greater than the throughput determined by the SLA, and a reward of "0" when the performance throughput is less than the throughput determined by the SLA.

[0034] 2 is a flowchart showing an example of the procedure of the communication control method according to this embodiment, which will be described with reference to FIG.

[0035] (Step S1) For each classification of the radio environment around the UE, the model learned by the model learning unit 10 is supplied to the model execution unit 30. Note that an initial model such as a neural network is set in advance in the communication control system 1. In the initial stage of model learning by the model learning unit 10, the initial model is supplied.

[0036] (Step S2) The model execution unit 30 corresponding to each classification of the radio environment around the UE executes its own trained model. As a result, the model execution unit 30 takes action on the target UE and obtains a state and a reward from the target UE as a result of the action. The model execution unit 30 uses the state and reward to obtain inference data from its own trained model. For each classification of the radio environment around the UE, the inference data is sent to the storage unit 20.

[0037] (Step S3) The storage unit 20 stores the inference data transmitted from the model execution unit 30 for each classification of the radio environment around the UE.

[0038] (Step S4) For each classification of the radio environment around the UE, the model learning unit 10 performs model learning (reinforcement learning) using the inference data stored in the storage unit 20.

[0039] (Step S5) If the process is to be completed, the process in Fig. 2 is terminated. On the other hand, if the process is to be continued, the process returns to step S1.

[0040] According to this embodiment, model learning is performed for each classification of the radio environment around the UE, thereby improving model learning efficiency. Also, since the same trained model is used for inference for each classification of the radio environment around the UE, inference accuracy is improved. This provides the effect of improving model learning efficiency and inference accuracy using trained models in RAN control.

[0041] [Second embodiment] Fig. 3 is a block diagram showing an example of the configuration of a communication control system according to a second embodiment. In Fig. 3, parts corresponding to those in Fig. 1 are given the same reference numerals, and their description will be omitted. In Fig. 3, the communication control system 1a further includes a terminal group organization unit 40 in addition to the communication control system 1 in Fig. 1.

[0042] The terminal group organizing unit 40 organizes UE groups for each classification of the wireless environment. For example, if the wireless environment around a UE is classified into two categories, "high" and "low," the terminal group organizing unit 40 organizes a UE group #1 corresponding to the "high" classification of the AP signal strength around the UE and a UE group #2 corresponding to the "low" classification of the AP signal strength around the UE.

[0043] The terminal group organizing unit 40 dynamically changes the organization of the UE groups in response to changes in the radio environment around each UE. For example, if the radio environment around the UE is classified into two categories, "high" and "low," the signal strength of the AP around the UE, the terminal group organizing unit 40 dynamically changes the organization of the UE group #1 corresponding to the category of "high" signal strength of the AP around the UE and the UE group #2 corresponding to the category of "low" signal strength of the AP around the UE in response to changes in the signal strength of the AP around the UE.

[0044] The UE groups organized by the terminal group organization unit 40 are used to associate the UEs with models for each classification of the wireless environment around the UE. For example, if the classification of the wireless environment around the UE is divided into two categories, "high" and "low," the signal strength of the AP around the UE is associated with the UE group #1 corresponding to the classification of "high" signal strength of the AP around the UE. On the other hand, the model learning unit 10 (Learner #2) corresponding to the classification of "low" signal strength of the AP around the UE is associated with the UE group #2 corresponding to the classification of "low" signal strength of the AP around the UE.

[0045] According to this embodiment, the composition of the UE group is dynamically changed in response to changes in the radio environment around each UE, and therefore, it is possible to adapt to changes in the radio environment around the UE, such as when the UE is moving at high speed.

[0046] [Third embodiment] Figure 4 is a block diagram showing an example of the configuration of a communication control system according to a third embodiment. In Figure 4, parts corresponding to those in Figure 1 are given the same reference numerals, and their description will be omitted. The communication control system 1b shown in Figure 4 is an example of the configuration when the communication control system 1 in Figure 1 is applied to a radio access network conforming to the O-RAN specifications.

[0047] In this embodiment, a model is learned by a "Non-RT RIC (Non-Real Time RAN Intelligent Controller)" 100 (first control unit), while an "Near-RT RIC (Near-Real Time RAN Intelligent Controller)" 200 (second control unit) performs inference using the learned model.

[0048] 4, the model learning unit 10 and the storage unit 20 are realized by an application 110 executed by the "Non-RT RIC" 100. Therefore, the model learning unit 10 and the storage unit 20 are realized by the "Non-RT RIC" 100 executing the application 110.

[0049] 4, the model execution unit 30 is realized by an application 210 executed by the "Near-RT RIC" 200. Therefore, the model execution unit 30 is realized by the "Near-RT RIC" 200 executing the application 210.

[0050] The "Non-RT RIC" 100 and the "Near-RT RIC" 200 communicate information using the A1 interface.

[0051] The model execution unit 30 transmits information (actions, states, rewards, etc.) to and from the UE via O-RAN base stations (O-CU (O-RAN Central Unit) and O-DU (O-RAN Distributed Unit)) 300 (300-1, 300-2).

[0052] Fig. 5 is a sequence diagram showing an example of the procedure of a communication control method according to this embodiment. The communication control method according to this embodiment will be described with reference to Fig. 5. In Fig. 5, a "Non-RT RIC" 100 and a "Near-RT RIC" 200 send and receive messages using the A1 interface.

[0053] (Step S11) The "Non-RT RIC" 100 transmits a "Model transfer" message to the "Near-RT RIC" 200. The "Near-RT RIC" 200 receives the "Model transfer" message. Based on the received "Model transfer" message, the "Near-RT RIC" 200 grasps the association between the UE and each trained model for the classification of the radio environment around the UE.

[0054] The "Model transfer" message is an example of a message that includes information (model association information) for associating each trained model of the classification of the radio environment around the UE with the UE. An example of the "Model transfer" message is shown below.

[0055] (Example 1 of "Model transfer" message) Example 1 of the "Model transfer" message is an example in the case where the "Non-RT RIC" 100 is provided with a function for associating each learned model of the classification of the radio environment around the UE with the UE. The "Model transfer" message includes information such as a model identifier (Model ID), a model learning unit identifier (Learner ID), model use information indicating the use of the model, and a UE identifier (UE ID) of the UE that uses the model for each "Model ID."

[0056] According to Example 1 of the "Model transfer" message, the UE is explicitly associated with each learned model of the classification of the radio environment around the UE.

[0057] (Example 2 of "Model transfer" message) Example 2 of the "Model transfer" message is an example in which the "Near-RT RIC" 200 is provided with a function for associating each learned model of the classification of the radio environment around the UE with the UE. The "Model transfer" message includes information such as a model identifier (Model ID), a model learning unit identifier (Learner ID), model use information indicating the use of the model, and model policy information indicating a model determination policy (model policy) for UEs that use a model for each "Model ID." An example of a model policy is a model determination policy that determines a model (Model ID) to be associated with a UE based on the RSRP value of APs around the UE. An example of a model policy is a model determination policy that determines a model (Model ID) to be associated with a UE based on the position of the AP to which the UE connects.

[0058] According to example 2 of the "Model transfer" message, the "Near-RT RIC" 200 associates the UE with each learned model of the classification of the radio environment around the UE according to the model policy.

[0059] (Step S12) The "Near-RT RIC" 200 transmits a "Model transfer response" message to the "Non-RT RIC" 100. The "Non-RT RIC" 100 receives the "Model transfer response" message. Based on the received "Model transfer response" message, the "Non-RT RIC" 100 grasps the association between the UE and each trained model of the classification of the radio environment around the UE in the "Near-RT RIC" 200. An example of the "Model transfer response" message is shown below.

[0060] (Example 1 of "Model transfer response" message) Example 1 of the "Model transfer response" message corresponds to a response (Ack) message corresponding to Example 1 of the "Model transfer" message. The "Model transfer response" message has information on the combination of "Model ID" and "UE ID" associated in the "Near-RT RIC" 200. The combination of "Model ID" and "UE ID" contained in the "Model transfer response" message is information on the "UE ID" for which inference will be performed or the "UE ID" for which inference has already been performed by the trained model of the "Model ID" in the "Near-RT RIC" 200.

[0061] (Example 2 of "Model transfer response" message) Example 2 of the "Model transfer response" message corresponds to the response (Ack) message corresponding to Example 2 of the "Model transfer" message. The "Model transfer response" message includes model policy information used by the "Near-RT RIC" 200 and information on the result of the association of the "Model ID" and the "UE ID" performed by the "Near-RT RIC" 200 in accordance with the model policy.

[0062] (Step S13) The "Non-RT RIC" 100 transmits a "Model change transfer" message to the "Near-RT RIC" 200. The "Near-RT RIC" 200 receives the "Model change transfer" message. Based on the received "Model change transfer" message, the "Near-RT RIC" 200 grasps the change in the association between the UE and each learned model of the classification of the radio environment around the UE.

[0063] The "Model change transfer" message is an example of a message that includes information (model association information) for associating each trained model of the classification of the radio environment around the UE with the UE. An example of the "Model change transfer" message is shown below.

[0064] (Example 1 of "Model change transfer" message) Example 1 of the "Model change transfer" message corresponds to the change notification message corresponding to Example 1 of the "Model transfer" message. The "Model change transfer" message includes information notifying changes such as "Model ID", "Learner ID", model usage information, and the "UE ID" of the UE that uses the model for each "Model ID".

[0065] According to Example 1 of the "Model change transfer" message, the association between the UE and each learned model of the classification of the radio environment around the UE is explicitly changed.

[0066] (Example 2 of "Model change transfer" message) Example 2 of the "Model change transfer" message corresponds to the change notification message corresponding to Example 2 of the "Model transfer" message. The "Model change transfer" message contains information notifying of the model policy information that has been changed for each "Model ID."

[0067] According to example 2 of the "Model transfer" message, the "Near-RT RIC" 200 associates the UE with each learned model of the classification of the radio environment around the UE according to the modified model policy.

[0068] (Step S14) The "Near-RT RIC" 200 transmits a "Model change transfer response" message to the "Non-RT RIC" 100. The "Non-RT RIC" 100 receives the "Model change transfer response" message. Based on the received "Model change transfer response" message, the "Non-RT RIC" 100 grasps the change in the association between the UE and each learned model of the classification of the radio environment around the UE in the "Near-RT RIC" 200. An example of the "Model change transfer response" message is shown below.

[0069] (Example 1 of "Model change transfer response" message) Example 1 of the "Model change transfer response" message corresponds to a response (Ack) message corresponding to Example 1 of the "Model change transfer" message. The "Model transfer response" message has information on the combination of "Model ID" and "UE ID" whose association has been changed in the "Near-RT RIC" 200.

[0070] (Example 2 of "Model change transfer response" message) Example 2 of the "Model change transfer response" message corresponds to a response (Ack) message corresponding to Example 2 of the "Model change transfer" message. The "Model transfer response" message includes information on the changed model policy used by the "Near-RT RIC" 200 and information on the result of the association of the "Model ID" and the "UE ID" performed by the "Near-RT RIC" 200 in accordance with the changed model policy.

[0071] According to this embodiment, when using a "Non-RT RIC" 100 and a "Near-RT RIC" 200, learning a model in the "Non-RT RIC" 100, and performing inference using the learned model in the "Near-RT RIC" 200, model association information for associating each learned model of the classification of the wireless environment with a UE can be transmitted from the "Non-RT RIC" 100 to the "Near-RT RIC" 200.

[0072] [Fourth Embodiment] Fig. 6 is a block diagram showing an example of the configuration of a communication control system according to a fourth embodiment. In Fig. 6, parts corresponding to those in Fig. 4 are assigned the same reference numerals, and their description will be omitted. In the communication control system 1c shown in Fig. 6, a "Data Pipeline" 220 (not shown in Fig. 4) that processes the learning data used for model learning narrows down the learning data. Examples of the learning data include RSRP values ​​(average, total, etc.) of APs around the UE and SINR values ​​(average, total, etc.) of APs around the UE.

[0073] 6, the "Data Pipeline" 220 is realized by an application 210 executed by the "Near-RT RIC" 200. Therefore, the "Data Pipeline" 220 is realized by the "Near-RT RIC" 200 executing the application 210.

[0074] 7 is a block diagram showing an example of the configuration of the "Data Pipeline" 220 according to this embodiment. In FIG. 7, the "Data Pipeline" 220 includes a data storage unit 221 and a data control unit 222.

[0075] The data storage unit 221 stores the inference data d1 acquired from each model execution unit 30 (Actor for UE1-UEk) in association with the "Model ID" of the inferred model. The data storage unit 221 stores all of the inference data d1 for each model execution unit 30 (Actor for UE1-UEk).

[0076] The data control unit 222 selects inference data to be used for model training from the inference data stored in the data storage unit 221 according to predetermined data selection conditions for each classification of the radio environment around the UE (i.e., for each model ("Model ID")).

[0077] The data selection conditions are transmitted from the "Non-RT RIC" 100 to the "Near-RT RIC" 200 using the A1 interface. Therefore, the A1 interface according to this embodiment is an interface that includes a message for transmitting the data selection conditions from the "Non-RT RIC" 100 to the "Near-RT RIC" 200.

[0078] The data selection conditions include, for each model ("Model ID"), whether data collection for re-learning is necessary, and the policy for determining whether data collection is necessary.

[0079] For example, the data selection condition is the data collection priority for each model ("Model ID"). The data control unit 222 extracts inference data d2 for relearning from the data storage unit 221 in accordance with the data collection priority for each model ("Model ID"). The data control unit 222 transmits the extracted inference data d2 for each model ("Model ID") to the "Non-RT RIC" 100 using the A1 interface.

[0080] The data collection priority is determined for each model ("Model ID") (i.e., for each model learning unit 10 (Learner #1-#N)) by determining the need for data for re-learning. Therefore, the application 110 according to this embodiment further realizes a function for determining data collection priority.

[0081] For example, for a model with insufficient training data as a data collection priority, all inference data of all UEs associated with the model ("Model ID") is extracted from the data storage unit 221. Therefore, for a model with insufficient training data, all inference data of all UEs associated with the model ("Model ID") in the data storage unit 221 is transmitted as inference data d2 to the "Non-RT RIC" 100 using the A1 interface.

[0082] On the other hand, for models with sufficient training data, inference data is not extracted from the data storage unit 221. Therefore, for models with sufficient training data, the inference data d2 is not transmitted to the "Near-RT RIC" 100 using the A1 interface. This reduces the load on the transmission path between the "Near-RT RIC" 200 and the "Non-RT RIC" 100. Furthermore, it is possible to prevent the model training unit 10 from using an excess of training data for model training, thereby improving the learning efficiency of the model training unit 10.

[0083] According to this embodiment, when using the "Non-RT RIC" 100 and the "Near-RT RIC" 200, learning a model in the "Non-RT RIC" 100, and performing inference using the learned model in the "Near-RT RIC" 200, it is possible to reduce the load on the transmission path between the "Near-RT RIC" 200 and the "Non-RT RIC" 100 and improve the learning efficiency in the model learning unit 10.

[0084] [Fifth Embodiment] Figure 8 is a block diagram showing an example of the configuration of a communication control system according to a fifth embodiment. In Figure 8, parts corresponding to those in Figure 4 are given the same reference numerals, and their description will be omitted. The communication control system 1d shown in Figure 8 acquires information about the radio environment around the UE (UE radio environment information) from a "Y1 Consumer," which is an external node (external node) of the O-RAN, using the Y1 interface. Therefore, the communication control system 1d includes a receiving unit (not shown) that receives the UE radio environment information using the Y1 interface.

[0085] In this embodiment, a model for each classification of the radio environment around the UE is associated with the UE based on the UE radio environment information. For example, the "Non-RT RIC" 100 or the "Near-RT RIC" 200 may associate a model for each classification of the radio environment around the UE with the UE based on the UE radio environment information. The "Non-RT RIC" 100 and the "Near-RT RIC" 200 use a combination of a model and a UE associated based on the UE radio environment information.

[0086] Examples of UE radio environment information include urban structure data, topographical data, UE-related data, and images (still images, video) taken by a camera located near the AP.

[0087] When urban structure data is used, the UE is associated with a model by, for example, classifying the wireless environment by clustering based on building height data, classifying the wireless environment by clustering based on satellite image data, classifying by building area ratio, or classifying by building occupancy rate around the UE based on point cloud data. The building occupancy rate around the UE is an indicator of whether the visibility around the UE is good.

[0088] When UE-related data is used, the UEs are associated with a model, for example, by classification according to the density of UE distribution, classification according to the density of AP distribution, or classification according to traffic volume.

[0089] An example of a model updating method according to this embodiment is shown below.

[0090] (Example 1 of Model Update Method) When a deterioration in the inference accuracy of a trained model is detected or predicted, a request may be made to a node outside the O-RAN to acquire the latest data on the radio environment around the UE in order to re-train the model, and the communication control system 1d may acquire the latest UE radio environment information from the node outside the O-RAN and update the model based on the acquired latest UE radio environment information. This enables model updating using the latest data on the radio environment around the UE between the node outside the O-RAN and the communication control system 1d.

[0091] (Example 2 of Model Update Method) When a node outside the O-RAN detects or predicts that the degree of update of the UE radio environment information will be equal to or greater than a threshold, the node may notify the communication control system 1d of a model update request, and when it determines that a model update is necessary, may transmit update data of the UE radio environment information to the communication control system 1d. This enables a model update that follows the latest data on the radio environment around the UE between the node outside the O-RAN and the communication control system 1d.

[0092] According to this embodiment, by using UE radio environment information obtained from a node outside the O-RAN, the radio environment around the UE can be classified in more detail, thereby improving the accuracy of the association between the model and the UE.

[0093] According to the above-described embodiments, it is possible to obtain an effect of improving the learning efficiency of a model and the inference accuracy of a trained model in controlling a RAN. Note that the above-described embodiments may be combined as appropriate.

[0094] This will enable, for example, improvements to the overall quality of services in wireless networks, which will contribute to Goal 9 of the United Nations-led Sustainable Development Goals (SDGs), which is to "Build resilient infrastructure, promote sustainable industrialization and foster innovation."

[0095] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and design changes and the like are also included within the scope that does not deviate from the gist of the present invention.

[0096] Furthermore, a computer program for implementing the functions of each of the above-described devices may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read and executed by a computer system. Note that the term "computer system" may also include hardware such as an OS and peripheral devices. Furthermore, the term "computer-readable recording medium" refers to a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a DVD (Digital Versatile Disc), or a storage device such as a hard disk built into a computer system.

[0097] Furthermore, the term "computer-readable recording medium" also includes a storage medium that stores a program for a certain period of time, such as a volatile memory (e.g., a dynamic random access memory (DRAM)) within a computer system that serves as a server or client when the program is transmitted via a network such as the Internet or a communication line such as a telephone line. The program may also be transmitted from a computer system that stores the program in a storage device or the like to another computer system via a transmission medium or by transmission waves within the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may also be a program that realizes part of the aforementioned functions. Furthermore, the program may be a so-called differential file (differential program) that can realize the aforementioned functions in combination with a program already stored in the computer system.

[0098] According to the present invention, it is possible to improve the learning efficiency of models and the inference accuracy of trained models in the control of radio access networks.

[0099] REFERENCE SIGNS LIST 1...Communication control system, 10...Model learning unit, 20...Storage unit, 30...Model execution unit, 40...Terminal group organization unit, 100...Non-RT RIC, 200...Near-RT RIC, 110, 210...Application, UE...User terminal, 220...Data Pipeline, 221...Data storage unit, 222...Data control unit

Claims

1. A communication control system in a radio access network conforming to the O-RAN (Open Radio Access Network) specifications, comprising: a first control unit that performs learning to infer predetermined communication control items for models provided corresponding to each classification of radio environments around a user terminal; and a second control unit that uses the model learned by the first control unit for a user terminal associated with the classification of the radio environment corresponding to the learned model, wherein the second control unit comprises a data control unit that selects inference data to be used for learning the model corresponding to the classification of the one radio environment from inference data based on a model corresponding to the classification of the one radio environment in accordance with predetermined data selection conditions; and the first control unit performs the learning for the model corresponding to the classification of the one radio environment using the inference data of the classification of the one radio environment selected by the data control unit.

2. The communication control system according to claim 1, wherein the first control unit is realized using a "Non-RT RIC (Non-Real Time RAN Intelligent Controller)" and the second control unit is realized using a "Near-RT RIC (Near-Real Time RAN Intelligent Controller)." 3. A communication control method in a radio access network conforming to O-RAN (Open Radio Access Network) specifications, comprising: a first control unit that performs learning to infer predetermined communication control items for models provided corresponding to each classification of radio environments around a user terminal; and a second control unit that uses the model learned by the first control unit for a user terminal associated with the classification of the radio environment corresponding to the learned model, wherein the second control unit selects inference data to be used for learning the model corresponding to a classification of the one radio environment from inference data based on a model corresponding to the classification of the one radio environment in accordance with predetermined data selection conditions; and the first control unit performs the learning for the model corresponding to the classification of the one radio environment using the inference data of the classification of the one radio environment selected by the second control unit.

Images