Control device, wireless communication system, control method, and control program

Abstract

To provide a controller, a radio communications system, a control method, and a control program that can improve communication quality.SOLUTION: A controller 10 comprises a prediction part 11 which uses a learning unit 20 having learnt based upon time-series data, having respective pieces of radio quality information on a plurality of decentralized and arranged antennas AP for radio communication with a terminal device 50 arrayed in time series, as learning data and allocation information, including information on the antennas AP allocated for radio communication a predetermined time later from the learning data, as a learning label, and predicts, based upon operation-time time-series data as time-series data acquired in operation, prediction information including the information on the antennas AP allocated the predetermined time later from the operation-time time-series data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a control device, a wireless communication system, a control method, a control program, and a storage medium storing the control program, and more particularly to an antenna control method in a distributed antenna type (including distributed MIMO, etc.) base station, and in particular to an antenna prediction method in high frequency bands (millimeter waves or higher) for Beyond 5G, 6G, etc. [Background technology]

[0002] In order to realize communication in the high frequency band, methods have been proposed that estimate and predict the position of a terminal device based on signal analysis (e.g., Patent Document 1), physical maps (e.g., Patent Document 2), camera images (e.g., Patent Document 3), etc., and then direct a directional beam there.

[0003] Furthermore, methods have been proposed for estimating obstacles that interfere with communication based on radio wave information according to the position of the terminal device (e.g., Patent Document 4), and for predicting communication quality by dynamically detecting objects using cameras (e.g., Non-Patent Document 1) and radar (e.g., Patent Document 5). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 2021-528010 [Patent Document 2] Special Publication No. 2017-532811 [Patent Document 3] Japanese Patent Application Laid-Open No. 2012-204955 [Patent Document 4] Special Publication No. 2020-507233 [Patent Document 5] Japanese Patent Application Laid-Open No. 2014-182554 [Non-patent literature]

[0005] [Non-Patent Document 1] T. Nishio, et al., “Proactive Received Power Prediction Using Machine Learning and Depth Images for mmWave Networks,” IEEE Journal on Selected Areas in Commun. no.37(11), pp.2413-2427, Nov. 2019. Summary of the Invention [Problem to be solved by the invention]

[0006] In high-frequency communications, radio wave beams tend to travel in a straight line, and obstructions can cause the beam to deviate from the terminal device. If the beam deviates during communication, it takes time to reconnect. Therefore, it is necessary to predict the movement of the terminal device and obstructions to the terminal device and select an appropriate antenna and beam.

[0007] However, it is difficult to predict both the movement of the terminal device and the presence of obstacles using signal analysis alone. Furthermore, it is difficult to recognize dynamic obstacles and control the beam using radio wave information (e.g., a radio wave map) based on the terminal device's location. Cameras and radars may be difficult to install in some environments, or may not be applicable or may not provide sufficient performance due to factors such as the viewing angle.

[0008] The purpose of the present disclosure has been made to solve such problems, and is to provide a control device, a wireless communication system, a control method, and a control program that can predict the movement of a terminal device and obstacles to the terminal device, and improve communication quality. [Means for solving the problem]

[0009] A control device according to one embodiment includes a prediction unit that, during learning, uses time series data in which wireless quality information of each of a plurality of antennas that are distributed and that communicate wirelessly with a terminal device is arranged in chronological order as learning data, and uses a learning device that has learned allocation information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, based on operation time series data, which is the time series data acquired during operation.The prediction unit predicts predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data.

[0010] A wireless communication system according to one embodiment includes a plurality of distributed antennas that perform wireless communication with a terminal device, and a control device having a prediction unit that, during learning, uses time series data in which wireless quality information of each of the plurality of antennas is arranged in chronological order as learning data, and uses allocation information including information of the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, based on operation time series data, which is the time series data acquired during operation. The prediction unit predicts predicted information including information of the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data.

[0011] In one embodiment of the control method, during learning, time series data in which the wireless quality information of each of a plurality of antennas that are distributed and that communicate wirelessly with a terminal device is arranged in chronological order is used as learning data, and allocation information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data is used as learning labels.This predicts predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data, based on operation time series data, which is the time series data obtained during operation.

[0012] A control program according to one embodiment, or a storage medium storing the control program, causes a computer to, during learning, use time series data in which wireless quality information of each of a plurality of distributed antennas that communicate wirelessly with a terminal device is arranged in chronological order as learning data, and use a learning device that has learned allocation information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, based on operation time series data, which is the time series data acquired during operation, to predict predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data. [Effects of the Invention]

[0013] According to the present disclosure, it is possible to provide a control device, a wireless communication system, a control method, and a control program that can improve communication quality. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 1 is a block diagram illustrating a control device according to an embodiment. [Figure 2] 1 is a configuration diagram illustrating a wireless communication system according to an overview of an embodiment. [Figure 3] 1 is a flowchart illustrating a wireless communication method according to an overview of an embodiment. [Figure 4] 1 is a block diagram illustrating a control device 10 according to an embodiment of the present invention. [Figure 5] 1 is a diagram illustrating a wireless communication system according to a first embodiment. [Figure 6] 2 is a diagram illustrating a plurality of beams input to or output from each antenna in the wireless communication system according to the first embodiment. FIG. [Figure 7] 3 is a diagram illustrating wireless quality information of each antenna in the wireless communication system according to the first embodiment. [Figure 8] 2 is a diagram illustrating a plurality of beams input to or output from each antenna in the wireless communication system according to the first embodiment. FIG. [Figure 9] 3 is a diagram illustrating allocation information including wireless quality information of each antenna and information on antennas allocated to wireless communication in the wireless communication system according to the first embodiment. FIG. [Figure 10] 2 is a diagram illustrating a plurality of beams input to or output from each antenna in the wireless communication system according to the first embodiment. FIG. [Figure 11] 3 is a diagram illustrating allocation information including wireless quality information of each antenna and information on antennas allocated to wireless communication in the wireless communication system according to the first embodiment. FIG. [Figure 12] 2 is a diagram illustrating a plurality of beams input to or output from each antenna in the wireless communication system according to the first embodiment. FIG. [Figure 13] 3 is a diagram illustrating allocation information including wireless quality information of each antenna and information on antennas allocated to wireless communication in the wireless communication system according to the first embodiment. FIG. [Figure 14] 1 is a block diagram illustrating a control device according to a first embodiment. [Figure 15] 4 is a diagram for explaining the operation of a learning device learning based on time-series data of wireless quality information of an antenna input during learning in the control device according to the first embodiment. FIG. [Figure 16] 4 is a diagram for explaining, in a schematic manner, the operation executed by a learning device based on time-series data of wireless quality information of an antenna input during operation in the control device according to the first embodiment. FIG. [Figure 17] 3 is a block diagram illustrating a learning data selection unit in the control device according to the first embodiment. FIG. [Figure 18] 10 is a block diagram illustrating another learning data selection unit in the control device according to the first embodiment. FIG. [Figure 19] 3 is a flowchart illustrating a control method according to the first embodiment. [Figure 20] 6 is a flowchart illustrating another control method according to the first embodiment. [Figure 21]FIG. 10 is a block diagram illustrating a control device according to a second embodiment. [Figure 22] 10 is a block diagram illustrating a learning data selection unit in a control device according to a second embodiment. FIG. [Figure 23] FIG. 10 is a diagram illustrating the flow of learning data and learning labels input to a learning device during learning in a control device according to a third embodiment. [Figure 24] FIG. 10 is a diagram illustrating an example of the flow of operation time-series data and movement information of a terminal device input to a prediction unit that predicts prediction information using a learning device during operation in a control device according to a third embodiment. [Figure 25] FIG. 10 is a diagram illustrating another example of the flow of learning data and learning labels input to a learning device during learning in a control device according to the third embodiment. [Figure 26] FIG. 10 is a diagram illustrating another example of the flow of operation time-series data input to a prediction unit that predicts prediction information using a learning device during operation in a control device according to the third embodiment. [Figure 27] FIG. 10 is a block diagram illustrating a wireless base station device in which a control device according to a fourth embodiment is arranged. [Figure 28] FIG. 10 is a block diagram illustrating an example of a RIC device in which a control device according to a fourth embodiment is arranged. DETAILED DESCRIPTION OF THE INVENTION

[0015] Hereinafter, embodiments will be described with reference to the drawings. For clarity of explanation, the following description and drawings have been omitted and simplified as appropriate. In addition, the same elements in each drawing are designated by the same reference numerals, and duplicate explanations have been omitted as necessary.

[0016] (Outline of the embodiment) First, a control device according to an outline of an embodiment will be described. FIG. 1 is a block diagram illustrating a control device according to an outline of an embodiment. As shown in FIG. 1, a control device 10 includes a prediction unit 11. The prediction unit 11 functions as a prediction means. The prediction unit 11 predicts information about antennas to be allocated for wireless communication with a terminal device by using a learning device.

[0017] Specifically, the prediction unit 11 uses a learning device to predict information about antennas that will be assigned after a predetermined time from the operation time-series information, based on operation time-series data, which is time-series data acquired during operation. Here, during learning, the learning device uses the time-series data as learning data and learns information about antennas that will be assigned to wireless communication after a predetermined time from the learning data as learning labels.

[0018] Note that data in which the wireless quality information of each of the multiple antennas that are distributed is arranged in time series is called time series data. Also, information about antennas assigned to wireless communication is called allocation information. Furthermore, information about antennas that will be assigned a predetermined time after the time series data predicted by the prediction unit 11 is called prediction information. Therefore, the allocation information includes information about antennas that will be assigned to wireless communication a predetermined time after the learning data. The prediction information includes information about antennas that will be assigned a predetermined time after the operational time series data. Also, the allocation information and prediction information may include information about beams that are formed and input / output using multiple antenna elements in the antenna.

[0019] Fig. 2 is a configuration diagram illustrating a wireless communication system according to an outline of an embodiment. As shown in Fig. 2, the wireless communication system 1 includes a plurality of antenna APs and the above-described control device 10. The plurality of antenna APs are connected to the control device 10 via wired or wireless communication lines. The wired communication line may be, for example, a line including optical fiber. The plurality of antenna APs communicate wirelessly with a terminal device 50.

[0020] The control device 10 may be provided in a radio base station device or in a location other than a radio base station device. For example, a radio base station device may include part or all of the control device 10. For example, at least one of a radio unit (RU), a distributed unit (DU), and a center unit (CU) in a radio base station device may include part or all of the control device 10.

[0021] Other than the wireless base station device, an external device such as a RAN Intelligent Controller (RIC) that controls the wireless base station device may include part or all of the control device 10. In this case, the wireless communication system 1 further includes a RIC device that controls one or more wireless base station devices. The control device 10 is provided in the RIC device and acquires wireless quality information from the wireless base station device. The relationship between the control device 10, the wireless base station device, and the RIC device will be described later.

[0022] Next, a control method will be described. Fig. 3 is a flowchart illustrating a control method according to an outline of an embodiment. As shown in step S11 of Fig. 3, the control method includes a step of predicting predicted information. In the step of predicting predicted information, during learning, time-series data is used as learning data, and the predicted information is predicted using a learning device that has learned allocation information after a predetermined time from the learning data as learning labels.

[0023] The control device 10 described above may be, for example, an information processing device such as a microcomputer, a personal computer, or a server. FIG. 4 is a block diagram illustrating the control device 10 according to an outline of an embodiment. As shown in FIG. 4, the control device 10 may include a processor 15, a memory 16, and a storage device 17. The storage device 17 may store the processes performed by each component of the control device 10 as a program. The processor 15 may also load the program from the storage device 17 into the memory 16 and execute the program. In this way, the processor 15 realizes the functions of each component of the control device 10, such as the prediction unit 11.

[0024] Each component of the control device 10 may be realized by dedicated hardware. Furthermore, some or all of the components may be realized by a general-purpose or dedicated circuit, a processor 15, or a combination thereof. These may be configured by a single chip, or by multiple chips connected via a bus. Some or all of the components may be realized by a combination of the above-mentioned circuits and programs. Furthermore, the processor 15 may be a central processing unit (CPU), a graphics processing unit (GPU), a field-programmable gate array (FPGA), a quantum processor (quantum computer control chip), or the like.

[0025] Furthermore, when some or all of the components of the control device 10 are realized by multiple information processing devices, circuits, etc., the multiple information processing devices, circuits, etc. may be centrally or decentralized. For example, the information processing devices, circuits, etc. may be realized in a form in which they are connected to each other via a communication network using a client-server system, a cloud computing system, etc. Furthermore, the functions of the control device 10 may be provided in a SaaS (Software as a Service) format.

[0026] The control device 10 and the wireless communication system 1 of this embodiment include a prediction unit 11 that predicts predicted information by using a learning device that has learned using time-series data and allocation information during learning. Therefore, it is possible to predict predicted information about antennas allocated to wireless communication and improve communication quality.

[0027] (Embodiment 1) Next, a control device 10 and a wireless communication system according to the first embodiment will be described. FIG. 5 is a diagram illustrating a wireless communication system 1 according to the first embodiment. As shown in FIG. 5, the wireless communication system 1 includes, for example, a wireless base station device 10a and a plurality of antenna APs. In this embodiment, the wireless base station device 10a may have the above-described control device 10 (prediction unit 11). The wireless communication system 1 controls a beam B that is input to or output from a plurality of distributed antenna APs of a distributed antenna type. The distributed antenna type is also called distributed MIMO (Multiple-Input and Multiple-Output). Note that in FIG. 5, some symbols are omitted to avoid cluttering the diagram.

[0028] The wireless communication system 1 controls, for example, an antenna AP and a beam B in a high frequency band (millimeter waves or higher) for Beyond 5G, 6G, etc. Note that the frequency band controlled by the wireless communication system 1 is not limited to Beyond 5G and 6G.

[0029] A plurality of antenna APs are distributed within an area including a route along which the terminal device 50 travels. Each antenna AP is connected to the wireless base station device 10a via at least one of a wired and wireless communication line. The terminal device 50 is, for example, a mobile terminal device such as a smartphone, a tablet, or a laptop computer. The terminal device 50 may also be a wearable device with a communication function, an information terminal such as AR / VR glasses, a game console, a camera, an automobile, an AGV (Automated Guided Vehicle), a robot, or other industrial equipment. The antenna APs are arranged on facilities such as buildings 61, streetlights 62, traffic lights 63, and telephone poles 64 that are provided within an area including a route along which the terminal device 50 travels.

[0030] The terminal device 50 moves while communicating wirelessly via an antenna AP of the wireless communication system 1. As the terminal device 50 moves, the antenna AP that communicates wirelessly with the terminal device 50 is switched. For example, when the terminal device 50 moves from position P1 to position P2, the antenna AP that communicates wirelessly with the terminal device 50 is switched from antenna AP1 to antenna AP2. In this specification, the symbol AP is used to refer to antennas AP collectively, and symbols such as AP1 and AP2 are used to refer to specific antennas AP.

[0031] Fig. 6 is a diagram illustrating a plurality of beams B input to or output from each antenna AP in the wireless communication system 1 according to the first embodiment. As shown in Fig. 6, antenna AP1 inputs or outputs beams B1 to B5. Antenna AP2 inputs or outputs beams B1 to B5. Antenna AP3 inputs or outputs beams B1 to B5. Beams B1 to B5 represent radio waves input and output using a plurality of antenna elements with different beam control settings. Therefore, antennas AP1 to AP3 input and output beams B1 to B5 through their respective antenna elements.

[0032] Fig. 7 is a diagram illustrating wireless quality information of each antenna in the wireless communication system 1 according to the first embodiment. As shown in Fig. 7, the wireless quality information of each beam B1 to B5 of each antenna AP1 to AP3 may be displayed using a bar graph. The horizontal axis of each antenna AP1 to AP3 indicates each beam B1 to B5, and the vertical axis indicates the wireless quality information. The wireless quality information may be displayed using values ​​such as RSRP (Reference Signal Received Power) of each beam B1 to B5.

[0033] Note that the wireless quality information is not limited to the RSRP value. For example, the wireless quality information may include at least one of signal power information such as RSRP and RSSI (Received Signal Strength Indicator) of a beam, and channel estimation information such as each frequency and each RB (Resource Block). Furthermore, the wireless quality information may be SINR (Signal-to-Interference plus Noise Ratio) or CSI (Channel State Information).

[0034] The wireless quality information may be a value obtained by channel estimation in the control device 10 of an uplink signal including an SRS (Sounding Reference Signal) and the like from the terminal device 50. Furthermore, the wireless quality information may be a value obtained by the control device 10 from information obtained by measuring a downlink signal including an SSB (Synchronization Signal Block), a CSI-RS (Channel State Information Reference Signal) and the like measured by the terminal device 50.

[0035] When the wireless quality information is expressed as a value, a value equal to or less than a certain threshold may be replaced with zero or a lower limit value. The wireless quality information of each of the beams B1 to B5 is also the wireless quality information of each channel.

[0036] 8 is a diagram illustrating a plurality of beams B1 to B5 input to or output from the antennas AP1 to AP3 in the wireless communication system 1 according to the first embodiment. As shown in FIG. 8, the terminal device 50 moves to positions P1 to P5 between times T1 and T5. In this case, the terminal device 50 switches the beam B used for wireless communication from the beam B of the antenna AP2 to the beam B of the antenna AP1. Note that when beams are referred to collectively, they are referred to as beam B, and when a specific beam B is referred to, a symbol such as beam B1 is used. The beam B may be input to the antenna AP or output from the antenna AP.

[0037] Specifically, at time T1, the terminal device 50 is located at position P1 and performs wireless communication using beam B3 of antenna AP2. At time T2, the terminal device 50 is located at position P2 and performs wireless communication using beam B2 of antenna AP2. At time T3, the terminal device 50 is located at position P3 and performs wireless communication using beam B5 of antenna AP1. At time T4, the terminal device 50 is located at position P4 and performs wireless communication using beam B4 of antenna AP1. At time T5, the terminal device 50 is located at position P5 and performs wireless communication using beam B3 of antenna AP1.

[0038] 9 is a diagram illustrating assignment information including wireless quality information of each antenna AP1 to AP3 and information on antennas AP assigned to wireless communication in the wireless communication system 1 according to the first embodiment. As shown in FIG. 9, at time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, 0, and low, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is low, medium, high, medium, and low, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, 0, medium, and low, respectively. Therefore, at time T1, antenna AP2 is assigned to wireless communication based on the wireless quality information. Specifically, beam B3 of antenna AP2 is assigned to wireless communication as beam B having the best (highest) wireless quality based on the wireless quality information. The allocation information includes antenna AP2 and beam B3 of antenna AP2.

[0039] At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, low level, and medium level, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is medium level, high level, medium level, low level, and 0, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, low level, medium level, and 0, respectively. Therefore, at time T2, antenna AP2 is assigned to wireless communication based on the radio quality information. Specifically, beam B2 of antenna AP2 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP2 and beam B2 of antenna AP2.

[0040] At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, low level, medium level, and high level, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is high level, medium level, low level, and 0, 0, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, low level, medium level, low level, and 0, respectively. Beam B5 of antenna AP1 and beam B1 of antenna AP2 are both high level, but beam B5 of antenna AP1 is slightly higher. Therefore, at time T3, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B5 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B5 of antenna AP1. In addition, when transmission and reception from multiple antennas AP are permitted, in addition to beam B5 of antenna AP1 with the best wireless quality, beam B1 of antenna AP2 with the next best wireless quality may also be assigned to wireless communication.

[0041] At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, low level, medium level, high level, and medium level, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is medium level, low level, 0, 0, 0, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, medium level, low level, 0, 0, respectively. Therefore, at time T4, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B4 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B4 of antenna AP1.

[0042] At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is low level, medium level, high level, medium level, and low level, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is low level, 0, 0, 0, 0, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is low level, medium level, and 0, 0, 0, respectively. Therefore, at time T5, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B3 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B3 of antenna AP1.

[0043] In this way, at each of the times T1 to T5, based on the wireless quality information, for example, the antenna AP and beam B with the greatest wireless quality information are allocated to wireless communication.

[0044] 10 is a diagram illustrating a plurality of beams B1 to B5 input to or output from the antennas AP1 to AP3 in the wireless communication system 1 according to the first embodiment. As shown in FIG. 10, the terminal device 50 moves to positions P1 to P5 between times T1 and T5. In the case of FIG. 10, a static shield 65 that blocks beam B is placed between the positions P2 to P5 of the terminal device 50 and the antenna AP2. Therefore, the terminal device 50 switches from beam B of antenna AP2 to beam B of antenna AP1 via beam B of antenna AP3.

[0045] Specifically, at time T1, the terminal device 50 is located at position P1 and performs wireless communication using beam B3 of antenna AP2. At time T2, the terminal device 50 is located at position P2 and performs wireless communication using beam B4 of antenna AP3. At time T3, the terminal device 50 is located at position P3 and performs wireless communication using beam B3 of antenna AP3. At time T4, the terminal device 50 is located at position P4 and performs wireless communication using beam B4 of antenna AP1. At time T5, the terminal device 50 is located at position P5 and performs wireless communication using beam B3 of antenna AP1.

[0046] 11 is a diagram illustrating assignment information including wireless quality information of each antenna AP1 to AP3 and information on antennas AP assigned to wireless communication in the wireless communication system 1 according to the first embodiment. As shown in FIG. 11, at time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, 0, 0, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, medium level, high level, medium level, and low level, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, low level, medium level, and low level, respectively. Therefore, at time T1, antenna AP2 is assigned to wireless communication based on the wireless quality information. Specifically, beam B3 of antenna AP2 is assigned to wireless communication as beam B with the best (highest) wireless quality based on the wireless quality information. The allocation information includes antenna AP2 and beam B3 of antenna AP2.

[0047] At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, low level, and medium level, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, 0, low level, and 0, 0, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, low level, medium level, and low level, respectively. Beam B5 of antenna AP1 and beam B4 of antenna AP3 are both medium level, but beam B4 of antenna AP3 is slightly higher. Therefore, at time T2, antenna AP3 is assigned to wireless communication based on the radio quality information. Specifically, beam B4 of antenna AP3 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP3 and beam B4 of antenna AP3. In addition, when transmission and reception from multiple antennas AP are permitted, in addition to beam B4 of antenna AP3, which has the best wireless quality, beam B5 of antenna AP1, which has the second best wireless quality, may also be assigned to wireless communication.

[0048] At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, low level, medium level, and high level, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, 0, 0, 0, and 0, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, low level, medium level, low level, and 0, respectively. Therefore, at time T3, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B5 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B5 of antenna AP1.

[0049] At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, low level, medium level, high level, and medium level, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, 0, 0, 0, and 0, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is low level, medium level, low level, and 0, 0, respectively. Therefore, at time T4, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B4 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B4 of antenna AP1.

[0050] At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is low level, medium level, high level, medium level, and low level, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, 0, 0, 0, 0, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is low level, medium level, low level, and 0, 0, respectively. Therefore, at time T5, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B3 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B3 of antenna AP1.

[0051] In this way, at each time, based on the wireless quality information, for example, antenna AP and beam B with the greatest wireless quality information are assigned to wireless communication. Since there is a static obstruction 65 between positions P2 to P5 and antenna AP2, antenna AP2 and beams B1 to B5 of antenna AP2 are not assigned to wireless communication at positions P2 to P5.

[0052] 12 is a diagram illustrating a plurality of beams B1 to B5 input to or output from the antennas AP1 to AP3 in the wireless communication system 1 according to the first embodiment. As shown in FIG. 12, the terminal device 50 moves to positions P1 to P5 between times T1 and T5. In the case of FIG. 12, dynamic shields 66 that block beam B are placed between position P2 of the terminal device 50 and antenna AP2, and between position P4 of the terminal device 50 and antenna AP1. Therefore, the terminal device 50 switches from beam B of antenna AP2 to beam B of antenna AP1 via beam B of antenna AP3.

[0053] Specifically, at time T1, the terminal device 50 is located at position P1 and performs wireless communication using beam B3 of antenna AP2. At time T2, the terminal device 50 is located at position P2 and performs wireless communication using beam B4 of antenna AP3. At time T3, the terminal device 50 is located at position P3 and performs wireless communication using beam B5 of antenna AP1. At time T4, the terminal device 50 is located at position P4 and performs wireless communication using beam B2 of antenna AP3. At time T5, the terminal device 50 is located at position P5 and performs wireless communication using beam B3 of antenna AP1.

[0054] 13 is a diagram illustrating allocation information including wireless quality information of each antenna AP1 to AP3 and information on antennas allocated to wireless communication in the wireless communication system 1 according to the first embodiment. As shown in FIG. 13, at time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, 0, and low level, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is low level, medium level, high level, medium level, and low level, respectively. At time T1, the wireless quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, low level, medium level, and low level, respectively. Therefore, at time T1, antenna AP2 is allocated to wireless communication based on the wireless quality information. Specifically, beam B3 of antenna AP2 is allocated to wireless communication as beam B having the best (highest) wireless quality based on the wireless quality information. The allocation information includes antenna AP2 and beam B3 of antenna AP2.

[0055] At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, 0, low level, and medium level, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is 0, 0, low level, and 0, 0, respectively. At time T2, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, 0, low level, medium level, and low level, respectively. Beam B5 of antenna AP1 and beam B4 of antenna AP3 are both medium level, but beam B4 of antenna AP3 is slightly higher. Therefore, at time T2, antenna AP3 is assigned to wireless communication based on the radio quality information. Specifically, beam B4 of antenna AP3 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP3 and beam B4 of antenna AP3. In addition, when transmission and reception from multiple antennas AP are permitted, in addition to beam B4 of antenna AP3, which has the best wireless quality, beam B5 of antenna AP1, which has the second best wireless quality, may also be assigned to wireless communication.

[0056] At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, low level, medium level, and high level, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is high level, medium level, low level, and 0, 0, respectively. At time T3, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is 0, low level, medium level, low level, and 0, respectively. Beam B5 of antenna AP1 and beam B1 of antenna AP2 are both high level, but beam B5 of antenna AP1 is slightly higher. Therefore, at time T3, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B5 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B5 of antenna AP1.

[0057] At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is 0, 0, low level, 0, 0, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is medium level, low level, 0, 0, 0, respectively. At time T4, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is low level, medium level, low level, 0, 0, respectively. Beam B1 of antenna AP2 and beam B2 of antenna AP3 are both medium level, but beam B2 of antenna AP3 is slightly higher. Therefore, at time T4, antenna AP3 is assigned to wireless communication based on the radio quality information. Specifically, beam B2 of antenna AP3 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP3 and beam B2 of antenna AP3. In addition, when transmission and reception from multiple antennas AP are permitted, in addition to beam B2 of antenna AP3, which has the best wireless quality, beam B1 of antenna AP2, which has the next best wireless quality, may also be assigned to wireless communication.

[0058] At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP1 is low level, medium level, high level, medium level, and low level, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP2 is low level, 0, 0, 0, 0, respectively. At time T5, the radio quality information of beams B1, B2, B3, B4, and B5 of antenna AP3 is low level, medium level, low level, and 0, 0, respectively. Therefore, at time T5, antenna AP1 is assigned to wireless communication based on the radio quality information. Specifically, beam B3 of antenna AP1 is assigned to wireless communication as beam B with the best (highest) radio quality based on the radio quality information. The assignment information includes antenna AP1 and beam B3 of antenna AP1.

[0059] In this way, at each time, based on the wireless quality information, for example, antenna AP and beam B with the greatest wireless quality information are assigned to wireless communication. Dynamic shielding objects 66 that block beam B are placed between position P2 of the terminal device 50 and antenna AP2, and between position P4 of the terminal device 50 and antenna AP1. For this reason, at time T2 and time T4, antennas AP1 and AP2 are not assigned to wireless communication, respectively.

[0060] Fig. 14 is a block diagram illustrating the control device 10 according to embodiment 1. As shown in Fig. 14, the control device 10 may further include a learning device 20 and a learning data selection unit 30. The learning device 20 and the learning data selection unit 30 function as a learning means and a learning data selection means, respectively.

[0061] The learning device 20 learns time-series data in which wireless quality information of each of a plurality of antenna APs that perform wireless communication with the terminal device 50 is arranged in time series as learning data. The learning device 20 may learn allocation information including information of antenna APs that are allocated to wireless communication after a predetermined time from the learning data as learning labels. The learning device 20 may be provided in the prediction unit 11. Note that the learning device 20 may be provided outside the control device 10.

[0062] The learning data selection unit 30 selects learning data for learning by the learning device 20. The learning data selection unit 30 is provided in the control device 10. Note that the learning data selection unit 30 may be provided outside the control device 10.

[0063] 15 is a diagram illustrating an example of an operation of the learning device 20 during learning in the control device 10 according to the first embodiment, in which the learning device 20 learns using time-series data of wireless quality information of the antennas AP1 to AP3 that is input during learning. As shown in FIG. 15, the time-series data of wireless quality information of the antennas AP1 to AP3 that is input during learning includes wireless quality information from time Tf to time Tfk, where T is the current time. That is, the time-series data includes wireless quality information from time Tfk to time Tf.

[0064] The learning device 20 uses the input time series data as learning data and learns allocation information including information about the antenna AP allocated for wireless communication after a predetermined time f has elapsed from the time series data as learning labels. The learning device 20 receives the learning data and learning labels from a learning data selection unit 30, which will be described later. Here, the current time T is the time after the time f has elapsed from the time series data.

[0065] The learning device 20 may include a CNN (Convolutional Neural Network) and an LSTM (Long Short Term Memory). The learning device 20 may extract at least one of a local maximum and a local minimum and time fluctuations using the CNN, estimate long-term changes in wireless quality due to the position and movement of the terminal device 50 using the LSTM (Long), and estimate short-term changes in wireless quality due to the shielding objects 65 and 66 using the LSTM (Short). Here, the short-term changes in wireless quality due to the shielding objects 65 and 66 include deterioration (decrease) of wireless quality due to radio wave shielding and fluctuations in wireless quality due to multipath (fading) that occurs when radio waves are reflected by the shielding objects.

[0066] FIG. 16 is a diagram illustrating an operation performed by the learning device 20 of the control device 10 according to the first embodiment, based on time-series data of wireless quality information of the antennas AP1 to AP3 input during operation. As shown in FIG. 16, the prediction unit 11 acquires time-series data of wireless quality information of the antennas AP1 to AP3 input during operation. The time-series data of wireless quality information during operation is referred to as operation-time time-series data. The operation-time time-series data includes wireless quality information from the current time T back to time Tk. In other words, the operation-time time-series data includes wireless quality information from time Tk to time T. Then, by using the above-described learning device 20, the prediction unit 11 predicts, based on the operation-time time-series data, prediction information including information about the antenna AP to be assigned to wireless communication after a predetermined time f has elapsed from the operation-time time-series data.

[0067] FIG. 17 is a block diagram illustrating a learning data selection unit 30 in the control device 10 according to Embodiment 1. As shown in FIG. 17, the control device 10 may further include a learning data selection unit 30. The learning data selection unit 30 includes a beam determination unit 31, a beam determination unit 32, a comparison unit 33, a comparison unit 34, a data storage unit 35, and a data selection unit 36. The beam determination unit 31, the beam determination unit 32, the comparison unit 33, the comparison unit 34, the data storage unit 35, and the data selection unit 36 each have a function as a beam determination means, a beam determination means, a comparison means, a comparison means, a data storage means, and a data selection means. The learning data selection unit 30 is input with time-series data of radio quality information and allocation information including information on an antenna assigned to wireless communication at time T. The allocation information includes an antenna AP and a beam B assigned to wireless communication based on the radio quality information at time T.

[0068] The beam determination unit 31 determines the antenna AP and the beam B assigned to wireless communication at each time based on the radio quality information at each time from time (T - f - k) to time (T - f - 1). The beam determination unit 32 determines the antenna AP and the beam B assigned to time (T - f) based on the radio quality information at time (T - f). The comparison unit 33 compares the antenna AP and the beam B assigned to each time from time (T - f - k) to time (T - f - 1) (for example, T - f - m, 1 ≤ m < k) with the antenna AP and the beam B assigned to the immediately previous time (T - f - m - 1). The comparison unit 34 compares the antenna AP and the beam B assigned to time (T - f) with the antenna AP and the beam B assigned to time T. The antenna AP and the beam B assigned to time T are the allocation information serving as a learning label.

[0069] The data storage unit 35 stores, in the first group, the antenna AP and beam B at the time (T-f) immediately before the learning label, which are the same as the antenna AP and beam B of the learning label. The data storage unit 35 stores, in the second group, those in which the beam B is different from the beam B of the learning label but the antenna AP is the same as the antenna AP of the learning label among the antenna AP and beam B at the time (T-f) immediately before the learning label. The data storage unit 35 stores, in the third group, those in which both the antenna AP and beam B at the time (T-f) immediately before the learning label are different from the antenna AP and beam B of the learning label.

[0070] The data selection unit 36 selects, without bias, from the first group, the second group, and the third group as learning data. For example, when the time-series data of the first group is overwhelmingly large and the time-series data of the second group and the third group are overwhelmingly small, the data selection unit 26 selects the learning data so that the ratio of the first group is small and the ratios of the second group and the third group are large.

[0071] In this way, the learning data selection unit 30 selects, as learning data, the time-series data classified (into the first group to the third group) according to the comparison result between the allocation information set as the learning label and the allocation information assigned to wireless communication based on the wireless quality information immediately before the learning label among the time-series data during learning. The learning data selection unit 30 outputs the selected time-series data to the learning device 20.

[0072] Also, the comparison unit 33 compares the antenna AP and beam B assigned to an arbitrary time (for example, T-f-n, 1 ≤ n < k) from time (T-f-k) to time (T-f-1) with the antenna AP and beam B assigned to the immediately preceding time (T-f-n-1).

[0073] The data storage unit 35 then stores in a fourth group those antenna APs and beam B at time (Tfn) that are the same as the antenna APs and beam B at time (Tfn-1). The data storage unit 35 stores in a fifth group those antenna APs and beam B at time (Tfn) that have a different beam B from the beam B at time (Tfn-1) but have the same antenna AP as the antenna AP at time (Tfn-1). The data storage unit 35 stores in a sixth group those antenna APs and beam B at time (Tfn) that are both different from the antenna APs and beam B at time (Tfn-1).

[0074] The data selection unit 36 ​​selects training data from the fourth, fifth, and sixth groups without bias. For example, if the time-series data of the fourth group is overwhelmingly large and the time-series data of the fifth and sixth groups is overwhelmingly small, the data selection unit 26 selects training data such that the proportion of the fourth group is small and the proportion of the fifth and sixth groups is large.

[0075] In this way, the learning data selection unit 30 selects, as learning data, time-series data classified (into the fourth to sixth groups) according to the comparison result between the pieces of allocation information assigned to wireless communication based on the pieces of wireless quality information in the time-series data. The learning data selection unit 30 outputs the selected time-series data to the learning device 20.

[0076] FIG. 18 is a block diagram illustrating another learning data selection unit in the control device according to the first embodiment. As shown in FIG. 18, the learning data selection unit 30a further includes a reconnection determination unit 37. The reconnection determination unit 37 functions as a reconnection determination means. The reconnection determination unit 37 determines whether a reconnection has occurred during operation. The reconnection determination unit 37 classifies reconnection time-series data, which is time-series data indicating when a reconnection has occurred during operation, and reconnection allocation information, which includes information about antennas allocated for wireless communication after a predetermined time from the reconnection time-series data. For example, the reconnection determination unit 37 attaches flags to the reconnection time-series data and the reconnection allocation information and stores them in the data storage unit 35.

[0077] The data storage unit 35 stores the classified reconnection time-series data and reconnection allocation information. The data selection unit 36 ​​preferentially selects the reconnection time-series data and reconnection allocation information as learning data and learning labels, respectively. In this way, the learning data selection unit 30a selects the reconnection time-series data and reconnection allocation information classified by the reconnection determination unit 37 as learning data and learning labels.

[0078] Cases where disconnection and reconnection occur during operation correspond to cases where the results of learning by the learning device 20 have not been reflected, or cases where data on cases (scenes) where the relevant disconnection or reconnection occurs have not been learned by the learning device 20. Therefore, by preferentially including in the learning data time-series data on cases where disconnection and reconnection occur during such operation, it becomes possible to deal with disconnection and reconnection.

[0079] Next, a control method according to the first embodiment will be described. FIG. 19 is a flowchart illustrating the control method according to the first embodiment. As shown in step S10 of FIG. 19, the control method of this embodiment further includes a learning step. In the learning step, time-series data in which wireless quality information of each of a plurality of antenna APs that are distributed and that perform wireless communication with the terminal device 50 is arranged in time series is used as learning data during learning. In addition, allocation information including information about antenna APs that are allocated for wireless communication after a predetermined time from the learning data is learned as a learning label.

[0080] When learning, time-series data classified according to a comparison result between allocation information set as a learning label and allocation information assigned to wireless communication based on wireless quality information immediately before the learning label in the time-series data may be selected as learning data. Furthermore, when selecting as learning data, time-series data classified according to a comparison result between each piece of allocation information assigned to wireless communication based on each piece of wireless quality information in the time-series data may be selected as learning data.

[0081] Fig. 20 is a flowchart illustrating another control method according to the first embodiment. As shown in step S9 of Fig. 20, the control method may determine whether a reconnection has occurred during operation. When determining whether a reconnection has occurred, reconnection time-series data, which is time-series data when a reconnection has occurred, and reconnection allocation information, which includes information about an antenna AP assigned to wireless communication after a predetermined time from the reconnection time-series data, are classified. Then, when learning in step S10, the reconnection time-series data and the reconnection allocation information may be selected as learning data and learning labels.

[0082] Next, the effects of this embodiment will be described. The control device 10 of this embodiment predicts prediction information using a learning device 20 that has learned time-series data of wireless quality information when the terminal device 50 moves or when a static obstruction 65 and a dynamic obstruction 66 occur. Therefore, even when the terminal device 50 moves or when the obstructions 65 and 66 occur, it is possible to predict an antenna AP and a beam B that can communicate. This makes it possible to suppress interruptions in wireless communication and improve communication quality.

[0083] Furthermore, the control device 10 of this embodiment predicts the forecast information using only the time-series data of the wireless quality information. Therefore, even when the terminal device 50 moves or when a static obstruction 65 or a dynamic obstruction 66 occurs, it is possible to eliminate the need for object detection means using a camera, radar, or the like other than the wireless base station.

[0084] Specifically, the prediction unit 11 in the control device 10 uses a learning device 20 that has trained using training data including cases of 1. movement of the terminal device 50, 2. static obstruction 65, 3. dynamic obstruction 66, etc. This allows the prediction unit 11 to improve prediction accuracy in such cases. In the case of 1. movement of the terminal device 50, the learning device 20 learns, for example, from the long-term relationship of the wireless quality information. On the other hand, in the cases of 2. static obstruction 65 and 3. dynamic obstruction 66, the learning device 20 learns, for example, from short-term changes in the wireless quality information.

[0085] Furthermore, even during online (real-time) learning, the learning data selection unit 30 learns without bias each scene where there is a high probability of beam switching, such as 1. movement of the terminal device 50, 2. static obstruction 65, and 3. dynamic obstruction 66. This makes it possible to improve the prediction accuracy in each scene.

[0086] Note that the prediction time f may be set to f=0.2 seconds or the like, taking into consideration the Fresnel zone (space showing radio wave visibility), distance, expected moving speed, etc. in the frequency band (28 GHz, etc.) of the wireless communication system 1. By setting such a prediction time f and performing learning and prediction, the next predicted information can be predicted when the obstructions 65 and 66 begin to appear.

[0087] For example, if a person carrying a terminal device 50 that communicates in the 28 GHz frequency band is traveling at a speed of 3 km / h (kilometers per hour; the same applies below), the influence of an obstruction 65 or the like will appear approximately 0.1 seconds in advance. The prediction time can be set taking into consideration such Fresnel zones, distance, expected traveling speed, etc. Here, for example, if the frequency band is greater than 28 GHz, the beam B will tend to travel more directly and will be more likely to be cut off. Therefore, the prediction time is set to be shorter than f=0.2 seconds.

[0088] As mentioned above, the radio quality information for each input beam may be not only the RSRP value, but also channel estimation information for each frequency or RB, information such as RSSI, SINR, CSI, or channel estimation information for uplink signals such as SRS, and information obtained by measuring downlink signals such as SSB and CSI-RS.

[0089] The prediction information (correct value information at the time of learning) of at least one of the antenna AP and beam B to be predicted may be information on the antenna number only, not the beam number. Alternatively, if the beam number is the same as the current communication beam, it may be the next best (second) beam number. In other words, the prediction unit 11 may learn and predict beams that are candidates for switching. Furthermore, effective beam sets, which are combinations of effective beams, may be learned and defined as multiple sets, and the effective beam set numbers may be used as correct values ​​(predicted values). Specifically, beam sets such as the first beam, second beam, and third beam may be predicted.

[0090] Therefore, the allocation information may include the first antenna and the second antenna that is next to the first antenna, which are assigned to the wireless communication based on the wireless quality information. The prediction information may include the third antenna and the fourth antenna that is next to the third antenna, which are assigned after a predetermined time from the operation time-series data. Furthermore, the allocation information may be an allocation set including at least the first antenna and the second antenna that is next to the first antenna, which are assigned to the wireless communication based on the wireless quality information, or the prediction information may be a prediction set including at least the third antenna and the fourth antenna that is next to the third antenna, which are assigned after a predetermined time from the operation time-series data. Furthermore, this is not limited to the antenna AP, but also applies to beam B. In other words, the allocation information may include the first beam and the second beam that is next to the first beam, which are assigned to the wireless communication based on the wireless quality information. The prediction information may include the third beam and the fourth beam that is next to the third beam, which are assigned after a predetermined time from the operation time-series data. In addition, the allocation information may be an allocation set including at least a first beam and a second beam that is next to the first beam, which are assigned to wireless communication based on wireless quality information, and the prediction information may be a prediction set including at least a third beam and a fourth beam that is next to the third beam, which are assigned after a predetermined time from the operational time series data.

[0091] During learning, the learning data selection unit 30 may prioritize time-series data in which disconnection and reconnection occurred during operation and include it in the learning data, thereby enabling the system to handle disconnection and reconnection.

[0092] For input wireless quality information, values ​​below a certain threshold may be converted to zero before input. This increases the likelihood that sparsity will exist in the input data, which has the advantage of making learning more efficient and enabling higher accuracy by applying a learning method specialized for sparsity.

[0093] Also, a function for arranging the beams of the input wireless quality information may be added. For example, by arranging the beams of the wireless quality information along the route on which the terminal device 50 moves, the learning efficiency of the learning device 20 and the prediction efficiency of the prediction unit 11 can be improved.

[0094] The wireless communication system 1 of this embodiment may be applied not only to millimeter wave wireless communication but also to terahertz wave (sub-terahertz wave) communication, optical space communication (free space optical communication), visible light communication (optical wireless communication), and the like.

[0095] (Embodiment 2) Next, a control device according to a second embodiment will be described. The control device according to this embodiment utilizes movement information of the terminal device 50. For example, the control device according to this embodiment calculates the movement speed of the terminal device 50 and outputs the calculated movement speed of the terminal device 50 to the learning data selection unit 30. FIG. 21 is a block diagram illustrating an example of a control device according to the second embodiment. As shown in FIG. 21, the control device 12 according to this embodiment includes a movement speed calculation unit 40. The movement speed calculation unit 40 functions as a movement speed calculation unit.

[0096] The moving speed calculation unit 40 calculates the moving speed of the terminal device 50. The moving speed calculation unit 40 performs a position estimation process using, for example, radio quality information. Then, the moving speed calculation unit 40 may calculate the moving speed of the terminal device 50 from a plurality of position estimation results. For example, the moving speed calculation unit 40 may calculate the moving speed of the terminal device 50 from a change in intensity of each beam B. Also, for example, the moving speed calculation unit 40 may calculate the moving speed of the terminal device 50 from the positional relationship between each assigned antenna AP and each beam B. Furthermore, the moving speed calculation unit 40 may calculate the moving speed of the terminal device 50 using the results of triangulation of a plurality of antenna APs. Also, the moving speed calculation unit 40 may estimate a Doppler frequency from radio quality information and calculate the moving speed of the terminal device 50 from the estimated Doppler frequency. The moving speed calculated by the moving speed calculation unit 40 is output to the learning data selection unit 30.

[0097] Fig. 22 is a block diagram illustrating a learning data selection unit in the control device 12 according to embodiment 2. As shown in Fig. 22, the learning data selection unit 30b of this embodiment further includes a speed determination unit 38. The speed determination unit 38 functions as a speed determination means.

[0098] The speed determination unit 38 determines the moving speed of the terminal device 50. For example, the speed determination unit 38 determines the moving speed of the terminal device 50 to one of categories such as low speed, medium speed, and high speed. The low speed category is, for example, a moving speed greater than 0 km / h and equal to or less than 3 km / h. The medium speed category is, for example, a moving speed greater than 3 km / h and equal to or less than 30 km / h. The high speed category is, for example, a moving speed greater than 30 km / h. Note that the number of categories and the speed values ​​of each category are not limited to the above and may be changed depending on the installation environment of the control device 12. The speed determination unit 38 classifies the time-series data into a plurality of moving speed categories using information on the moving speed of the terminal device 50 calculated from the wireless quality information.

[0099] The data storage unit 35 classifies the time-series data into a plurality of movement speed categories and stores the data. The data selection unit 36 ​​may select data for each movement speed category so that the data is learned without bias. In this way, the learning data selection unit 30 selects, as learning data, time-series data classified according to the movement information of the terminal device 50 calculated from the time-series data. Specifically, the learning data selection unit 30 uses the terminal movement information calculated from the wireless quality information to classify the data into a plurality of movement speed categories, such as low speed, medium speed, and high speed, and selects the time-series data for each category so that the data is learned without bias when selecting learning data.

[0100] According to this embodiment, the control device 12 determines the movement speed of the terminal device 50 and selects time-series data of the determined movement speed category as learning data without bias. Therefore, since learning is performed taking into account the movement speed of the terminal device 50, it is possible to improve the prediction accuracy of the prediction information. Other configurations and effects are included in the description of embodiment 1.

[0101] (Embodiment 3) Next, a control device according to embodiment 3 will be described. The control device of this embodiment is similar to embodiment 2 in that it utilizes movement information of the terminal device 50. However, the control device of this embodiment calculates the movement speed of the terminal device 50 and outputs the calculated movement speed of the terminal device 50 to the learning device 20 in the prediction unit 11.

[0102] FIG. 23 is a diagram illustrating the flow of learning data and learning labels input to the learning device 20 during learning in a control device according to the third embodiment. As shown in FIG. 23, during learning, a movement speed calculation unit 40 in the control device 13 calculates the movement speed of the terminal device 50 from time-series data of wireless quality information. The movement speed calculation unit 40 inputs movement information including the calculated movement speed of the terminal device 50 to the learning device 20 as learning data. The learning device 20 learns the movement speed of the terminal device 50 as learning data in addition to the time-series data. That is, the learning device 20 includes the movement information of the terminal device 50 calculated from the time-series data in the learning data. In this case, the learning device 20 may learn allocation information as learning labels.

[0103] Fig. 24 is a diagram illustrating the flow of operation time-series data and movement information of the terminal device 50 input to the prediction unit 11 that predicts prediction information using the learning device 20 during operation in the control device 13 according to the third embodiment. As shown in Fig. 24, during operation, the movement speed calculation unit 40 calculates the movement speed of the terminal device 50 from the operation time-series data of the radio quality information. The movement speed calculation unit 40 inputs the calculated movement speed of the terminal device 50 to the prediction unit 11. By using the learning device 20, the prediction unit 11 predicts prediction information based on the movement speed of the terminal device 50 in addition to the operation time-series data.

[0104] FIG. 25 is a diagram illustrating another flow of learning data and learning labels input to the learning device 20 during learning in the control device 13 according to the third embodiment. As shown in FIG. 25, during learning, the movement speed calculation unit 40 calculates the movement speed of the terminal device 50 from time-series data of wireless quality information. The movement speed calculation unit 40 may input the calculated movement speed of the terminal device 50 as a learning label to the learning device 20. The learning device 20 learns the movement speed of the terminal device 50 as a learning label in addition to the allocation information. Therefore, the learning device 20 may perform multitask learning. The learning device 20 learns time-series data as learning data.

[0105] Fig. 26 is a diagram illustrating another example of the flow of operation time-series data input to the prediction unit 11, which predicts prediction information using the learning device 20 during operation, in the control device 13 according to the third embodiment. As shown in Fig. 26, during operation, the prediction unit 11 predicts prediction information based on the operation time-series data by using the learning device 20. Note that in this case, although not shown in Fig. 26, it is also possible to predict and output the travel speed by using the learning device 20 that has undergone the above-mentioned multitask learning.

[0106] According to this embodiment, the learning device 20 performs learning using the movement speed of the terminal device 50 as learning data or learning labels, thereby improving the prediction accuracy of the prediction information. Furthermore, the prediction unit 11 predicts the prediction information based on the movement speed of the terminal device 50 in addition to the operation time-series data, thereby improving the prediction accuracy of the prediction information. Other configurations and effects are included in the descriptions of the first and second embodiments.

[0107] (Embodiment 4) Next, a control device according to a fourth embodiment will be described. This embodiment provides details of a case where the control device is disposed in a radio base station device and a case where the control device is disposed in a RIC device. FIG. 27 is a block diagram illustrating a radio base station device in which a control device according to the fourth embodiment is disposed. As shown in FIG. 27, the control device 14 may be disposed in the radio base station device 10a. As described above, the control device 14 may be disposed in the RU, DU, CU, etc. of the radio base station device 10a. The control device 14 may further include a radio quality estimator 70 and an antenna controller 80 in addition to the predictor 11, the learner 20, and the training data selector 30. The radio quality estimator 70 and the antenna controller 80 function as radio quality estimation means and antenna control means.

[0108] The wireless quality estimator 70 estimates wireless quality information of each of a plurality of distributed antenna APs that perform wireless communication with the terminal device 50. For example, the wireless quality estimator 70 measures the above-mentioned RSRP or the like of each antenna and estimates the wireless quality information. The wireless quality estimator 70 outputs the estimated wireless quality information to the antenna control unit 80. The wireless quality estimator 70 also outputs the estimated wireless quality information to the learning data selector 30. The wireless quality estimator 70 further outputs the estimated wireless quality information to the prediction unit 11. Here, the wireless quality estimator 70 may use the estimated wireless quality information to determine whether wireless communication with the terminal device 50 has been disconnected or reconnected. Alternatively, the wireless quality estimator 70 may obtain information from the terminal device 50 indicating whether reconnection has occurred. The reconnection information may then be output to the learning data selector 30.

[0109] The antenna control unit 80 receives wireless quality information from the wireless quality estimation unit 70. The antenna control unit 80 controls the antenna AP to be assigned to the wireless communication based on the wireless quality information. Specifically, the antenna control unit 80 determines the antenna AP to be assigned to the wireless communication. In addition to the antenna AP to be assigned to the wireless communication, the antenna control unit 80 may also determine a beam B to be assigned to the wireless communication. For example, the antenna control unit 80 may select candidates for the antenna AP and beam B to be assigned to the wireless communication from the wireless quality information and determine the antenna AP and beam B to be actually assigned from among the candidate antenna APs and beam Bs. Furthermore, the antenna control unit 80 may determine candidate antenna APs and candidate beam Bs for which wireless quality is measured as candidates to be assigned, in addition to the antenna AP and beam B to be actually assigned. In this way, the antenna control unit 80 determines allocation information including information on antennas to be assigned to the wireless communication based on the wireless quality information. The antenna control unit 80 outputs the allocation information to the learning data selection unit 30.

[0110] The learning data selection unit 30 receives wireless quality information from the wireless quality estimation unit 70. The learning data selection unit 30 selects learning data from the time-series data of the wireless quality information received from the wireless quality estimation unit 70. The learning data selection unit 30 outputs the selected learning data to the learner 20.

[0111] Furthermore, the training data selection unit 30 receives allocation information from the antenna control unit 80. The training data selection unit 30 selects training labels from the allocation information received from the antenna control unit 80. The training data selection unit 30 outputs the selected training labels to the learning device 20. The learning device 20 performs training using the training data and training labels received from the training data selection unit 30.

[0112] The prediction unit 11 uses the learning device 20 to predict prediction information (the antenna AP or beam B to be assigned after a predetermined time from the operation time-series data) based on the time-series data of the operation-time wireless quality information received from the wireless quality estimation unit 70. The prediction unit 11 outputs the prediction information to the antenna control unit 80.

[0113] The antenna control unit 80 may add the antenna AP or beam B in the prediction information predicted by the prediction unit 11 to the candidate antenna AP and beam B to be assigned to wireless communication. Also, the antenna control unit 80 may add the antenna AP or beam B in the prediction information predicted by the prediction unit 11 to the candidate antenna AP and candidate beam B for measuring wireless quality.

[0114] FIG. 28 is a block diagram illustrating an example of a RIC device in which a control device 14 according to the fourth embodiment is disposed. As shown in FIG. 28, in the wireless communication system 4 of this embodiment, the control device 14 may be disposed in a RIC device 10b. The control device 14 disposed in the RIC device 10b includes a prediction unit 11, a learning device 20, and a learning data selection unit 30. The RIC device 10b is connected to a wireless base station device 10a. There may be one or more wireless base station devices 10a connected to the RIC device 10b. The wireless base station device 10a includes a wireless quality estimation unit 70 and an antenna control unit 80.

[0115] The wireless quality estimator 70 of the wireless base station device 10a outputs the estimated wireless quality information to the antenna controller 80 of the wireless base station device 10a. The wireless quality estimator 70 also outputs the estimated wireless quality information to the learning data selector 30 and the predictor 11 in the RIC device 10b. The wireless quality estimator 70 may also output the determined or acquired reconnection information to the learning data selector 30.

[0116] The antenna control unit 80 of the wireless base station device 10a outputs the determined allocation information to the learning data selection unit 30 in the RIC device 10b.

[0117] The learning data selection unit 30 of the RIC device 10b receives wireless quality information from the wireless quality estimation unit 70 of the wireless base station device 10a and receives allocation information from the antenna control unit 80. The learning data selection unit 30 selects learning data and learning labels from the received time-series data of wireless quality information and allocation information. The learning data selection unit 30 outputs the selected learning data and learning labels to the learning device 20. The learning device 20 of the RIC device 10b performs learning using the learning data and learning labels received from the learning data selection unit 30.

[0118] The prediction unit 11 of the RIC device 10b predicts prediction information (the antenna AP or beam B to be assigned a predetermined time after the operation time-series data) based on the time-series data of the operation-time radio quality information received from the radio quality estimation unit 70 in the radio base station device 10a, by using the learning device 20. The prediction unit 11 outputs the prediction information to the antenna control unit 80 in the radio base station device 10a.

[0119] The antenna control unit 80 may add the antenna AP or beam B in the prediction information predicted by the prediction unit 11 in the RIC device 10b to the candidate antenna AP and beam B to be assigned to wireless communication. Also, the antenna control unit 80 may add the antenna AP or beam B in the prediction information predicted by the prediction unit 11 to the candidate antenna AP and candidate beam B for measuring wireless quality.

[0120] Note that the learning device 20 during training may be implemented in an external processing device other than the radio base station device 10a and the RIC device 10b. In that case, during training, pre-stored training data is output from the training data selector 30 to the learning device 20 of the processing device, and training is performed in the processing device. Alternatively, part or all of the training data selector 30 may be implemented in the processing device. Then, during operation, the trained learning device 20 is transferred from the processing device to the predictor 11 implemented in the radio base station device 10a or the like, and prediction is performed.

[0121] According to this embodiment, the control device 14 may be arranged in the wireless base station device 10a or in the RIC device 10b. This improves the degree of freedom in the configuration of the wireless communication system 4. Furthermore, by arranging the control device 14 in the RIC device 10b, it is possible to control multiple wireless base station devices 10a, thereby improving efficiency. Other configurations and effects are included in the descriptions of the first to third embodiments.

[0122] The present invention is not limited to the above-described embodiments, and can be modified as appropriate without departing from the spirit of the invention. For example, embodiments that combine the configurations of Embodiment 1, Modifications 1 to 3, and Embodiment 22 are also included within the scope of the technical idea. 。

[0123] The program includes a set of instructions (or software code) that, when loaded into a control device 10 including a computer, causes the computer to perform one or more functions described in the embodiments. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, computer-readable media or tangible storage media include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technology, CD-ROM, digital versatile disc (DVD), Blu-ray (registered trademark) disc or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. The program may also be transmitted on a transitory computer-readable medium or communication medium. By way of example and not limitation, transitory computer-readable media or communication media include electrical, optical, acoustic, or other forms of propagated signals.

[0124] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0125] (Appendix A1) A control device having a prediction unit that, during learning, uses time series data in which wireless quality information of each of a plurality of antennas that are distributed and that communicate wirelessly with a terminal device is arranged in chronological order as learning data, and uses a learning device that has learned allocation information including information of the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, based on operation time series data, which is the time series data acquired during operation, to predict predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data. (Appendix A2) a learning data selection unit that selects, during the learning, the time-series data classified according to a comparison result between the allocation information set as the learning label and the allocation information assigned to the wireless communication based on the wireless quality information immediately before the learning label in the time-series data, as the learning data; 10. The control device according to claim A1. (Appendix A3) the learning data selection unit selects, as the learning data, the time-series data classified according to a comparison result between pieces of allocation information assigned to the wireless communication based on each piece of wireless quality information in the time-series data. 10. The control device according to claim A2. (Appendix A4) a reconnection determination unit that determines whether a reconnection has occurred during the operation; the reconnection determination unit classifies reconnection time-series data, which is the time-series data in which the reconnection occurred, and reconnection assignment information, which includes information on the antennas assigned to the wireless communication after the predetermined time from the reconnection time-series data; the learning data selection unit selects the reconnection time-series data and the reconnection allocation information as the learning data and the learning label. 10. A control device according to claim A2 or A3. (Appendix A5) the learning data selection unit selects, as the learning data, the time series data classified according to movement information of the terminal device calculated from the time series data; The control device according to any one of appendices A2 to A4. (Appendix A6) an antenna control unit that controls the antennas allocated to the wireless communication based on the wireless quality information; the antenna control unit adds the prediction information predicted by the prediction unit to a candidate antenna that is a candidate for the antenna to be assigned based on the wireless quality information. A control device according to any one of appendices A1 to A5. (Appendix A7) The allocation information and the prediction information include information on beams input and output using a plurality of antenna elements in the antenna. The control device according to any one of appendices A1 to A6. (Appendix A8) The wireless quality information includes at least one of signal power information of RSRP and RSSI of the antenna and channel estimation information of each frequency and each RB. A control device according to any one of appendices A1 to A7. (Appendix A9) the allocation information includes a first antenna and a second antenna next to the first antenna allocated to the wireless communication based on the wireless quality information, The prediction information includes a third antenna and a fourth antenna that is next in rank to the third antenna and that will be assigned after the predetermined time from the operational time-series data. The control device according to any one of appendices A1 to A8. (Appendix A10) the allocation information is an allocation set including at least a first antenna and a second antenna next to the first antenna that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third antenna and a fourth antenna next to the third antenna, which will be assigned after the predetermined time from the operational time-series data. The control device according to any one of appendices A1 to A8. (Appendix A11) the allocation information includes a first beam and a second beam next to the first beam allocated to the wireless communication based on the wireless quality information, The prediction information includes a third beam to be assigned after the predetermined time from the operation time-series data and a fourth beam next to the third beam. 10. The control device according to claim A7. (Appendix A12) the allocation information is an allocation set including at least a first beam and a second beam next to the first beam that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third beam and a fourth beam next to the third beam, which will be assigned after the predetermined time from the operation time-series data. 10. The control device according to claim A7. (Appendix A13) the learning device includes, in the learning data, movement information of the terminal device calculated from the time-series data. The control device according to any one of appendices A1 to A12. (Appendix A14) The wireless quality information includes information in which values ​​equal to or less than a certain threshold are replaced with zero or a lower limit value and input. A control device according to any one of appendices A1 to A13. (Appendix B1) a plurality of antennas arranged in a distributed manner for wireless communication with a terminal device; a control device having a prediction unit that, during learning, uses time-series data in which wireless quality information of each of the plurality of antennas is arranged in chronological order as learning data, and predicts predicted information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data based on operation-time time-series data that is the time-series data acquired during operation, by using a learner that has learned allocation information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the learning data as learning labels; A wireless communication system comprising: (Appendix B2) the control device further includes a learning data selection unit that selects, during the learning, the time-series data classified according to a comparison result between the allocation information set as the learning label and the allocation information assigned to the wireless communication based on the wireless quality information immediately before the learning label in the time-series data, as the learning data. 10. A wireless communication system as described in Appendix B1. (Appendix B3) the learning data selection unit selects, as the learning data, the time-series data classified according to a comparison result between pieces of allocation information assigned to the wireless communication based on each piece of wireless quality information in the time-series data. 10. A wireless communication system as described in Appendix B2. (Appendix B4) the control device further includes a reconnection determination unit that determines whether a reconnection has occurred during operation, the reconnection determination unit classifies reconnection time-series data, which is the time-series data in which the reconnection occurred, and reconnection assignment information, which includes information on the antennas assigned to the wireless communication after the predetermined time from the reconnection time-series data; the learning data selection unit selects the reconnection time-series data and the reconnection allocation information as the learning data and the learning label. 10. A wireless communication system according to claim 8, wherein the wireless communication system is a wireless communication system according to claim 9. (Appendix B5) the learning data selection unit selects, as the learning data, the time series data classified according to movement information of the terminal device calculated from the time series data; A wireless communication system according to any one of appendices B2 to B4. (Appendix B6) The allocation information and the prediction information include information on beams input and output using a plurality of antenna elements in the antenna. A wireless communication system according to any one of appendices B1 to B5. (Appendix B7) The wireless quality information includes at least one of signal power information of RSRP and RSSI of the antenna and channel estimation information of each frequency and each RB. A wireless communication system according to any one of appendices B1 to B6. (Appendix B8) the allocation information includes a first antenna and a second antenna next to the first antenna allocated to the wireless communication based on the wireless quality information, The prediction information includes a third antenna and a fourth antenna that is next in rank to the third antenna and that will be assigned after the predetermined time from the operational time-series data. A wireless communication system according to any one of appendices B1 to B7. (Appendix B9) the allocation information is an allocation set including at least a first antenna and a second antenna next to the first antenna that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third antenna and a fourth antenna next to the third antenna, which will be assigned after the predetermined time from the operational time-series data. A wireless communication system according to any one of appendices B1 to B7. (Appendix B10) the allocation information includes a first beam and a second beam next to the first beam allocated to the wireless communication based on the wireless quality information, The prediction information includes a third beam to be assigned after the predetermined time from the operation time-series data and a fourth beam next to the third beam. 10. A wireless communication system as described in Appendix B6. (Appendix B11) the allocation information is an allocation set including at least a first beam and a second beam next to the first beam that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third beam and a fourth beam next to the third beam, which will be assigned after the predetermined time from the operation time-series data. 10. A wireless communication system as described in Appendix B6. (Appendix B12) the learning device includes, in the learning data, movement information of the terminal device calculated from the time-series data. A wireless communication system according to any one of appendices B1 to B11. (Appendix B13) The wireless quality information includes information in which values ​​equal to or less than a certain threshold are replaced with zero or a lower limit value and input. A wireless communication system according to any one of appendices B1 to B12. (Appendix B14) the control device further includes an antenna control unit that controls the antennas allocated to the wireless communication based on the wireless quality information, the antenna control unit adds the prediction information predicted by the prediction unit to a candidate antenna that is a candidate for the antenna to be assigned based on the wireless quality information. A wireless communication system according to any one of appendices B1 to B13. (Appendix B15) The control device is provided in a radio base station device. A wireless communication system according to any one of appendices B1 to B14. (Appendix B16) a RIC for controlling one or more radio base station devices; The control device provided in the RIC, acquiring the radio quality information from at least one of the radio base station devices; outputting the prediction information to at least one of the wireless base station devices; A wireless communication system according to any one of appendices B1 to B13. (Appendix C1) During learning, time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with a terminal device is arranged in chronological order is used as learning data, and a learner that has learned allocation information including information of the antennas that will be assigned to the wireless communication after a predetermined time from the learning data is used as learning labels is used to predict predicted information including information of the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data, based on operation time series data that is the time series data acquired during operation. Control method. (Appendix C2) During the learning, time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with the terminal device is arranged in time series is used as learning data, and allocation information including information of the antennas that are assigned to the wireless communication after a predetermined time from the learning data is learned as a learning label; When learning the above, selecting, as the learning data, the time-series data classified according to a comparison result between the allocation information set as the learning label and the allocation information assigned to the wireless communication based on the wireless quality information immediately before the learning label in the time-series data; A control method as described in Appendix C1. (Appendix C3) When selecting the training data, selecting, as the learning data, the time-series data classified according to a comparison result between pieces of allocation information assigned to the wireless communication based on each piece of wireless quality information in the time-series data; A control method as described in Appendix C2. (Appendix C4) determining whether a reconnection has occurred during the operation; When determining whether the reconnection has occurred, classifying reconnection time-series data, which is the time-series data in which the reconnection occurred, and reconnection assignment information, which includes information on the antennas assigned to the wireless communication after the predetermined time from the reconnection time-series data; selecting the reconnection time-series data and the reconnection allocation information as the learning data and the learning label; 10. The control method according to claim C2 or C3. (Appendix C5) When selecting the training data, selecting, as the learning data, the time-series data classified according to movement information of the terminal device calculated from the time-series data; 10. The control method according to claim C2 or C3. (Appendix C6) controlling the antennas to be allocated based on the wireless quality information; When controlling the antenna, adding the predicted prediction information to a candidate antenna that is a candidate for the antenna to be assigned based on the wireless quality information; A control method according to any one of appendices C1 to C3. (Appendix C7) The allocation information and the prediction information include information on beams input and output using a plurality of antenna elements in the antenna. A control method according to any one of appendices C1 to C6. (Appendix C8) The wireless quality information includes at least one of signal power information of RSRP and RSSI of the antenna and channel estimation information of each frequency and each RB. A control method according to any one of appendices C1 to C7. (Appendix C9) the allocation information includes a first antenna and a second antenna next to the first antenna allocated to the wireless communication based on the wireless quality information, The prediction information includes a third antenna and a fourth antenna that is next in rank to the third antenna and that will be assigned after the predetermined time from the operational time-series data. A control method according to any one of appendices C1 to C8. (Appendix C10) the allocation information is an allocation set including at least a first antenna and a second antenna next to the first antenna that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third antenna and a fourth antenna next to the third antenna, which will be assigned after the predetermined time from the operational time-series data. A control method according to any one of appendices C1 to C8. (Appendix C11) the allocation information includes a first beam and a second beam next to the first beam allocated to the wireless communication based on the wireless quality information, The prediction information includes a third beam to be assigned after the predetermined time from the operation time-series data and a fourth beam next to the third beam. A control method as described in Appendix C7. (Appendix C12) the allocation information is an allocation set including at least a first beam and a second beam next to the first beam that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third beam and a fourth beam next to the third beam, which will be assigned after the predetermined time from the operation time-series data. A control method as described in Appendix C7. (Appendix C13) the learning device includes, in the learning data, movement information of the terminal device calculated from the time-series data. A control method according to any one of appendices C1 to C12. (Appendix C14) The wireless quality information includes information in which values ​​equal to or less than a certain threshold are replaced with zero or a lower limit value and input. A control method according to any one of appendices C1 to C13. (Appendix D1) During learning, time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with a terminal device is arranged in chronological order is used as learning data, and allocation information including information about the antennas that are assigned to the wireless communication after a predetermined time from the learning data is used as learning labels, thereby predicting predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data based on operation time series data, which is the time series data acquired during operation. A control program that causes a computer to execute a task. (Appendix D2) During the learning, time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with the terminal device is arranged in time series is used as learning data, and allocation information including information of the antennas that are assigned to the wireless communication after a predetermined time from the learning data is learned as a learning label; When the learning is performed, selecting, as the learning data, the time-series data classified according to a comparison result between the allocation information set as the learning label and the allocation information assigned to the wireless communication based on the wireless quality information immediately before the learning label in the time-series data; The control program according to appendix D1, which causes a computer to execute the above. (Appendix D3) When selecting the learning data, selecting, as the learning data, the time-series data classified according to a comparison result between pieces of allocation information assigned to the wireless communication based on pieces of wireless quality information in the time-series data; The control program according to appendix D2, which causes a computer to execute the above. (Appendix D4) determining whether a reconnection has occurred during the operation; When determining whether or not the reconnection has occurred, classifying reconnection time-series data, which is the time-series data in which the reconnection occurred, and reconnection assignment information, which includes information on the antennas assigned to the wireless communication after the predetermined time from the reconnection time-series data; selecting the reconnection time-series data and the reconnection allocation information as the learning data and the learning label; The control program according to appendix D2 or D3, which causes a computer to execute the above. (Appendix D5) When selecting the learning data, selecting, as the learning data, the time-series data classified according to movement information of the terminal device calculated from the time-series data; The control program according to appendix D2 or D3, which causes a computer to execute the above. (Appendix D6) controlling the antennas to be allocated based on the wireless quality information; When controlling the antenna, adding the predicted prediction information to a candidate antenna that is a candidate for the antenna to be assigned based on the wireless quality information; The control program according to any one of appendices D1 to D3, which causes a computer to execute the above. (Appendix D7) The allocation information and the prediction information include information on beams input and output using a plurality of antenna elements in the antenna. A control program according to any one of appendices D1 to D6. (Appendix D8) The wireless quality information includes at least one of signal power information of RSRP and RSSI of the antenna and channel estimation information of each frequency and each RB. A control program according to any one of appendices D1 to D7. (Appendix D9) the allocation information includes a first antenna and a second antenna next to the first antenna allocated to the wireless communication based on the wireless quality information, The prediction information includes a third antenna and a fourth antenna that is next in rank to the third antenna and that will be assigned after the predetermined time from the operational time-series data. A control program according to any one of appendices D1 to D8. (Appendix D10) the allocation information is an allocation set including at least a first antenna and a second antenna next to the first antenna that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third antenna and a fourth antenna next to the third antenna, which will be assigned after the predetermined time from the operational time-series data. A control program according to any one of appendices D1 to D8. (Appendix D11) the allocation information includes a first beam and a second beam next to the first beam allocated to the wireless communication based on the wireless quality information, The prediction information includes a third beam to be assigned after the predetermined time from the operation time-series data and a fourth beam next to the third beam. The control program described in Appendix D7. (Appendix D12) the allocation information is an allocation set including at least a first beam and a second beam next to the first beam that are allocated to the wireless communication based on the wireless quality information, The prediction information is a prediction set including at least a third beam and a fourth beam next to the third beam, which will be assigned after the predetermined time from the operation time-series data. The control program described in Appendix D7. (Appendix D13) the learning device includes, in the learning data, movement information of the terminal device calculated from the time-series data. A control program according to any one of appendices D1 to D12. (Appendix D14) The wireless quality information includes information in which values ​​equal to or less than a certain threshold are replaced with zero or a lower limit value and input. A control program according to any one of appendices D1 to D13. [Explanation of symbols]

[0126] 1, 4 Wireless communication system 10, 12, 13, 14 Control device 10a Wireless base station equipment 10b RIC device 11 Prediction Department 15 processors 16 memory 17 Storage device 20 Learning Units 30, 30a, 30b Learning data selection unit 31, 32 Beam determination unit 33, 34 Comparison section 35 Data storage section 36 Data selection section 37 Reconnection determination unit 38 Speed ​​judgment section 40 Movement speed calculation section 50 Terminal Equipment 61 Building 62 Streetlight 63 Traffic Light 64 Telegraph pole 65, 66 Shield 70 Radio quality estimation section 80 Antenna control unit AP, AP1, AP2, AP3 antennas B, B1, B2, B3, B4, B5 beams P1, P2, P3, P4, P5 position

Claims

1. A control device having a prediction unit that, during learning, uses time series data in which wireless quality information of each of a plurality of antennas that are distributed and that communicate wirelessly with a terminal device is arranged in chronological order as learning data, and uses a learning device that has learned allocation information including information of the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, based on operation time series data, which is the time series data acquired during operation, to predict predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data.

2. a learning data selection unit that selects, during the learning, the time-series data classified according to a comparison result between the allocation information set as the learning label and the allocation information assigned to the wireless communication based on the wireless quality information immediately before the learning label in the time-series data, as the learning data; The control device according to claim 1 .

3. the learning data selection unit selects, as the learning data, the time-series data classified according to a comparison result between pieces of allocation information assigned to the wireless communication based on each piece of wireless quality information in the time-series data. The control device according to claim 2 .

4. a reconnection determination unit that determines whether a reconnection has occurred during the operation; the reconnection determination unit classifies reconnection time-series data, which is the time-series data in which the reconnection occurred, and reconnection assignment information, which includes information on the antennas assigned to the wireless communication after the predetermined time from the reconnection time-series data; the learning data selection unit selects the reconnection time-series data and the reconnection allocation information as the learning data and the learning label. The control device according to claim 2 or 3.

5. the learning data selection unit selects, as the learning data, the time series data classified according to movement information of the terminal device calculated from the time series data; The control device according to claim 2 or 3.

6. an antenna control unit that controls the antennas allocated to the wireless communication based on the wireless quality information; the antenna control unit adds the prediction information predicted by the prediction unit to a candidate antenna that is a candidate for the antenna to be assigned based on the wireless quality information. The control device according to any one of claims 1 to 3.

7. The allocation information and the prediction information include information on beams input and output using a plurality of antenna elements in the antenna. The control device according to any one of claims 1 to 3.

8. The wireless quality information includes at least one of signal power information of RSRP and RSSI of the antenna, and channel estimation information of each frequency and each RB. The control device according to any one of claims 1 to 3.

9. the allocation information includes a first antenna and a second antenna next to the first antenna allocated to the wireless communication based on the wireless quality information, The prediction information includes a third antenna and a fourth antenna that is next in rank to the third antenna and that will be assigned after the predetermined time from the operation time-series data. The control device according to any one of claims 1 to 3.

10. the allocation information is an allocation set including at least a first antenna and a second antenna next to the first antenna that are allocated to the wireless communication based on the wireless quality information, the prediction information is a prediction set including at least a third antenna and a fourth antenna next to the third antenna, which will be assigned after the predetermined time from the operational time-series data; The control device according to any one of claims 1 to 3.

11. the allocation information includes a first beam and a second beam next to the first beam allocated to the wireless communication based on the wireless quality information, The prediction information includes a third beam to be assigned after the predetermined time from the operation time-series data and a fourth beam next to the third beam. The control device according to claim 7.

12. the allocation information is an allocation set including at least a first beam and a second beam next to the first beam that are assigned to the wireless communication based on the wireless quality information, the prediction information is a prediction set including at least a third beam and a fourth beam next to the third beam, which will be assigned after the predetermined time from the operation time-series data; The control device according to claim 7.

13. the learning device includes, in the learning data, movement information of the terminal device calculated from the time-series data. The control device according to any one of claims 1 to 3.

14. The wireless quality information includes information in which a value equal to or less than a certain threshold is replaced with zero or a lower limit value and input. The control device according to any one of claims 1 to 3.

15. a plurality of antennas arranged in a distributed manner for wireless communication with a terminal device; a control device having a prediction unit that, during learning, uses time-series data in which wireless quality information of each of the plurality of antennas is arranged in chronological order as learning data, and predicts predicted information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data based on operation-time time-series data that is the time-series data acquired during operation, by using a learner that has learned allocation information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the learning data as learning labels; A wireless communication system comprising:

16. The control device is provided in a radio base station device.

16. The wireless communication system of claim 15.

17. The radio communication system further includes a RIC that controls one or more radio base station devices, The control device provided in the RIC, acquiring the radio quality information from at least one of the radio base station devices; outputting the prediction information to at least one of the wireless base station devices; 16. The wireless communication system of claim 15.

18. An information processing device including a computer uses time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with a terminal device is arranged in chronological order as learning data during learning, and uses a learner that has learned allocation information including information about the antennas that will be assigned to the wireless communication after a predetermined time from the learning data as learning labels, thereby predicting predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data based on the operation time series data, which is the time series data obtained during operation. Control method.

19. During learning, time series data in which wireless quality information of each of a plurality of antennas that are distributed and that perform wireless communication with a terminal device is arranged in chronological order is used as learning data, and allocation information including information about the antennas that are assigned to the wireless communication after a predetermined time from the learning data is used as learning labels, thereby predicting predicted information including information about the antennas that will be assigned to the wireless communication after the predetermined time from the operation time series data based on operation time series data, which is the time series data acquired during operation. A control program that causes a computer to execute a task.

Images