Control board and program

Abstract

Provided is a control board comprising: a training data storage unit for storing communication status training data, including disaster log data for recording a disaster that occurred in the past, disaster-stricken area log data for recording a disaster-stricken area, and communication status log data for recording the communication status in the disaster-stricken area; and an execution unit having a model generation function for generating, through machine learning, a communication status estimation model that estimates the communication status in the disaster-stricken area from disaster identification data identifying the disaster and disaster-stricken area identification data identifying the disaster-stricken area by using a plurality of pieces of the communication status training data stored in the training data storage unit as teacher data, and a RAN control function for controlling an RAN function covering the disaster-stricken area by executing an AI process (RAN Intelligent Controller (RIC)) or the like on the basis of a communication status estimation result obtained by estimating the communication status in the disaster-stricken area from the disaster identification data and the disaster-stricken area identification data using the communication status estimation model.

Description

Control infrastructure and program ,

[0004] ,

[0005]

[0001] The present invention relates to a control infrastructure and a program.

[0002] In Patent Document 1, there is described an in-region information estimation model for estimating information portions of event information and related information within a certain region, which can prepare training (learning) data necessary for construction and can also reflect various related information in the estimation. There are also described an in-region information estimation apparatus and method using the model. In Patent Document 2, there is described a disaster-time traffic prediction method and system for accurately calculating the estimated traffic volume of network nodes for each disaster scale. In Patent Document 3, there is described a technique for visualizing disaster risks and regions with high disaster risks. [Prior art documents] [Patent documents] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2022-178385 [Patent Document 2] Japanese Unexamined Patent Application Publication No. 2010-062681 [Patent Document 3] Japanese Unexamined Patent Application Publication No. 2018-194968

[0003] When a large-scale disaster occurs, a communication environment that is significantly different from normal and changes moment by moment occurs due to damage to wireless base stations caused by the disaster and sudden changes in the flow of people. Even in conventional communication systems, various measures have been taken to optimize the communication environment in the event of a large-scale disaster, but it is an optimization of the communication environment mainly based on information obtained from communication facilities.

[0004] In the system according to the present embodiment, for example, the function of the RAN (Radio Access Network) is made to run on a high-performance GPU (Graphics Processing Unit) server instead of a general-purpose machine server, so that the surplus computing resources can be utilized in the form of AI (Artificial Intelligence) processing. As types of AI processing, there are AI processing related to RAN control (sometimes described as RAN control AI processing) and AI processing not related to RAN control (sometimes described as non-RAN control AI processing).

[0005] As an example of RAN control AI processing, RIC (RAN Intelligent Controller) can be cited. RIC is a technology that uses AI to optimize RAN radio resources and automate RAN operation. RIC includes Non-RT RIC and Near-RT RIC (Near-Real Time RIC). Non-RT RIC is sometimes called Centralized RIC. Non-RT RIC is located inside SMO (Service Management and Orchestration) that performs RAN management and orchestration. Non-RT RIC generates and notifies policies related to RAN control and transmits information to Near-RT RIC. For example, Non-RT RIC generates a learned model related to RAN control by executing machine learning using data collected from RAN and transmits it to Near-RT RIC. Near-RT RIC is sometimes called Distributed RIC. Near-RT RIC is located closer to RAN nodes (RU (Radio Unit), DU (Distributed Unit), CU (Central Unit)) compared to Non-RT RIC and performs control of RAN nodes, resource control, etc. Near-RT RIC performs processing with higher real-time performance compared to Non-RT RIC. Near-RT RIC, for example, executes inference processing related to RAN control using the learned model obtained from Non-RT RIC. RAN control AI processing is not limited to RIC.

[0006] Non-RAN control AI processing may correspond to so-called MEC (Multi-access Edge Computing) applications. Examples of non-RAN control AI processing include, but are not limited to, monitoring AI execution processing for determining the situation within the imaging range of an input imaging image, answer AI execution processing for outputting an answer to an inquiry by an input user, etc.

[0007] In the system according to this embodiment, for example, based on information obtained from communication equipment such as damage and restoration of communication environments and communication facilities during past major disasters, and information from external institutions that cannot be obtained from communication equipment such as changes in the flow of people, changes in demand content, and risks based on hazard maps, a mechanism for learning simulations is adopted. Furthermore, in the system according to this embodiment, a mechanism is adopted to integrate the simulation results with communication data obtained from the disaster area and infer and optimize changes in the communication environment. As a result, the system according to this embodiment forms a top-down simulation like the human brain, and by multiplying in real time the multimodal bottom-up information coming from communication equipment and various institutions that arise during a disaster, plastic traffic control can be realized. In addition, the system according to this embodiment can provide a more flexible communication environment by utilizing communication situations in anticipation of an increase in IoT (Internet of Things) terminals, an increase in mobile terminals possessed by non-humans such as autonomous vehicles, and the like.

[0008] According to an embodiment of the present invention, a control board is provided. The control board may include a learning data storage unit that stores communication situation learning data including disaster log data that records disasters that occurred in the past, disaster area log data that records disaster areas damaged by the disasters, and communication situation log data that records communication situations in the disaster areas after the disasters occurred. The control board uses the plurality of communication situation learning data stored in the learning data storage unit as teacher data, and generates a communication situation estimation model for estimating the communication situation in the disaster area after the disaster occurred by machine learning from disaster identification data that identifies disasters and disaster area identification data that identifies the disaster areas damaged by the disasters. The control board also has an execution unit having a RAN control function that controls the function of the RAN that covers the disaster area based on a communication situation estimation result obtained by estimating the communication situation in the disaster area after the disaster occurred from the disaster identification data that identifies disasters and the disaster area identification data that identifies the disaster areas damaged by the disasters using the communication situation estimation model.

[0009] In the control infrastructure, the RAN control function controls the functions of the RAN so that the radio resources of the RAN are preferentially allocated to the communication related to the disaster.

[0010] In any of the control infrastructures, the learning data storage unit may further store disaster situation learning data including disaster log data recording disasters that occurred in the past, disaster area log data recording disaster areas affected by the disaster, hazard map log data recording hazard maps covering the disaster areas, and disaster situation log data recording the disaster situations of the disaster areas. The model generation function may further generate a disaster situation estimation model for estimating the disaster situation of the disaster area by machine learning from a plurality of the disaster situation learning data stored in the learning data storage unit as teacher data, disaster identification data for identifying disasters, disaster area identification data for identifying the disaster areas affected by the disasters, and hazard map data covering the disaster areas. The RAN control function may control the functions of the RAN covering the disaster area based on the disaster situation estimation result obtained by estimating the disaster situation of the disaster area from the disaster identification data, the disaster area identification data, and the hazard map data covering the disaster area using the disaster situation estimation model.

[0011] In any of the control infrastructures, the RAN control function may control the functions of the RAN so that the power consumption consumed by a radio base station estimated to have a longer operating time using battery power after the occurrence of the disaster is preferentially reduced among a plurality of radio base stations constituting the RAN based on at least any of the communication situation estimation result and the disaster situation estimation result.

[0012] In any of the control boards, based on at least any one of the communication situation estimation result and the disaster situation estimation result, the RAN control function may control the function of the RAN so that the coverage area of one radio base station constituting the RAN, which is estimated to have a normally functioning wireless communication function, covers at least a part of the coverage area of another radio base station constituting the RAN, which is estimated to have a non-normally functioning wireless communication function.

[0013] In any of the control boards, based on at least any one of the communication situation estimation result and the disaster situation estimation result, the RAN control function may control the function of the RAN so that the coverage area of one radio base station mounted on a mobile body and constituting the RAN covers at least a part of the coverage area of another radio base station constituting the RAN, which is estimated to have a non-normally functioning wireless communication function.

[0014] In any of the control boards, the learning data storage unit may further store crowd flow situation learning data including disaster log data recording disasters that occurred in the past, disaster area log data recording disaster areas damaged by the disasters, evacuation shelter map log data recording evacuation shelter maps covering the disaster areas, and crowd flow situation log data recording the crowd flow situation in the disaster areas. The model generation function may further generate a crowd flow situation estimation model for estimating the crowd flow situation in the disaster area by machine learning from a plurality of the crowd flow situation learning data stored in the learning data storage unit as teacher data, disaster identification data for identifying disasters, disaster area identification data for identifying disaster areas damaged by the disasters, and evacuation shelter map data covering the disaster areas. The RAN control function may further control the function of the RAN covering the disaster area based on the crowd flow situation estimation result obtained by estimating the crowd flow situation in the disaster area from the disaster identification data, the disaster area identification data, and the evacuation shelter map data covering the disaster area using the crowd flow situation estimation model.

[0015] In any of the control platforms, the RAN control function may control the function of the RAN such that the radio resources of the RAN are preferentially allocated to the communication in an area where a larger number of people per unit area is estimated within the disaster area, based on at least one of the communication situation estimation result and the people flow situation estimation result.

[0016] Any of the control platforms may further include an acquisition unit that acquires real-time information indicating at least one of the communication situation, the disaster situation, and the people flow situation in the disaster area and having real-time property, and the RAN control function may control the function of the RAN covering the disaster area based on the real-time information.

[0017] In any of the control platforms, the acquisition unit may acquire SNS (Social Networking Service) messages as the real-time information, the control platform may further include a selection unit that selects SNS messages passing through radio base stations constituting the RAN covering the disaster area from among a plurality of the SNS messages, and the RAN control function may control the function of the RAN covering the disaster area based on the SNS messages selected by the selection unit.

[0018] According to an embodiment of the present invention, a control board is provided. The control board may include a learning data storage unit that stores comprehensive situation learning data including disaster log data recording disasters that occurred in the past, disaster area log data recording disaster areas affected by the disasters, hazard map log data recording hazard maps covering the disaster areas, communication situation log data recording communication situations in the disaster areas after the disasters occurred, and disaster situation log data recording disaster situations in the disaster areas. The control board uses the plurality of comprehensive situation learning data stored in the learning data storage unit as teacher data to generate, by machine learning, a comprehensive situation estimation model that estimates the comprehensive situation of the disaster area, including the communication situation and the disaster situation in the disaster area after the disaster occurred, from disaster identification data for identifying disasters, disaster area identification data for identifying the disaster areas affected by the disasters, and hazard map data covering the disaster areas. The control board may include an execution unit having a RAN control function that controls the function of a RAN covering the disaster area based on a comprehensive situation estimation result obtained by estimating the comprehensive situation of the disaster area after the disaster occurred from the disaster identification data for identifying disasters, the disaster area identification data for identifying the disaster areas affected by the disasters, and the hazard map data covering the disaster areas, using the comprehensive situation estimation model.

[0019] According to an embodiment of the present invention, a program for causing a computer to function as any one of the control boards is provided when executed by the computer.

[0020] Note that the above summary of the invention does not list all the necessary features of the present invention. Also, sub-combinations of these feature groups can also be inventions.

[0021] FIG. 1 is a schematic diagram illustrating an example of a system 10. FIG. 2 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 3 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 4 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 5 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 6 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 7 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 8 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 9 is an explanatory diagram illustrating an example of a process in which the control board 100 controls the functions of the RAN. FIG. 10 is an explanatory diagram illustrating an example of a functional configuration of the control board 100. FIG. 11 is an explanatory diagram illustrating an example of a processing flow of the system 10. FIG. 11 is a schematic diagram illustrating an example of the hardware configuration of a computer 1200 that functions as the control board 100.

[0022] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention as claimed. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0023] 1 schematically illustrates an example of a system 10. The system 10 may include a control platform 100. The system 10 may include a data management device 200. The system 10 may include a model generation device 300. In the system 10 according to this embodiment, for example, the control platform 100 may control the RAN and perform AI processing.

[0024] The RAN may be a virtualized vRAN (Virtual RAN), and the control infrastructure 100 may control the vRAN. The RAN may be a physical RAN, and the control infrastructure 100 may control the physical RAN. In this embodiment, a case where the RAN is a vRAN will be mainly described as an example.

[0025] The AI ​​processing performed by the control infrastructure 100 may include RAN-controlled AI processing (sometimes referred to as RAN_AI). The AI ​​processing performed by the control infrastructure 100 may include non-RAN-controlled AI processing (sometimes referred to as non-RAN_AI).

[0026] The control infrastructure 100 may be a data center located in various locations. The control infrastructure 100 may be configured with multiple devices. The control infrastructure 100 may be realized on a virtualization infrastructure consisting of multiple devices. The control infrastructure 100 may also be realized by a single device. In other words, the control infrastructure 100 may be a control device.

[0027] FIG. 1 illustrates an example in which the system 10 includes one control board 100. The system 10 may also include multiple control boards 100. In this case, the multiple control boards 100 are arranged hierarchically, for example. For example, when multiple control boards 100 are arranged in two layers, the control board 100 in the upper layer may be called a Core Brain, and the control board 100 in the lower layer may be called a Regional Brain. For example, when multiple control boards 100 are arranged in three layers, the control board 100 in the upper layer may be called a Core Brain, the control board 100 in the middle layer may be called a Regional Brain, and the control board 100 in the lower layer may be called a Sub-Regional Brain.

[0028] The control board 100 may be provided with one or more central processing units (CPUs). The control board 100 may be provided with one or more GPUs. The control board 100 may be provided with multiple super chips, each of which has a CPU and a GPU connected via an interconnect. The interconnect may have memory coherency and may be capable of achieving high bandwidth and low latency. In this way, the control board 100 may have CPU resources and GPU resources as computational resources.

[0029] The control board 100 receives various data used in the learning process from, for example, a data management device 200 that manages various data. The control board 100 receives various data used in the learning process from, for example, the data management device 200 via a network 20.

[0030] The network 20 may include a core network provided by a telecommunications carrier. The core network may conform to, for example, a 5th Generation (5G) communication system. The core network may conform to a 6th Generation (6G) communication system or a later mobile communication system. The core network may conform to a 3rd Generation (3G) communication system. The core network may conform to a Long Term Evolution (LTE) communication system. The network 20 may include the Internet.

[0031] The control infrastructure 100 is, for example, located on a core network. The term "on the core network" includes both inside and outside the core network. If the control infrastructure 100 is a control device, the control infrastructure 100 may be located on the Internet.

[0032] The data management device 200 manages, for example, disaster log data that records disasters that have occurred in the past. The disaster is, for example, an earthquake. The disaster is, for example, a tsunami. The disaster is, for example, a typhoon. The disaster is, for example, a flood. The disaster is, for example, a high tide. The disaster is, for example, heavy rain. The disaster is, for example, heavy snow. The disaster is, for example, a landslide. The disaster is, for example, a volcanic eruption. The disaster may be any other natural disaster. The disaster may be a man-made disaster such as a fire.

[0033] The disaster log data includes, for example, disaster type log data that records the type of disaster. The disaster log data includes, for example, disaster scale log data that records the scale of the disaster. The disaster log data includes, for example, disaster occurrence location log data that records the location where the disaster occurred. The disaster log data includes, for example, disaster occurrence time log data that records the time when the disaster occurred. In the case of a disaster whose disaster center moves, such as a typhoon, the disaster log data may include disaster movement path log data that records the movement path of the disaster center. The disaster movement path log data may associate the location of the disaster center with the time when the disaster center was located at that location.

[0034] The data management device 200 manages, for example, disaster area log data that records disaster areas affected by disasters that have occurred in the past. The disaster area log data corresponds to, for example, disaster log data. The correspondence of the disaster area log data to the disaster log data may mean that the disaster area recorded by the disaster area log data is the disaster area of ​​the disaster recorded by the disaster log data.

[0035] The disaster area log data may record, for example, one disaster area, but may also record multiple disaster areas.

[0036] The data management device 200 is managed by, for example, a public institution. The data management device 200 is managed by, for example, an administrative agency. The data management device 200 may also be managed by a private business operator.

[0037] Here, an example will be described in which the control board 100 executes the learning process using disaster log data and disaster area log data. Here, the control board 100 starts in a state in which it does not store any learning data to be used in the learning process.

[0038] The control infrastructure 100 receives disaster log data from, for example, the data management device 200. The control infrastructure 100 receives disaster area log data corresponding to the disaster log data from, for example, the data management device 200.

[0039] The control board 100 acquires communication status log data that records the communication status in the disaster-affected area recorded in the disaster area log data after the disaster recorded in the disaster log data occurs. The communication status log data records the communication status in the disaster-affected area from when the disaster occurred to when the disaster-affected area was restored.

[0040] The communication status log data includes, for example, RAN demand log data that records the demand status of the RAN covering the disaster area. The communication status log data includes, for example, communication terminal number log data that records the status of the number of communication terminals in the RAN covering the disaster area. The communication status log data includes, for example, radio resource consumption log data that records the status of the consumption of radio resources in the RAN covering the disaster area. The communication status log data includes, for example, traffic log data that records the status of traffic, which is the amount of packet data transferred per unit time on the RAN covering the disaster area. The communication status log data includes, for example, congestion log data that records the status of congestion in the RAN covering the disaster area. The communication status log data includes, for example, computational resource consumption log data that records the status of the consumption of computational resources in the control infrastructure 100 that controls the RAN covering the disaster area. The communication status log data includes, for example, radio base station log data that records the status of the radio base station 40 that constitutes the RAN covering the disaster area. The status of the radio base station 40 includes, for example, whether or not the wireless communication function is functioning normally. The status of the radio base station 40 includes, for example, the time when the radio communication function stopped functioning normally. The status of the radio base station 40 includes, for example, the operating time during which the radio base station 40 is operating using battery power. The status of the radio base station 40 includes, for example, the status of the radio resources allocated to the radio base station 40. The status of the radio base station 40 includes, for example, the status of the coverage area 140 of the radio base station 40.

[0041] The communication status log data is managed by, for example, a telecommunications carrier. The communication status log data may be managed by the data management device 200. In this case, the control board 100 may receive the communication status log data from the data management device 200.

[0042] The control infrastructure 100 stores communication status learning data including, for example, disaster log data, disaster area log data, and communication status log data. For example, the control infrastructure 100 uses the stored communication status learning data as training data to generate, through machine learning, a communication status estimation model that estimates the communication status in a disaster-affected area identified by the disaster area identification data after the occurrence of a disaster identified by the disaster identification data, from disaster identification data that identifies a disaster and disaster area identification data corresponding to the disaster identification data. Note that the correspondence of the disaster area identification data to the disaster identification data may mean that the disaster area identified by the disaster area identification data is an area affected by the disaster identified by the disaster identification data.

[0043] The disaster identification data includes disaster type identification data that identifies the type of disaster. The disaster identification data includes, for example, disaster scale identification data that identifies the scale of the disaster. The disaster identification data includes, for example, disaster occurrence location identification data that identifies the location where the disaster occurred. The disaster identification data includes, for example, disaster occurrence time identification data that identifies the time when the disaster occurred. In the case of a disaster whose disaster center moves, such as a typhoon, the disaster identification data may include disaster movement path identification data that identifies the movement path of the disaster center.

[0044] The disaster area identification data may, for example, identify one disaster area, or may identify multiple disaster areas.

[0045] The control infrastructure 100, for example, uses the generated communication situation estimation model to obtain a communication situation estimation result that estimates the communication situation in the disaster area identified by the disaster area identification data after the occurrence of a disaster identified by the disaster identification data, from the disaster identification data and the disaster area identification data corresponding to the disaster identification data. The control infrastructure 100 receives the disaster identification data and the disaster area identification data from the data management device 200, for example.

[0046] The communication status includes, for example, the demand status of the RAN covering the affected area. The communication status includes, for example, the status of the number of communication terminals in the RAN. The communication status includes, for example, the status of the consumption of radio resources in the RAN. The communication status includes, for example, the status of traffic transferred on the RAN. The communication status includes, for example, the congestion status of the RAN. The communication status includes, for example, the status of the consumption of computational resources of the control infrastructure 100 that controls the RAN. The communication status includes, for example, the status of the radio base stations 40 that make up the RAN.

[0047] The control infrastructure 100 controls the functions of the RAN, for example, based on the acquired communication situation estimation result. The control infrastructure 100 controls the functions of the RAN, for example, to control the number of communication terminals in the RAN. The control infrastructure 100 controls the functions of the RAN, for example, to control the amount of radio resources consumed by the RAN. The control infrastructure 100 controls the functions of the RAN, for example, to control traffic of the RAN. The control infrastructure 100 controls the functions of the RAN, for example, to control the amount of computational resources consumed by the control infrastructure 100 that controls the RAN. The control infrastructure 100 controls the functions of the RAN, for example, to control each radio base station 40 of the multiple radio base stations 40 that make up the RAN. A specific control method for controlling the functions of the RAN will be described later.

[0048] The control platform 100 includes, for example, an execution unit having a model generation function that generates various estimation models through machine learning and a RAN control function that controls the functions of the RAN. The model generation function and the RAN control function may use the same computational resources. The computational resources of the control platform 100 may be the computational resources used for the model generation function and the RAN control function.

[0049] The control infrastructure 100 may receive, from a model generation device 300 that generates various estimation models, various estimation models generated by the model generation device 300. The control infrastructure 100 receives the various estimation models from the model generation device 300, for example, via the network 20. The control infrastructure 100 may control the functions of the RAN using the various estimation models received from the model generation device 300.

[0050] 1 illustrates an example in which the system 10 includes one data management device 200. The system 10 may include multiple data management devices 200. In this case, each of the multiple data management devices 200 may manage different data.

[0051] Conventionally, when a disaster occurs, telecommunications carriers improve the communication conditions in the affected area based on information indicating the communication conditions in the affected area after the disaster and their experience in dealing with similar disasters in the past. However, when a disaster occurs, people in the affected area tend to gather in places where they would not normally be present, such as evacuation shelters, so the communication conditions in the affected area may differ from the communication conditions in the absence of the disaster. Furthermore, the communication conditions in the affected area may vary even more depending on the type and scale of the disaster. Furthermore, wireless base stations constituting the RAN covering the affected area may be damaged by the disaster and may no longer function properly. Furthermore, the communication conditions in the affected area may change from moment to moment depending on the recovery status of the affected area. Therefore, when a disaster occurs, it is desirable to estimate the communication conditions in the affected area in advance, taking into account the type and scale of the disaster, and to maximize the use of limited resources based on the estimation results.

[0052] Furthermore, the radio base stations that make up the RAN must meet very high SLAs (Service Level Agreements), and it is necessary to create mechanisms, such as redundancy, to prevent service disruptions. As a result, only a few tenths of the computational resources on the execution platform that runs the RAN service can be used. In other words, there are many unused computational resources. Furthermore, at night, even though the demand for RAN services decreases, computational resources are wasted in an attempt to maintain the same level of service as during the daytime. In other words, the system is designed to accommodate peak demand.

[0053] In contrast, according to the system 10 of the present embodiment, the control infrastructure 100 uses multiple pieces of communication situation learning data as training data to generate a communication situation estimation model by machine learning, based on disaster identification data and disaster area identification data corresponding to the disaster identification data, to estimate the communication situation in the disaster area identified by the disaster identification data after the occurrence of the disaster identified by the disaster identification data. The control infrastructure 100 then uses the communication situation estimation model to estimate the communication situation in the disaster area identified by the disaster area identification data corresponding to the disaster identification data after the occurrence of the disaster identified by the disaster identification data. The control infrastructure 100 then controls the function of the RAN covering the disaster area based on the communication situation estimation result. The communication situation estimation model generated by the control infrastructure 100 is an estimation model that uses disaster identification data, including data such as the type and scale of the disaster, as input data, and outputs the communication situation in the disaster area, including the demand for a RAN covering the disaster area identified by the disaster area identification data corresponding to the disaster identification data, the number of wireless base stations in the disaster area whose wireless communication function is not functioning properly, and so on. Therefore, when a disaster occurs, the control board 100 uses the communication status estimation model to control the functions of the RAN that covers the affected area, so that the system 10 according to this embodiment can estimate the communication status of the affected area in advance, taking into account the type and scale of the disaster, and can maximize the use of limited resources based on the estimation results. As a result, the system 10 according to this embodiment can more appropriately improve the communication status of the affected area.

[0054] Furthermore, according to the system 10 of this embodiment, the control infrastructure 100 includes an execution unit having a model generation function and a RAN control function, and therefore the control infrastructure 100 can execute generation of an estimation model and control of RAN functions using the same computational resources. As a result, for example, during a time period when RAN service demand is relatively low, the control infrastructure 100 can execute the learning process of the estimation model within a range in which the SLAs of the radio base stations constituting the RAN are maintained. Therefore, the system 10 of this embodiment can improve the utilization efficiency of the computational resources of the control infrastructure while maintaining high SLAs of the radio base stations constituting the RAN.

[0055] 2 to 8 are explanatory diagrams for explaining an example of a process in which the control board 100 controls the functions of the RAN. Fig. 2 mainly shows an example of information used by the control board 100 to control the functions of the RAN, and Figs. 3 to 8 mainly show a control method in which the control board 100 controls the functions of the RAN.

[0056] 2 , the control platform 100 receives, for example, disaster log data, disaster area log data corresponding to the disaster log data, hazard map log data recording a hazard map covering the disaster area recorded in the disaster area log data, evacuation shelter map log data recording a shelter map covering the disaster area, disaster situation log data recording the damage situation in the disaster area, and people flow situation log data recording the people flow situation in the disaster area from the data management device 200. The control platform 100 also acquires, for example, communication situation log data recording the communication situation in the disaster area after the disaster recorded in the disaster log data occurs from a telecommunications carrier. The control platform 100 stores comprehensive situation learning data including the disaster log data, the disaster area log data, the hazard map log data, the evacuation shelter map log data, the communication situation log data, the disaster situation log data, and the people flow situation log data.

[0057] The disaster situation log data includes, for example, radio base station log data that records the status of radio base stations within the disaster area. The disaster situation log data includes, for example, traffic log data that records the traffic status within the disaster area. The traffic status includes, for example, the status of transportation facilities. The traffic status includes, for example, the status of roads. The traffic status includes, for example, the status of traffic congestion. The traffic status includes, for example, the status of the actual passage of cars, trucks, etc. The disaster situation log data includes, for example, building log data that records the status of buildings within the disaster area. The disaster situation log data includes, for example, weather log data that records the weather status within the disaster area. The disaster situation log data includes, for example, water level log data that records the water level status of rivers within the disaster area.

[0058] For example, the control platform 100 uses a plurality of comprehensive situation learning data as training data to generate a comprehensive situation estimation model through machine learning, which estimates the comprehensive situation of a disaster-stricken area after a disaster identified by the disaster identification data occurs, including the communication situation in the disaster-stricken area, the damage situation in the disaster-stricken area, and the human flow situation in the disaster-stricken area, from disaster identification data, disaster area identification data corresponding to the disaster identification data, hazard map data covering the disaster area identified by the disaster area identification data, and evacuation shelter map data covering the disaster area. The control platform 100 may generate the comprehensive situation estimation model through machine learning by performing pre-learning before a disaster occurs.

[0059] The damage situation includes, for example, the status of wireless base stations in the disaster area. The damage situation includes, for example, the traffic situation in the disaster area. The damage situation includes, for example, the status of buildings in the disaster area. The damage situation includes, for example, the weather situation in the disaster area. The damage situation includes, for example, the water level situation of rivers in the disaster area.

[0060] For example, when a disaster occurs, the control platform 100 uses the comprehensive situation estimation model to estimate the comprehensive situation of an area affected by the disaster. For example, the control platform 100 receives disaster identification data, disaster area identification data corresponding to the disaster identification data, a hazard map covering the disaster area identified by the disaster area identification data, and an evacuation shelter map covering the disaster area from the data management device 200. The control platform 100 may input the disaster identification data, the disaster area identification data, the hazard map, and the evacuation shelter map as input data into the comprehensive situation estimation model.

[0061] The data management device 200 manages, for example, a hazard map for each type of disaster. The data management device 200 manages, for example, an evacuation shelter map for each type of disaster.

[0062] The control platform 100, for example, inputs input data into a comprehensive situation estimation model to obtain a comprehensive situation estimation result that estimates the comprehensive situation of the affected area that has been affected by the disaster. The control platform 100 controls the functions of the RAN that covers the affected area based on the obtained comprehensive situation estimation result.

[0063] The control platform 100 acquires real-time information indicating, for example, at least one of the communication status, the damage status, and the human flow status in the disaster-stricken area, and may control the functions of the RAN covering the disaster-stricken area based on the real-time information.

[0064] 3 and 4, when a disaster occurs, the control platform 100 determines the allocation of radio resources of a RAN covering the disaster area based on an overall situation estimation result obtained by estimating the overall situation of the disaster-affected area using an overall situation estimation model and real-time information of the disaster area. The control platform 100 may control the function of the RAN to allocate the radio resources of the RAN in accordance with the determined allocation.

[0065] In the example shown in Figure 3, the control infrastructure 100 determines the allocation of radio resources of a RAN based on, for example, the number of communication terminals within the coverage area of ​​a radio base station constituting a RAN that covers a disaster area. For example, the control infrastructure 100 estimates the number of communication terminals within the coverage area of ​​each of multiple radio base stations constituting the RAN. Next, the control infrastructure 100 ranks the multiple radio base stations in descending order of the number of communication terminals within their coverage areas. Thereafter, the control infrastructure 100 determines the allocation of radio resources of the RAN so as to allocate more radio resources of the RAN to radio base stations with higher rankings.

[0066] 3 , the control infrastructure 100 estimates the number of communication terminals 50 within the coverage area 142 of the radio base station 42, the number of communication terminals 50 within the coverage area 144 of the radio base station 44, and the number of communication terminals 50 within the coverage area 146 of the radio base station 46. Next, the control infrastructure 100 ranks the radio base stations 42, 44, and 46 in descending order of the number of communication terminals within their coverage areas that are estimated to be the largest. Thereafter, the control infrastructure 100 determines the allocation of radio resources of the RAN so as to allocate the largest number of radio resources of the RAN to the radio base station 42, the second largest number of radio resources of the RAN to the radio base station 46, and the smallest number of radio resources of the RAN to the radio base station 44.

[0067] The communication terminal 50 may be any communication terminal capable of wireless communication with a wireless base station. For example, the communication terminal 50 may be a mobile phone such as a smartphone, a tablet terminal, a wearable terminal, or the like. The communication terminal 50 may be a PC (Personal Computer). The communication terminal 50 may be an IoT (Internet of Things) terminal. The communication terminal 50 may include anything that falls under the category of IoE (Internet of Everything).

[0068] In an example shown in Figure 4, the control infrastructure 100 determines the allocation of radio resources of a RAN based on, for example, the number of people within the coverage area of ​​a radio base station constituting a RAN that covers a disaster area. For example, the control infrastructure 100 estimates the number of people within the coverage area of ​​each of multiple radio base stations constituting the RAN. Next, the control infrastructure 100 ranks the multiple radio base stations in descending order of the number of people estimated to be within their coverage areas. Thereafter, the control infrastructure 100 determines the allocation of radio resources of the RAN so as to allocate more radio resources of the RAN to radio base stations with higher rankings.

[0069] 4 , the control infrastructure 100 estimates the number of people within the coverage area 142 of the radio base station 42, the number of people within the coverage area 144 of the radio base station 44, and the number of people within the coverage area 146 of the radio base station 46. Next, the control infrastructure 100 ranks the radio base stations 42, 44, and 46 in descending order of the number of people estimated to be within their coverage areas. Thereafter, the control infrastructure 100 determines the allocation of radio resources of the RAN so as to allocate the largest number of radio resources of the RAN to the radio base station 42, the second largest number of radio resources to the radio base station 46, and the smallest number of radio resources of the RAN to the radio base station 44.

[0070] 5 and 6 , when a disaster occurs, the control platform 100 reduces the power consumption of the radio base stations that constitute the RAN covering the disaster area based on the overall situation estimation result obtained by estimating the overall situation of the disaster-affected area using the overall situation estimation model and real-time information on the disaster area. The control platform 100 controls the functions of the RAN so as to reduce the power consumption of the radio base stations.

[0071] 5 , the control infrastructure 100 sets priorities for communications, for example. The control infrastructure 100 sets priorities for communications so that communications related to disasters have a higher priority than communications not related to disasters. The control infrastructure 100 reduces power consumption in the wireless base station, for example, by maintaining communications with a higher priority and temporarily suspending communications with a lower priority.

[0072] Disaster-related communications include, for example, communications between communication terminals owned by disaster victims, communications for obtaining information about the damage situation in a disaster-affected area, and communications related to public institutions in the disaster-affected area.

[0073] In the example shown in Figure 5, communication from communication terminal 50 and communication from automobile 60 are communication from communication terminals owned by victims of the disaster. Communication from meter 70 is communication for obtaining information on the damage situation in the disaster-stricken area. Meter 70 may be a water meter. Meter 70 may be an electricity meter. Meter 70 may be a gas meter. Communication for registering evacuees using My Number card 80 is communication related to public institutions in the disaster-stricken area.

[0074] 6, the control infrastructure 100 determines, for example, a radio base station for which power consumption reduction is to be prioritized among multiple radio base stations constituting the RAN. For example, the control infrastructure 100 prioritizes reducing the power consumption of a radio base station that is estimated to operate using battery power for a longer period after a disaster occurs. For example, the radio base station can operate for approximately 72 hours using battery power.

[0075] 6 , the control board 100 prioritizes reducing the power consumption of the wireless base station 42 that has the longest estimated operating time while operating using battery power after the occurrence of a disaster. The control board 100 prioritizes reducing the power consumption of the wireless base station 46 that has the second longest estimated operating time while operating using battery power after the occurrence of a disaster. The control board 100 prioritizes reducing the power consumption of the wireless base station 44 that has the third longest estimated operating time while operating using battery power after the occurrence of a disaster.

[0076] 7 and 8 , when a disaster occurs, the control platform 100 determines a recovery method for recovering the coverage area of ​​a radio base station whose wireless communication function is estimated to be malfunctioning, based on an overall situation estimation result obtained by estimating the overall situation of an affected area affected by the disaster using an overall situation estimation model and real-time information on the affected area. The control platform 100 may control the function of the RAN so that the coverage area is recovered using the determined recovery method.

[0077] 7 , the control board 100 recovers the coverage area of ​​a radio base station whose wireless communication function is estimated to be malfunctioning by, for example, expanding the coverage area of ​​a radio base station whose wireless communication function is estimated to be functioning normally. In this case, the radio base station whose wireless communication function is estimated to be functioning normally may be a radio base station adjacent to the radio base station whose wireless communication function is estimated to be malfunctioning.

[0078] 7 , the control board 100 recovers the coverage area of ​​the wireless base station 44 whose wireless communication function is estimated to be malfunctioning by expanding the coverage area 142 of the wireless base station 42 whose wireless communication function is estimated to be functioning normally. Furthermore, the control board 100 recovers the coverage area of ​​the wireless base station 44 by expanding the coverage area 146 of the wireless base station 46 whose wireless communication function is estimated to be functioning normally.

[0079] In the example shown in FIG. 8 , the control board 100 recovers the coverage area of ​​a wireless base station whose wireless communication function is estimated to be malfunctioning, for example, by using a mobile body equipped with a wireless base station. The mobile body may be any mobile body that can be equipped with a wireless base station. The mobile body may be, for example, a vehicle. The mobile body may be, for example, a ship. The mobile body may be, for example, an aircraft. The aircraft may be, for example, an unmanned aircraft. The aircraft may be, for example, a drone. The aircraft may be, for example, a HAPS (High Altitude Platform Station). The aircraft may also be a manned aircraft.

[0080] 8 , the control board 100 recovers the coverage area of ​​the wireless base station 44 whose wireless communication function is estimated to be malfunctioning, for example, by forming a coverage area 148 using the wireless base station 48 mounted on the mobile object 400. The control board 100 determines the location where the mobile object 400 will be placed based on a hazard map, for example, and instructs the mobile object 400 to be placed at the determined location.

[0081] 9 shows an example of the functional configuration of the control platform 100. The control platform 100 includes an acquisition unit 102, a training data storage unit 104, an execution unit 106, a model storage unit 110, and a selection unit 112. Note that it is not essential for the control platform 100 to have all of these components.

[0082] The acquisition unit 102 acquires various types of information. The acquisition unit 102 acquires various types of information by receiving the various types of information via the network 20, for example.

[0083] The acquisition unit 102 acquires various types of information from, for example, the data management device 200. The acquisition unit 102 acquires disaster log data from, for example, the data management device 200. The acquisition unit 102 acquires disaster area log data corresponding to the disaster log data from, for example, the data management device 200. The acquisition unit 102 acquires hazard map log data from, for example, the data management device 200, which records a hazard map covering the disaster area recorded in the disaster area log data. The acquisition unit 102 acquires evacuation shelter map log data from, for example, the data management device 200, which records an evacuation shelter map covering the disaster area. The acquisition unit 102 acquires disaster situation log data from, for example, the data management device 200, which records the disaster situation in the disaster area. The acquisition unit 102 acquires people flow situation log data from, for example, the data management device 200, which records the people flow situation in the disaster area. The acquisition unit 102 acquires, for example, from a telecommunications carrier, communication status log data that records the communication status in the disaster-stricken area after the disaster recorded in the disaster log data occurs.

[0084] The acquisition unit 102 stores communication situation learning data including, for example, the disaster log data, the disaster area log data, and the communication situation log data in the learning data storage unit 104. The acquisition unit 102 stores disaster situation learning data including, for example, the disaster log data, the disaster area log data, the hazard map log data, and the disaster situation log data in the learning data storage unit 104. The acquisition unit 102 stores people flow situation learning data including, for example, the disaster log data, the disaster area log data, the evacuation shelter map log data, and the people flow situation log data in the learning data storage unit 104. The acquisition unit 102 stores first comprehensive situation learning data including, for example, the disaster log data, the disaster area log data, the hazard map log data, the communication situation log data, and the disaster situation log data in the learning data storage unit 104. The acquisition unit 102 stores second comprehensive situation learning data, which includes, for example, the disaster log data, the disaster area log data, the evacuation shelter map log data, the communication status log data, and the people flow status log data, in the learning data storage unit 104. The acquisition unit 102 stores third comprehensive situation learning data, which includes, for example, the disaster log data, the disaster area log data, the hazard map log data, the evacuation shelter map log data, the communication status log data, the disaster status log data, and the people flow status log data, in the learning data storage unit 104.

[0085] The execution unit 106 executes various processes. The execution unit 106 includes, for example, a model generation function 107 and a RAN control function 109. The model generation function 107 and the RAN control function 109 may use the same computational resources.

[0086] The model generation function 107 generates various estimation models. For example, the model generation function 107 generates various estimation models by machine learning. The model generation function 107 may store the generated various estimation models in the model storage unit 110.

[0087] The model generation function 107 generates various estimation models, for example, by executing an application. The model generation function 107 generates various estimation models, for example, by executing an AI application. The model generation function 107 generates various estimation models, for example, by executing a non-RAN AI application.

[0088] The model generation function 107 determines the timing for generating various estimation models based on, for example, the demand of the RAN that is the control target of the RAN control function 109. The model generation function 107 generates various estimation models, for example, at a timing when the demand of the RAN is low.

[0089] The model generation function 107, for example, uses a plurality of communication situation learning data stored in the learning data storage unit 104 as training data to generate, through machine learning, a communication situation estimation model that estimates the communication situation in a disaster-affected area after the occurrence of a disaster, based on disaster identification data that identifies a disaster and disaster area identification data that identifies a disaster-affected area affected by the disaster. The model generation function 107, for example, uses a plurality of disaster situation learning data stored in the learning data storage unit 104 as training data to generate, through machine learning, a disaster situation estimation model that estimates the damage situation in a disaster-affected area, based on disaster identification data that identifies a disaster, disaster area identification data that identifies a disaster-affected area affected by the disaster, and hazard map data covering the disaster area. The model generation function 107, for example, uses a plurality of people flow situation learning data stored in the learning data storage unit 104 as training data to generate, through machine learning, a people flow situation estimation model that estimates the people flow situation in a disaster-affected area, based on disaster identification data that identifies a disaster, disaster area identification data that identifies a disaster-affected area affected by the disaster, and evacuation shelter map data covering the disaster area.

[0090] The model generation function 107 generates, by machine learning, a first comprehensive situation estimation model that estimates a first comprehensive situation of a disaster-stricken area, including the communication situation in the disaster-stricken area and the damage situation in the disaster-stricken area after the occurrence of a disaster, from, for example, disaster identification data that identifies a disaster, disaster area identification data that identifies a disaster-stricken area affected by the disaster, and hazard map data covering the disaster-stricken area, using, for example, a plurality of first comprehensive situation learning data stored in the learning data storage unit 104 as training data.The model generation function 107 generates, by machine learning, a second comprehensive situation estimation model that estimates a second comprehensive situation of a disaster-stricken area, including the communication situation in the disaster-stricken area and the human flow situation in the disaster-stricken area after the occurrence of a disaster, from, for example, disaster identification data that identifies a disaster, disaster area identification data that identifies a disaster-stricken area affected by the disaster, and evacuation shelter map data covering the disaster-stricken area, using, for example, a plurality of second comprehensive situation learning data stored in the learning data storage unit 104 as training data. The model generation function 107, for example, uses a plurality of third comprehensive situation learning data stored in the learning data storage unit 104 as training data and generates by machine learning a third comprehensive situation estimation model that estimates the third comprehensive situation of the disaster-stricken area, including the communication situation in the disaster-stricken area after the occurrence of the disaster, the damage situation in the disaster-stricken area, and the human flow situation in the disaster-stricken area, from disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster-stricken area affected by the disaster, hazard map data that covers the disaster-stricken area, and evacuation shelter map data that covers the disaster-stricken area.

[0091] The acquisition unit 102 acquires, for example, disaster identification data that identifies a disaster from the data management device 200. The acquisition unit 102 acquires, for example, disaster area identification data that identifies a disaster area affected by the disaster from the data management device 200. The acquisition unit 102 acquires, for example, hazard map data covering the disaster area from the data management device 200. The acquisition unit 102 acquires, for example, evacuation shelter map data covering the disaster area from the data management device 200.

[0092] The acquiring unit 102 acquires various estimation models from, for example, the model generating device 300. The acquiring unit 102 may store the acquired various estimation models in the model storage unit 110.

[0093] The acquisition unit 102 acquires a communication situation estimation model from, for example, the model generation device 300. The acquisition unit 102 acquires a damage situation estimation model from, for example, the model generation device 300. The acquisition unit 102 acquires a people flow situation estimation model from, for example, the model generation device 300. The acquisition unit 102 acquires a first overall situation estimation model from, for example, the model generation device 300. The acquisition unit 102 acquires a second overall situation estimation model from, for example, the model generation device 300. The acquisition unit 102 acquires a third overall situation estimation model from, for example, the model generation device 300.

[0094] The RAN control function 109 controls the functions of the RAN. For example, the RAN control function 109 controls the functions of the RAN by executing the RAN_AI. The RAN control function 109 may also control the functions of the RAN by executing any other processing.

[0095] The RAN control function 109 controls the functions of the RAN based on, for example, various pieces of information acquired by the acquisition unit 102. The RAN control function 109 controls the functions of the RAN using, for example, various estimation models stored in the model storage unit 110.

[0096] The RAN control function 109 controls the function of the RAN covering the disaster-stricken area based on a communication situation estimation result obtained by estimating the communication situation in the disaster-stricken area after the occurrence of a disaster, from disaster identification data that identifies the disaster and disaster area identification data that identifies the disaster-stricken area, using, for example, a communication situation estimation model stored in the model storage unit 110. The RAN control function 109 controls the function of the RAN covering the disaster-stricken area based on a disaster situation estimation result obtained by estimating the damage situation in the disaster-stricken area, from disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster-stricken area, and hazard map data that covers the disaster-stricken area, using, for example, a damage situation estimation model stored in the model storage unit 110. The RAN control function 109 controls the functions of the RAN covering the disaster-stricken area based on the results of estimating the human flow situation in the disaster-stricken area, which are obtained by, for example, using a human flow situation estimation model stored in the model storage unit 110, estimating the human flow situation in the disaster-stricken area from disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster-stricken area affected by the disaster, and evacuation shelter map data that covers the disaster-stricken area.

[0097] The RAN control function 109 controls the function of the RAN covering the disaster-stricken area based on a first comprehensive situation estimation result obtained by estimating a first comprehensive situation of the disaster-stricken area after the occurrence of a disaster from disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster-stricken area affected by the disaster, and hazard map data that covers the disaster-stricken area, using, for example, a first comprehensive situation estimation model stored in the model storage unit 110. The RAN control function 109 controls the function of the RAN covering the disaster-stricken area based on a second comprehensive situation estimation result obtained by estimating a second comprehensive situation of the disaster-stricken area after the occurrence of a disaster from disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster-stricken area affected by the disaster, and evacuation shelter map data that covers the disaster-stricken area, using, for example, a second comprehensive situation estimation model stored in the model storage unit 110. The RAN control function 109 controls the functions of the RAN covering the disaster-stricken area based on the third comprehensive situation estimation result, which estimates the third comprehensive situation of the disaster-stricken area after the occurrence of the disaster from, for example, disaster identification data that identifies the disaster, disaster area identification data that identifies the disaster area affected by the disaster, hazard map data that covers the disaster area, and evacuation shelter map data that covers the disaster area, using the third comprehensive situation estimation model stored in the model storage unit 110.

[0098] The first overall situation estimation result may include a communication situation estimation result and a disaster situation estimation result. The second overall situation estimation result may include a communication situation estimation result and a people flow situation estimation result. The third overall situation estimation result may include a communication situation estimation result, a disaster situation estimation result, and a people flow situation estimation result.

[0099] The RAN control function 109 controls the functions of the RAN so that, for example, based on the communication situation estimation result, the radio resources of the RAN covering the affected area are preferentially allocated to communications in an area within the affected area where the number of communication terminals per unit area is estimated to be greater.The RAN control function 109 controls the functions of the RAN so that, for example, based on the communication situation estimation result, the radio resources of the RAN are preferentially allocated to communications in a radio base station, among multiple radio base stations whose coverage area includes the affected area, where the number of communication terminals within the coverage area is estimated to be greater.

[0100] The RAN control function 109 controls the functions of the RAN, for example, based on at least one of a communication situation estimation result and a people flow situation estimation result, so that radio resources of the RAN are preferentially allocated to communications in an area in the disaster-stricken area where the number of people per unit area is estimated to be larger. The RAN control function 109 controls the functions of the RAN, for example, based on at least one of a communication situation estimation result and a people flow situation estimation result, so that radio resources of the RAN are preferentially allocated to communications in a radio base station in a coverage area where the number of people in the coverage area is estimated to be larger, among a plurality of radio base stations whose coverage area includes the disaster-stricken area.

[0101] The RAN control function 109 controls the functions of the RAN, for example, so that radio resources of the RAN are preferentially allocated to disaster-related communications. For example, the RAN control function 109 sets priorities for communications so that disaster-related communications have a higher priority than non-disaster-related communications. The RAN control function 109 controls the functions of the RAN so that high-priority communications are maintained and low-priority communications are temporarily stopped. This allows the RAN control function 109 to reduce the power consumption of radio base stations whose coverage areas include the disaster-stricken area.

[0102] The RAN control function 109 controls the functions of the RAN, for example, based on at least one of the communication situation estimation result and the disaster situation estimation result, so that power consumption of a radio base station that is operating using battery power after the occurrence of a disaster and that is estimated to have a longer operating time is preferentially reduced among multiple radio base stations constituting the RAN. The RAN control function 109 preferentially reduces the power consumption of a radio base station that is estimated to have a longer operating time, for example, by preferentially reducing the coverage area of ​​the radio base station that is estimated to have a longer operating time. The RAN control function 109 preferentially reduces the power consumption of a radio base station that is estimated to have a longer operating time, for example, by preferentially reducing traffic of the radio base station that is estimated to have a longer operating time.

[0103] The RAN control function 109 controls the functions of the RAN, for example, based on at least one of the communication situation estimation result and the damage situation estimation result, so that the coverage area of ​​one radio base station constituting the RAN, which is estimated to have a wireless communication function that functions normally, covers at least a part of the coverage area of ​​another radio base station constituting the RAN, which is estimated to have a wireless communication function that does not function normally.The RAN control function 109 controls the functions of the RAN, for example, based on at least one of the communication situation estimation result and the damage situation estimation result, so that the coverage areas of multiple radio base stations constituting the RAN, which are estimated to have a wireless communication function that functions normally, cover at least a part of the coverage area of ​​another radio base station constituting the RAN, which is estimated to have a wireless communication function that does not function normally.

[0104] The RAN control function 109 controls the function of the RAN, for example, based on at least one of the communication situation estimation result and the disaster situation estimation result, so that the coverage area of ​​one radio base station mounted on the mobile body and constituting the RAN covers at least a part of the coverage area of ​​another radio base station constituting the RAN whose wireless communication function is estimated not to function normally. For example, the RAN control function 109 determines a position at which the mobile body 400 will be placed based on a hazard map. The RAN control function 109 instructs the mobile body 400 to be placed at the determined position. For example, if the mobile body 400 is an unmanned mobile body, the RAN control function 109 instructs the mobile body 400. For example, if the mobile body 400 is a manned mobile body, the RAN control function 109 instructs the driver of the mobile body 400.

[0105] For example, when there are multiple other radio base stations, the RAN control function 109 controls the function of the RAN so that the coverage area of ​​the one radio base station mounted on the mobile object preferentially covers at least a part of the coverage area of ​​the other radio base station that is estimated to have had a longer elapsed time since the time the wireless communication function stopped functioning normally. For example, when there are multiple other radio base stations, the RAN control function 109 controls the function of the RAN so that the coverage area of ​​the one radio base station mounted on the mobile object preferentially covers at least a part of the coverage area of ​​the other radio base station that is estimated to have a larger number of people per unit area based on the people flow situation estimation result.

[0106] The acquisition unit 102 acquires, for example, real-time information. The acquisition unit 102 acquires, for example, SNS messages as the real-time information. The RAN control function 109 may control the functions of the RAN further based on the real-time information.

[0107] The selection unit 112 selects real-time information to be used for the functions of the RAN. For example, when the real-time information is an SNS message, the selection unit 112 selects, as the SNS message to be used for the functions of the RAN, an SNS message that has been transmitted via a radio base station that constitutes the RAN from among the multiple SNS messages acquired by the acquisition unit 102. The RAN control function 109 may control the functions of the RAN based on the SNS message selected by the selection unit 112.

[0108] 10 is an explanatory diagram illustrating an example of the processing flow of the system 10. Here, an example will be mainly described in which the control board 100 controls the RAN 500 using the third comprehensive situation estimation model when a disaster occurs within the coverage area of ​​the RAN 500. Note that a state in which the control board 100 has not generated the third comprehensive situation estimation model is defined as the starting state.

[0109] In step (sometimes abbreviated as S) 102, the acquisition unit 102 acquires various log data included in the third comprehensive situation learning data from the data management device 200. 120 acquires communication situation log data included in the third comprehensive situation learning data from the telecommunications carrier.

[0110] In S104, the model generation function 107 generates a third comprehensive situation estimation model using, as training data, a plurality of third comprehensive situation learning data including the various log data acquired by the acquisition unit 102 in S102. The model generation function 107 may store the generated third comprehensive situation estimation model in the model storage unit 110.

[0111] In S106, a disaster occurs within the coverage area of ​​the RAN 500. The acquisition unit 102 acquires a notification indicating that a disaster has occurred within the coverage area of ​​the RAN 500, for example, from a public institution or the like.

[0112] In S108, the acquisition unit 102 acquires real-time information on the disaster area within the coverage area of ​​the RAN 500. In S110, the RAN control function 109 controls the functions of the RAN 500 based on the real-time information on the disaster area acquired by the acquisition unit 102 in S108.

[0113] In S112, the acquisition unit 102 acquires input data for the third overall situation estimation model generated by the model generation function 107 in S104 from the data management device 200. In S114, the model generation function 107 estimates a third overall situation of the affected area from the input data acquired by the acquisition unit 102 in S112 using the third overall situation estimation model. In S116, the RAN control function 109 controls the functions of the RAN 500 based on the third overall situation estimation result that estimates the third overall situation of the affected area. Thereafter, the affected area is restored, and the processing of the system 10 ends.

[0114] 11 schematically illustrates an example of the hardware configuration of a computer 1200 functioning as the control board 100. A program installed on the computer 1200 can cause the computer 1200 to function as one or more "parts" of the apparatus according to the above-described embodiments, or can cause the computer 1200 to execute operations associated with the apparatus according to the above-described embodiments or one or more "parts," and / or can cause the computer 1200 to execute a process or steps of the process according to the above-described embodiments. Such a program can be executed by the CPU 1212 to cause the computer 1200 to perform specific operations associated with some or all of the blocks in the flowcharts and block diagrams described herein.

[0115] The computer 1200 according to this embodiment includes a CPU 1212, a RAM 1214, and a graphics controller 1216, which are interconnected by a host controller 1210. The computer 1200 also includes input / output units such as a communication interface 1222, a storage device 1224, a DVD drive 1226, and an IC card drive, which are connected to the host controller 1210 via an input / output controller 1220. The DVD drive 1226 may be a DVD-ROM drive, a DVD-RAM drive, or the like. The storage device 1224 may be a hard disk drive, a solid-state drive, or the like. The computer 1200 also includes legacy input / output units such as a ROM 1230 and a keyboard 1242, which are connected to the input / output controller 1220 via an input / output chip 1240.

[0116] The CPU 1212 operates according to programs stored in the ROM 1230 and the RAM 1214, thereby controlling each unit. The graphics controller 1216 acquires image data generated by the CPU 1212 into a frame buffer or the like provided in the RAM 1214 or into the graphics controller 1216 itself, and causes the image data to be displayed on the display device 1218.

[0117] The communication interface 1222 communicates with other electronic devices via a network. The storage device 1224 stores programs and data used by the CPU 1212 in the computer 1200. The DVD drive 1226 reads programs or data from a DVD-ROM 1227 or the like and provides them to the storage device 1224. The IC card drive reads programs and data from an IC card and / or writes programs and data to an IC card.

[0118] The ROM 1230 stores therein a boot program or the like that is executed by the computer 1200 upon activation, and / or programs that depend on the hardware of the computer 1200. The input / output chip 1240 may also connect various input / output units to the input / output controller 1220 via a USB port, a parallel port, a serial port, a keyboard port, a mouse port, etc.

[0119] The programs are provided by a computer-readable storage medium such as a DVD-ROM 1227 or an IC card. The programs are read from the computer-readable storage medium, installed in the storage device 1224, RAM 1214, or ROM 1230, which are also examples of computer-readable storage media, and executed by the CPU 1212. Information processing described in these programs is read by the computer 1200, and brings about cooperation between the programs and the various types of hardware resources described above. An apparatus or a method may be configured by implementing operations or processing of information in accordance with the use of the computer 1200.

[0120] For example, when communication is performed between computer 1200 and an external device, CPU 1212 may execute a communication program loaded into RAM 1214 and instruct communication interface 1222 to perform communication processing based on the processing described in the communication program. Under the control of CPU 1212, communication interface 1222 reads transmission data stored in a transmission buffer area provided in RAM 1214, storage device 1224, DVD-ROM 1227, or a recording medium such as an IC card, and transmits the read transmission data to a network, or writes received data received from the network to a reception buffer area or the like provided on the recording medium.

[0121] Furthermore, the CPU 1212 may cause all or a necessary portion of a file or database stored in an external recording medium such as the storage device 1224, the DVD drive 1226 (DVD-ROM 1227), an IC card, etc. to be read into the RAM 1214, and may perform various types of processing on the data on the RAM 1214. The CPU 1212 may then write back the processed data to the external recording medium.

[0122] Various types of information, such as various types of programs, data, tables, and databases, may be stored on the recording medium and may undergo information processing. The CPU 1212 may perform various types of processing on data read from the RAM 1214, including various types of operations, information processing, conditional judgment, conditional branching, unconditional branching, information search / replacement, etc., as described throughout this disclosure and specified by the instruction sequences of the programs, and write the results back to the RAM 1214. The CPU 1212 may also search for information in a file, database, etc. on the recording medium. For example, if multiple entries, each having an attribute value of a first attribute associated with an attribute value of a second attribute, are stored on the recording medium, the CPU 1212 may search for an entry whose attribute value of the first attribute matches a specified condition from among the multiple entries, read the attribute value of the second attribute stored in the entry, and thereby obtain the attribute value of the second attribute associated with the first attribute that satisfies a predetermined condition.

[0123] The above-described programs or software modules may be stored in a computer-readable storage medium on or near the computer 1200. A recording medium such as a hard disk or RAM provided in a server system connected to a dedicated communication network or the Internet can also be used as a computer-readable storage medium, thereby providing the programs to the computer 1200 via the network.

[0124] The blocks in the flowcharts and block diagrams in the present embodiments may represent stages of a process in which an operation is performed or "parts" of a device responsible for performing the operation. Particular stages and "parts" may be implemented by dedicated circuitry, programmable circuitry provided with computer-readable instructions stored on a computer-readable storage medium, and / or a processor provided with computer-readable instructions stored on a computer-readable storage medium. The dedicated circuitry may include digital and / or analog hardware circuits, and may include integrated circuits (ICs) and / or discrete circuits. The programmable circuitry may include reconfigurable hardware circuits, such as field programmable gate arrays (FPGAs) and programmable logic arrays (PLAs), including AND, OR, XOR, NAND, NOR, and other logical operations, flip-flops, registers, and memory elements.

[0125] A computer-readable storage medium may include any tangible device capable of storing instructions that are executed by an appropriate device, such that a computer-readable storage medium having instructions stored thereon comprises an article of manufacture, including instructions that can be executed to create means for performing the operations specified in the flowcharts or block diagrams. Examples of computer-readable storage media may include electronic, magnetic, optical, electromagnetic, and semiconductor storage media. More specific examples of computer-readable storage media may include floppy disks, diskettes, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), electrically erasable programmable read-only memories (EEPROMs), static random access memories (SRAMs), compact disc read-only memories (CD-ROMs), digital versatile discs (DVDs), Blu-ray discs, memory sticks, integrated circuit cards, and the like.

[0126] The computer readable instructions may include either assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state-setting data, or source or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk®, JAVA®, C++, etc., and conventional procedural programming languages ​​such as the “C” programming language or similar programming languages.

[0127] Computer-readable instructions may be provided locally or over a local area network (LAN), a wide area network (WAN) such as the Internet, to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, or programmable circuitry, such that the processor or programmable circuitry executes the computer-readable instructions to generate means for performing the operations specified in the flowcharts or block diagrams. Examples of processors include computer processors, processing units, microprocessors, digital signal processors, controllers, microcontrollers, etc.

[0128] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be apparent to those skilled in the art that various modifications and improvements can be made to the above embodiments. It is clear from the claims that such modifications and improvements can also be included within the technical scope of the present invention.

[0129] It should be noted that the execution order of each process, such as operations, procedures, steps, and stages, in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not specifically stated as "before," "prior to," etc., and that the processes can be performed in any order unless the output of a previous process is used in a subsequent process. Even if the operational flow in the claims, specifications, and drawings is described using "first," "next," etc. for convenience, this does not mean that the processes must be performed in this order.

[0130] 10 System, 20 Network, 40 Wireless base station, 42 Wireless base station, 44 Wireless base station, 46 Wireless base station, 48 Wireless base station, 50 Communication terminal, 60 Automobile, 70 Meter, 80 My Number card, 100 Control platform, 102 Acquisition unit, 104 Learning data storage unit, 106 Execution unit, 107 Model generation function, 109 RAN control function, 110 Model storage unit, 112 Selection unit, 140 Coverage area, 142 Coverage area, 144 Coverage area, 146 Coverage area, 148 Coverage area, 200 Data management device, 300 Model generation device, 400 Mobile object, 500 RAN, 1200 Computer, 1210 Host controller, 1212 CPU, 1214 RAM, 1216 Graphics controller, 1218 Display device, 1220 Input / output controller, 1222 communication interface, 1224 storage device, 1226 DVD drive, 1227 DVD-ROM, 1230 ROM, 1240 input / output chip, 1242 keyboard

Claims

1. A learning data storage unit that stores communication situation learning data including disaster log data that records past disasters, disaster area log data that records disaster areas affected by the disasters, and communication situation log data that records communication situations in the disaster areas after the disasters occur; a model generation function that generates, by machine learning, a communication situation estimation model for estimating a communication situation in a disaster area after a disaster occurs from disaster identification data for identifying a disaster and disaster area identification data for identifying a disaster area affected by the disaster, using the plurality of communication situation learning data stored in the learning data storage unit as teacher data; and an execution unit having a RAN (Radio Access Network) control function that controls the function of the RAN that covers the disaster area based on a communication situation estimation result obtained by estimating the communication situation in the disaster area after the disaster occurs from the disaster identification data for identifying a disaster and the disaster area identification data for identifying a disaster area affected by the disaster, using the communication situation estimation model.

2. The control base according to claim 1, wherein the RAN control function controls the function of the RAN so that radio resources of the RAN are preferentially allocated to communication related to the disaster.

3. The learning data storage unit further stores disaster situation learning data including disaster log data recording disasters that occurred in the past, disaster area log data recording disaster areas affected by the disasters, hazard map log data recording hazard maps covering the disaster areas, and disaster situation log data recording the disaster situation of the disaster areas. The model generation function uses the plurality of disaster situation learning data stored in the learning data storage unit as teacher data to further generate, by machine learning, a disaster situation estimation model for estimating the disaster situation of the disaster area from disaster identification data for identifying disasters, disaster area identification data for identifying the disaster areas affected by the disasters, and hazard map data covering the disaster areas. The RAN control function controls the function of the RAN covering the disaster area based on the disaster situation estimation result obtained by estimating the disaster situation of the disaster area from the disaster identification data, the disaster area identification data, and the hazard map data covering the disaster area using the disaster situation estimation model. The control base according to claim 1 or 2.

4. The RAN control function controls the function of the RAN such that, based on at least one of the communication situation estimation result and the disaster situation estimation result, the power consumption consumed by a radio base station estimated to have a longer operating time using battery power after the occurrence of the disaster among the plurality of radio base stations constituting the RAN is preferentially reduced. The control base according to claim 3.

5. The RAN control function controls the function of the RAN such that, based on at least one of the communication situation estimation result and the disaster situation estimation result, the coverage area of one radio base station constituting the RAN estimated to have a normally functioning wireless communication function covers at least a part of the coverage area of another radio base station constituting the RAN estimated to have an abnormally functioning wireless communication function. The control base according to claim 3.

6. The RAN control function controls the function of the RAN such that, based on at least one of the communication situation estimation result and the disaster situation estimation result, the coverage area of one radio base station mounted on the mobile body and constituting the RAN covers at least a part of the coverage area of another radio base station constituting the RAN for which it is estimated that the wireless communication function does not function properly. The control base according to claim 3.

7. The learning data storage unit further stores crowd flow situation learning data including disaster log data recording disasters that occurred in the past, disaster area log data recording disaster areas affected by the disasters, evacuation shelter map log data recording evacuation shelter maps covering the disaster areas, and crowd flow situation log data recording the crowd flow situation in the disaster areas. The model generation function further generates a crowd flow situation estimation model for estimating the crowd flow situation in the disaster area by machine learning from a plurality of the crowd flow situation learning data stored in the learning data storage unit as teacher data, disaster identification data for identifying disasters, disaster area identification data for identifying the disaster areas affected by the disasters, and evacuation shelter map data covering the disaster areas. The RAN control function controls the function of the RAN covering the disaster area further based on the crowd flow situation estimation result obtained by estimating the crowd flow situation in the disaster area from the disaster identification data, the disaster area identification data, and the evacuation shelter map data covering the disaster area using the crowd flow situation estimation model. The control base according to claim 1 or 2.

8. The RAN control function controls the function of the RAN such that, based on at least one of the communication situation estimation result and the crowd flow situation estimation result, the radio resources of the RAN are preferentially allocated to communication in an area where the number of people per unit area is estimated to be larger in the disaster area. The control base according to claim 7.

9. The control base further includes an acquisition unit that acquires real-time information indicating at least one of the communication situation, the disaster situation, and the crowd flow situation in the disaster area and having real-time property. The RAN control function controls the function of the RAN covering the disaster area further based on the real-time information. The control base according to claim 1 or 2.

10. The acquisition unit acquires an SNS (Social Networking Service) message as the real-time information, and the control chassis further includes a selection unit that selects an SNS message that passes through a radio base station constituting the RAN that covers the disaster area from among the plurality of SNS messages, and the RAN control function controls the function of the RAN that covers the disaster area based on the SNS message selected by the selection unit. The control chassis according to claim 9.

11. A learning data storage unit that stores comprehensive situation learning data including disaster log data that records disasters that occurred in the past, disaster area log data that records the disaster areas affected by the disasters, hazard map log data that records hazard maps that cover the disaster areas, communication situation log data that records the communication situation in the disaster areas after the disasters occurred, and disaster situation log data that records the disaster situations of the disaster areas; A model generation function that generates, by machine learning, a comprehensive situation estimation model that estimates the comprehensive situation of the disaster area, including the communication situation and the disaster situation in the disaster area after the disaster occurred, from the disaster identification data that identifies the disaster, the disaster area identification data that identifies the disaster area affected by the disaster, and the hazard map data that covers the disaster area, using the plurality of comprehensive situation learning data stored in the learning data storage unit as teacher data; and an execution unit having a RAN control function that controls the function of the RAN that covers the disaster area based on a comprehensive situation estimation result obtained by estimating the comprehensive situation of the disaster area after the disaster occurred from the disaster identification data that identifies the disaster, the disaster area identification data that identifies the disaster area affected by the disaster, and the hazard map data that covers the disaster area, using the comprehensive situation estimation model. A control chassis.

12. A program for causing a computer to function as the control chassis according to claim 1 or 11 when executed by the computer.

Images