Wireless communication system

The wireless communication system employs machine learning to predict communication quality by processing terminal and environmental data, addressing the challenge of deteriorated communication due to mobile scenarios and environmental obstructions, achieving high accuracy and cost-effectiveness.

WO2024232005A1PCT designated stage expired Publication Date: 2024-11-14NT T INC +1

Patent Information

Application Number
PCT/JP2023/017400
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-09
Publication Date
2024-11-14

AI Technical Summary

Technical Problem

Current wireless communication systems face challenges in accurately predicting communication quality, especially in mobile scenarios where radio waves can be blocked by surrounding objects, leading to deteriorated communication quality, particularly at higher frequency bands.

Method used

A wireless communication system that utilizes machine learning to predict communication quality by processing terminal information such as position, orientation, and surrounding environment data, reducing the number of machine learning parameters to be optimized and sharing model coefficients across devices for efficient prediction model construction.

Benefits of technology

This approach enables high-accuracy prediction of communication quality, reducing costs associated with model learning and usage, and improving communication stability by considering both terminal and environmental factors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2023017400_14112024_PF_FP_ABST
    Figure JP2023017400_14112024_PF_FP_ABST
Patent Text Reader

Abstract

A wireless communication system 1 comprises at least two devices among a plurality of devices that include a base station device and one or more mobile wireless communication devices, wherein at least one device of said two devices comprises a first processing unit 11, 21A, 21B that processes device information including at least one among the location, the orientation, the speed, an operation, a control state, and a use state of the own device or the counterpart device, a second processing unit 12, 22A, 22B that processes surrounding environment information relating to the surrounding environment of the device, a third processing unit 13, 23A, 23B that processes communication information relating to the wireless communication of the device, a generation unit 14, 24A, 24B that performs machine learning on the relationship between the device information, the surrounding environment information, and the communication information, and generates a predictive model for predicting communication information relating to wireless communication on the basis of device information and surrounding environment information, and a prediction unit 15, 25A, 25B that uses the predictive model to predict communication information relating to wireless communication on the basis of inputted device information and surrounding environment information.
Need to check novelty before this filing date? Find Prior Art

Description

wireless communication system

[0001] The present disclosure relates to wireless communication systems.

[0002] Currently, wireless communication is used for various purposes, but there are cases where wireless communication cannot meet certain communication quality requirements depending on the service.

[0003] For example, when a wireless communication terminal moves, the direction of the antenna directivity and the radio wave propagation environment change, which results in a decrease in the communication quality of wireless communication. Therefore, Non-Patent Document 1 discloses a technology for predicting the communication quality of wireless communication using distance information between the terminal and a base station.

[0004] For example, in a high frequency band such as 28 GHz, radio waves have a high tendency to propagate in a straight line, and therefore, not only the movement of the wireless communication terminal but also the movement of surrounding objects can cause radio wave blocking, resulting in a decrease in communication quality of wireless communication. Therefore, Non-Patent Document 2 discloses a technology for determining whether radio waves are blocked by fixing a wireless communication terminal.

[0005] Christopher J. Lowrance, et al., “An active and incremental learning framework for the online prediction of link quality in robot networks”, Engineering Applications of Artificial Intelligence 77, 2019, p.197-p.211 Takayuki Nishio, et al., “Proactive Received Power Prediction Using Machine Learning and Depth Images for mmWave Networks”, IEEE Journal on Selected Areas in Communications, vol.37, no.11, November 2019, p.2413-p.2427

[0006] However, in a mobile wireless communication terminal that is in motion, it is difficult to accurately determine whether radio waves are being blocked by surrounding objects.

[0007] The present disclosure has been made in consideration of the above circumstances, and an object of the present disclosure is to provide a technology that can predict the communication quality of wireless communication with high accuracy and reduce the costs associated with learning and using a prediction model.

[0008] A wireless communication system according to one aspect of the present disclosure is a wireless communication system having at least two devices among a plurality of devices including a base station device and one or more mobile wireless communication devices, wherein at least one of the two devices comprises a first processing unit that processes device information including at least one of the position, orientation, speed, operation, control status, and usage status of the device or the other device, a second processing unit that processes surrounding environment information related to the device's surrounding environment, a third processing unit that processes communication information related to the device's wireless communication, a generation unit that performs machine learning on the relationship between the device information and the surrounding environment information and the communication information, and generates a prediction model that predicts communication information related to wireless communication based on the device information and the surrounding environment information, and a prediction unit that uses the prediction model to predict communication information related to wireless communication based on input device information and surrounding environment information.

[0009] According to the present disclosure, it is possible to provide a technology that can predict the communication quality of wireless communication with high accuracy and reduce the costs associated with learning and using a prediction model.

[0010] FIG. 1 is a diagram showing the configuration of a wireless communication system. FIG. 2 is a diagram showing a method for generating a prediction model. FIG. 3A is a diagram showing an example of surrounding environment information. FIG. 3B is a diagram showing an example of surrounding environment information. FIG. 4A is a diagram showing an example of generating a prediction model. FIG. 4B is a diagram showing an example of generating a prediction model. FIG. 5 is a diagram showing a method for predicting target information. FIG. 6 is a top view showing an indoor experimental environment and terminal movement. FIG. 7A is a diagram showing an example of throughput prediction. FIG. 7B is a diagram showing an example of throughput prediction. FIG. 7C is a diagram showing an example of throughput prediction. FIG. 7D is a diagram showing an example of throughput prediction. FIG. 8 is a diagram showing the R2 score of each throughput prediction result. FIG. 9 is a diagram showing the hardware configuration of a base station and a mobile wireless communication terminal.

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the description of the drawings, the same parts are designated by the same reference numerals and the description thereof will be omitted.

[0012] [Summary of the present disclosure] Depending on the state (e.g., position, posture, movement) of a wireless communication terminal or the state of static or dynamic objects existing around the wireless communication terminal, the radio wave propagation environment between the communication party and the terminal may change, degrading the communication quality of wireless communication and significantly affecting the services provided by wireless communication. In particular, the higher the frequency used, the more directional the radio waves tend to be, and the more susceptible the communication quality of wireless communication may be to be affected.

[0013] Therefore, in order for a wireless communication terminal to perform stable wireless communication and to manage the communication quality of wireless communication more accurately, measures are needed to deal with fluctuations in communication quality due to the state of the wireless communication terminal and the state of the wireless communication terminal's surrounding environment.

[0014] In this regard, modeling the relationship between communication quality and environmental information using machine learning is considered to be a useful method. If a prediction model is constructed using machine learning, it will be possible to output predicted values ​​for wireless communication quality and control information for the wireless communication system using input signals obtained from the state of the wireless communication terminal and environmental information.

[0015] Therefore, the present disclosure predicts the communication quality of wireless communication using a machine learning model by using physical space information related to various physical spaces, including terminal information and surrounding environment information, as input information. For example, to predict the communication quality of wireless communication, the present disclosure inputs terminal information including at least one of the position, orientation, speed, operation, control status, and usage status of the terminal performing wireless communication, and surrounding environment information related to the terminal's surrounding environment, into a prediction model. This enables the communication quality of wireless communication to be predicted with high accuracy.

[0016] In particular, by using terminal information including at least one of the terminal's position, orientation, speed, operation, control status, and usage status, the number of machine learning parameters to be optimized is reduced, thereby reducing the costs associated with learning and using the predictive model.

[0017] Furthermore, when generating a predictive model that outputs communication-related parameters, machine learning blocks constructed with the same coefficients and structures are shared, allowing predictive models to be constructed efficiently and reducing costs associated with the predictive models.

[0018] [Configuration of Wireless Communication System] FIG. 1 is a diagram showing the configuration of a wireless communication system according to this embodiment.

[0019] The wireless communication system 1 includes a base station 10 connected to a predetermined communication network, and a first mobile wireless communication terminal 20A and a second mobile wireless communication terminal 20B that perform wireless communication with the base station 10 and / or with each other. The number of mobile wireless communication terminals may be one or more.

[0020] [Configuration of Base Station] The base station 10 includes a first processing unit 11, a second processing unit 12, a third processing unit 13, a generation unit 14, a prediction unit 15, a first communication unit 16a, and a second communication unit 16b. The number of communication units may be one or more. The base station 10 may be a wireless LAN router or the like.

[0021] The first processing unit 11 has a function of processing (generating and acquiring) terminal information (device information) related to a terminal to be predicted. The terminal to be predicted is, for example, the first mobile wireless communication terminal 20A, which is a partner terminal in wireless communication.

[0022] The terminal information includes, for example, the position, orientation, speed, and operation (e.g., rotation of the terminal, change of antenna angle) of the first mobile wireless communication terminal 20A, which is the other terminal with which wireless communication is performed, the control status (e.g., presence or absence of control commands for circuits, control command information, total amount of execution commands for application programs), and the usage status (e.g., CPU usage rate, data communication volume). The first processing unit 11 processes the terminal information including at least one of these.

[0023] The second processing unit 12 has a function of processing (generating and acquiring) surrounding environment information relating to the surrounding environment of the base station 10 and the partner terminal.

[0024] The surrounding environment information is, for example, image information captured by a built-in camera of the base station 10 or a camera installed in the surrounding environment, sensor information sensed by a sensor installed in the surrounding environment, information extracted from the image information and sensor information, information about other terminals present in the vicinity of the base station 10 (for example, position information, orientation information, speed information, terminal information, and control information of the second mobile wireless communication terminal 20B that is not the other terminal with which wireless communication is performed), and information other than other terminals present in the vicinity of the base station 10 (for example, position information, orientation information, and speed information of passersby). In other words, the surrounding environment information is information about objects (for example, terminals, passersby, vehicles, and buildings) present in the vicinity of the base station 10 or the other terminal.

[0025] The surrounding environment information may be information generated based on the above information. For example, the second processing unit 12 may generate the surrounding environment information based on object distribution data obtained from sensor information (object distribution data including the base station 10, its manager, the manager's belongings, the first mobile wireless communication terminal 20A, its owner / occupant, and the owner's / occupant's belongings / occupied items, etc.).

[0026] The third processing unit 13 has a function of processing (acquiring, generating) communication information related to wireless communication between the base station 10 and the partner terminal. For example, the third processing unit 13 acquires communication information related to wireless communication between the base station 10 and the first mobile wireless communication terminal 20A.

[0027] The communication information related to wireless communication is information related to wireless communication signals and wireless communication data obtained from the first communication unit 16 a and the second communication unit 16 b, such as received signal power, signal-to-noise power ratio, signal-to-interference-plus-noise power ratio, RSSI (Received Signal Strength Indication), RSRQ (Received Signal Reference Quality), packet error rate, number of arriving bits, bit error rate, number of arriving bits per unit time, MCS (Modular Code Index), number of retransmissions, packet arrival delay time, error correction technology settings, user contract information of the mobile wireless communication terminal, differential information of these values, indices calculated from these values ​​using a predetermined formula, frequency conditions such as the frequency of the wireless communication system and the bandwidth of the resources used, and setting items of the wireless communication system that affect these indices.

[0028] The generation unit 14 has a function of inputting terminal information, surrounding environment information, and communication information related to wireless communication as learning data, and using an arbitrary machine learning algorithm to learn the relationship between the terminal information, surrounding environment information, and communication information related to wireless communication, and generating a prediction model that predicts communication information related to wireless communication as target information based on the terminal information and surrounding environment information.

[0029] The generation unit 14 can also perform learning by adding wireless communication system information. In this case, the generation unit 14 generates a prediction model that predicts communication information related to wireless communication as target information based on the terminal information, surrounding environment information, and wireless communication system information.

[0030] The prediction unit 15 has a function of predicting communication information related to wireless communication as target information based on the input terminal information and surrounding environment information using a prediction model. When wireless communication system information is available, the prediction unit 15 predicts communication information related to wireless communication as target information based on the input terminal information, surrounding environment information, and wireless communication system information.

[0031] The target information is, for example, communication information related to wireless communication and parameters for controlling wireless communication. The content of the target information can be selected arbitrarily. The parameters are, for example, information about the movement of the base station 10 or the remote terminal, control information in the OSI reference model from the physical layer to the application layer (e.g., mode, communication speed, communication destination, communication path, communication method), and control information about structures, metamaterials, and dielectrics that affect the radio wave propagation environment (e.g., control information about position, movement, and settings).

[0032] The first communication unit 16a has a function of performing wireless communication with the first mobile wireless communication terminal 20A.

[0033] The second communication unit 16b has a function of performing wireless communication with the second mobile wireless communication terminal 20B.

[0034] [Configuration of First Mobile Wireless Communication Terminal] The first mobile wireless communication terminal 20A is, for example, a mobile phone terminal or a smartphone terminal, and includes a first processing unit 21A, a second processing unit 22A, a third processing unit 23A, a generation unit 24A, a prediction unit 25A, and a communication unit 26A.

[0035] The first processing unit 21A has a function of processing (acquiring, generating) terminal information related to the terminal to be predicted, similar to the first processing unit 11 of the base station 10. The terminal to be predicted may be the base station 10, the second mobile wireless communication terminal 20B, or the first mobile wireless communication terminal 20A itself.

[0036] The second processing unit 22A has a function of processing (acquiring, generating) surrounding environment information related to the surrounding environment of the first mobile wireless communication terminal 20A, similar to the second processing unit 12 of the base station 10. The second processing unit 22A may generate the surrounding environment information based on object distribution data obtained from sensor information (object distribution data including the first mobile wireless communication terminal 20A, its owner / occupant, and the possessions / occupancies of the owner / occupant, etc.).

[0037] Similar to the third processing unit 13 of the base station 10, the third processing unit 23A has the function of processing (acquiring, generating) communication information relating to wireless communication between the first mobile wireless communication terminal 20A and the partner terminal.

[0038] Similar to the generation unit 14 of the base station 10, the generation unit 24A has a function of inputting terminal information, surrounding environment information, and communication information related to wireless communication as learning data, performing machine learning on the relationship between the terminal information, surrounding environment information, and communication information related to wireless communication using an arbitrary machine learning algorithm, and generating a prediction model that predicts communication information related to wireless communication as target information based on the terminal information and surrounding environment information. The generation unit 24A may perform machine learning by adding wireless communication system information, and generate a prediction model that includes the wireless communication system information in the input information.

[0039] The prediction unit 25A has a function of predicting communication information related to wireless communication as target information based on input terminal information and surrounding environment information using a prediction model, similar to the prediction unit 15 of the base station 10. When wireless communication system information is available, the prediction unit 15 predicts communication information related to wireless communication as target information based on the input terminal information, surrounding environment information, and wireless communication system information.

[0040] The communication unit 26A has a function of performing wireless communication with the first communication unit 16a of the base station 10 and performing wireless communication with the communication unit 26B of the second mobile wireless communication terminal 20B.

[0041] [Configuration of Second Mobile Wireless Communication Terminal] The second mobile wireless communication terminal 20B is, for example, a mobile phone terminal or a smartphone terminal. The second mobile wireless communication terminal 20B includes a first processing unit 21B, a second processing unit 22B, a third processing unit 23B, a generation unit 24B, a prediction unit 25B, and a communication unit 26B. The second mobile wireless communication terminal 20B has the same functions as the first mobile wireless communication terminal 20A. A description thereof will be omitted.

[0042] [Variant example of generation unit] The generation units 14, 24A, 24B that generate the prediction model may be provided in only one of the base station 10, the first mobile radio communication terminal 20A, and the second mobile radio communication terminal 20B, or may be provided in each terminal.

[0043] When only one terminal has the generation unit, it is preferable that the base station 10 that accommodates multiple mobile wireless communication terminals has the generation unit. This is because more wireless communication information can be collected and more training data can be used, which leads to improved accuracy of the prediction model. In this case, terminals other than the base station 10 use the prediction model generated by the base station 10.

[0044] When a generation unit is provided for each terminal, the prediction model may be optimized for each terminal using techniques such as transfer learning and fine tuning so that the prediction model is optimized for each terminal.

[0045] [Specific examples of wireless communication] Examples of wireless communication include wireless LAN (Local Area Network) defined by the standard IEEE802.11, Bluetooth (registered trademark), cellular communication using LTE or 5G, low power wide area (LPWA) communication for IoT, electronic toll collection system (ETC) used for vehicle communication, vehicle information and communication system (VICS), and ARIB-STD-T109.

[0046] [Method of Generating a Prediction Model] Fig. 2 is a diagram showing a method of generating a prediction model. A case where a first mobile radio communication terminal 20A generates a prediction model will be described. It is assumed that the first mobile radio communication terminal 20A is performing wireless communication with the base station 10.

[0047] Step S101: The first processing section 21A acquires the location information of the first mobile radio communication terminal 20A as terminal information.

[0048] Step S102: The second processing unit 22A acquires point cloud data relating to the positions of objects in the surrounding environment from a LiDAR (Laser Imaging Detection and Ranging) device installed in the surrounding environment of the first mobile wireless communication terminal 20A, converts the point cloud data into an image, and acquires the image as surrounding environment information. Examples of such images are shown in Figures 3A and 3B. The white areas represent the positions of the walls that form the experimental environment as viewed from above.

[0049] Step S103: The third processing section 23A acquires the throughput of the wireless communication carried out between the first mobile wireless communication terminal 20A and the base station 10 as communication information related to the wireless communication.

[0050] Step S104: As shown in FIG. 4A, the generation unit 24A inputs terminal information (location information), surrounding environment information (image), and communication information related to wireless communication (wireless communication throughput) as learning data, performs machine learning on the relationship between the terminal information and surrounding environment information and the communication information related to wireless communication, and generates a prediction model that predicts communication information related to wireless communication (wireless communication throughput) as target information based on the terminal information and surrounding environment information.

[0051] As shown in FIG. 4B, the generation unit 14 may perform learning by adding the wireless communication system information and generate a prediction model that also uses the wireless communication system information as input information.

[0052] [Method for predicting target information] Fig. 5 is a diagram showing a method for predicting target information. A case where target information is predicted by the first mobile radio communication terminal 20A will be described. It is assumed that the first mobile radio communication terminal 20A is performing radio communication with the base station 10.

[0053] Step S201: The first processing section 21A acquires the current location information of the first mobile radio communication terminal 20A as terminal information.

[0054] Step S202: The second processing unit 22A acquires point cloud data regarding the current positions of objects in the surrounding environment from a LiDAR installed in the surrounding environment of the first mobile wireless communication terminal 20A, converts the point cloud data into an image, and acquires the image as surrounding environment information.

[0055] Step S203: The prediction unit 15 uses the prediction model generated in step S104 to predict and output communication information related to wireless communication (throughput of wireless communication) as target information based on the terminal information (current location information) acquired in step S201 and the surrounding environment information (image related to point cloud data related to the current location) acquired in step S202.

[0056] [Specific Example of Prediction Model] A prediction model is formed based on input information so as to output communication information related to wireless communication or some parameter as target information.

[0057] The prediction model may be constructed so that the output value as target information is a highly accurate predicted value of some parameter, or it may output a strategy (improvement proposal) for changing the position of each communication unit to improve parameters related to wireless communication, or for the terminal components, surrounding environment, or from the application layer to the physical layer in the OSI reference model.

[0058] In this case, the prediction model may be constructed by reinforcement learning so as to maximize parameters related to wireless communication. For example, when predicting the received power at a future time, the received power at a future time is used as a parameter of the target information, and the prediction model is constructed so as to reduce the error between the predicted value of the received power that is output and the actual received power.

[0059] On the other hand, when outputting the X coordinate velocity, Y coordinate position velocity, and rotation command of the terminal so as to maximize the received power, a reward is set for maximizing the received power using reinforcement learning, and a prediction model is formed so as to output the X coordinate velocity, Y coordinate position velocity, and rotation command that will increase the reward.

[0060] Additionally, when generating a predictive model that outputs communication-related parameters, machine learning blocks constructed with the same coefficients and structure may be shared. For example, if a predictive model generated by one terminal and a predictive model generated by another terminal exist in the same environment, their structures will be identical or similar to each other. Therefore, the predictive model may be generated using the same machine learning blocks, or the other terminal's predictive model may be used as is, or may be generated by modifying part of the other terminal's predictive model. This allows predictive models to be constructed efficiently and reduces the costs associated with the predictive models.

[0061] [Experimental Results] Experiments were conducted to demonstrate the effects of this embodiment.

[0062] Figure 6 is a top view showing the indoor experimental environment and the movement of the terminal. Two autonomous mobile robots equipped with a phantom simulating a human body and a device for measuring position information, etc., were driven.

[0063] One of the robots was designated as the target robot 300, and it was made to move around in the experimental environment with a first mobile wireless communication terminal 20A capable of 5G communication carried on its back, and UDP communication was carried out between the first mobile wireless communication terminal 20A and the base station 10, which is a 5G base station. The throughput of the wireless communication between the first mobile wireless communication terminal 20A and the base station 10 was measured.

[0064] The other robot was designated as a blocking robot 500, which was made to travel within the experimental environment as a robot that interferes with wireless communication between the first mobile wireless communication terminal 20A and the base station 10. The blocking robot 500 interrupts the wireless communication path between the target robot 300 and the base station 10, thereby affecting the wireless communication therebetween.

[0065] The first mobile wireless communication terminal 20A obtained, as terminal information, the position, orientation, forward speed, and rotation speed of the target robot 300 from the target robot 300. The first mobile wireless communication terminal 20A also obtained, as surrounding environment information, the position, orientation, forward speed, and rotation speed of the blocking robot 500 from the blocking robot 500. The first mobile wireless communication terminal 20A then used the obtained terminal information and surrounding environment information as input information and predicted the wireless communication throughput as target information.

[0066] Here, the throughput after one second was predicted as the target information. A random forest was used as the prediction model algorithm, which was composed of 100 decision trees. The results of evaluation using four types of input information for the prediction are shown in Figures 7A to 7D. The horizontal axis is time, and the vertical axis is throughput. The solid lines in Figures 7A to 7D represent the correct throughput to be predicted, which was actually observed.

[0067] The dashed line in FIG. 7A represents the throughput prediction result when past throughput (two values ​​in total: the current throughput value and the throughput value five seconds ago) is used as input information.

[0068] The dashed line in FIG. 7B represents the throughput prediction result when the position information of the target robot 300 (two values ​​in total: X coordinate value and Y coordinate value on a two-dimensional plane) is used as input information.

[0069] The dashed line in Figure 7C shows the throughput prediction result when the position information, orientation information, and speed information of the target robot 300 (a total of six values: X coordinate value, Y coordinate value, sin(θ / 2) and cos(θ / 2) of the angle θ from the X coordinate axis, and the speed in the direction of travel and the rotational speed) are used as input information.

[0070] The dashed line in Fig. 7D shows the throughput prediction result when the position information, orientation information, and speed information of the target robot 300 and the position information, orientation information, and speed information of the blocking robot 500 (12 values ​​in total) are used as input information. The dashed line in Fig. 7D shows the prediction result of this embodiment.

[0071] 7C and 7D, it can be seen that the prediction result in Fig. 7D, which is made using both the terminal information and the surrounding environment information, is closer to the throughput indicated by the solid line than the prediction result in Fig. 7C, which is made using only the terminal information, in the time period indicated by symbol A. From this result, it can be said that by making a prediction using both the terminal information and the surrounding environment information, it is possible to predict the communication quality of wireless communication with high accuracy.

[0072] Furthermore, each throughput prediction result was evaluated using the evaluation function of the R2 score shown in equation (1).

[0073]

[0074] C[t] is the throughput shown by the solid line. ave [t] is the average throughput of the solid line, and C'[t] is the throughput prediction result of the dashed line.

[0075] The R2 score for each throughput prediction result is shown in Figure 8. The R2 score when past throughput was used was 0.61, and when prediction was made simply from the position information of the target robot (T robot) 300, it was 0.43, which is lower than the prediction when past throughput information was used. However, when orientation information and speed information of the target robot 300 were also used, it was 0.60, which means that prediction can be made with a high level of accuracy equivalent to that of predictions using past throughput information.

[0076] Furthermore, when the method of this embodiment further uses the position information, orientation information, and speed information of the blocking robot (robot B) 500 as surrounding environment information, the result is 0.68, which confirms that even higher prediction accuracy is obtained. From this result, it can be said that by making predictions using both terminal information and surrounding environment information, it is possible to predict the communication quality of wireless communication with high accuracy.

[0077] [Effect] According to this embodiment, communication information related to wireless communication is predicted using a prediction model that predicts communication information related to wireless communication based on terminal information of the terminal and surrounding environment information related to the terminal's surrounding environment, so that the communication quality of wireless communication can be predicted with high accuracy.

[0078] Furthermore, according to this embodiment, terminal information including at least one of the terminal's position, orientation, speed, operation, control status, and usage status is used, thereby reducing the number of machine learning parameters that need to be optimized and reducing the costs associated with learning and using the predictive model.

[0079] [Others] The present disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope of the gist of the present disclosure.

[0080] The base station 10 of the present embodiment described above can be realized, for example, by using a general-purpose computer system including a CPU 901, a memory 902, a storage 903, a communication device 904, an input device 905, and an output device 906, as shown in Fig. 9. The memory 902 and the storage 903 are storage devices. In this computer system, the CPU 901 executes a predetermined program loaded onto the memory 902, thereby realizing each function of the base station 10.

[0081] The base station 10 may be implemented by one computer. The base station 10 may be implemented by multiple computers. The base station 10 may be a virtual machine implemented on a computer. The base station 10 may also be realized using hardware such as a programmable logic device (PLD) or a field programmable gate array (FPGA).

[0082] The program for the base station 10 can be stored in a computer-readable recording medium such as a HDD, SSD, USB memory, CD, or DVD. The computer-readable recording medium is, for example, a non-transitory recording medium. The program for the base station 10 can also be distributed via a communication network such as the Internet or a telephone line.

[0083] The first mobile radio communication terminal 20A and the second mobile radio communication terminal 20B also have the same hardware configuration as the base station 10.

[0084] 1 Wireless communication system 10 Base station (base station device) 20A First mobile wireless communication terminal (mobile wireless communication device) 20B Second mobile wireless communication terminal (mobile wireless communication device) 11, 21A, 21B First processing unit 12, 22A, 22B Second processing unit 13, 23A, 23B Third processing unit 14, 24A, 24B Generation unit 15, 25A, 25B Prediction unit 16a First communication unit 16b Second communication unit 26A, 26B Communication unit 300 Target robot 500 Blocking robot 901 CPU 902 Memory 903 Storage 904 Communication device 905 Input device 906 Output device

Claims

1. A wireless communication system having at least two devices out of a plurality of devices including a base station device and one or more mobile wireless communication devices, wherein at least one of the two devices comprises: a first processing unit that processes device information including at least one of the position, orientation, speed, operation, control status, and usage status of the own device or a partner device; a second processing unit that processes surrounding environment information relating to the device's surrounding environment; a third processing unit that processes communication information related to the wireless communication of the device; a generation unit that performs machine learning on the relationship between the device information and the surrounding environment information and the communication information, and generates a prediction model that predicts communication information related to wireless communication based on the device information and the surrounding environment information; and a prediction unit that predicts communication information related to wireless communication based on input device information and surrounding environment information using the prediction model.

2. The wireless communication system of claim 1, wherein the surrounding environment information is any one of image information captured around the device, sensor information sensed around the device, information about other devices in the vicinity of the device, and information about devices other than other devices in the vicinity of the device.

3. The wireless communication system according to claim 1, wherein the second processing unit generates the surrounding environment information based on object distribution data including the device, the user of the device, and items possessed by the user obtained from sensor information.

Citation Information

Patent Citations

  • Communication system and base station

    WO2020217457A1

  • Communication terminal and communication quality prediction method

    WO2021064849A1

Cited By

  • Object detection system and object detection method

    US12669630B2

  • Object detection system and object detection method

    US20240036227A1