Communication Quality Prediction Model Updating Device, Communication Quality Prediction Model Updating Method, and Communication Quality Prediction Model Updating Program

Abstract

To provide a communication quality prediction model update device, a communication quality prediction model update method, and a communication quality prediction model update program that improve prediction accuracy of future communication quality in a radio communication area.SOLUTION: A communication quality prediction model update device 1 comprises: a physical space information acquisition unit 11 that acquires physical space information that is physical information within a radio communication area AR; a communication information acquisition unit 12 that acquires communication information of current radio communication in the radio communication area; a data storage unit 13 that stores physical space information and communication information; a communication quality prediction model storage unit 14 that stores a communication quality prediction model for predicting future communication quality related to radio communication within the radio communication area; an update rule determination unit 15 that determines the update frequency or update timing of the communication quality prediction mode; and a model updating unit 16 that updates a communication quality prediction model using physical space information and communication information on the basis of the update frequency or the update timing.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a communication quality prediction model update device, a communication quality prediction model update method, and a communication quality prediction model update program.

Background Art

[0002] With the increase in Internet, IoT (Internet of Things), and M2M (Machine to Machine) traffic, the use of higher frequencies has been under consideration. In next-generation mobile communications such as 5G and 6G, the realization of high-speed large-capacity communication using frequencies of 30 GHz or higher, called millimeter waves, is expected.

[0003] On the other hand, in wireless communication using high frequencies such as sub6GHz, millimeter waves, and terahertz waves, it is known that the communication quality deteriorates sharply due to the strong influence of the surrounding environment and the shielding of the wireless communication path by the human body or the like (Non-Patent Document 1).

[0004] Such a sharp change in communication quality is a factor that greatly degrades the perceived communication quality. Therefore, there is a need for a device that predicts the shielding of a wireless communication path in advance and performs communication control to reduce its influence.

[0005] For example, a device is known that acquires physical space information such as depth images and point cloud information of a wireless communication area using an RGB camera, a depth camera, and LiDAR (Light Detection and Ranging), learns the correspondence with the communication quality value of wireless communication by machine learning using the physical space information as input, predicts the future communication quality using a communication quality prediction model that maps from the physical space information to the communication quality value, and performs handover control and transmission power control based on the prediction result (Non-Patent Document 2).

Prior Art Documents

Non-Patent Documents

[0006]

Non-Patent Document 1

[0007] However, in Non-Patent Document 2, since it is assumed that the communication quality prediction model once learned by machine learning is repeatedly used, there is a problem that as time passes, differences gradually occur between the time of machine learning and the current state of the communication environment, and the prediction accuracy of the communication quality decreases.

[0008] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a technology capable of improving the prediction accuracy of future communication quality in a wireless communication area. [Means for Solving the Problems]

[0009] The communication quality prediction model updating device according to one aspect of the present invention includes a first acquisition unit that acquires physical space information, which is physical information within a wireless communication area, a second acquisition unit that acquires communication information currently being wirelessly communicated within the wireless communication area, a first storage unit that stores the physical space information and the communication information, a second storage unit that stores a communication quality prediction model for predicting future communication quality related to wireless communication within the wireless communication area, a determination unit that determines an update frequency or update timing of the communication quality prediction model, and an update unit that updates the communication quality prediction model using the physical space information and the communication information based on the update frequency or the update timing.

[0010] The communication quality prediction model updating method according to one aspect of the present invention is a communication quality prediction model updating method performed by a communication quality prediction model updating device, and includes steps of acquiring physical space information, which is physical information within a wireless communication area, acquiring communication information currently being wirelessly communicated within the wireless communication area, storing the physical space information and the communication information, determining an update frequency or update timing of a communication quality prediction model for predicting future communication quality related to wireless communication within the wireless communication area, and updating the communication quality prediction model using the physical space information and the communication information based on the update frequency or the update timing.

[0011] The communication quality prediction model updating program according to one aspect of the present invention causes a computer to function as the above communication quality prediction model updating device.

Advantages of the Invention

[0012] According to the present invention, it is possible to provide a technique capable of improving the prediction accuracy of future communication quality in a wireless communication area.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Embodiments for Carrying Out the Invention

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the description of the drawings, the same reference numerals are given to the same parts and the description thereof is omitted.

[0015] [Summary of the Invention] The present invention discloses a technique for updating a communication quality prediction model. Specifically, physical space information included in a wireless communication area and communication information currently in wireless communication are acquired, and the communication quality prediction model is updated based on these information.

[0016] At this time, the update frequency or update timing is set (updated) based on the positional relationship between the installation locations of cameras and LiDAR and the installation locations of wireless communication access points, and the calculation cost and data collection cost required for updating the communication quality prediction model are reduced.

[0017] In the present invention, since the communication quality prediction model is updated, it is possible to avoid a situation where a deviation occurs between the predicted value of communication quality by the communication quality prediction model and the current measured value over time, and it is possible to improve the prediction accuracy.

[0018] [Overall Configuration of the System] Figure 1 is a diagram showing the overall configuration of the system according to the present embodiment. The system according to the present embodiment includes a communication quality prediction model update device 1, a communication quality prediction device 2, and a communication control device 3.

[0019] The communication quality prediction device 2 is a device that predicts the future communication quality of each wireless communication path R formed between the access point 4 that performs wireless communication in the wireless communication area AR and each user terminal 5 by using a communication quality prediction model and physical space information obtained from the camera 6. Here, the camera 6 may be a sensor such as a LiDAR.

[0020] The communication control device 3 is a device that controls the communication of each wireless communication path R based on the prediction result of the future communication quality of each wireless communication path R predicted by the communication quality prediction device 2. For example, the communication control device 3 performs handover control and transmission power control on the wireless communication path R predicted to have a degraded communication quality due to shielding by a human body or the like.

[0021] The communication quality prediction model update device 1 is a device that updates the communication quality prediction model used by the communication quality prediction device 2. Specifically, the communication quality prediction model update device 1 acquires physical space information and current communication information during wireless communication included in the wireless communication area AR, appropriately updates the communication quality prediction model based on those pieces of information, and transmits the updated communication quality prediction model to the communication quality prediction device 2.

[0022] For example, as shown in FIG. 1, the communication quality prediction model update device 1 includes a physical space information acquisition unit 11, a communication information acquisition unit 12, a data storage unit 13, a communication quality prediction model storage unit 14, an update rule determination unit 15, and a model update unit 16.

[0023] The physical space information acquisition unit (first acquisition unit) 11 has a function of acquiring, as image data or point cloud data, physical space information regarding physical information such as the position and presence of an object in the wireless communication area AR from a camera 6 or a sensor installed near the wireless communication area AR.

[0024] For example, the physical space information acquisition unit 11 acquires, in time series, image data of physical space information captured by an RGB camera, a depth camera, etc., and point cloud data of physical space information detected by a LiDAR, etc. from the camera 6. The RGB camera, the depth camera, and the LiDAR are examples of collection means for collecting physical space information within the wireless communication area AR.

[0025] Physical space information refers to, for example, dynamic objects such as moving people, vehicles, communication devices, growing plants, weather, buildings under construction, rivers with fluctuating water levels, etc., and static objects such as luggage, benches, traffic lights, etc.

[0026] Note that since the point cloud data obtained by the LiDAR is the light reflection point information seen from the LiDAR, the physical space information acquisition unit 11 may convert the point cloud data of the physical space information acquired from the LiDAR into physical space information in a coordinate system recognizable by the communication quality prediction model update device 1.

[0027] For example, the physical space information acquisition unit 11 may perform conversion processing such as converting the obtained physical space information into a multi-dimensional coordinate system or voxels, and signal processing such as value normalization and extraction of difference information, and then output it to the data storage unit 13. A voxel is a data format for representing a three-dimensional object, and by assigning values (for example, 0 or 1, or a coefficient from 0 to 1) to cubes obtained by dividing a three-dimensional space into a grid, it represents whether there is an object in a small area of the cube and the characteristics of the existing object.

[0028] The communication information acquisition unit (second acquisition unit) 12 has a function of acquiring communication information currently being performed in the wireless communication between the access point 4 installed in the wireless communication area AR and each user terminal 5 from the access point 4. The access point 4 and the user terminal 5 are examples of communication devices (transceiver stations, transmitting stations, receiving stations).

[0029] Communication information includes, for example, received signal power, signal-to-noise power ratio (SNR), signal-to-interference-plus-noise power ratio (SINR), RSSI (Received Signal Strength Indication), RSRQ (Received Signal Reference Quality), packet error rate, number of bits received, number of bits received per unit time, MCS (Modulation and Coding Scheme index), number of retransmissions, delay time, error correction method, frequency of the communication system, frequency conditions such as the bandwidth of the resources used, traffic of the communication network, etc. It may also be differential information of these values, an index calculated by substituting these values into a predetermined calculation formula, a system setting item that affects the index, and the like.

[0030] The data storage unit (first storage unit) 13 has a function of storing the physical space information acquired by the physical space information acquisition unit 11 and the communication information during the current wireless communication acquired by the communication information acquisition unit 12. The data storage unit 13 may perform preprocessing such as normalization of absolute values on these pieces of information, or may select the data to be stored.

[0031] The communication quality prediction model storage unit (second storage unit) 14 has a function of storing a communication quality prediction model for predicting the future communication quality related to the wireless communication within the wireless communication area AR.

[0032] When there are a plurality of cameras 6, the communication quality prediction model may be prepared for each camera, or one may be prepared for a plurality of cameras. The case where there are a plurality of cameras 6 is, for example, the case where there are a plurality of cameras at different positions for one wireless communication area, or the case where there is one or more cameras for each of a plurality of wireless communication areas.

[0033] The update rule determination unit (determination unit) 15 has a function of determining an update rule for the communication quality prediction model. The update rule is, for example, an update period, an update frequency, an update timing, an update condition, and an update method. The update rule determination unit 15 may determine the same update rule for a plurality of or all communication quality prediction models, or may determine different update rules for each communication quality prediction model.

[0034] The model update unit (update unit) 16 has a function of updating the communication quality prediction model stored in the communication quality prediction model storage unit 14 by using the physical space information and communication information stored in the data storage unit 13 based on the update rule determined by the update rule determination unit 15.

[0035] For example, the model update unit 16 updates the communication quality prediction model currently in use in the communication quality prediction apparatus 2 by performing fine-tuning, transfer learning, etc. using more recent physical space information and communication information in terms of time. The updated communication quality prediction model is input to the communication quality prediction apparatus 2 and used for wireless communication, control of the user terminal 5, etc.

[0036] [Update method for communication quality prediction model] (First update method) FIG. 2 is a flowchart showing a first update method for the communication quality prediction model.

[0037] Step S101; First, the physical space information acquisition unit 11 acquires the physical space information in the wireless communication area AR from the camera 6 and sensors installed in the vicinity of the wireless communication area AR in time series and stores it in the data storage unit 13.

[0038] Also, the communication information acquisition unit 12 acquires the communication information during the current wireless communication between the access point 4 and each user terminal 5 from the access point 4 installed in the wireless communication area AR in time series and stores it in the data storage unit 13.

[0039] Step S102; Next, the model update unit 16 determines whether or not the update frequency or update timing defined in the update rule of the update rule determination unit 15 is satisfied. For example, when the current time exceeds the time obtained by adding the update period to the previous update time, the model update unit 16 determines that the update frequency is satisfied.

[0040] Step S103; If the update frequency or update timing defined in the update rule is satisfied, the model update unit 16 reads the latest physical space information and communication information (= Training data) from the data storage unit 13, and reads the verification data (= Validation data) for tuning the machine learning parameters from a predetermined database. Using the read information and data, machine learning such as fine-tuning and transfer learning is performed according to the update conditions and update methods predetermined in the update rule, thereby updating the current communication quality prediction model. Thereafter, the model update unit 16 returns to step S102.

[0041] On the other hand, if the update frequency or update timing defined in the update rule is not satisfied, the model update unit 16 repeats step S102 without updating the communication quality prediction model.

[0042] (Second update method) FIG. 3 is a flowchart showing a second update method of the communication quality prediction model.

[0043] Since the first update method remains at the update process of the communication quality prediction model, it is unclear whether the communication quality prediction model, update conditions, etc. are appropriate. The second update method improves this point.

[0044] Steps S201 to S202; Steps S101 to S102 are the same.

[0045] Step S203; When the update frequency or update timing defined in the update rule is satisfied, the model update unit 16 updates the current communication quality prediction model in the same manner as in step S103, and stores (maintains) the communication quality prediction model before the update.

[0046] Step S204; Next, the model update unit 16 evaluates the updated communication quality prediction model and the communication quality prediction model before the update using the test data. More specifically, the model update unit 16 calculates the improvement degree of the communication quality prediction by the updated communication quality prediction model with respect to the communication quality prediction model before the update using the test data.

[0047] The test data is the physical space information and communication information stored in the data storage unit 13 so far, and is the data used to test (evaluate) whether the communication quality prediction model is operating correctly.

[0048] For example, the model update unit 16 inputs the same physical space information and communication information into the updated communication quality prediction model and the communication quality prediction model before the update respectively, and calculates how much the accuracy rate of the prediction result by the updated communication quality prediction model has improved compared to the accuracy rate by the communication quality prediction model before the update.

[0049] The model update unit 16 may test (evaluate) that the communication quality prediction model has not changed significantly using test data from quite a long time ago, such as several months ago, several years ago, or several decades ago, or may test (evaluate) the performance for the current communication situation using test data from several weeks ago, several days ago, or several hours ago. The model update unit 16 may use both past test data and test data from a time relatively close to that past. The communication quality prediction model before the update and the past communication quality prediction models can be held in the communication quality prediction model storage unit 14.

[0050] Step S205; Next, based on the evaluation result of the updated communication quality prediction model, the model update unit 16 evaluates whether the updated communication quality prediction model has been improved from the previous communication quality prediction model, and notifies the update rule determination unit 15 of the evaluation result of the effect. More specifically, the model update unit 16 determines whether the improvement degree of the updated communication quality prediction model calculated in step S204 is equal to or greater than a predetermined improvement degree.

[0051] Step S206; If the updated communication quality prediction model has not been improved from the previous communication quality prediction model, the update rule determination unit 15 updates (changes) the update conditions, update frequency, update timing, etc. of the model update based on the effect evaluation of the model update input from the model update unit 16.

[0052] For example, when the improvement degree of the prediction result by the updated communication quality prediction model is less than the predetermined improvement degree, the update rule determination unit 15 updates the update rule so that the update process of the communication quality prediction model is not frequently performed, such as reducing the update frequency defined in the update rule or increasing the update timing. Then, it returns to step S202.

[0053] At this time, the model update unit 16 may discard (delete) the updated communication quality prediction model with no improvement effect following the update of the update rule by the update rule determination unit 15, and transmit the previous communication quality prediction model that was maintained to the communication quality prediction device 2.

[0054] On the other hand, if the updated communication quality prediction model has been improved from the previous communication quality prediction model, the update rule determination unit 15 does not update the update rule. At this time, if the improvement in prediction accuracy is significant, the update rule determination unit 15 may update the update rule so that the update process of the communication quality prediction model is frequently performed, such as increasing the update frequency or reducing the update timing. Then, it returns to step S202.

[0055] Regarding the update rules and timing of step S102 and step S202, the update rule determination unit 15 may be determined based on, for example, a preset time interval, time, the accuracy of the communication quality prediction model, the amount or quality or both of the training data for updating the communication quality prediction model. For example, the update rule determination unit 15 can set cases such as when the specified number of days interval, day of the week / date / time is reached, or when the sample for the training data can be obtained in an amount equal to or more than a preset amount when the reference accuracy of the communication quality prediction model is not met.

[0056] As described above, the first update method and the second update method have been explained.

[0057] The first update method and the second update method can be performed for each communication quality prediction model corresponding to each camera (sensor) installed at different positions in one or more wireless communication areas. As confirmed by the experiments described later, since the required update frequency or update timing of the communication quality prediction model varies depending on the position of the installed camera (sensor), by using an optimal update rule for each camera (sensor) at each position, the prediction results by the communication quality prediction model can be optimized, and the management of the communication quality prediction model can be appropriately and efficiently performed.

[0058] That is, in the communication quality prediction model update device 1, the update rule determination unit 15 determines the update frequency or update timing for each of a plurality of communication quality prediction models corresponding to a plurality of cameras (sensors) installed at different positions or combinations thereof, and the model update unit 16 updates each of the plurality of communication quality prediction models based on the frequency or timing determined for each communication quality prediction model.

[0059] [Method for Determining Update Frequency or Update Timing] Next, the method for determining the update frequency or update timing will be described.

[0060] The update frequency or update timing of the communication quality prediction model is determined based on the positional relationship between the positions of environmental sensors such as camera 6 and LiDAR, and the positions of access point 4 and user terminal 5.

[0061] Specifically, taking the straight line connecting the position of access point 4 and the position of user terminal 5 as the reference line (ax + by + c = 0; a, b, c are arbitrary initial values), draw a perpendicular line from the position (x1, y1) of the environmental sensor to the reference line, and calculate the length of the distance d between the intersection point and the position of the environmental sensor based on the "formula for the distance from a point to a line" in Equation (1).

[0062]

Equation

[0063] The larger this distance d is, the higher the update frequency is set and the smaller the update timing is.

[0064] For example, measure the period during which the communication quality prediction model was not updated and the decrease in the prediction accuracy of the communication quality, and based on the measurement results, consider an appropriate update frequency according to the distance d, and create a correspondence table between the distance d or the range of the distance d and the update frequency.

[0065] Regarding the method of grasping the decrease in prediction accuracy, prepare the above-mentioned test data at different timings, obtain the prediction results of the communication quality for each, and calculate the difference (error) with respect to the correct communication quality indicating the correct answer.

[0066] For example, acquire and store the test data every month, and determine the decrease in prediction accuracy from the change in the prediction error based on the prediction results of the communication quality using the test data for each month with the communication quality prediction model updated 5 months ago.

[0067] More specifically, if the prediction accuracy starts to decrease by a certain amount or more (e.g., 10%) within a distance d of 30 cm in one month, the update frequency for "d ≦ 30 cm" is set to "one month". If the distance d exceeds 30 cm and is within 1 m, and the prediction accuracy starts to decrease by a certain amount or more (e.g., 10%) in one week, the update frequency for "30 cm < d ≦ 1 m" is set to "one week".

[0068] [Method for Generating Dataset for Model Update] Next, a method for generating a dataset for model update will be described. The dataset for model update referred to here is a dataset based on physical space information, communication information (= Training data), and verification data (= Validation data).

[0069] At the update frequency or update timing determined by the above determination method, a dataset for model update is generated using the information in the data storage unit 13. First, N pieces of communication information are retrieved in descending order of the time stamp. Next, for each of the N pieces of retrieved communication information, the physical space information acquired at the time closest to the time τ seconds before the acquisition time of each communication information is retrieved from the data storage unit 13. Then, the N pieces of communication information are associated with the N pieces of physical space information using the N pieces of communication information as labels. Associating with the physical space information τ seconds before is to apply Temporal Difference Labeling (Non-Patent Document 2) for label assignment with a time shift in order to predict future communication quality.

[0070] Subsequently, predetermined preprocessing is performed on the retrieved communication information and physical space information. This preprocessing is, for example, normalization.

[0071] For communication information, multiply by weights or add offsets so that its distribution approximates an average value of 0 and a variance of 1. The weights and offsets may be set so that the values are distributed between 0 and 1. Emphasize fixing the weights and offset values, and make the value distribution approximately between -1 and 1, and cap values that deviate significantly with values such as 1 or -1.

[0072] For physical space information, perform processing to match the input of the currently used communication quality prediction model. For example, if it is image information, adjust the resolution to the same resolution as the input of the communication quality prediction model and normalize the values within the range of 0 to 1. If it is point cloud information, after removing outliers and downsampling, convert it into a 2D map or voxels.

[0073] The following shows the specific procedure for point cloud preprocessing. Point cloud preprocessing is performed by the physical space information acquisition unit 11, the data storage unit 13, or the model update unit 16.

[0074] First, remove outliers (noise points) that enter the point cloud data during measurement. Removing the noise points enables accurate prediction. Examples of noise point removal methods include calculating the distances from a certain point to a certain number of each neighboring point and removing points with distances greater than a threshold, and removing points with a small number of points contained within a sphere centered on the point.

[0075] Next, reduce (downsample) the number of points in the point cloud data. Reducing the number of points can reduce the input dimensionality to the communication quality prediction model and suppress the computational complexity of machine learning. Examples of downsampling methods include randomly thinning the indices and uniformly thinning at regular intervals.

[0076] Next, the three-dimensional space of the point cloud data is converted into an arbitrary two-dimensional plane. By converting the point cloud data into a two-dimensional plane, the capacity of the point cloud data can be further reduced. As a method for converting into a two-dimensional plane, for example, there is a method of determining one two-dimensional plane from among a plurality of two-dimensional planes forming the three-dimensional space of the point cloud data and mapping the point cloud data onto that two-dimensional plane.

[0077] Next, the point cloud data is converted into voxel data. By converting the point cloud data into voxel data, the capacity of the point cloud data can be further reduced. As a method for converting into voxel data, according to the number of points in the voxel obtained by dividing the three-dimensional space of the point cloud data into a grid, the value of the voxel space is set to 0 or 1. For example, for each voxel, if a certain number (at least one) of point clouds exist in the space corresponding to the voxel, the value of that voxel is set to 1, and if no certain number of points exist in that space, the value of that voxel is set to 0.

[0078] As another method for converting into voxel data, there is a method of setting the voxel value with the number of points in the space corresponding to the voxel as the density. For example, if the boundary coordinates of the voxel (two diagonal points out of the eight vertices of the voxel) are (x1, y1, z1) and (x2, y2, z2) respectively, if points satisfying “x1≦x≦x2”, “y1≦y≦y2”, and “z1≦z≦z2” exist, 1 or the number of those points is substituted as the value of the voxel, and if they do not exist, it is set to 0.

[0079] Finally, the preprocessed data for each of the communication information and physical space information serving as labels is used as a dataset for model update.

[0080] [Method for Updating Communication Quality Prediction Model] The model update unit 16 updates the communication quality prediction model using the model update dataset generated by the above method. Specifically, the model update unit 16 divides the model update dataset into Training data and Validation data, and repeats the process of updating the communication quality prediction model with the Training data and the process of verifying the performance of the communication quality prediction model with the Validation data until no improvement in the model performance (prediction result) is observed.

[0081] [Measurement results of prediction error] FIG. 4 is a diagram showing the prediction error based on the prediction result of the communication quality by the updated communication quality prediction model. For two cameras A and B installed at different positions, the prediction error after a certain period of time is shown. The distance d serving as a criterion for the update frequency or update timing is larger for camera A.

[0082] The used communication quality prediction model is a mathematical model in which processing units for linearly transforming the input are connected in a network. 3D Convolutional Neural Networks were used, which constructed a three-dimensional Neural Network that is the basis of deep learning and includes a Convolution layer for performing convolution.

[0083] The prediction error before model update is when the data amount on the horizontal axis is 0. From FIG. 4, it can be understood that the prediction error becomes smaller by performing model update. Also, it can be seen that camera A with a larger distance d has a larger prediction error compared to camera B. That is, it can be understood that the decrease in prediction accuracy due to the passage of time is large. Also, this tendency depends on the communication quality prediction model used. It was shown that by performing model update with a certain amount or more of data for this communication quality prediction model, the prediction error can be reduced to a certain level (set to 4 dB this time) or less, and the prediction accuracy can be improved.

[0084] [Industrial application] The present invention relates to a system used for predicting communication quality or control based on communication quality in a device equipped with a wireless LAN, particularly a millimeter-wave communication function. The invention appropriately updates a communication quality model, reduces the computational load, reduces the amount of data prepared in a database, and manages the communication quality model with stable performance over a long period. The invention is particularly useful in applications that require a highly reliable and stable wireless network, such as improving the reliability of wireless communication using high frequencies such as millimeter waves where communication quality prediction is important.

[0085] [Effects of the Embodiment] According to the present embodiment, since the communication quality prediction model is updated, a technique capable of improving the prediction accuracy of future communication quality in a wireless communication area can be provided.

[0086] Also, according to the present embodiment, the update frequency or update timing is determined for each communication quality prediction model, and a plurality of communication quality prediction models are updated based on the update frequency or update timing for each communication quality prediction model. Therefore, the prediction results by the communication quality prediction model can be optimized, and the management of the communication quality prediction model can be performed appropriately and efficiently.

[0087] Also, according to the present embodiment, the update frequency or update timing is determined according to the improvement degree of the communication quality prediction accuracy by the updated communication quality prediction model with respect to the communication quality prediction model before the update. Therefore, a technique capable of further improving the prediction accuracy of communication quality can be provided.

[0088] Also, according to the present embodiment, the update frequency or update timing is determined based on the positional relationship between the positions of cameras and sensors that collect physical space information and the positions of access points and user terminals that perform wireless communication within the wireless communication area. Therefore, a technique capable of further improving the prediction accuracy of communication quality can be provided.

[0089] [Others] The present invention is not limited to the above-described embodiments. The present invention can be variously modified within the scope of the gist of the present invention. The communication quality prediction model updating device 1 according to the present embodiment can also be realized by a computer and a program, and it is also possible to record the program on a recording medium or provide it through a network.

[0090] For example, as shown in FIG. 5, the communication quality prediction model updating device 1 according to the present embodiment can be realized 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. The memory 902 and the storage 903 are storage devices. In the computer system, each function of the communication quality prediction model updating device 1 is realized by the CPU 901 executing a predetermined program loaded on the memory 902.

[0091] The communication quality prediction model updating device 1 may be implemented by one computer. The communication quality prediction model updating device 1 may be implemented by a plurality of computers. The communication quality prediction model updating device 1 may be a virtual machine implemented on a computer.

[0092] The program for the communication quality prediction model updating device 1 can be stored in a computer-readable recording medium such as an HDD, an SSD, a USB memory, a CD, or a DVD. The program for the communication quality prediction model updating device 1 can also be distributed via a communication network.

Explanation of Reference Numerals

[0093] 1... Communication quality prediction model updating device 11... Physical space information acquisition unit 12... Communication information acquisition unit 13... Data storage unit 14... Communication quality prediction model storage unit 15... Update rule determination unit 16... Model update unit 2... Communication quality prediction device 3... Communication control device 4…Access Point 5…User Terminal 6…Camera 901…CPU 902…Memory 903…Storage 904…Communication Device 905…Input Device 906…Output Device

Claims

1. A first acquisition unit that acquires physical space information, which is physical information within a wireless communication area; A second acquisition unit that acquires communication information that is currently in wireless communication within the wireless communication area; A first storage unit that stores the physical space information and the communication information; A second storage unit that stores a communication quality prediction model for predicting future communication quality related to wireless communication within the wireless communication area; A determination unit that determines an update frequency or an update timing of the communication quality prediction model; An update unit that updates the communication quality prediction model using the physical space information and the communication information based on the update frequency or the update timing, wherein the determination unit determines the update frequency or the update timing such that, based on the positional relationship between the position of the collection means that collects the physical space information and the positions of a plurality of communication devices that perform wireless communication within the wireless communication area, the greater the length of the distance between the position of the collection means and the intersection point when a perpendicular line is dropped to the straight line connecting the positions of two of the communication devices from the position of the collection means, the higher the update frequency or the smaller the update timing. A communication quality prediction model update device.

2. The communication quality prediction model is a plurality of communication quality prediction models corresponding to a plurality of collection means installed at different positions that collect the physical space information, wherein the determination unit determines the update frequency or the update timing for each communication quality prediction model, and the update unit updates each of the plurality of communication quality prediction models based on the update frequency or the update timing for each communication quality prediction model. The communication quality prediction model update device according to claim 1.

3. The determination unit determines the update frequency or the update timing according to the degree of improvement in communication quality prediction accuracy by the updated communication quality prediction model with respect to the communication quality prediction model before the update. The communication quality prediction model update device according to claim 1 or 2.

4. The determination unit determines the update frequency or the update timing of the communication quality prediction model based on the amount of training data for training the communication quality prediction model. The communication quality prediction model update device according to any one of claims 1 to 3.

5. In a communication quality prediction model update method performed by a communication quality prediction model update device, a step of acquiring physical space information, which is physical information within a wireless communication area; a step of acquiring communication information that is currently in wireless communication within the wireless communication area; storing the physical space information and the communication information; determining an update frequency or an update timing of a communication quality prediction model for predicting future communication quality related to wireless communication within the wireless communication area; updating the communication quality prediction model using the physical space information and the communication information based on the update frequency or the update timing; and in the determining step, based on a positional relationship between a position of a collection unit that collects the physical space information and positions of a plurality of communication devices that perform wireless communication within the wireless communication area, the greater the length of a distance between the position of the collection unit and an intersection point when a perpendicular line is drawn to a straight line connecting positions of two of the communication devices from the position of the collection unit, the higher the update frequency or the smaller the update timing is determined so as to determine the update frequency or the update timing; A communication quality prediction model update method. Claims 6 A communication quality prediction model update program for causing a computer to function as the communication quality prediction model update device according to any one of Claims 1 to 4.

Images