Communication quality prediction device and information amount reduction device

Abstract

To provide a technique that can appropriately reduce physical space information acquired by an environmental sensor.SOLUTION: A communication quality prediction device 1 includes a physical space information acquisition unit 11 that acquires physical space information representing the surrounding environment of a wireless communication path from an environmental sensor, an information amount reduction unit 12 that reduces the amount of information of the physical space information so as to retain features that affect the communication quality of the wireless communication path, and a communication quality prediction unit 14 that calculates the future communication quality of the wireless communication path based on the physical space information after the information amount reduction.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a communication quality prediction device and an information amount reduction device. [Background technology]

[0002] Wireless communication is strongly influenced by the surrounding environment. In wireless communication using millimeter waves or terahertz waves, the communication quality drops sharply due to obstructions such as human bodies (Non-Patent Document 1). Such a sharp change in communication quality is a factor that significantly reduces the perceived communication quality.

[0003] Therefore, there is a technology in which current physical space information is acquired using an environmental sensor such as an RGB camera, and in a communication quality prediction device, machine learning is performed from the current physical space information to predict the future state of the wireless communication channel, and handover control and transmission power control of the wireless communication channel are performed based on the prediction results (Patent Document 1, Non-Patent Documents 2 and 3). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2023-26789 A [Non-patent literature]

[0005] [Non-Patent Document 1] Sylvain Collonge and others, “Influence of the Human Activity on Wide-Band Characteristics of the 60 GHz Indoor Radio Channel”, IEEE Transactions on Wireless Communications, Vol. 3, No. 6, November 2004, p. 2396-p. 2406 [Non-Patent Document 2] Shoki Ohta and 3 others, “Millimeter-wave Received Power Prediction Using Point Cloud Data”, Proceedings of IEEE VTC-Spring, Helsinki, Finland, June 2022 [Non-Patent Document 3] Takayuki Nishio and 7 others, “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 [Non-Patent Document 4] Shoki Ohta and 4 others, "Fundamental Study on Reducing Point Cloud Data Amount for Predicting Millimeter-Wave Communication Quality Using Point Cloud," Institute of Electronics, Information and Communication Engineers, IEICE Technical Report, SeMI Study Group Conference Proceedings, January 2022 [Non-Patent Document 5] Shunyao Wu and others “LiDAR-Aided Mobile Blockage Prediction in Real-World Millimeter Wave Systems”, Proc. IEEE WCNC, Austin, TX, USA, 2022, p.2631-p.2636 Summary of the Invention [Problem to be solved by the invention]

[0006] However, since the physical space information collected by the environmental sensor, which contains a large amount of information, was transmitted directly to the communication quality prediction device, there were problems in that it put a strain on the bandwidth of the communication path from the environmental sensor to the communication quality prediction device, it took a long time for the communication quality prediction device to predict the communication quality, and it had a negative effect on programs other than the communication quality prediction program.

[0007] To address this issue, it is conceivable to apply a technique that applies random downsampling to point cloud data (physical space information) (Non-Patent Document 4) or a technique that converts point cloud data into a heat map (Non-Patent Document 5). However, because part of the point cloud data is deleted regardless of the features in the physical space information, there is a possibility that important features that affect communication quality will be deleted.

[0008] The present disclosure has been made in consideration of the above circumstances, and aims to provide a technology capable of appropriately reducing physical space information acquired by an environmental sensor. [Means for solving the problem]

[0009] A communication quality prediction device of one embodiment of the present disclosure includes an acquisition unit that acquires physical spatial information representing the surrounding environment of a wireless communication path from an environmental sensor, a reduction unit that reduces the amount of information of the physical spatial information so as to retain features that affect the communication quality of the wireless communication path, and a prediction unit that calculates the future communication quality of the wireless communication path based on the physical spatial information after the amount of information has been reduced.

[0010] An information amount reduction device of one embodiment of the present disclosure includes a reduction unit that reduces the amount of information of physical spatial information representing the surrounding environment of a wireless communication path so as to retain features that affect the communication quality of the wireless communication path. Effect of the Invention

[0011] According to the present disclosure, a technique can be provided that can appropriately reduce physical space information acquired by an environmental sensor. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of a communication quality predicting device. [Diagram 2] FIG. 2 is a diagram illustrating an example of the configuration of a communication quality prediction system. [Diagram 3] FIG. 3 is a diagram illustrating an example of the configuration of an information amount reduction device. [Figure 4] FIG. 4 is a diagram showing an example of a laser ring. [Diagram 5] FIG. 5 shows the distance function for one laser ring. [Figure 6] FIG. 6 is a diagram showing the distance function before and after application of the synthesis function. [Figure 7] FIG. 7 shows the results of applying a fast Fourier transform to a distance function. [Figure 8] FIG. 8 is a diagram showing distance functions before and after application of a high-pass filter. [Figure 9] FIG. 9 is a diagram showing the distance function after the sampling theorem is applied. [Figure 10] FIG. 10 is a diagram showing a flow of the information amount reduction method. [Figure 11] FIG. 11 is a diagram showing measured and predicted values ​​of throughput for each reduction technique. [Figure 12] FIG. 12 is a diagram illustrating a hardware configuration of the communication quality predicting device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0014] [overview] The present disclosure discloses a technique for reducing the amount of physical space information used in predicting communication quality. By providing an information amount reduction unit in addition to a communication quality prediction device, the amount of physical space information is reduced without losing spatial features necessary for predicting communication quality.

[0015] Specifically, an information amount reduction unit that reduces the amount of physical space information collected by the environmental sensor is disposed in front of the communication quality prediction device. For example, in the communication quality prediction device, the information amount reduction unit is disposed in front of the communication quality prediction unit. For example, the communication quality prediction device is disposed between the environmental sensor and the communication quality prediction device.

[0016] In parallel with the communication quality prediction device performing communication quality prediction, the information amount reduction means reduces the amount of physical space information collected by the environmental sensor while maintaining information necessary for communication quality prediction, thereby reducing the amount of physical space information for the communication quality prediction device. At this time, an appropriate reduction rate is set based on the positional relationship between the installation environment of the environmental sensor and the communication station and the original information amount, thereby reducing the traffic involved in transmission and the data collection cost.

[0017] According to the present disclosure, even in situations where the amount of physical space information obtained from environmental sensors is large, it is possible to predict communication quality with minimal impact on other applications and systems, without putting a strain on the bandwidth when transmitting to a communication quality prediction device.

[0018] [Example of the configuration of a communication quality prediction device] 1 is a diagram showing an example of the configuration of a communication quality prediction device according to the present embodiment. The communication quality prediction device 1 includes a physical space information acquisition unit 11, an information amount reduction unit 12, a terminal position acquisition unit 13, a communication quality prediction unit 14, and a communication control unit 15.

[0019] The physical space information acquisition unit (acquisition unit) 11 has a function of acquiring, from the environmental sensor 2, physical space information that indicates the surrounding environment of the wireless communication path.

[0020] The information amount reduction unit (reduction unit) 12 has a function of reducing the amount of physical space information acquired by the physical space information acquisition unit 11 so as to retain features that affect the communication quality of the wireless communication path.

[0021] The terminal location acquisition unit 13 has a function of acquiring location information of the user terminal 3 .

[0022] The communication quality prediction unit (prediction unit) 14 has a function of calculating the future communication quality of the wireless communication channel by machine learning using the physical space information after the amount of information reduction and the position information of the user terminal 3, etc.

[0023] The communication control unit 15 has a function of controlling wireless communication of the communication station 4 based on the prediction result of the future communication quality of the wireless communication path.

[0024] [Example of communication quality prediction system configuration] 1 illustrates a case where the information amount reduction unit 12 is located inside the communication quality prediction device 1. In the following, a case where the information amount reduction unit 12 is assumed to be an information amount reduction device 12 and is located between the environmental sensor 2 and the communication quality prediction device 1 will be described.

[0025] 2 is a diagram showing an example of the configuration of a communication quality prediction system according to the present embodiment. The communication quality prediction system includes an environmental sensor 2, a plurality of user terminals 3, a communication station 4, a communication quality prediction device 1, and an information amount reduction device 12.

[0026] The environmental sensor 2 is a sensor that collects physical space information in a wireless communication area. For example, it is an RGB camera, a depth camera, or LiDAR (Laser Imaging Detection and Ranging). The physical space information is, for example, an RGB image taken by an RGB camera, a depth image taken by a depth camera, point cloud data of an object measured by LiDAR, and the position, speed, and acceleration of the object.

[0027] The user terminal 3 is a wireless terminal located in a wireless communication area. For example, it is a personal computer, a smartphone, a mobile phone, or a tablet terminal. The communication station 4 is a station that provides wireless communication to each of the multiple user terminals 3. For example, it is a base station or a wireless LAN router.

[0028] (Example of configuration of communication quality prediction device 1) The communication quality prediction device 1 includes a physical space information acquisition unit 11, a terminal position acquisition unit 13, a communication quality prediction unit 14, a communication control unit 15, and a communication information collection unit 16.

[0029] The communication information collecting unit 16 has a function of measuring the communication quality (for example, received power, throughput) of wireless communication currently being performed by the communication station 4 that is the prediction target, and collecting the measured communication quality information.

[0030] The physical space information acquisition unit 11 has a function of collecting physical space information after the amount of information has been reduced (for example, an RGB image, point cloud data) from the environmental sensor 2 via the information amount reduction device 12.

[0031] The terminal location acquisition unit 13 has a function of acquiring location information of the user terminal 3 .

[0032] The communication quality prediction unit 14 has a function of calculating the future communication quality of the wireless communication path between the communication station 4 and the user terminal 3 through machine learning, using the physical space information after the amount of information has been reduced, the position information of the user terminal 3, and communication quality information for the user terminal 3, the communication station 4, and the correct data acquired at the time of learning.

[0033] The communication control unit 15 has a function of causing the communication station 4 to execute handover control of the wireless communication path, transmission power control, etc., based on the communication quality information collected by the communication information collection unit 16 and the prediction result of the future communication quality predicted by the communication quality prediction unit 14.

[0034] (Example of the configuration of the information amount reduction device 12) The information amount reduction device 12 is disposed between the environmental sensor 2 and the communication quality prediction device 1. The information amount reduction device 12 is a device that acquires physical space information (e.g., RGB image, point cloud data) of a wireless communication area from the environmental sensor 2, and reduces the amount of information of the physical space information while maintaining information necessary for communication quality prediction.

[0035] 3 is a diagram showing an example of the configuration of the information amount reduction device 12. The information amount reduction device 12 includes a scalar function transformation unit 121, a frequency domain processing unit 122, and a reduction processing unit 123.

[0036] The scalar function conversion unit (reduction unit) 121 has a function of reducing the amount of information of the physical space information so as to leave the features that affect the communication quality of the wireless communication channel. Specifically, the amount of information of the physical space information is reduced by reducing the number of dimensions of the physical space information acquired from the environmental sensor 2 and converting the physical space information after the number of dimensions reduction into a scalar function of distance.

[0037] For example, for laser rings in the elevation angle direction based on point cloud data, which is physical space information, the number of dimensions of the physical space information is reduced by extracting laser rings corresponding to a predetermined altitude from among multiple laser rings.The physical space information is further reduced by converting the extracted point cloud data into a scalar function related to the distance from the environmental sensor 2.At this time, if there is a missing value in the scalar function, the distance information for the missing azimuth angle is complemented with a predetermined value.

[0038] The frequency domain processor 122 has a function of performing a frequency domain analysis on a scalar function.

[0039] The reduction processing unit (reduction unit) 123 also has a function of reducing the amount of information of the physical space information so as to leave the features that affect the communication quality of the wireless communication channel. Specifically, the amount of information of the physical space information is further reduced based on the frequency domain analysis performed on the scalar function. For example, the amount of information of the physical space information is reduced by applying a high-pass filter or sampling processing to the scalar function so as to leave the features based on the results of the frequency domain analysis.

[0040] [Specific functions of the information amount reduction device 12] When a rotating LiDAR is used as the environmental sensor 2, measurements are taken in all directions (360 degrees) horizontally from the LiDAR. In addition, the LiDAR emits 16 to 128 beams of light in the elevation direction and detects objects by observing the reflected light.

[0041] One laser in this elevation direction is called a laser ring. A laser ring contains information on reflection points for 360 degrees at a certain elevation angle. An example of a laser ring is shown in Figure 4. The upper diagram in Figure 4 shows 16 laser rings with elevation angles ranging from -15 degrees to 15 degrees in 2 degree increments. The elevation angle at which the laser rings are extracted can be determined according to the environment, such as the location where the LiDAR is installed and the size and movement range of objects that affect communication quality.

[0042] The scalar function conversion unit 121 and the frequency domain processing unit 122 are applied to each of these laser rings. In an environment where there are stronger restrictions on the amount of information or it is difficult to secure calculation resources, it is also possible to apply it to only one representative laser ring. For example, from among multiple laser rings in the elevation angle direction, one laser ring in which an object (peripheral object) that affects wireless communication is moving is selected. In this case, multiple laser rings may be selected.

[0043] This is because objects that affect communication quality, such as human bodies and vehicles, change little in the elevation angle direction, so using only one elevation angle has little effect on prediction accuracy. When using all laser rings, compatibility with the use of one laser ring is achieved by taking the average of each application result. The following example shows the case where only one laser ring is used, as shown in the lower diagram of Figure 4.

[0044] By using laser ring, it is possible to convert it into a scalar function. Furthermore, in the scalar function, it is possible to efficiently extract the parts that affect communication quality by using the property that the frequency of the parts corresponding to the human body and the vehicle is different from that of the background object such as a wall in the frequency domain.

[0045] Each point included in the laser ring has information about the distance from the LiDAR. By regarding this as a function of the horizontal angular axis direction, the laser ring can be converted into a scalar function. Therefore, the scalar function conversion unit 121 converts one laser ring into a distance function. Figure 5(a) shows an example of converting one laser ring into a simple distance function.

[0046] In addition, since the angles where the LiDAR could not acquire points are discontinuous, they are processed to be continuous by interpolation. In other words, the scalar function conversion unit 121 performs interpolation on the discontinuous parts of the distance function. Figure 5(b) shows an example in which linear interpolation is performed on the null points near -50 degrees and 70 degrees in Figure 5(a) and converted to a continuous function.

[0047] In the example of Fig. 5(b), a distance change due to human body shielding, that is, a state where a person is moving, appears as a wave in the vicinity from 60 degrees to 130 degrees. Since the dimensionality of the laser ring is reduced, it can be said that the object is easy to recognize.

[0048] On the other hand, if simply using the distance function, it is conceivable that extremely far points that have little impact on communication quality in an outdoor environment or the like are observed, resulting in outliers. In such a case, it is highly likely to affect the prediction. Therefore, the scalar function conversion unit 121 processes the distance function by applying a composite function so that large outliers do not occur. Examples of the composite function include, for example, the reciprocal function, the exponential function, and the logarithmic function.

[0049] In the reciprocal function, by returning 1 / d for the distance d, it corresponds to a large distance. However, when 1 < d, a constant 1 is returned. Fig. 6(a) shows examples before and after synthesizing the reciprocal function.

[0050] In the exponential function, by returning 10 1-d for the distance d, it corresponds to a large distance. However, when 1 < d, a constant 1 is returned. Fig. 6(b) shows examples before and after synthesizing the exponential function.

[0051] In the logarithmic function, by returning logd for the distance d, it corresponds to a large distance. However, when 1 < d, a constant 0 is returned.

[0052] So far, the scalar function conversion unit 121 has been described. As described above, the scalar function conversion unit 121 converts one laser ring into a scalar function by using the distance function or the composite function of the distance function.

[0053] Next, the frequency domain processing unit 122 will be described.

[0054] The frequency domain processing unit 122 performs frequency domain analysis on the scalar function using a Fourier transform. Large structures such as walls have a long fluctuation period in distance space. In other words, they are composed of low frequency components. On the other hand, objects that affect communication quality such as human bodies and vehicles are smaller than structures and are composed of high frequency components. Therefore, the reduction processing unit 123 at the subsequent stage further reduces the amount of information by deleting frequency components corresponding to large structures that do not fluctuate.

[0055] FIG. 7 shows an example of applying a fast Fourier transform to a distance function. The unit of frequency on the horizontal axis is "1 / degree." To remove the DC component, the average is calculated and then the fast Fourier transform is applied. From this result, it is possible to understand the frequency characteristics of the distance function. Based on these characteristics, the reduction processing unit 123 at the downstream stage determines the reduction rate and reduction method.

[0056] Next, the reduction processing unit 123 will be described.

[0057] The reduction processing unit 123 reduces the amount of information of the scalar function based on the frequency characteristics of the scalar function. The information amount reduction method consists of two steps: feature extraction processing using a high-pass filter and data amount deletion processing based on the sampling theorem. Each processing may be performed independently.

[0058] These two processes are applied to the scalar function calculated by the scalar function conversion unit 121 based on the frequency characteristic information obtained by the frequency domain processing unit 122, thereby reducing the amount of information. In addition, since the scalar function itself is a target for reducing the amount of information, it is also possible to use the scalar function as it is. Each of these will be described in detail below.

[0059] First, the feature extraction process using a high-pass filter will be described.

[0060] Objects such as human bodies and vehicles that affect communication quality are relatively small in both horizontal and vertical directions compared to surrounding walls and buildings. By utilizing this characteristic and leaving only high components in the frequency domain, it is possible to reduce the amount of information while leaving only factors that affect communication quality such as human bodies and vehicles.

[0061] The cutoff frequency of the high-pass filter is twice the frequency at which the amplitude reaches its maximum value. Figure 8 shows an example of applying a high-pass filter to the synthesis of the inverse function of the distance function. Based on Figure 7, the cutoff frequency was set to 1. The upper diagram in Figure 8 shows the result before the application of the high-pass filter (synthesis of the inverse function of the distance function), and the lower diagram in Figure 8 shows the result after the application of the high-pass filter.

[0062] As can be seen from the lower diagram of Fig. 8, only the 5rad part where the human body appears is left and the other parts are deleted, and this processing makes it possible to extract features. By deleting horizontal angles with values ​​less than 0.1 times the maximum value of the distance function after applying the high-pass filter, the amount of data in the point cloud data can be reduced.

[0063] Next, the data amount reduction process based on the sampling theorem will be described.

[0064] The sampling theorem states that "if sampling is performed at a frequency that is at least twice the maximum frequency of the original signal, the original waveform can be completely reconstructed." Based on this, by resampling the horizontal axis at a frequency that is about twice the maximum frequency for the frequency characteristics obtained by the frequency domain processing unit 122, it is possible to reduce the amount of information while retaining the features that affect the communication quality of human bodies, vehicles, etc.

[0065] The frequency for resampling is twice the maximum frequency that exceeds the average amplitude value. Figures 9(a) and (b) show an example of application to a distance function. As shown in Figure 7, the maximum frequency is set to 2 (1 / degree), so resampling is performed every 4 degrees. The area around 100 degrees represents a human body, and the characteristics of the human body remain even after resampling, indicating that it is possible to reduce the amount of information while retaining the characteristics of the human body.

[0066] [Example of operation of the information amount reduction device 12] FIG. 10 is a diagram showing a flow of the information amount reduction method.

[0067] First, the scalar function conversion unit 121 reduces the number of dimensions of point cloud data measured by a rotary LiDAR by using, for example, only one laser ring from among multiple laser rings, and converts the one laser ring into a scalar function (step S1). If the scalar function is discontinuous, it is made continuous by performing interpolation. A reciprocal function, exponential function, or logarithmic function may be applied to the scalar function.

[0068] Next, the frequency domain processing unit 122 performs a frequency domain analysis on the scalar function using a Fourier transform (step S2).

[0069] Finally, based on the analysis result of the frequency domain of the scalar function, the reduction processing unit 123 further reduces the amount of information by applying a high-pass filter and / or further by applying the sampling theorem to the scalar function in order to leave only features that affect communication quality (step S3).

[0070] [Communication quality prediction results] Point cloud data of an indoor environment was measured using LiDAR, and the amount of data of the point cloud data was reduced by an information amount reduction device 12, after which the communication quality was predicted by a communication quality prediction device 1. The communication quality prediction device 1 predicted the communication quality 500 ms ahead using a communication quality prediction model using a gradient boosting decision tree.

[0071] Table 1 shows the data reduction results for each reduction method for the experimental data used in this study.

[0072] [Table 1]

[0073] It can be seen that all of the reduction methods reduce the amount of data compared to the original point cloud data. It can also be seen that by applying each reduction method in turn, the amount of data is gradually reduced.

[0074] Figure 11 shows the measured and predicted throughput for each reduction method. Figure 11(a) shows the prediction using the distance function (RMSE, 0.293). Figure 11(b) shows the prediction using the distance function after downsampling based on the sampling theorem (RMSE, 0.296). Figure 11(c) shows the prediction using the distance function after applying a high-pass filter (RMSE, 0.308).

[0075] For all reduction methods, in places where communication quality significantly attenuates in the measured (actual) values, the predicted values ​​also significantly attenuate. From these experimental results, it can be said that all reduction methods were able to predict attenuation of communication quality due to human body obstruction. From the above, it was shown that by using this reduction method, the amount of physical space information can be reduced while maintaining the accuracy of communication quality prediction.

[0076] [effect] According to this embodiment, the amount of physical space information is reduced so as to leave characteristics that affect the communication quality of the wireless communication path, so that the amount of physical space information acquired by the environmental sensor can be appropriately reduced.

[0077] In other words, even in situations where the amount of physical space information obtained from the environmental sensors is large, it is possible to predict communication quality with little impact on other applications and systems, without putting a strain on the bandwidth when transmitting to the communication quality prediction device.

[0078] According to this embodiment, a laser ring corresponding to a specified altitude is extracted from multiple laser rings, and the extracted laser ring is converted into a scalar function related to the distance from the environmental sensor, thereby more appropriately reducing the physical space information acquired by the environmental sensor.

[0079] According to this embodiment, when there is a deficiency in the scalar function, the distance information for the missing azimuth angle is complemented with a predetermined value, so that the physical space information acquired by the environmental sensor can be reduced more appropriately.

[0080] According to this embodiment, a high-pass filter is applied to the scalar function, so that the physical space information acquired by the environmental sensor can be reduced more appropriately.

[0081] According to this embodiment, since the sampling process is applied to the scalar function, it is possible to more appropriately reduce the physical space information acquired by the environmental sensor.

[0082] [others] The present disclosure is not limited to the above-described embodiment, and various modifications are possible within the scope of the present disclosure.

[0083] The present disclosure is a technology for stabilizing communication quality in wireless LAN, particularly in devices equipped with millimeter wave communication functions, and is a particularly useful invention for applications requiring high-capacity, stable wireless networks, such as sensing data collection in factories and VR (Virtual Reality) / AR (Augmented Reality) applications.

[0084] The communication quality prediction device 1 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. 12.

[0085] The memory 902 and the storage 903 are storage devices. In the computer system, the CPU 901 executes a predetermined program loaded onto the memory 902, whereby each function of the communication quality prediction device 1 is realized.

[0086] The communication quality predicting device 1 may be implemented by one computer. The communication quality predicting device 1 may be implemented by multiple computers. The communication quality predicting device 1 may be a virtual machine implemented in a computer.

[0087] The program for the communication quality prediction device 1 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 communication quality prediction device 1 can also be distributed via a communication network.

[0088] The information amount reduction device 12 also has a hardware configuration similar to that of the communication quality prediction device 1 shown in FIG. [Explanation of symbols]

[0089] 1. Communication quality prediction device 2. Environmental Sensors 3. User terminal 4. Communications Bureau 11 Physical space information acquisition unit 12 Information reduction department, information reduction device 13 Terminal location acquisition unit 14 Communication Quality Prediction Department 15 Communication control section 16 Communications Information Collection Department 121 Scalar function conversion section 122 Frequency domain processing section 123 Reduction Processing Unit 901 CPU 902 Memory 903 Storage 904 Communication equipment 905 Input Device 906 Output Device

Claims

1. an acquisition unit that acquires physical space information representing a surrounding environment of a wireless communication path from an environmental sensor; a reduction unit that reduces an amount of the physical space information so as to leave characteristics that affect communication quality of the wireless communication path; a prediction unit that calculates a future communication quality of the wireless communication channel based on the physical space information after the amount of information reduction; A communication quality prediction device comprising:

2. The environmental sensor is a rotating sensor. The physical space information is point cloud data representing objects in the surrounding environment by points, The reduction section is The communication quality prediction device of claim 1, wherein a laser ring corresponding to a predetermined altitude is extracted from among multiple laser rings in the elevation angle direction from the point cloud data, and the extracted laser ring is converted into a scalar function related to the distance from the environmental sensor.

3. The reduction section is The communication quality prediction device according to claim 2 , wherein when the scalar function has a missing value, distance information for the missing azimuth angle is complemented with a predetermined value.

4. The reduction section is The communication quality prediction device according to claim 2 , further comprising a step of applying a high-pass filter to the scalar function.

5. The reduction section is The communication quality prediction device according to claim 2 , wherein a sampling process is applied to the scalar function.

6. a reduction unit that reduces an amount of physical space information representing a surrounding environment of a wireless communication path so as to retain features that affect communication quality of the wireless communication path; An information amount reduction device comprising:

Images