Radio wave propagation estimation device and radio wave propagation estimation method

The radio wave propagation estimation device uses CNN and FNN to process two-dimensional map data, addressing errors and complexity in existing methods, achieving efficient and accurate propagation estimation.

JP7590371B2Active Publication Date: 2024-11-26NTT DOCOMO INC

Patent Information

Application Number
JP2022060306
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2021-06-25
Filing Date
2022-03-31
Publication Date
2024-11-26
Estimated Expiration
2042-03-31

AI Technical Summary

Technical Problem

Existing radio wave propagation estimation methods, such as ray tracing and neural networks, face challenges in accurately calculating power and reducing computational complexity due to errors in power calculation and increased processing demands, especially when considering multiple reflections and diffractions.

Method used

A radio wave propagation estimation device and method that utilizes a convolutional neural network (CNN) and fully connected neural network (FNN) to process two-dimensional map data, identifying structures visible from transmission and reception points, and estimating propagation characteristics based on these structures, reducing the need for complex three-dimensional calculations.

Benefits of technology

This approach significantly reduces computational load while enhancing accuracy in estimating radio wave propagation by simplifying calculations and improving estimation precision, even when radio waves scatter in various directions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007590371000002
    Figure 0007590371000002
  • Figure 0007590371000003
    Figure 0007590371000003
  • Figure 0007590371000004
    Figure 0007590371000004
Patent Text Reader

Abstract

To provide a radio wave propagation estimation device and a radio wave propagation estimation method capable of effectively reducing errors and an amount of calculation in power calculation of radio waves arriving at a receiving point.SOLUTION: A radio wave propagation estimation device uses identification information that can identify an object including a structure and map data including the structure, and based on the identification information, extracts the structure of the object visible from both of a direct or simulated radio wave transmitting point and a direct or simulated radio wave receiving point. The radio wave propagation estimation device uses information indicating features of the extracted structure of the object and estimates radio wave propagation characteristics.SELECTED DRAWING: Figure 16
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to a radio wave propagation estimation device and a radio wave propagation estimation method for estimating radio wave propagation characteristics in a wireless communication system. [Background technology]

[0002] Ray tracing methods based on the geometrical optics theory of diffraction are widely used to estimate the radio wave propagation characteristics in wireless communication systems such as mobile communication systems. The ray tracing method is a method of tracing the path of rays from the transmitting point to the receiving point, and the image method and ray launching method are known.

[0003] The image method is a method of searching for the shortest path from the transmitting point to the receiving point through structures that reflect and diffract rays, by extracting structures in advance. The image method has the advantage of being able to precisely find the ray from the transmitting point to the receiving point, but it has the problem that the amount of calculation increases exponentially when the number of surfaces (building walls, etc.) and edges of the structures to be considered, and the maximum number of reflections and diffractions, are increased.

[0004] Therefore, in order to reduce the amount of computation required for the image method, it has been proposed and demonstrated that rays are traced by searching for combinations of faces and edges for only structures within an ellipse whose foci are the transmitting and receiving points, or that buildings that are visible from the transmitting or receiving point are searched for and rays are traced (Patent Document 1).

[0005] In addition, methods using neural networks or deep learning have been developed to improve the accuracy of estimating radio wave propagation characteristics (Non-Patent Documents 1, 2). [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2019-122008 A [Non-patent literature]

[0007] [Non-Patent Document 1] Tetsuro Imai, Yukihiko Okumura, "A Study on Radio Wave Propagation Estimation Using Deep Learning", IEICE Technical Report, AP2017-165, pp81-86, January 2018 [Non-Patent Document 2] Tetsuro Imai, Takahiro Asai, "A Study on Estimation of Radio Wave Propagation Using Deep Learning with CNN -Relationship between Input Map Parameters and Estimation Accuracy-" IEICE Technical Report, vol.118, no.246, AP2018-89, pp.1-6, October 2018 Summary of the Invention [Problem to be solved by the invention]

[0008] However, the image method still has the following problems: For example, in tracing a single reflection ray, if the ray Tx-Ir-Rx is discarded, such as when the regular reflection point Ir does not belong to the building surface, an error occurs in the calculation of the power of the radio wave arriving at the receiving point Rx from the transmitting point Tx.

[0009] In addition, in tracing the reflected rays, the incident angle of the radio wave or the visible area of ​​the building surface may be reduced by the first Fresnelson. In this case, the radio wave is scattered in various directions, not just in the normal direction, causing an error in the calculation of the power of the radio wave arriving at the receiving point Rx.

[0010] On the other hand, in order to reduce such errors, it is possible to trace the rays of diffraction caused by multiple edges of the target building and collect the power arriving at the reception point Rx. However, this increases the amount of calculation required to calculate the power at the reception point Rx.

[0011] Therefore, the following disclosure has been made in consideration of the above circumstances, and aims to provide a radio wave propagation estimation device and a radio wave propagation estimation method that can effectively reduce errors and the amount of calculations in power calculation of radio waves arriving at a reception point. [Means for solving the problem]

[0012] One aspect of the radio wave propagation estimation device includes an extraction unit that extracts a structure of an object visible from both the reception point and the transmission point based on identification information from map data, the structure including a reception point of radio waves transmitted from a transmission point and including identification information for the object, and an estimation unit that estimates propagation characteristics of the radio waves from structures in the surrounding environment that may cause scattering of the radio waves based on the structure extracted by the extraction unit.

[0013] Moreover, one aspect of the radio wave propagation estimation device includes an extraction unit that extracts two-dimensional data from three-dimensional map data including reception points of radio waves transmitted from transmission points, and an estimation unit that estimates propagation characteristics of the radio waves based on the two-dimensional data extracted by the extraction unit, involving diffraction calculation of the radio waves in the surrounding environment.

[0014] One aspect of the radio wave propagation estimation device includes an extraction unit that applies a convolutional neural network to map data including a reception point of radio waves transmitted from a transmission point and including at least one of identification information and distance information for objects including structures, to extract a first parameter from the map data, the first parameter representing a structure that may cause scattering of the radio waves in a surrounding environment including the reception point, and an estimation unit that estimates propagation characteristics of the radio waves by applying a fully connected neural network to the first parameter extracted by the extraction unit and a second parameter representing a configuration of a wireless communication system that performs wireless communication between the transmission point and the reception point.

[0015] One aspect of the radio wave propagation estimation device includes an extraction unit that uses identification information capable of identifying an object including a structure and map data including the object, and extracts a structure of the object that is visible from both a direct or pseudo transmission point of the radio waves and a direct or pseudo reception point of the radio waves based on the identification information, and an estimation unit that estimates the propagation characteristics of the radio waves using information indicating the characteristics of the structure extracted by the extraction unit. [Brief description of the drawings]

[0016] [Figure 1] FIG. 1 is a functional block diagram showing an example of the configuration of a radio wave propagation estimating device according to the first embodiment. [Diagram 2] FIG. 2 is a diagram for explaining the problems with the conventional ray tracing method. [Diagram 3] FIG. 3 is a diagram for explaining the problems with the conventional ray tracing method. [Figure 4] FIG. 4 is a diagram for explaining the problems with the conventional ray tracing method. [Diagram 5] FIG. 5 is a diagram for explaining the problems with the conventional ray tracing method. [Figure 6] FIG. 6 is a diagram for explaining a problem with the conventional ray tracing method. [Figure 7] FIG. 7 is a diagram for explaining the problem with the conventional ray tracing method. [Figure 8] FIG. 8 is a diagram for explaining a problem with the conventional ray tracing method. [Figure 9] FIG. 9 is a diagram for explaining a problem with the conventional ray tracing method. [Figure 10] FIG. 10 is a diagram showing the identification numbers for buildings and the positional relationship between reception points and transmission points. [Figure 11] FIG. 11 is a diagram showing a two-dimensional image seen from a transmitting point Tx and a two-dimensional image seen from a receiving point Rx in FIG. [Figure 12] FIG. 12 is a diagram comparing calculations by a conventional 3D ray tracing method with the method of this embodiment. [Figure 13] FIG. 13 is a functional block diagram of the first embodiment. [Figure 14] FIG. 14 is a diagram showing an example of an "image seen from a transmitting point" and an "image seen from a receiving point." [Figure 15] FIG. 15 is a diagram illustrating an example of the configuration of a DNN applied to the city structure parameter extraction unit and the propagation calculation unit 102f in the first embodiment. [Figure 16] FIG. 16 is a diagram showing an example of an image viewed from a transmission point Tx and a reception point Rx according to the second embodiment. [Figure 17] FIG. 17 is a diagram showing an image of a wall surface commonly seen from a transmission point Tx and a reception point Rx according to the second embodiment. [Figure 18] FIG. 18 is a diagram showing an example of grouping of radio wave scattering objects. [Figure 19] FIG. 19 shows an example in which one color is assigned to an entire face, and an example in which a face is divided into a plurality of regions (meshes) and a different color is assigned to each mesh. [Figure 20] FIG. 20 is a diagram showing an example in which radio waves are reflected once between a transmission point Tx (antenna) and a reception point Rx (antenna), and an example in which radio waves are reflected multiple times between a transmission point Tx (antenna) and a reception point Rx (antenna). [Figure 21A] FIG. 21A is a diagram showing an example of an image of a receiving point, the ground, a structure, and the like as viewed from a transmitting point. [Figure 21B] FIG. 21B is a diagram showing an example of determining visibility using a conventional ray tracing method. [Figure 22] FIG. 22 is a diagram showing an example of a method for calculating a diffracted wave. [Figure 23A] FIG. 23A is a diagram illustrating an example of the configuration of a radio wave propagation estimating device according to the second embodiment. [Figure 23B] FIG. 23B is a diagram illustrating an example of the configuration of a radio wave propagation estimating device according to the second embodiment. [Figure 24] FIG. 24 is a diagram showing the operation flow of the preprocessing unit 231 of the calculation unit 230. [Diagram 25] FIG. 25 is a diagram showing an operation flow of the post-processing unit 232 when a past pre-processing result is used. [Figure 26A] FIG. 26A is a diagram showing the operation flow of the pre-processing unit 231 and the post-processing unit 232 in the case where calculation is performed directly from the beginning. [Figure 26B] FIG. 26B is a diagram showing the operation flow of the pre-processing unit 231 and the post-processing unit 232 in the case where calculation is performed directly from the beginning. [Figure 27] FIG. 27 is a diagram showing an example of calculation of received power and propagation loss by radio wave propagation estimating apparatus 200. In FIG. [Figure 28] FIG. 28 is a diagram showing an example of calculation of channel characteristics by the radio wave propagation estimating apparatus 200. In FIG. [Figure 29] FIG. 29 is a diagram showing an operational flow in which the radio wave propagation estimating apparatus 200 captures an object that is not present in the radio wave scattering contributing object DB. [Figure 30A] FIG. 30A is a diagram showing examples of objects that do not exist in the radio wave scattering contributing object DB, and examples of a car and human body model. [Figure 30B] FIG. 30B is a diagram showing examples of objects that do not exist in the radio wave scattering contributing object DB, and examples of a car and human body model. [Figure 30C] FIG. 30C is a diagram showing examples of objects that do not exist in the radio wave scattering contributing object DB, and examples of a car and human body model. [Figure 31A] FIG. 31A is a diagram showing an image of control of propagation characteristics at a receiving point by a RIS or the like. [Figure 31B] FIG. 31B is a diagram showing an image of control of propagation characteristics at a receiving point by a RIS or the like. [Diagram 32] FIG. 32 is a block diagram illustrating an example of a hardware configuration of a radio wave propagation estimating device according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Some embodiments of the present invention will be described in detail below with reference to the drawings. Note that the following description of the embodiments is not limited to the invention described in the claims.

[0018] In addition, the following multiple embodiments can be implemented in any combination.

[0019] [First embodiment] 1. Overview of the embodiment In the technical field of wireless communication such as mobile communication, an accurate estimation formula for propagation loss (sometimes called "path loss") is required for efficient cell design. Approaches for obtaining the estimation formula include, for example, an approach based on electromagnetic field theory and an approach based on measurement data (which may also be called "actual data" or "actual values"). Of these approaches, the latter approach, i.e., the approach based on measurement data, is often used in the technical field of wireless communication.

[0020] For example, within the range assumed in the wireless communication system, the frequency, base station antenna height, distance between transmitter and receiver, and / or propagation environment are changed and propagation loss is measured, and the relationship between these parameters and the data obtained by the measurement is analyzed to create an estimation equation for radio wave propagation.

[0021] When multiple regression analysis is used to analyze the relationship between parameters and measurement data, the "input parameters" and "function form" constituting the estimation formula are determined. The "input parameters" refer to information related to the propagation environment, for example, urban structure in an urban area. For convenience, the "information related to urban structure" may be referred to as urban structure parameters or urban structure related parameters. Examples of urban structure related parameters include building occupancy rate, road width, road angle, roadside building height, average building height, and building height near the base station.

[0022] Here, the determination of the parameters and function form of the estimation formula largely depends on the sense of the person who creates the estimation formula and the data used. In addition, the parameters and function form of the estimation formula must be determined for each different relationship between the surrounding building height and the antenna height of the base station and / or mobile station. For example, the parameters and function form used in the estimation formula may change for each cell deployment scenario, such as the relationship between macro cells and small cells. Therefore, it is not easy to create a unified estimation formula by any of the above-mentioned approaches.

[0023] Therefore, in one aspect of this embodiment, the use of a deep neural network (DNN), which is an example of a neural network (NN) for "deep learning," is considered for determining the parameters and function form used in the estimation equation for radio wave propagation.

[0024] "Deep learning" is a methodology of machine learning that uses a multilayered neural network. For example, by utilizing abundant computer resources and big data on the Internet in a multilayered neural network, it is possible to improve the processing performance of speech recognition, image recognition, natural language processing, and the like. One of the reasons that it is possible to improve processing performance is that deep learning has the ability to learn the very features used to solve problems from input data.

[0025] As mentioned above, DNN can learn the features required for problem solving from the input data, so for example, by using map data such as residential maps as input data for DNN, the city structure parameters used for radio wave propagation estimation can be learned by DNN. In addition, by using DNN to determine the "function form" that constitutes the estimation formula, it is possible to obtain an estimation formula with higher accuracy than existing estimation formulas.

[0026] For learning the DNN, data transmitted (or reported) from individual mobile stations (MS) such as mobile phones to a communication network can be used. One example of data reported from an MS to a communication network is information that explicitly or implicitly (or directly or indirectly) indicates the received power of the MS, the location of the MS, and / or the radio wave propagation environment of a base station (BS) to which the MS is connected (accessed). The MS is an example of user equipment (UE).

[0027] Data reported from a mobile station to a communication network may be collected, for example, by a base station constituting the communication network and / or a host station of the base station. A server computer (which may be called a "cloud computer") in a data center may correspond to the host station. For example, the data collected in the data center may be stored and managed in a storage device in the data center as so-called "big data."

[0028] Therefore, "data reported from MSs is used to train the DNN" may be read as "big data on radio wave propagation environments is used to train the DNN." Note that data that cannot be obtained from the network may be supplemented using, for example, simulation results by ray tracing.

[0029] Here, when map data such as residential maps are used as input data for DNN to learn urban structure parameters to be used in radio wave propagation estimation, many iterations are required to obtain the desired parameters, and there is an issue that convergence is difficult.As an example, when multiple map data viewed from radio wave transmission points, reception points, and the sky are used as input data, there is a problem that convergence is difficult until learning results are obtained.

[0030] In one aspect of this embodiment, the map data includes identification information (such as color and identification number) and / or distance information for objects including structures. More specifically, a two-dimensional image in which a color and an identification number are attached to the faces (and / or edges) of buildings, etc. is utilized, and the faces and edges of the buildings that are visible from the transmitting point and the receiving point are extracted and used for propagation estimation (calculation). Furthermore, a two-dimensional image (map) in which distance information to the faces and edges of the buildings that are visible from the transmitting point and the receiving point is calculated is also utilized, and the distances to the faces and edges that are visible are extracted and used for propagation estimation (calculation). As a result, even when multiple map data are used, identification information and distance information are given to the same structure, so that the iterations of learning are easily converged, and it is possible to obtain radio wave propagation estimation results with higher accuracy than before when compared with the same resources.

[0031] In addition, in one aspect of the present embodiment, not only when the above-mentioned neural network is used, but also when performing radio wave propagation estimation calculation, the map data includes identification information (such as color and identification number) for objects including structures. Then, from the map data, the structure (such as surface and edge) of the object visible from both the receiving point and the transmitting point is extracted based on the identification information. Then, based on the extracted structure (such as surface and edge), the structure that may cause radio wave scattering in the surrounding environment is estimated, and the propagation characteristics of the radio wave are estimated. As a result, the structure (such as surface and edge) of the object visible from both the receiving point and the transmitting point, that is, the structure that may cause radio wave scattering in the surrounding environment can be easily estimated based on the identification information, so that the amount of calculation can be reduced and the speed can be increased, and thus a radio wave propagation estimation result with higher accuracy than before can be obtained compared to the case of the same resource.

[0032] In addition, in one aspect of the present embodiment, not only when the above-mentioned neural network is used, but also when performing radio wave propagation estimation calculation, two-dimensional data is extracted from map data, and radio wave propagation characteristics are estimated with calculation of radio wave diffraction in the surrounding environment. Therefore, the amount of calculation can be significantly reduced compared to three-dimensional diffraction calculation. In addition, since the diffraction calculation can be simplified to perform calculations of reflection, transmission, etc., it is expected that the radio wave propagation estimation result as a whole can be obtained with higher accuracy than the conventional case, compared to the case of the same resources.

[0033] 2. Example of radio wave propagation estimation device configuration 1 is a functional block diagram showing an example of the configuration of a radio wave propagation estimation device 100 according to an embodiment. The term "estimation device" may be replaced with "simulator," "model," or "tool," for example.

[0034] 1, the radio wave propagation estimation apparatus 100 may, for example, include a control unit 102 that extracts parameters of urban structure and performs propagation estimation calculations, a storage unit 106 that includes various parameters and a database, and an output unit 108 that outputs calculation results of propagation characteristics, etc. As shown in the figure, the control unit 102 includes a setting unit 102a, a surface edge identification unit 102b, a transmitting point view search unit 102c, a receiving point view search unit 102d, a propagation calculation information extraction unit 102e, and a propagation calculation unit 102f.

[0035] The setting unit 102a is a setting means for setting various data and parameters for estimating and calculating propagation propagation characteristics. For example, the setting unit 102a may set identification information and distance information for objects including structures in map data including transmission points and reception points of radio waves. As an example, the setting unit 102a may set a calculation target area, a propagation calculation parameter setting (e.g., setting of a frequency band of radio waves to be estimated, etc.), and a setting rule for identification information (e.g., identification color or identification number) for faces and edges of buildings, etc. Note that the setting of identification information and distance information in map data, etc. may be performed manually, or may be automatically assigned from a color image or a 3D image read by a camera or a distance sensor, etc.

[0036] The surface edge identifying unit 102b is an identifying means for identifying the surfaces and / or edges of a building, etc. Here, Figures 2 to 9 are diagrams for explaining problems with the conventional ray tracing method.

[0037] Various methods have been developed to reduce the amount of calculations. For example, as shown in Figures 2 and 3, in order to reduce the amount of calculations in the image method, methods have already been proposed and demonstrated, such as tracing rays by finding combinations of faces and edges only for structures within an ellipse with the transmitting point and receiving point as foci, or searching for buildings that are visible from the transmitting point or receiving point and tracing rays (ACE tool).

[0038] In addition, the conventional method has problems such as errors in estimation accuracy. For example, in tracing the reflection ray of the ray tracing method, for example, in the case of a single reflection, an image point Tx' of the transmission point Tx is created for the face of a building, and then a line is drawn between the image point Tx' and the reception point Rx to derive the regular reflection point Ir of that face. If the regular reflection point Ir does not belong to the face, or if the regular reflection point Ir belongs to the face but the ray Tx-Ir-Rx is blocked by the face of another building, the ray Tx-Ir-Rx is discarded.

[0039] However, in the real world, the waves arriving at Rx may be scattered in various directions, not just the normal direction, due to the surfaces that are visible to both Tx and Rx. Therefore, discarding the ray Tx-Ir-Rx would cause an error in the calculation of the power of the waves arriving at Rx.

[0040] In addition, in tracing the reflection rays of the ray tracing method, as shown in Figure 6, the part of the building surface visible from both the transmitting point Tx and the receiving point Rx is considered to be larger than the first Fresnelson, and the reflection level in the regular reflection direction is calculated.

[0041]

number

[0042] However, in the real world, as shown in Figure 7, the incidence angle and the visible area of ​​the building surface may be reduced by the first Fresnelson. In such a case, the light is scattered in various directions, not just in the normal direction, and in the normal reflection direction, the scattering level is smaller than the reflection level obtained by the ray tracing method, resulting in an error.

[0043] Furthermore, conventionally, as shown in Figures 8 and 9, if the regular reflection point Ir does not lie within the plane, or if the regular reflection point Ir is within the plane but the ray Tx-Ir-Rx is blocked by the surface of another building, the ray Tx-Ir-Rx is discarded. However, in reality, radio waves are scattered by building surfaces, which results in calculation errors.

[0044] Therefore, in the past, in order to reduce calculation errors, calculations were performed to trace the rays of diffraction caused by multiple edges of the target building and gather the power arriving at the receiving point, but this had the disadvantage of increasing the amount of calculations.In addition, it was time-consuming to create image points and reflection points, determine the conditions of the reflection points, and trace the rays, and similarly, it was time-consuming to find diffraction points, trace the rays, and perform calculations using complex formulas for diffraction.

[0045] Therefore, in this embodiment, the following estimation method is proposed, which can significantly reduce the amount of calculation (time) compared to the conventional ray tracing method while taking into consideration the calculation accuracy.

[0046] That is, the surface edge identification unit 102b performs linking (association) using identification information (e.g., color or number) so that each surface and edge of each building can be identified in the two-dimensional image. Here, FIG. 10 is a diagram showing the identification numbers for buildings and the positional relationship between the reception point and the transmission point. As shown in FIG. 10, after the surfaces (and edges) of the building, etc. are assigned with their respective identification colors (or identification numbers), a two-dimensional image of the surfaces and edges of the building, etc. as seen (projected) from the transmission point Tx and the reception point Rx is created in a later process, and the color (or number) of each element of the image from the transmission point Tx and the image from the reception point Rx is examined, so that the surfaces (and edges) of the building as seen from the transmission point Tx or the reception point Rx can be identified.

[0047] Here, the identification color may be, for example, a hexadecimal number of RGB. Note that the identification color is not limited to being assigned by a human being. For example, the color of the building may be used from stereoscopic photograph data of the building stored in the topography and structure database of the storage unit 106, or a computer such as an artificial intelligence may assign colors and numbers to the stereoscopic structure data of the building. The face edge identification unit 102b stores in the storage unit 106 a database of the faces and edges of the target structure to which identification information has been assigned.

[0048] The transmission point view search unit 102c is a search means for searching for surfaces and edges of buildings and the like that are visible from the transmission point. Here, as shown in the upper diagram of FIG. 11, after drawing a two-dimensional image seen from the transmission point Tx in FIG. 10, the image is stored in the storage unit 106. Also, as shown in the lower left diagram of FIG. 14, a diagram (map) of the distance to the building seen from the transmission point Tx in FIG. 10 is also created, and the image (map) is stored in the storage unit 106. Note that these two-dimensional images may be automatically acquired by a camera or a distance sensor, and the image may be stored in the storage unit 106. Furthermore, the color (or number) of each element of these images is examined, and stored in a "database of surfaces and edges visible from the transmission point."

[0049] The reception point view search unit 102d is a search means for searching for surfaces and edges of buildings and the like that are visible from the reception point. Here, as shown in the lower diagram of FIG. 11, after drawing a two-dimensional image seen from the reception point Rx in FIG. 10, the image is stored in the storage unit 106. Also, as shown in the lower right diagram of FIG. 14, a diagram (map) of the distance to the building seen from the reception point Rx in FIG. 10 is also created, and the image (map) is stored in the storage unit 106. Note that a two-dimensional image may be automatically acquired by a camera or a distance sensor, and the image may be stored in the storage unit 106. Furthermore, the color (or number) of each element of these images is examined, and stored in a "database of surfaces and edges visible from the reception point."

[0050] Here, the propagation calculation information extraction unit 102e extracts and processes combinations of visible surfaces and edges based on information on the surfaces (and edges) of buildings visible from the transmitting point Tx or the receiving point Rx, in accordance with the conditions and information required for propagation calculation. Therefore, it is possible to specify the area of ​​the visible surface portion and the visible edges required for the calculation of scattered (reflected, diffracted) waves by the propagation calculation unit 102f described later. Note that the area of ​​the visible portion can be calculated using the total number of image elements (number of pixels) corresponding to that portion and the positional relationship to the radio wave irradiation origin or the observation point.

[0051] In addition, for the visible edge, the radio wave irradiation origin and the observation point are projected onto a plane perpendicular to the edge, and the irradiation angle and diffraction angle are calculated. In this way, in the diffraction calculation of the edge with the propagation calculation by the propagation calculation unit 102f described later, the obtained irradiation angle and diffraction angle can be substituted into a simple relational expression determined in advance to obtain the amount of attenuation due to diffraction. Note that, in order to perform high-speed calculation, a uniform amount of attenuation may be given in advance without obtaining the irradiation angle and diffraction angle.

[0052] Here, all or some of the setting unit 102a, surface edge identification unit 102b, transmission point view search unit 102c, reception point view search unit 102d, and propagation calculation information extraction unit 102e function as an "extraction unit".

[0053] The propagation calculation unit 102f functions as an "estimation unit" that estimates the propagation characteristics of radio waves based on the data extracted by the "extraction unit." For example, the propagation calculation unit 102f can estimate structures that may cause radio wave scattering in the surrounding environment based on the extracted structures (surfaces, edges, etc.), and estimate the propagation characteristics of radio waves.

[0054] More specifically, the propagation calculation unit 102f can estimate the propagation characteristics of radio waves for the extraction result of the combination of visible surfaces and edges that satisfy the propagation conditions. In addition, the propagation calculation unit 102f can simplify the calculation of the area, irradiation angle, and scattering angle of only the surface part that is commonly visible from the radio wave irradiation origin (transmission point) and the observation point (reception point) for the scattered wave calculation. In addition, the propagation calculation unit 102f calculates the irradiation angle and diffraction angle of the edge of the building that is commonly visible from the radio wave irradiation origin (transmission point) and the observation point (reception point) for the diffracted wave calculation. In addition, the propagation calculation unit 102f calculates the received power of each ray (path) and calculates the total received power.

[0055] The estimation method of the present embodiment simplifies the calculation, so that it is not necessary to obtain an image point in reflection, calculate a reflection point, or determine a reflection point. It is also not necessary to obtain a diffraction point in diffraction, or to obtain various conditions required for a complex calculation formula. Then, by simply processing the projected image after applying a color (or an identification number) to the surface or edge, it is possible to quickly extract the conditions and information required for the propagation calculation. Also, for a surface that is visible from both the radio wave irradiation origin (transmission point) and the observation point (reception point), it is possible to eliminate the need to calculate the reflection of the surface and the diffraction of the edge by simply calculating the scattering level using a radar equation or the like. Therefore, the amount of calculation can be reduced, and high-speed estimation is possible.

[0056] In addition, the estimation method of this embodiment uses radar equations or the like to calculate the scattering level for the surface portion visible from both the radio wave irradiation origin (transmission point) and the observation point (reception point), making it possible to calculate the scattering not only in the normal direction but also in any direction. In addition, even if the surface portion visible from both is smaller than the first Fresnel radius, it is possible to calculate the scattering level (the above conditions cannot be calculated in the case of the ray tracing method). Therefore, it is expected that radio wave propagation estimation results with higher accuracy than conventional methods can be obtained.

[0057] In addition, the propagation calculation unit 102f may estimate the propagation characteristics of the radio wave, which involves the calculation of the diffraction of the radio wave in the surrounding environment, based on the extracted two-dimensional data (e.g., data on the vertical surface of the edge commonly seen from Tx and Rx). Here, FIG. 12 is a diagram comparing the calculation by the conventional ray tracing method in three dimensions with the method of this embodiment. As shown in FIG. 12, in the conventional ray tracing method, the calculation of diffraction in three-dimensional space is performed for buildings and the like from the transmitting point to the receiving point, which is complicated. In this embodiment, as described above, the calculation of diffraction in two-dimensional space is performed using two-dimensional map data of the vertical surface of the edge commonly seen from Tx and Rx, so that the calculation is simplified. That is, the edge is projected onto a vertical surface, and substituted into a simple relational expression that determines the irradiation angle and the diffraction angle to obtain the amount of attenuation. Since the wave level of diffraction is weak to begin with, the estimation accuracy can be maintained even if the above-mentioned simple relational expression is used. Furthermore, the amount of calculation can be reduced, and high-speed estimation is possible.

[0058] Furthermore, the propagation calculation unit 102f may estimate the propagation characteristics of the radio wave by using a neural network, etc. In particular, in this embodiment, the propagation calculation unit 102f may function as an "estimation unit" that estimates the propagation characteristics of the radio wave by applying a fully connected neural network to the first parameters extracted by the "extraction unit" and the second parameters that represent the configuration of a wireless communication system that performs wireless communication between a transmission point and a reception point.

[0059] The above-mentioned "extraction unit" can receive map data such as a residential map and extract urban structure parameters used for propagation estimation from the map data using a neural network (NN).

[0060] The urban structure parameters may include, by way of non-limiting example, one or more parameters selected from the aforementioned building occupancy ratio, road width, road angle, road edge building height, average building height, and building height near the base station.

[0061] The urban structure parameter may take different values ​​depending on the relationship between the positions of the base station and / or mobile station and the positions of surrounding buildings. In other words, the urban structure parameter is an example of a first parameter that represents a structure that may cause scattering of radio waves in a surrounding environment including a receiving point (mobile station or base station) of radio waves transmitted from a transmitting point (base station or mobile station).

[0062] As a non-limiting example, a convolutional neural network (CNN) may be applied as the NN.

[0063] CNN has a structure in which a convolutional layer that locally extracts features from input data (e.g., an image) and a pooling layer that locally collects features are repeated, as described later in Fig. 13. The pooling layer (or subsampling layer), for example, compresses (sometimes called "downsampling") the input data from the convolutional layer into an easier-to-process format.

[0064] In CNN, the parameters of the convolution filter (hereinafter sometimes abbreviated as "filter") are shared at all locations in the input image. Therefore, the number of parameters can be reduced compared to a simple fully connected neural network (FNN).

[0065] In addition, the pooling layer can further reduce the number of parameters and also gradually impart invariance to translation of the input image.

[0066] When the input data is image data, CNN can be conceptually interpreted as a network that calculates the co-occurrence of adjacent features at different scales while gradually reducing the resolution of the image data, and selectively transmits information that is useful for classification or identification to the next layer (higher layer).

[0067] The propagation calculation unit 102f receives, for example, the city structure parameters extracted by the city structure parameter extraction unit and the system parameters, and estimates radio wave propagation in the wireless communication system by a neural network using both parameters. The system parameters are an example of second parameters that represent the configuration of a wireless communication system that performs wireless communication between a transmitting point and a receiving point. The "estimation" of radio wave propagation may be interpreted as "analysis," "analysis," or "simulation," etc.

[0068] The system parameters may include, for example, information indicating one or more of the following: the transmission / reception distance, which is the distance between the transmitting point and the receiving point; the frequency of the radio wave for which radio wave propagation is to be estimated; the antenna height of the base station; the antenna height of the mobile station; and whether or not there is a line of sight between the transmitting and receiving points.

[0069] The system parameters may be determined according to the system and / or scenario for which radio wave propagation is to be estimated. For example, when a macrocell environment is to be estimated, it may be assumed that the base station antenna is installed at a higher position than surrounding buildings (i.e., base station antenna height>surrounding building height).

[0070] Furthermore, since the "frequency" is determined, for example, according to the placement mode (in other words, cell design) of the base station corresponding to either the transmitting point or the receiving point, it may be considered to be included in an example of a system parameter determined according to the positional relationship between the transmitting point and the receiving point.

[0071] For example, FNN is applied to the NN in the propagation calculation unit 102f to express a function that estimates the propagation loss from the input parameters. In the FNN, the nodes in each layer are connected to all the nodes in the next layer. The FNN plays the role of a discrimination unit that classifies the combinations of features detected up to the lower layers into specific prediction results. In the NN, the "nodes" in each layer are sometimes called "neurons" or "units."

[0072] The propagation calculation unit 102f may include an error calculation unit and a weight update unit (not shown). The error calculation unit and the weight update unit are used to learn weight coefficients (which may be referred to as "weight parameters") of the DNN formed by, for example, the CNN and the FNN.

[0073] In a CNN, the weight parameters can be considered to correspond to, for example, the filter coefficients of a convolution filter applied in a convolution layer (C layer). In an FNN, the weight parameters can be considered to correspond to the coefficients assigned to the weight connections between layers (in other words, between nodes in different layers) that constitute the FNN. For example, back propagation may be applied to learn the weight parameters.

[0074] The backpropagation method is a method in which, for example, the amount of update for the weight parameters of each layer is calculated using values ​​obtained from the higher layer (output layer), and the values ​​that determine the amount of update for the weight parameters of each layer are calculated and propagated toward the lower layer (input layer).

[0075] For example, the error calculation unit calculates the error between the estimated value of the radio wave propagation characteristics (hereinafter sometimes abbreviated as "propagation characteristics") obtained by the propagation calculation unit 102f and the radio wave propagation characteristics obtained from the actual measurement value (or may be a theoretical value). The weight update unit repeatedly updates the weight parameters of the DNN until the calculated error converges to a given convergence condition.

[0076] The error calculated by the error calculation unit may be expressed by, for example, an error function that uses weight parameters as arguments. The error function may also be called a loss function, a cost function, or an (estimated) objective function.

[0077] For example, the Gradient Descent Method may be applied to update the weight parameters. In the Gradient Descent Method, for example, the weight parameters of the DNN are iteratively updated by making the gradient value of the error function approach zero, in other words, by making the partial differential value of the error function approach zero.

[0078] The middle part of Fig. 13 is a functional block diagram illustrating the functional block diagram illustrated in the upper part of Fig. 13 in a more abstract manner than the upper part of Fig. 13. In the middle part of Fig. 13, the weight parameters of the FNN and CNN may be sequentially updated from the upper layer to the lower layer (from right to left in the middle part of Fig. 13) by the above-mentioned backpropagation method in each layer of the DNN.

[0079] For learning weight parameters using the gradient descent method, batch learning or mini-batch learning may be applied. In mini-batch learning, learning is performed in units of data used for learning (which may be called "learning data", "teacher data", or "training data") divided into multiple parts.

[0080] Mini-batch learning can shorten the learning time, improve the efficiency of computer resource utilization, and / or suppress the decrease in learning efficiency, as compared to batch learning. For example, when the above-mentioned big data on radio wave propagation environment is used as learning data, it is useful to apply mini-batch learning. The unit into which the learning data is divided is sometimes called an "epoch."

[0081] Examples of input data used in the radio wave propagation estimation device 100 of this embodiment and calculation methods using neural networks and the like can be appropriately introduced from the items described in JP 2019-122008 A.

[0082] The map data on the left side in the lower part of Fig. 13 is a diagram showing an example of data (e.g., map data) used to estimate propagation characteristics for a pair of a certain BS position and a certain MS position. In this figure, the BS position corresponds to the transmission point, and the MS position corresponds to the reception point.

[0083] Map data of a specific area (for example, a rectangular area of ​​128 m×128 m) including the position of the MS corresponding to the reception point is input to the urban structure parameter extraction unit, as shown in the lower part of FIG.

[0084] The specific region including the MS position is not limited to a rectangular region, and may be a circle, an ellipse, or another shape. The size of the specific region including the MS position may be adjusted appropriately depending on, for example, the expected load of the calculation.

[0085] Furthermore, the map data input to the urban structure parameter extraction unit may be divided into meshes, for example, as shown in the lower part of Fig. 13. The size of one mesh, which is the unit of division, is, for example, 2m x 2m. However, the unit of division is not limited to 2m x 2m, and may be appropriately adjusted depending on, for example, the load expected for calculation.

[0086] Furthermore, the map data input to the urban structure parameter extraction unit may include "building map data", "line-of-sight map data", "images seen from a transmitting point", and "images seen from a receiving point". Hereinafter, "building map data" and "line-of-sight map data" may be abbreviated to "building map" and "line-of-sight map", respectively.

[0087] A "building map" is, for example, "map data consisting of information on the height of buildings in each mesh," and meshes with no buildings are defined as 0m. On the other hand, a "line-of-sight map" is, for example, "map data consisting of information on the presence or absence of line-of-sight for each mesh to the BS." Here, FIG. 14 is a diagram showing examples of "images seen from a transmitting point" and "images seen from a receiving point."

[0088] As shown in FIG. 14, the map data of this embodiment may be (1) an image seen from a transmission point with structures given a distinguishing color, (2) an image seen from a transmission point with structures given distance information, (3) an image seen from a reception point with structures given a distinguishing color, or (4) an image seen from a reception point with structures given distance information.

[0089] In the prior art, the estimation accuracy of the propagation characteristics (propagation loss) was poor (RMS error was about 10 dB). This is mainly due to the fact that the range handled in the calculation was limited, making it impossible to handle all buildings that contribute to the scattered waves, and that accurate scattering power could not be calculated without using information on the size of the contributing building faces. According to the present embodiment, in order to improve the accuracy, the above-mentioned images (1) to (4) below are input to CNN as shown in FIG. 15, and the control unit 102 (extraction unit) extracts features (the size of the wall and edge parts visible from Tx and Rx, and the distance to those parts) using CNN, and then inputs them to the FNN. (1) 2D image (colored image) of the faces and edges of buildings, etc. as seen from the transmitting point Tx (2) The distance from the transmission point Tx to the position corresponding to each element of the image in (1) (this may be the average distance for each face or edge) (3) 2D image (colored image) of the faces and edges of buildings, etc. as seen from the receiving point Rx (4) The distance from the receiving point Rx to the position corresponding to each element of the image in (3) (sometimes the average distance per face or edge) 3. Example of operation When the operation of the radio wave propagation estimation device 100 is divided into a learning stage and an estimation stage after the learning is completed, the big data described above may be used for learning the CNN and FNN in the learning stage. Note that the word "stage" may be replaced with other terms such as "phase," "stage," "step," or "procedure."

[0090] The system parameters input to the propagation calculation unit 102f in the learning and estimation stages are, for example, information indicating one or both of the “transmission distance D” (meters) corresponding to the distance between the BS and the MS, and “whether or not there is line of sight.”

[0091] The information indicating the "transmission distance D (m)" may be the logarithm of the transmission distance (logD). The information indicating the "presence or absence of line of sight" may be the line of sight flag (LOS). The line of sight flag is an example of a parameter that indicates the presence or absence of line of sight with a numerical value (e.g., "1" and "0").

[0092] The input data described with reference to FIG. 3 is the input data when the MS corresponds to the reception point. Also, for the case where the BS corresponds to the reception point, by inputting map data of a specific area including the position of the BS, which is the reception point, to the urban structure parameter extraction unit (see FIG. 15 etc.), urban structure parameters for the surrounding environment including the BS, which is the reception point, can be extracted. Therefore, the radio wave propagation characteristics of the surrounding environment including the BS can be estimated.

[0093] <Configuration example of DNN> Next, a configuration example of the DNN applied to the urban structure parameter extraction unit and the propagation calculation unit 102f of the present embodiment will be described with reference to FIG. 15. As shown in FIG. 15, for example, a CNN is applied to the urban structure parameter extraction unit, and for example, an FNN is applied to the propagation calculation unit 102f.

[0094] (Configuration example of CNN) The CNN exemplarily has a convolutional layer (C layer), a pooling layer (P layer), and a fully connected layer (F layer).

[0095] The C layer extracts feature amounts of local patterns in the input data, for example, by performing a convolution operation using a convolution filter on the input data.

[0096] The P layer pools (in other words, aggregates) the values of a small area (sometimes referred to as a "local area") in the input data into one value. Some positional deviation is allowed in the pooling.

[0097] The F layer performs classification (or identification) based on the feature amounts obtained in the C layer and the P layer.

[0098] 15, the urban structure parameter extraction unit has, as a non-limiting example, four C layers (C0 to C3), two P layers (P0 and P1), and two F layers (F0 and F1). However, the numbers of C layers, P layers, and F layers are not limited to the numbers illustrated in FIG. 15, and may be appropriately adjusted depending on, for example, the load expected for the calculation.

[0099] As shown in FIG. 15, the urban structure parameter extraction unit receives input data of, for example, a size of "256×256". Note that "256×256" represents the number of samples per map. Also, "4" added to the input data of the size of "256×256" represents the number of maps. For example, in the case of a transmission point, the four maps may be considered to correspond to "a map of the R component of the color of the image of the building seen from the transmission point", "a map of the G component of the color of the image of the building seen from the transmission point", "a map of the B component of the color of the image of the building seen from the transmission point", and "a map of the distance to the building seen from the transmission point". In the case of a reception point, they may be considered to correspond to "a map of the R component of the color of the image of the building seen from the reception point", "a map of the G component of the color of the image of the building seen from the reception point", "a map of the B component of the color of the image of the building seen from the reception point", and "a map of the distance to the building seen from the reception point".

[0100] The urban structure parameter extraction unit performs a convolution operation on input data of size “256×256×number of maps” in the C0 layer, for example, with a filter size of “5×5×number of maps” and a filter number of “48”.

[0101] For example, the urban structure parameter extraction unit multiplies the element values ​​of a filter represented by a three-dimensional array (in other words, a matrix) of "5 x 5 x number of maps" by the element values ​​of the input data represented by a matrix of "256 x 256 x number of maps" (in other words, sample values), and adds (sums) the multiplication results.

[0102] By performing such multiplication and addition while sliding the filter so as to cover a range of "256 x 256" in two dimensions, a convolution operation result for a two-dimensional array of "252 x 252" is obtained. For example, this convolution operation produces map data having a size of "252 x 252" and a number equal to the number of filters = 48.

[0103] A filter that slides over input data is sometimes called a "sliding window," a "kernel," or a "feature detector."

[0104] Next, in the P0 layer, the urban structure parameter extraction unit performs a pooling operation on the output of the C0 layer (i.e., 48 map data pieces each having a size of 252 x 252), for example, with the size of the local region being 4 x 4.

[0105] By the pooling operation of the P0 layer, 48 pieces of map data are obtained, in which the size of "252 x 252" is compressed to the size of "63 x 63". In addition, either Max pooling, which selects the maximum value in a local region, or Average pooling, which obtains the average value in a local region, may be applied to the pooling operation. For convenience, an example in which Average pooling is applied will be described below.

[0106] Furthermore, the urban structure parameter extraction unit performs a convolution operation on the output of the P0 layer (i.e., 48 map data pieces having a size of "63 x 63") in the C1 layer with a filter size of "4 x 4 x 48" and a number of filters of "24". The convolution operation in the C1 layer results in map data pieces having a size of "60 x 60" and a number of filters equal to 24.

[0107] Next, in the P1 layer, the urban structure parameter extraction unit performs a pooling operation on the output of the C1 layer (i.e., 24 map data pieces each having a size of "60x60"), for example, with a local area size of "4x4". The pooling operation in the P1 layer results in 24 map data pieces, with the size of "60x60" compressed to the size of "15x15".

[0108] Furthermore, the urban structure parameter extraction unit performs a convolution operation on the output of the P1 layer (i.e., 24 map data pieces each having a size of "15x15") in the C2 layer with a filter size of "4x4x24" and a filter count of "12". The convolution operation in the C2 layer results in map data pieces each having a size of "12x12" and a number of filters equal to 12.

[0109] Next, the urban structure parameter extraction unit performs a convolution operation on the output of the C2 layer (i.e., 12 map data pieces having a size of "12x12") in the C3 layer with a filter size of "4x4x12" and a filter number of "6". The convolution operation in the C3 layer results in map data pieces having a size of "9x9" and a number of filters equal to 6.

[0110] The urban structure parameter extraction unit performs the above-mentioned convolution and pooling operations on each of the two map sets (four maps), namely, "map of structures as seen from the transmitting point Tx" and "map of structures as seen from the receiving point Rx", and then sums up the 486 (9 x 9 x 6) samples in each of the outputs of the C3 layer to obtain 972 (9 x 9 x 6 x 2) data.

[0111] The urban structure parameter extraction unit then inputs 972 sample data into the F0 layer, which is fully connected to 972 F1 layers, and outputs 972 pieces of data (in other words, 972 types of data) from the F1 layer as the final output from the CNN.

[0112] As described above, the urban structure parameter extraction unit can find the values ​​of 972 types of urban structure parameters from map data (number of samples = 256 x 256 x number of maps x 2 sets).

[0113] The input data size to the CNN, and the filter sizes and / or numbers of filters applied to the C layer (C0 to C3), P layer (P0 and P1), and F layer (F0 and F1) are not limited to the above values. For example, these values ​​may be adjusted appropriately depending on the expected load of the calculation.

[0114] (Example of FNN configuration) On the other hand, the FNN applied to the propagation calculation unit 102f may be configured with three layers, that is, an input layer, a hidden layer, and an output layer, as illustrated in FIG.

[0115] The input layer receives the city structure parameters obtained by the city structure parameter extraction unit and the above-mentioned system parameters (e.g., the distance between the transmitter and receiver: logD, and the presence or absence of line of sight: LOS). Therefore, the number of input parameters to the FNN constituting the propagation calculation unit 102f is "974". The number of nodes in the intermediate layer and the output layer are "3" and "1", respectively.

[0116] In this FNN configuration, an estimate of the propagation characteristics (eg, propagation loss) is obtained by two fully-connected NNs using two F layers (F2 and F3).

[0117] (Activation function) Next, the activation function f(x) applied to the C layer and the F layer will be described. Note that "x" represents the index of the local region. For example, the ReLU function may be applied to each of the C layers (C0 to C3). Note that "ReLU" is an abbreviation for "Rectified Linear Unit".

[0118] In contrast to this, for example, a sigmoid function may be applied to each of the F0 to F2 layers among the F layers (F0 to F3).

[0119] For the F3 layer, for example, the identity function: f(x) = x may be applied.

[0120] <DNN Learning Method> As described above, the error backpropagation method may be applied to the learning of the weight parameters. Here, the larger the scale of the DNN and the larger the amount of learning data, the longer the time required for the learning of the DNN tends to be.

[0121] Therefore, by applying the mini-batch gradient descent method to the learning of the DNN, the learning time of the DNN can be shortened. The mini-batch gradient descent method is a method of updating the weight parameters by the gradient descent method for every n (<N) samples (in other words, mini-batch) out of N samples of learning data. In order to enhance the convergence of the weight parameters, Adam (adaptive moment estimation), which is one of the optimization methods, may be used in combination.

[0122] <Pooling Operation of P Layer> As described above, the P layer downsamples and compresses information in order to transform the input data into a more manageable format. By compressing the information, effects such as improvement in robustness against minute positional changes, suppression of overfitting, and / or reduction of computational cost can be expected.

[0123] In MAX pooling, the maximum value is selected for a local region. For example, in a DNN, among a plurality of nodes in the lower layer connected to a (plurality of) nodes in the upper layer, the value of the node having the maximum value becomes the output value of the node in the upper layer.

[0124] In other words, in MAX pooling, the features of a plurality of nodes in the lower layer are treated as the "maximum value" among those plurality of nodes, and subsampling is performed. From the perspective of neuron pooling, it can be said that MAX pooling is a structure that "intensively" outputs the maximum value of the neuron pool composed of a plurality of neurons in the lower layer, and reproduces the concentration of the neuron pool.

[0125] For example, if the data set has the characteristics that most of the values ​​of the nodes in the lower layer are "0" and only a few have valid values, MAX pooling can separate useless inputs more efficiently than average pooling. In other words, MAX pooling has higher performance as a noise filter than average pooling.

[0126] On the other hand, in average pooling, the value output from multiple nodes in the lower layer to the nodes in the upper layer is the average value, not the maximum value. From the perspective of the neuron pool, if there is no fluctuation in the sum of the neurons in the lower layer, or if the fluctuation is negligible, it functions sufficiently as a noise filter.

[0127] In average pooling, unlike MAX pooling, neuron values ​​smaller than the maximum value are not ignored, so compared to MAX pooling, the possibility of falling into a local minimum problem can be reduced. Note that a "local minimum problem" refers to a phenomenon in which the system converges to a partial optimal solution and does not reach the true optimal solution.

[0128] The operation of the radio wave propagation estimation device 100 having the above-described configuration, particularly the learning and estimation calculations, can use known methods described in, for example, JP 2019-122008 A.

[0129] As an example, the radio wave propagation estimation device 100 calculates urban structure parameters from the acquired map data using, for example, the urban structure parameter extraction unit, and calculates the propagation characteristics using the urban structure parameters and system parameters, etc., as described in FIG. 15 using the propagation calculation unit 102f.

[0130] When the propagation characteristics are calculated, the radio wave propagation estimation device 100 may, for example, use an error calculation unit to calculate an error between the calculated estimated value of the propagation characteristics (for example, propagation loss) and the propagation characteristics obtained based on actual measurement values ​​(or theoretical values) of the propagation characteristics. Note that the actual measurement values ​​(or theoretical values) of the propagation characteristics may be obtained based on information (in other words, big data) on the radio wave propagation environment between the BS and the MS, stored in the above-mentioned center DB 400.

[0131] Then, the error calculation unit determines whether the calculated error has converged to the given convergence condition. If the calculated error has not converged to the given convergence condition, for example, the weight update unit updates the weight parameters of the FNN and CNN that constitute the DNN by the error backpropagation method.

[0132] After updating the weighting parameters, the city structure parameters and the propagation characteristics are recalculated in the city structure parameter extracting section and the propagation calculation section 102f, respectively, and the error is calculated again in the error calculation section, and the convergence condition is judged.

[0133] The above process is repeated until it is determined that the convergence condition is satisfied, thereby updating the weight parameters of the FNN and CNN that constitute the DNN. When it is determined that the convergence condition is satisfied, the radio wave propagation estimation apparatus 100 may end the process of the learning stage.

[0134] In the estimation stage, for example, the radio wave propagation estimation device 100 sets estimation conditions for radio wave propagation. The estimation conditions may be set in the city structure parameter extraction unit, for example.

[0135] The configuration data of the estimation conditions may include, for example, information on a transmission point corresponding to a BS to which the MS can connect, and information on a reception point corresponding to the MS.

[0136] Furthermore, the radio wave propagation estimating apparatus 100 acquires map data (for example, a building map and a visibility map) to be input to the CNN as illustrated in FIG. 15, for example, from a map data DB.

[0137] Furthermore, the radio wave propagation estimation device 100 sets parameters (for example, weight parameters) used in radio wave propagation estimation for each layer constituting the DNN. The setting may be performed, for example, in an urban structure parameter extraction unit.

[0138] In other words, the urban structure parameter extraction unit may have a function as a setting unit that sets parameters used in radio wave propagation estimation.

[0139] Next, the radio wave propagation estimation apparatus 100 calculates city structure parameters from the acquired map data, for example, by the city structure parameter extraction unit that has been trained in the learning stage, as described with reference to FIG.

[0140] When the city structure parameters are calculated, the radio wave propagation estimation device 100 calculates propagation characteristics as described in FIG. 15, for example, by the propagation calculation unit 102f that has been trained in the learning stage, using the city structure parameters and system parameters.

[0141] As described above, according to the radio wave propagation estimation device 100 of one embodiment, the city structure parameter extraction unit using CNN can extract and determine city structure parameters that contribute to radio wave propagation estimation from map data such as a residential map, limited to structures visible from a receiving point and / or a transmitting point. Therefore, it is possible to simplify the determination of city structure parameters to be used for radio wave propagation estimation, shorten the time required for radio wave propagation estimation, and thereby achieve more accurate estimation compared to calculations using the same resources.

[0142] In addition, the accuracy of learning the DNN (CNN and FNN) used in the urban structure parameter extraction unit and the propagation calculation unit 102f can be improved by using, for example, information (e.g., big data) about the radio wave propagation environment reported from the MS to the network. Since the accuracy of learning the DNN can be improved, the accuracy of estimating the propagation characteristics can also be improved.

[0143] [Second embodiment] Next, an embodiment in which identification information (identification number or color) is assigned to the surface (and edge) of a structure and used in propagation calculation will be further described. Below, differences from the first embodiment will be mainly described, and similarities will not be described.

[0144] Specifically, a radio wave propagation estimation device (and a radio wave propagation estimation method) capable of calculating reflected (scattered) waves in any direction on the wall surface of a structure or the like will be described.

[0145] (Feature 1) In order to search for the walls of structures and other objects that reflect (scatter) radio waves in any direction, different identification numbers or colors are assigned to the surfaces of all objects, including structures (which may include road signs, utility poles, streetlights, fixtures, etc.).

[0146] The radio wave propagation estimation device according to this embodiment draws images of a structure or the like as seen from both the radio wave transmission point and the radio wave reception point, and counts the number of pixels of each color (corresponding to a wall) that are commonly seen. The number of counted pixels corresponds to the area of ​​the wall that is commonly seen from the transmission point Tx and the reception point Rx, and varies depending on the distance and angle between the transmission point Tx and the reception point Rx and the wall. Therefore, the result of multiplying the number of pixels of the images of both the transmission point Tx and the reception point Rx is proportional to the strength of the received radio wave (received power) at the wall.

[0147] FIG. 16 shows an example of an image seen from a transmitting point Tx and a receiving point Rx. FIG. 17 shows an image of a wall surface commonly seen from the transmitting point Tx and the receiving point Rx. The image may be created using a computer, but it may also be a still image or a video of an object such as a structure captured using a camera or the like. This may be done.

[0148] 16 and 17, the radio wave propagation estimating device detects a color that is commonly seen from both the transmitting point Tx and the receiving point Rx (assuming here that it is a wall surface of a structure such as a building) and counts the number of pixels of the detected commonly seen color. The number of pixels seen from the transmitting point Tx and the receiving point Rx (corresponding to the area of ​​the visible part) may be expressed as follows:

[0149] When viewed from the transmitting point Tx: Tx_pixels When viewed from the receiving point Rx: Rx_pixels As shown in Fig. 17, a surface of a structure (such as a wall of a building) that has the same color is commonly seen from a transmitting point Tx and a receiving point Rx. The strength of the radio wave (which may be interpreted as the received power) received at the receiving point Rx through such a wall can be expressed as follows:

[0150] Received signal strength ∝Tx_pixels * Rx_pixels Alternatively, the strength of the received radio wave may be expressed as follows:

[0151] Received signal strength Pr = k * Tx_pixels * Rx_pixels The proportionality coefficient (k) is not limited to a constant, and may be a function of the antenna gain, distance, angle, a constant of the material of the scattering object, frequency, or the like.

[0152] Such a method of calculating (estimating) the intensity of a received radio wave using a color that can uniquely identify the surface of a structure that is commonly seen from the transmitting point Tx and the receiving point Rx is referred to here as the color image method.

[0153] The identification information for identifying an object including a structure (specifically, a surface of the object) may be a color or a number. In addition, the color representation format is typically RBG, but other formats (e.g., CYMK) may also be used.

[0154] Furthermore, the term "structure" typically refers to a building (architecture) including a building, but is not limited to this. Specifically, it may also refer to all objects that contribute to the reflection (scattering) of radio waves, such as road signs, utility poles, streetlights, and fixtures. Furthermore, it may also refer to not only stationary objects, but also moving objects, such as vehicles.

[0155] Fig. 18 shows an example of grouping of radio wave scattering objects. As shown in Fig. 18, the identification information (a set of numbers or colors) may be classified into the ground and radio wave scattering objects, and the radio wave scattering objects may be classified into stationary objects and moving objects.

[0156] Stationary objects may include the above-mentioned structures, utility poles, street lights, fixtures, etc., and moving objects may include vehicles, etc. (which may include trains).

[0157] The method of assigning the identification number or color (identification information) may be a method of randomly assigning to each surface. Alternatively, the radio wave scattering objects may be grouped according to their type, and the identification information may be randomly assigned to each surface of the objects belonging to the same group. Note that the identification information does not necessarily have to be assigned randomly.

[0158] Generally, one identification number or color is assigned to an entire surface, but in some cases, one surface may be divided into a plurality of regions (eg, mesh-like) and an identification number or color may be assigned to each mesh.

[0159] FIG. 19 shows an example in which one color is assigned to an entire face, and an example in which a face is divided into a plurality of regions (meshes) and a different color is assigned to each mesh.

[0160] As shown in Fig. 19 (left side), a specific color may usually be assigned to an entire face, but as shown in Fig. 19 (right side), a face may be divided into multiple regions (meshes) and a different color may be assigned to each mesh.

[0161] When a specific color is assigned to an entire surface, because the entire surface is painted with a single color, even if the parts seen from the transmitting point Tx and the receiving point Rx do not actually overlap, the entire surface is determined to be seen as being common, resulting in an error in the calculation of the radio wave scattering level.

[0162] On the other hand, if the surface is divided into multiple regions (meshes), when viewed in terms of the colors of the meshes, there is no mesh that is commonly seen from both the transmitting point Tx and the receiving point Rx (right side of Figure 19), so no error occurs in the calculation of the radio wave scattering level as described above.

[0163] The transmitting point Tx and the receiving point Rx typically refer to the installation positions of the transmitting antenna and the receiving antenna, but are not necessarily limited to the installation positions of the antennas. For example, when radio waves are reflected (scattered) multiple times, they may include pseudo transmitting points and receiving points.

[0164] FIG. 20 shows an example in which radio waves are reflected once between the transmission point Tx (antenna) and the reception point Rx (antenna), and an example in which radio waves are reflected multiple times between the transmission point Tx (antenna) and the reception point Rx (antenna).

[0165] As shown in FIG. 20, when radio waves (which may be interpreted as wireless signals) reach a receiving point Rx (antenna) from a transmitting point Tx (antenna) via multiple walls, wall 1, wall N-1, etc. may be interpreted as pseudo receiving points and / or transmitting points.

[0166] (Feature 2) In order to quickly determine the visibility between the transmission point Tx and the reception point Rx, an identification number or color may be assigned to the ground. Specifically, the radio wave propagation estimation device creates an image of the reception point Rx, the ground, structures, and the like as viewed from the transmission point Tx, and may determine that the visibility between the transmission point Tx and the reception point Rx exists if the color (or identification number) of the pixel to which the reception point Rx belongs matches the color (or identification number) of the visible ground.

[0167] Fig. 21A shows an example of an image of a receiving point, the ground, a structure, etc. as seen from a transmitting point, and Fig. 21B shows an example of determining visibility using a conventional ray tracing method.

[0168] Note that instead of assigning a single identification number or color to the ground, the ground may be divided into meshes and an identification number or color may be assigned to each mesh. This makes it possible to determine the visibility of the receiving points in the entire evaluation area by analyzing only one image seen from the transmitting point.

[0169] As shown in Fig. 21A, the ground may be divided into a plurality of regions, and each region is assigned a uniquely identifiable color (or an identification number). In Fig. 21A, the same color (here, black) different from the ground is assigned to the structures and the background.

[0170] FIG. 21A shows an image (topographical map) seen from the transmission point, and the reception point shown in FIG. 21A is determined to be in line of sight with the transmission point.

[0171] On the other hand, as shown in FIG. 21B, in the visibility determination using the conventional ray tracing method, a line is connected between each receiving point (RX1, 2, 3) and the transmitting point, and it is checked successively whether or not the receiving point is blocked by each structure, which takes time to make a determination.

[0172] (Feature 3) Considering time and other factors, the maximum number of reflections (scatterings) that the radio wave propagation estimation device can calculate is naturally limited (maximum N times). Therefore, it is possible that there will be a reception point where the reception strength cannot be calculated even after N reflections (scatterings), i.e., no commonly visible object surfaces can be found. To deal with such cases, the radio wave propagation estimation device may calculate propagation characteristics such as reception strength using a complementary calculation method.

[0173] Specifically, the radio wave propagation estimating device may use the following complementary calculation method.

[0174] (1) Set the default reception strength (2) The height of the transmitting point is changed (raised) so that the target receiving point can be calculated, and the statistics of the difference in the calculation results at other receiving points that can be calculated both before and after the change, that is, the average value and standard deviation value, are calculated, and the receiving strength before the change is calculated (estimated) using the average value and standard deviation value and the calculation result after the change at the receiving point. (3) Calculate the diffracted waves (4) Using the ray tracing method (5) Using other computable receiving points as pseudo transmitting points, calculate the power at the target receiving point. FIG. 22 shows an example of a method for calculating diffracted waves. To begin with, the power of diffracted waves is very weak, so a simple calculation method can be applied. As shown in FIG. 22, the diffracted waves are projected onto the vertical plane of the edge of the structure, and the irradiation angle of the radio wave (θ t ) and the diffraction angle (θ r ) may be substituted into a simple relational equation to obtain the amount of attenuation.

[0175] (Example of the configuration of a radio wave propagation estimation device) 23A and 23B show an example of the configuration of a radio wave propagation estimating device according to this embodiment. As shown in FIG. 23A and FIG. 23B, the radio wave propagation estimating device 200 includes an external input unit 210, a storage unit 220, a calculation unit 230, and an output unit 240.

[0176] The external input unit 210 can input information that is not stored (stored) in the storage unit 220 from the outside to the storage unit 220. For example, the external input unit 210 can input image data acquired by a camera or a laser scanner to the storage unit 220.

[0177] The storage unit 220 can store various databases (DBs) such as map data, etc. For example, the storage unit 220 can store a geographic information DB, a radio wave scattering contributing object DB, an identification number / color DB, a reception point DB, a transmission point DB, and the like.

[0178] The calculation unit 230 includes a pre-processing unit 231 and a post-processing unit 232. The pre-processing unit 231 can set reception points in a specified range and create and analyze images of the reception points in advance. Specifically, the pre-processing unit 231 may have the following functions.

[0179] (1) Specify the evaluation area, interval between reception points, database to be used, and other conditions. (2) Using a DB for acquiring reception points, for example a road network, reception points are calculated based on the evaluation area and the interval between reception points. (3) The surfaces of objects that contribute to radio wave scattering are colored or assigned an identification number. (4) Create an image seen from the receiving point (5) Analyze the image seen from the receiving point (6) The image seen from the receiving point and the analysis results are stored in the database. In this embodiment, the preprocessing unit 231 may function as a part of the extraction unit. Specifically, as described above, the preprocessing unit 231 uses identification information capable of identifying an object including a structure and map data including the structure, and can extract the structure of the object visible from a direct or pseudo reception point of the radio wave based on the identification information.

[0180] The structure of an object may be the external shape of a structure, and may particularly include the surface configuration of a building or the like that is made up of a plurality of surfaces (which may be flat surfaces or may be curved surfaces).

[0181] As described above, the preprocessing unit 231 (extraction unit) can extract the structure (surfaces) of an object based on the color (or number) that identifies the surface of the object. Furthermore, the surface of the object may be divided into a plurality of regions (meshes) as shown in Fig. 19. Different identification information (number or color) may be assigned to each of the regions (meshes).

[0182] Furthermore, the structure of the object may include, in a broad sense, the ground surface around the structure. The ground surface may be read as the ground. The ground may be interpreted as an area where no structures such as buildings are installed.

[0183] The post-processing unit 232 can set transmission points, create and analyze images seen from the transmission points, etc. The post-processing unit 232 can also compare and integrate the analysis results of the images of the reception points, and calculate and combine the radio wave scattering levels on each surface of the structure. Specifically, the post-processing unit 232 may have the following functions.

[0184] (1) Set the post-processing conditions (number of scatterings, transmission point specifications, antenna pattern, etc.) (2) Create an image seen from the transmitting point (3) Analyze the image seen from the transmitting point (4) Comparing and integrating the analysis results of the image seen from the transmitting point with the analysis results of the image seen from the receiving point (e.g., counting the number of pixels on a common surface seen from the transmitting point and the receiving point) (5) Calculation and synthesis of radio wave scattering levels on each surface of a structure In this embodiment, the functions (1) to (4) of the post-processing unit 232 may function as part of an extraction unit. Specifically, as described above, the functions (1) to (4) of the post-processing unit 232 use identification information capable of identifying an object including a structure and map data including the structure, and can extract the structure of an object visible from a direct or pseudo transmission point of radio waves based on the identification information. Therefore, the functions (1) to (4) of the post-processing unit 232 and the pre-processing unit 231 may function as an extraction unit. The propagation calculation unit of the post-processing unit 232 may function as an estimation unit. Specifically, the propagation calculation unit can estimate the propagation characteristics of radio waves using information indicating the structural features of an object such as a structure extracted by the pre-processing unit 231 and the functions (1) to (4) (extraction unit) of the post-processing unit 232. The structural features may include the type of object, the configuration of the external surface (number, size, direction, position (distance), area of ​​the visible part, etc.), the material of the surface, etc.

[0185] The propagation calculation unit (estimation unit) may estimate the propagation characteristics using the number of pixels (corresponding to the area of ​​the visible portion) constituting the surface of an object visible from both the transmitting point and the receiving point, as described above. Specifically, the propagation calculation unit may calculate the reception strength from the number of pixels of the image of the surface commonly visible from the transmitting point and the receiving point, as shown in Fig. 16 and Fig. 17, and estimate the propagation characteristics.

[0186] If no object is found that is visible from both the transmitting point and the receiving point, the propagation calculation unit may change the height of the transmitting point. Generally, a higher height is effective, but a lower height may also be included.

[0187] As described above, the propagation calculation unit may estimate the propagation characteristics before the height is changed based on the difference between the calculation results of the reception characteristics at all other reception points before and after the height is changed. Specifically, the height of the transmission point may be changed (raised) so that the target reception point can be calculated, and the statistics of the difference between the calculation results at other reception points that can be calculated both before and after the change, that is, the average value and standard deviation value, may be obtained, and the reception strength before the change may be calculated (estimated) using the average value and standard deviation value and the calculation result after the change at the reception point.

[0188] Furthermore, when an object visible from both the transmitting point and the receiving point cannot be found, the propagation calculation unit may estimate the propagation characteristics by setting a default value of the reception strength at the receiving point, calculating the diffracted waves of the radio wave, or using a ray tracing method, as described above. Alternatively, when an object visible from both the transmitting point and the receiving point cannot be found, the propagation calculation unit may use another calculable reception point as a pseudo transmitting point and estimate the propagation characteristics at the reception point to be estimated.

[0189] The propagation calculation unit may estimate at least one of the received power at the receiving point, the propagation loss between the transmitting point and the receiving point, and the channel characteristics between the transmitting point and the receiving point. The channel characteristics may include the power, delay time, and arrival angle of each ray. A specific estimation method will be described later.

[0190] The propagation calculation unit may externally acquire information on objects that contribute to scattering of radio waves (radio wave scattering contributing objects). Specifically, the propagation calculation unit may acquire information on the radio wave scattering contributing objects via the external input unit 210. The propagation calculation unit may further use the acquired information to estimate the above-mentioned propagation characteristics.

[0191] The object may include a structure, and may further include a reflector capable of dynamically controlling reflected waves. Such a reflector may be called a Reconfigurable Intelligent Surface (RIS). The propagation calculation unit may control the characteristics of the reflector and estimate the above-mentioned propagation characteristics.

[0192] The output unit 240 can display the calculation result (reception level) by the calculation unit 230 on a color map or output it to a file. The output unit 240 can also display or output the image data shown in Figs. 16 and 21, etc.

[0193] (Example of operation) Next, an operation example of the radio wave propagation estimation apparatus 200 according to this embodiment will be described. Fig. 24 shows the operation flow of the pre-processing unit 231 of the calculation unit 230. As shown in Fig. 24, the pre-processing unit 231 sets calculation conditions, processes topographical information, calculates reception points, associates colors or identification numbers, creates images seen from reception points, and performs analysis.

[0194] Fig. 25 shows an operation flow of the post-processing unit 232 when using past pre-processing results. As shown in Fig. 25, the post-processing unit 232 uses past pre-processing results, sets post-processing conditions, creates an image seen from the transmission point, analyzes the image seen from the transmission point, compares and integrates the analysis results of the image seen from the transmission point and the analysis results of the image seen from the reception point, calculates and combines the radio wave scattering level on each surface of the structure, and outputs the calculation results for each reception point to the output unit 240.

[0195] 26A and 26B show the operation flows of the pre-processing unit 231 and the post-processing unit 232 in the case of directly starting calculation from the beginning. As shown in Fig. 26A and 26B, the pre-processing unit 231 may execute an operation generally similar to that shown in Fig. 24, and the post-processing unit 232 may execute an operation generally similar to that shown in Fig. 25.

[0196] (obtainable propagation characteristics) Next, examples of propagation characteristics that can be acquired by the radio wave propagation estimation device 200 according to this embodiment will be described.

[0197] Fig. 27 shows an example of calculation of received power and propagation loss by radio wave propagation estimation apparatus 200. As shown in Fig. 27, if the received power of scattering on each plane i is Pri, the total received power Pr at the reception point can be expressed by the formula shown in Fig. 27.

[0198] 28 shows an example of calculation of channel characteristics by the radio wave propagation estimation apparatus 200. As described above, the channel characteristics may include the power, delay time, and angle of arrival of each ray.

[0199] Ray i (path) on each plane i has the following information, so that the channel characteristics at the receiving point can be calculated.

[0200] Received power Pri ·Emission direction θ ti , angle of arrival θ ri , delay time T i Moreover, the radio wave propagation estimation device 200 includes the external input unit 210, and can utilize external input as described above. The purpose of the external input is to incorporate objects that do not exist in the radio wave scattering contributing object DB (for example, vehicles, telephone booths, human bodies, etc.) and use them in the calculation of propagation characteristics.

[0201] Fig. 29 shows an operational flow for the radio wave propagation estimation apparatus 200 to capture an object not present in the radio wave scattering contributing object DB. Figs. 30A to 30C show examples of objects not present in the radio wave scattering contributing object DB, and examples of car and human body models.

[0202] As shown in FIG. 29, when incorporating an object that is not present in the radio wave scattering contributing object DB, the radio wave propagation estimating apparatus 200 may use a rough model (see FIG. 30B) as shown in methods 1 to 3, or may use a detailed model (see FIG. 30C) as shown in methods 2 and 3.

[0203] The radio wave propagation estimation device 200 can use such a model to create a model of the objects contributing to radio wave scattering and store it in the radio wave scattering contributing object DB. This makes it possible to realistically reproduce the surrounding environment, improve the estimation accuracy, and evaluate the effect on the channel characteristics of fluctuations in the surrounding environment due to moving objects.

[0204] (Example of change) The radio wave propagation estimating apparatus 200 may be used in combination with a reconfigurable intelligent surface (RIS) or the like.

[0205] 31A and 31B show an image of control of propagation characteristics at a reception point by a RIS, etc. When a RIS is combined, radio wave propagation estimating apparatus 200 may operate as follows.

[0206] (1) Assigning an identification number or color to RIS, etc. (2) Create and analyze images from the transmitting and receiving points, and count the number of pixels visible from both. (3) Calculate the received signal strength (Pr = k * Tx_pixels * Rx_pixels) (4) Adjust and control the incidence angle and reflection angle of the RIS etc. to maximize the received power. (Action and effect) The radio wave propagation estimation device 200 according to this embodiment provides the following actions and effects. Specifically, the radio wave propagation estimation device 200 uses identification information (color or number) capable of identifying an object including a structure and map data including the structure, and extracts the structure (external surface, etc.) of the object visible from both the direct or pseudo transmission point of the radio wave and the direct or pseudo reception point of the radio wave based on the identification information. Furthermore, the radio wave propagation estimation device 200 estimates the propagation characteristics of the radio wave using information indicating the characteristics of the extracted structure.

[0207] Therefore, even if the regular reflection point Ir does not belong to the plane of the building, it is possible to effectively reduce the error and the amount of calculation in the power calculation of the radio wave arriving at the reception point. Furthermore, since the radio wave is scattered in various directions, not just in the regular direction, the radio wave propagation estimation device 200 can effectively reduce the error and the amount of calculation in the power calculation of the radio wave arriving at the reception point.

[0208] In this embodiment, the radio wave propagation estimation apparatus 200 extracts the above-mentioned structure based on the color that identifies the surface of the object, and estimates the propagation characteristics using the number of pixels that make up the surface of the object that is visible from both the transmitting point and the receiving point. This makes it possible to estimate the propagation characteristics with higher accuracy while suppressing the processing load.

[0209] In this embodiment, the surface of the object may be divided into a plurality of regions (meshes), and different identification information may be assigned to each of the regions. This allows the surface visible from both the transmitting point and the receiving point to be determined with higher accuracy, and therefore the propagation characteristics can be estimated with higher accuracy.

[0210] In this embodiment, the object may include the ground surface around the structure. This allows the visibility of the receiving points in the entire evaluation area to be determined by analyzing only one image seen from the transmitting point, and allows efficient estimation of the propagation characteristics.

[0211] In this embodiment, the radio wave propagation estimation device 200 can estimate at least one of the received power at the reception point, the propagation loss between the transmission point and the reception point, and the channel characteristics between the transmission point and the reception point, and can therefore easily provide the required propagation characteristics.

[0212] In this embodiment, the radio wave propagation estimation device 200 can externally acquire information on objects contributing to radio wave scattering and estimate the propagation characteristics using the acquired information. This makes it possible to estimate the propagation characteristics with high accuracy even when an object that does not exist in the radio wave scattering contributing object DB is included.

[0213] In this embodiment, the object may include a reflector (such as a RIS) that can dynamically control the reflected wave, and the radio wave propagation estimation device 200 can control the characteristics of the reflector and estimate the propagation characteristics. Therefore, it is possible to control the propagation characteristics at the receiving point with high accuracy by combining a RIS.

[0214] [Other embodiments] The present invention has been described above with reference to the embodiments. However, the present invention is not limited to these embodiments, and it will be obvious to those skilled in the art that various modifications and improvements are possible.

[0215] The functional blocks (components) in the radio wave propagation estimation device 100 illustrated in Fig. 1 and the radio wave propagation estimation device 200 illustrated in Figs. 23A and 23B may be realized by any combination of hardware and / or software. Furthermore, the means for realizing each functional block is not limited to a specific means. For example, each functional block may be realized by one device that is physically and / or logically combined, or may be realized by directly and / or indirectly (for example, by wire and / or wirelessly) connecting multiple devices that are physically and / or logically separated.

[0216] 32 shows an example of a hardware configuration of the radio wave propagation estimating apparatus 100 and the radio wave propagation estimating apparatus 200. The radio wave propagation estimating apparatus 100 may be configured by a computer such as a personal computer (PC) or a server computer. The computer is an example of an information processing device.

[0217] The radio wave propagation estimating apparatus 100 (and the radio wave propagation estimating apparatus 200, hereinafter the same) may be configured with one computer or with multiple computers. The multiple computers may distribute the load of the processing executed by the radio wave propagation estimating apparatus 100.

[0218] As shown in FIG. 16, the radio wave propagation estimating device 100 may include, for example, a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, and a bus 1007.

[0219] In the following description, the term "apparatus" may be interpreted as a circuit, a device, a unit, etc. The hardware configuration of the radio wave propagation estimation apparatus 100 may be configured to include one or more of the devices illustrated in Fig. 10, or may be configured to exclude some of the devices.

[0220] For example, although only one processor 1001 is illustrated in FIG. 32, multiple processors may be provided in the radio wave propagation estimation apparatus 100. Furthermore, the processes in the radio wave propagation estimation apparatus 100 may be executed by one processor 1001 or multiple processors. In one or multiple processors, multiple processes may be executed simultaneously, in parallel, or sequentially, or may be executed by other methods. The processor 1001 may be a single-core processor or a multi-core processor. The processor 1001 may be implemented using one or more chips.

[0221] One or more functions of the radio wave propagation estimation apparatus 100 are realized, for example, by loading predetermined software into hardware such as the processor 1001 and the memory 1002. Note that the "software" may also be referred to as a "program," an "application," or a "software module."

[0222] For example, the processor 1001 reads and executes a program by controlling reading and / or writing of data stored in the memory 1002 and / or the storage 1003. The program may be transmitted from a network by communication via a telecommunication line by the communication device 1004.

[0223] The program may be a program that causes a computer to execute all or part of the processing in the radio wave propagation estimating apparatus 100. One or more functions of the radio wave propagation estimating apparatus 100 are realized in response to execution of program code included in the program. All or part of the program code may be stored in the memory 1002 or the storage 1003, or may be written as part of an operating system (OS).

[0224] For example, the program may include program code that embodies the functional blocks illustrated in Fig. 1 and the like, and may also include program code that executes the illustrated flow. A program including such program code may be called a "radio wave propagation estimation program."

[0225] The processor 1001 is an example of a processing unit, and controls the entire computer by running an OS, for example. The processor 1001 may be configured using a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, and the like.

[0226] Furthermore, the processor 1001 reads, for example, one or both of a program and data from one or both of the storage 1003 and the communication device 1004 into the memory 1002, and executes various processes according to the read program and / or data.

[0227] The memory 1002 is an example of a computer-readable recording medium, and may be configured using at least one of, for example, ROM, EPROM, EEPROM, RAM, SSD, etc. Note that "ROM" is an abbreviation for "Read Only Memory", "EPROM" is an abbreviation for "Erasable Programmable ROM", "EEPROM" is an abbreviation for "Electrically Erasable Programmable ROM", "RAM" is an abbreviation for "Random Access Memory", and "SSD" is an abbreviation for "Solid State Drive".

[0228] The memory 1002 may be called a register, a cache, a main memory, a work memory, a main storage device, etc. The memory 1002 stores a program executable to implement a radio wave propagation estimation method according to an embodiment of the present invention.

[0229] Storage 1003 is an example of a computer-readable recording medium, and may be configured using at least one of an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive (HDD), a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a flexible disk, a magnetic strip, and the like. Storage 1003 may be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other suitable medium including one or both of memory 1002 and storage 1003.

[0230] The communication device 1004 is hardware (which may also be called a “transceiver device”) for communicating between computers via one or both of a wired network and a wireless network. The "communication device" may be referred to as, for example, a network device, a network controller, a network card, a communication module, or the like.

[0231] The input device 1005 is an example of an input device that accepts an input from outside the radio wave propagation estimation apparatus 100. For example, the input device 1005 may include one or more of a keyboard, a mouse, a microphone, a switch, a button, and a sensor.

[0232] The output device 1006 is an example of an output device (output unit 108) that performs output to the outside of the radio wave propagation estimation apparatus 100. For example, one or more of a display, a speaker, an LED lamp, and the like may be included in the output device 1006. Note that "LED" is an abbreviation of "light emitting diode."

[0233] The input device 1005 and the output device 1006 may be separate devices, or may be integrated into one device, such as a touch panel.

[0234] Furthermore, each device such as the processor 1001 and the memory 1002 may be communicatively connected by a bus 1007. The devices may be connected by a single bus 1007, or may be connected using a plurality of buses.

[0235] The radio wave propagation estimation apparatus 100 may be configured to include hardware such as a microprocessor, a DSP, an ASIC, a PLD, and an FPGA. For example, the processor 1001 may be implemented by at least one of these pieces of hardware. Some or all of the functional blocks illustrated in Fig. 1 and Fig. 2 may be realized by the hardware.

[0236] "DSP" is an abbreviation for "Digital Signal Processor," "ASIC" is an abbreviation for "Application Specific Integrated Circuit," "PLD" is an abbreviation for "Programmable Logic Device," and "FPGA" is an abbreviation for "Field Programmable Gate Array."

[0237] (Adaptive Systems) The embodiments described herein may be applied to systems utilizing LTE (Long Term Evolution), LTE-Advanced (LTE-A), SUPER 3G, IMT-Advanced, 4G, 5G, FRA (Future Radio Access), W-CDMA (registered trademark), GSM (registered trademark), CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), IEEE 802.20, UWB (Ultra-WideBand), Bluetooth (registered trademark), or other suitable systems and / or next generation systems enhanced therewith.

[0238] (Processing procedures, etc.) The steps, sequences, flow charts, etc. of each aspect / embodiment described herein may be reordered unless inconsistent, For example, the methods described herein present elements of various steps in an example order, and are not limited to the particular order presented.

[0239] (Handling of input and output information, etc.) Input and output information, etc. may be stored in a specific location (e.g., memory) or may be managed in a management table. Input and output information, etc. may be overwritten, updated, or added. Output information, etc. may be deleted. Input information, etc. may be transmitted to another device.

[0240] (Judgment method) The determination may be based on a value represented by a single bit (0 or 1), a Boolean (true or false) value, or a numerical comparison (e.g., comparison with a predetermined value).

[0241] (software) Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0242] Additionally, software, instructions, etc. may be transmitted or received over a transmission medium. For example, if the software is transmitted from a website, server, or other remote source using wired technologies, such as coaxial cable, fiber optic cable, twisted pair, and Digital Subscriber Line (DSL), and / or wireless technologies, such as infrared, radio, and microwave, these wired and / or wireless technologies are included within the definition of transmission media.

[0243] (Information, Signals) The information, signals, etc. described herein may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0244] In addition, terms explained in this specification and / or terms that aid in the understanding of this specification may be replaced with terms having the same or similar meanings.

[0245] (Parameter) Furthermore, the information, parameters, etc. described in this specification may be expressed as absolute values, may be expressed as relative values ​​from a predetermined value, or may be expressed as other corresponding information.

[0246] The names used for the above parameters are not intended to be limiting in any way, and the mathematical expressions using these parameters may differ from those explicitly disclosed herein.

[0247] (base station) A base station can accommodate one or more (e.g., three) cells (also called sectors). When a base station accommodates multiple cells, the entire coverage area of ​​the base station can be divided into multiple smaller areas, and each smaller area can also provide communication services by a base station subsystem (e.g., a small base station for indoor use, RRH: Remote Radio Head). The term "cell" or "sector" refers to a part or the entire coverage area of ​​a base station and / or a base station subsystem that provides communication services in this coverage. Furthermore, the terms "base station", "BS", "eNB", "gNB", "cell", and "sector" may be used interchangeably in this specification. A base station may also be referred to by terms such as fixed station, NodeB, eNodeB (eNB), gNodeB, access point, small cell, etc. A small cell is an example of a cell with a smaller coverage than a macro cell. A small cell may be called differently depending on its coverage area. For example, a small cell may be referred to as a "femtocell," a "picocell," a "microcell," a "nanocell," a "metrocell," a "home cell," etc. The term "cell" or "sector" may mean an individual geographical area in which a base station provides wireless service, as well as a part of a communication function managed by the base station to communicate with terminals in that individual geographical area.

[0248] (UE) The UE may also be referred to by those skilled in the art as a user terminal, terminal, mobile station, subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, UE, or some other suitable term. The UE may be a fixed terminal whose location does not change, or a mobile terminal whose location changes. As a non-limiting example, the UE may be a mobile terminal such as a mobile phone, a smartphone, or a tablet terminal. The UE may also be an IoT (Internet of Things) terminal. Various "things" may be equipped with a communication function by the IoT. Various "things" equipped with a communication function can connect to the Internet, a wireless access network, or the like to communicate. For example, an IoT terminal may include a sensor device or a meter (measuring instrument) equipped with a wireless communication function. Some monitoring devices such as a surveillance camera or a fire alarm equipped with a sensor device or a meter may correspond to the terminal. Wireless communication between a UE, which is an IoT terminal such as a monitoring device, and a base station is sometimes called MTC (Machine Type Communications). Therefore, a UE, which is an IoT terminal, is sometimes called an "MTC device."

[0249] Functions include, but are not limited to, judgment, determination, judgment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, deeming, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs the function of transmission is called a transmitting unit or a transmitter. As described above, there is no particular limitation on the method of realization of any of these.

[0250] In the present disclosure, a specific operation performed by a base station may be performed by its upper node in some cases. In a network consisting of one or more network nodes having a base station, it is clear that various operations performed for communication with a terminal may be performed by at least one of the base station and other network nodes other than the base station (e.g., MME or S-GW, etc., but are not limited to these). Although the above example illustrates a case in which there is one other network node other than the base station, it may also be a combination of multiple other network nodes (e.g., MME and S-GW).

[0251] Information, signals (information, etc.) may be output from a higher layer (or a lower layer) to a lower layer (or a higher layer), and may be input and output via multiple network nodes.

[0252] Each aspect / embodiment described in this disclosure may be used alone, in combination, or switched according to execution. In addition, notification of specific information (e.g., notification that "X is the case") is not limited to being done explicitly, but may be done implicitly (e.g., not notifying the specific information).

[0253] As used in this disclosure, the terms "system" and "network" are used interchangeably.

[0254] The terms "connected," "coupled," or any variation thereof, refer to any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are "connected" or "coupled" to each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, "connected" may be read as "access." As used in this disclosure, two elements may be considered to be "connected" or "coupled" to each other using at least one of one or more wires, cables, and printed electrical connections, as well as electromagnetic energy having wavelengths in the radio frequency range, microwave range, and optical (both visible and invisible) range, as some non-limiting and non-exhaustive examples.

[0255] The reference signal may also be abbreviated as Reference Signal (RS), and may be called a pilot depending on the applicable standard.

[0256] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0257] Any reference to an element using a designation such as "first," "second," etc., used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient way to distinguish between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed therein or that the first element must precede the second element in some way.

[0258] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, as is the term "comprising." Further, as used in this disclosure, the term "or" is not intended to be an exclusive or.

[0259] In this disclosure, where articles have been added due to translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0260] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, search, inquiry (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in a memory), and the like. In addition, "judgment" and "decision" can include considering resolving, selecting, choosing, establishing, comparing, etc., to be a "judgment" or "decision." In other words, "judgment" and "decision" can include considering some action to be a "judgment" or "decision." In addition, "judgment" can be interpreted as "assuming," "expecting," "considering," etc.

[0261] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." In addition, the term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may be interpreted in the same manner as "different."

[0262] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described in the present disclosure. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure. [Explanation of symbols]

[0263] 100 Radio wave propagation estimation device 102 Control section 102a Setting section 102b Surface edge identification unit 102c Transmission point view search section 102d Reception point view search section 102e Information extraction unit for propagation calculation 102f Propagation Calculation Section 106 Storage section 108 Output section 200 Radio wave propagation estimation device 210 External input section 220 Storage section 230 Calculation Department 231 Pretreatment section 232 Post-processing section 240 Output section 1001 Processor 1002 Memory 1003 Storage 1004 Communication equipment 1005 Input Device 1006 Output device 1007 Bus

Claims

1. an extraction unit that uses identification information capable of identifying an object including a structure and map data including the object, and extracts a structure of the object visible from both a direct or pseudo transmission point of radio waves and a direct or pseudo reception point of the radio waves based on the identification information; an estimation unit that estimates a propagation characteristic of the radio wave by using information indicating a structural feature of the object extracted by the extraction unit; Equipped with The extraction unit extracts the structure based on a color or an identification number that identifies a surface of the object; The estimation unit estimates the propagation characteristics using the number of pixels constituting a surface of the object visible from both the transmitting point and the receiving point. Radio wave propagation estimation device.

2. A surface of the object is divided into a plurality of regions; The radio wave propagation estimation device according to claim 1 , wherein different identification information is assigned to each of the regions.

3. The radio wave propagation estimation device according to claim 1 , wherein the object includes the ground surface around the structure.

4. The estimation unit is 2. The radio wave propagation estimation device according to claim 1, wherein, when the object visible from both the transmitting point and the receiving point is not found, the propagation characteristics are estimated by either setting a default value for the reception strength at the receiving point, calculating diffracted waves of the radio wave, or using a ray tracing method.

5. The radio wave propagation estimation device according to claim 1 , wherein the estimation unit estimates at least one of a received power at the reception point, a propagation loss between the transmission point and the reception point, and a channel characteristic between the transmission point and the reception point.

6. The estimation unit is acquiring information on the object that contributes to scattering of the radio waves from an external source; The radio wave propagation estimation device according to claim 1 , wherein the propagation characteristics are estimated using the information.

7. The object includes a reflector capable of dynamically controlling a reflected wave; The radio wave propagation estimation device according to claim 1 , wherein the estimation unit controls a characteristic of the reflector and estimates the propagation characteristic.

8. A step of extracting a structure of the object visible from both a direct or quasi-transmitting point of radio waves and a direct or quasi-receiving point of the radio waves based on identification information capable of identifying the object including a structure and map data including the object; estimating a propagation characteristic of the radio wave using information indicating the extracted structural features; Including, The step of extracting includes extracting the structure based on a color or an identification number that identifies a surface of the object; the estimating step includes a step of estimating the propagation characteristics using a number of pixels constituting a surface of the object visible from both the transmitting point and the receiving point; Radio wave propagation estimation method.

9. An extraction unit that uses identification information capable of identifying an object including a structure and map data including the object, and extracts a structure of the object visible from both a direct or pseudo transmission point of radio waves and a direct or pseudo reception point of the radio waves based on the identification information; an estimation unit that estimates a propagation characteristic of the radio wave by using information indicating a structural feature of the object extracted by the extraction unit; Equipped with The estimation unit is if the object is not found that is visible from both the transmitting point and the receiving point, changing the height of the transmitting point; Estimating the propagation characteristics before the height is changed based on a difference between propagation characteristics at other reception points that can be calculated both before and after the height change, and the propagation characteristics at the reception point after the height change. Radio wave propagation estimation device.

10. A step of using identification information capable of identifying an object including a structure and map data including the object, and extracting a structure of the object visible from both a direct or pseudo transmission point of radio waves and a direct or pseudo reception point of the radio waves based on the identification information; estimating a propagation characteristic of the radio wave using information indicating the extracted structural features; Including, The estimating step includes: changing the height of the transmitting point if the object visible from both the transmitting point and the receiving point is not found; A step of estimating the propagation characteristics before the height is changed based on a difference between propagation characteristics at other reception points that can be calculated both before and after the height change, and the propagation characteristics at the reception point after the height change; Including, Radio wave propagation estimation method.

11. An extraction unit that uses identification information capable of identifying an object including a structure and map data including the object, and extracts a structure of the object visible from both a direct or pseudo transmission point of radio waves and a direct or pseudo reception point of the radio waves based on the identification information; an estimation unit that estimates a propagation characteristic of the radio wave by using information indicating a structural feature of the object extracted by the extraction unit; Equipped with The estimation unit is If the object is not found that is visible from both the transmitting point and the receiving point, using other computable reception points as pseudo transmission points to estimate the propagation characteristics at the reception point of interest; Radio wave propagation estimation device.

12. A step of using identification information capable of identifying an object including a structure and map data including the object, and extracting a structure of the object visible from both a direct or simulated transmission point of radio waves and a direct or simulated reception point of the radio waves based on the identification information; estimating a propagation characteristic of the radio wave using information indicating the extracted structural features; Including, The estimating step includes: If the object is not found that is visible from both the transmitting point and the receiving point, Using other calculable reception points as pseudo transmission points, estimating the propagation characteristics at the reception point of interest. Radio wave propagation estimation method.

Citation Information

Patent Citations

  • Device, method, program for estimating and calculating reception electric field intensity and recording medium

    JP2003318811A

  • Radio station position estimation device and method

    JP2005274205A

  • Network service information providing system, apparatus and method, and operation control method therefor

    JP2005341621A

  • Device, method and program for estimating radio wave propagation

    JP2019122008A

  • Radio-wave propagation estimation apparatus and radio-wave propagation estimation

    JP2019205127A

Cited By

  • Network node and control method

    JP2026178012A

  • Network node and control method

    JP2026178013A