Irs control device, irs control method, and program

Abstract

An IRS control device according to the present invention comprises a storage unit that stores a radio wave map for a user of radio waves for a prescribed area for when an intelligent reflecting surface in the prescribed area has been set to a plurality of reflection coefficients, a reception unit that receives a requested map for the prescribed area as in a radio wave state requested by the user, an inference unit that infers an optimal reflection coefficient pattern for the IRS that satisfies the requested map on the basis of the radio wave map stored at the storage unit, and a reflection coefficient control unit that controls the reflection coefficients of the IRS to the optimal reflection coefficient pattern inferred by the inference unit.

Description

IRS control device, IRS control method and program

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 447,892, filed February 24, 2023, the contents of which are incorporated herein by reference.

[0002] Advances in transmission and reception technologies, such as antennas and access methods, have dramatically improved the performance of wireless communication technology. However, it is said that the growth of such wireless communication is gradually reaching its limits. One of the factors hindering this growth is the physical constraint that communication becomes unstable when there is an obstacle between a base station and a communication terminal. Intelligent Reflecting Surfaces (IRS) have attracted attention as a solution to this problem (see, for example, Patent Documents 1 to 5). IRS is an electromagnetic wave reflector composed of a large number of passive reflecting elements. The reflecting elements are made of metamaterials whose properties can be dynamically changed. By appropriately changing the reflection properties of each element to design an ideal radio wave propagation path, it is possible to ensure a radio wave propagation route that bypasses obstacles, thereby enabling communication that is not affected by obstacles.

[0003] On the other hand, IRS requires a size of about 100 wavelengths of the carrier wave to ensure sufficient signal power through reflection. Therefore, if each telecommunications carrier installs its own communications infrastructure, there are concerns that the increase in communications equipment and wiring will damage the landscape and the environment. There are also concerns about economic impacts such as increased resource and energy consumption. Therefore, there is a demand for telecommunications carriers to improve the efficiency of communications infrastructure by sharing communications equipment such as IRS, thereby reducing the impact on the environment and the economy.

[0004] However, it is not easy for each telecommunications carrier to share communication facilities. One method is to share base stations. This involves sharing the device that transmits from the base station, and this can be done by splitting the signals from each telecommunications carrier. In this case, the electrical signals from each carrier are input to the radiating antenna from the same antenna, and no communication control processing is performed in the shared section.

[0005] Another possible approach is to share an IRS. This involves sharing an IRS device and its control function. However, each telecommunications carrier is assigned a different communication frequency, and current IRSs cannot set different phases for each frequency. Therefore, the control unit that controls the IRS must also be shared, and the IRS must be controlled so that each telecommunications carrier can achieve an acceptable radio wave condition.

[0006] US Patent Application Publication No. 2022 / 0216908 International Publication No. 2021 / 207748 US Patent Application Publication No. 2023 / 0047558 ​​US Patent Application Publication No. 2022 / 0216908 European Patent Application Publication No. 4099574

[0007] The problem to be solved is to provide good quality radio wave conditions to all users within a given area while sharing an IRS in a communication system.

[0008] One aspect of the present invention is an IRS control device that includes a memory unit that stores a radio wave map of a specified area of ​​a user who uses radio waves when an IRS (Intelligent Reflecting Surface) within the specified area is set with multiple reflection coefficients, a reception unit that receives a requested map when the specified area achieves the radio wave conditions requested by the user, an estimation unit that estimates an optimal reflection coefficient pattern of the IRS that satisfies the requested map based on the radio wave map stored in the memory unit, and a reflection coefficient control unit that controls the reflection coefficient of the IRS so that the optimal reflection coefficient pattern estimated by the estimation unit is achieved.

[0009] According to the present invention, it is possible to provide users within a predetermined area with good quality radio wave conditions while sharing an IRS in a communication system.

[0010] 1 is a diagram showing the configuration of a communication system according to the present embodiment; FIG. 2 is a diagram showing the configuration of an IRS control device; FIG. 3 is a flowchart showing the operation of an IRS control device; FIG. 4 is a diagram showing the configuration of an estimation model generation device; FIG. 5 is a flowchart showing the operation of an estimation model generation device; FIG. 6 is a diagram showing a measured radio wave map and an estimated radio wave map; FIG. 7 is a diagram showing the communication capacity of a terminal when the number of MNOs is changed; and FIG. 8 is a diagram showing the communication capacity of a terminal when the number of terminals (Users) is changed.

[0011] 1 is a diagram showing the configuration of a communication system 1 according to this embodiment. The communication system 1 includes a terminal 2, an MNO (Mobile Network Operator) 3, an IRS (intelligent reflecting surface) 4, and an IRS control device 5.

[0012] The multiple terminals 2 are terminals of users who receive communication services within a predetermined area controlled by the IRS 4, and communicate via base stations owned by the MNO 3. The terminals 2 are, for example, mobile phones, smartphones, personal computers, etc. The multiple terminals 2 may receive services from the same MNO 3 or from different MNOs.

[0013] There are multiple MNOs 3, and in this case, each MNO 3 is assigned a different communication frequency. The base station owned by the MNO 3 relays communication data received from the terminal 2 to the IRS control device 5 or other networks. The MNO 3 includes, for example, a RAN (Radio Access Network) and a core network. The terminal 2 may connect to a different MNO 3 depending on the radio wave conditions that are affected by the location, etc. The base station owned by the MNO 3 notifies the IRS control device 5 of a radio wave map (hereinafter referred to as a "request map") that requests that the communication environment of the terminal 2 be in an optimal state within a predetermined area controlled by the IRS 4. The radio wave map is a two-dimensional map that shows the radio wave strength within a predetermined range. The radio wave map can be, for example, a J ｘ ×J ｙ It is shown by the matrix:

[0014] The request map is, for example, a diagram showing the current location of a terminal 2 communicating with a base station owned by an MNO 3 within a predetermined area controlled by an IRS 4. The request map, for example, shows whether a two-dimensional plane is divided into grids and whether the terminal 2 is located in each grid. The request map is, for example, a J-ary where an element corresponding to the grid in which the terminal 2 is located is set to 1 and other elements are set to 0. ｘ ×J ｙ The request map notified by MNO3-u (u=1 to U) is hereinafter referred to as R ｕ It is defined as:

[0015] IRS4 is a device that integrates metamaterial elements, which are minute structures that can freely control electromagnetic properties. IRS4 sets a reflection coefficient for each part of the device based on the state of radio waves within a specified area. The reflection coefficient is determined, for example, by setting the phase. Below, the reflection coefficient pattern of IRS4-c (c = 1 to C) is referred to as b ｃ Reflection coefficient pattern b ｃ is a vector indicating the parameters β and φ of the structures constituting, for example, IRS4. When there are N structures, the reflection coefficient pattern b ｃ Ha b ｃ = {φ ｘ,１ , φ ｙ,１ , β ｘ,１ , β ｙ,１ ...φ ｘ,Ｎ , φ ｙ,Ｎ , β ｘ,Ｎ , β ｙ,Ｎ}, where φ is the first-order term of the reflection coefficient, and β is the second-order term of the reflection coefficient. All reflection coefficient patterns of IRS4 are defined as B = {b １ , ..., b Ｃ} is defined as

[0016] FIG. 2 is a diagram showing the configuration of the IRS control device 5. The IRS control device 5 includes a reception unit 51, an estimation unit 52, a control unit 53, and a storage unit 54. The storage unit 54 stores an estimation model for estimating a radio wave map based on the reflection coefficient pattern of the IRS. The reception unit 51 receives requested maps from multiple MNOs 3. The estimation unit 52 estimates a reflection coefficient pattern of the IRS 4 that satisfies the requested maps of the multiple MNOs 3 using the estimation model stored in the storage unit 54. An example of a reflection coefficient pattern that satisfies the requested map is a reflection coefficient pattern that maximizes the total value of the communication capacity of terminals 2 associated with all MNOs 3 within a predetermined area controlled by the IRS 4. The communication capacity of the terminal 2 is the communication capacity of the grid in which the terminal 2 is located.

[0017] The method of the present invention will now be described in detail.

[0018] Each of the multiple MNOs 3 has a different communication frequency, and when IRS4 is set to a certain reflection coefficient, the radio wave map of each operator is different. A radio wave map for each MNO 3 when IRS4 is set to a certain reflection coefficient is calculated. The memory unit 54 learns an output model of the radio wave map when the MNO 3 sets IRS4 to a certain reflection coefficient, estimates a radio wave map when the MNO 3 sets IRS4 to a reflection coefficient not included in the learning data, and stores the learned output model of the radio wave map. The memory unit 54 is not limited to a storage medium such as a memory, but may also be a location where data can be stored, such as the cloud.

[0019] The IRS control device 5 searches for an IRS reflection coefficient pattern that satisfies the requirements of each MNO 3 in the estimation unit 52. Using the learned radio wave map of each MNO 3, reflection coefficient patterns that can be set by the IRS are collected from the storage unit 54.

[0020] Reflection coefficient pattern b ｃ Based on this, we use an estimation model to estimate the radio wave map by MNO3-u (u = 1 to U). ｕ Also, f ｕ Estimated radio wave map X, which is the result of the estimation ｕ ~(b) is defined by the following formula:

[0021] The reception unit 51 receives the request map R from each MNO 3. ｕ Then, the estimation unit 52 receives the estimation model f ｕ and requirement map R ｕ Based on this, the estimation unit 52 calculates an optimal reflection coefficient pattern B in which reflection coefficients that maximize the communication capacity of the entire system including MNO3 are set. The estimation unit 52 calculates the optimal reflection coefficient pattern B by, for example, calculating equation (2).

[0022] Equation (2) is an equation for calculating the communication capacity by the Shannon-Hartley theorem and calculating the reflection coefficient pattern B when the total value of the communication capacity in each grid is maximized. The control unit 53 controls IRS4 to have reflection coefficient pattern B, which is the optimal reflection coefficient pattern estimated by the estimation unit 52.

[0023] By the control unit 53 controlling the IRS4 to have reflection coefficient pattern B, the ISR4 can provide high-quality radio wave conditions to all users, including terminals 2 with different MNOs 3, within a specified area controlled by the IRS4.

[0024] 3 is a flowchart showing the operation of the IRS control device 5. The reception unit 51 receives a request map R ｕ (Step S101). The estimation unit 52 uses the learned radio wave maps of each MNO3 stored in the storage unit 54 to estimate the reflection coefficient pattern of IRS4 that satisfies the required map, and the reflection coefficient patterns that can be set by the IRS to estimate the reflection coefficient that maximizes the communication capacity of the entire system including MNO3 (Step S102). The control unit 53 controls the reflection coefficient pattern B of IRS4 so that it becomes the optimal reflection coefficient pattern estimated by the estimation unit 52 (Step S103).

[0025] The estimation unit 52 will be described in more detail below. The estimation model used by the estimation unit 52 is generated by an estimation model generation device 6. FIG. 4 is a diagram showing the configuration of the estimation model generation device 6. The estimation model generation device 6 includes a data acquisition unit 61, a learning unit 62, and an output unit 63. The estimation model generation device 6 may be included in the communication system 1.

[0026] The data acquisition unit 61 acquires data indicating the relationship between the reflection coefficient pattern of IRS4 and the learned radio wave map including the radio wave map not included in the learning data for each MNO3. (bi) u Data showing the relationship between is prepared for each IRS 4 and each MNO 3. The radio wave map may be converted to ground truth, for example, by equation (3).

[0027] Here, X ｍａｘ Radio Map X ｕ ^ (bi) is the maximum value of the elements of

[0028] The data acquisition unit 61 acquires the estimated model f ｕ The reflection coefficient pattern of IRS4 input to the estimation model f ｕ The radio wave map for MNO3 output from the UE 100 may be acquired. This makes it possible to increase the amount of data used for learning, which will be described later.

[0029] The learning unit 62 generates an estimation model f that estimates the radio wave map based on data indicating the relationship between the reflection coefficient pattern of the IRS 4 and the radio wave map in the MNO 3. ｕ The estimation model is, for example, a neural network, and estimates the radio wave map by a machine learning technique. The learning unit 62 updates the estimation model f ｕ By inputting the reflection coefficient pattern b into the ｕ ~(b) The learning unit 62 estimates the measured radio wave map X , which includes the measurement results of the actually measured radio wave map and the learned unknown radio wave map. ｕ (b) and Estimated Radio Wave Map X ｕ~(b) A loss function is calculated between the mean square error (MSE) and the mean square error (MSE).

[0030] The learning unit 62 uses the measured radio wave map X ｕ (b) and Estimated Radio Wave Map X ｕ ~(b) In calculating the loss function between ｕ (b) The learning unit 62 calculates the loss function only from the grids where the measured radio wave map X ｕ (b) The map W is a map in which the value of the grid where ｕ Use X ｕ (b) ・W ｕ and X ｕ ~(b) Calculate the loss function between

[0031] The learning unit 62 learns the estimation model f so that the calculated value of the loss function becomes smaller. ｕ Update.

[0032] The output unit 63 outputs the updated estimation model f ｕ is output to the IRS control device 5.

[0033] 5 is a flowchart showing the operation of the estimation model generating device 6. The data acquiring unit 61 acquires data showing the relationship between the reflection coefficient pattern of IRS4 and the radio wave map in MNO3 (step S201). The learning unit 62 generates an estimation model f that estimates the radio wave map based on the data showing the relationship between the reflection coefficient pattern of IRS4 and the radio wave map in MNO3. ｕ (step S202). After that, the output unit 63 updates the updated estimation model f ｕ The estimation model generating device 6 executes step S202 multiple times to generate the estimation model f ｕ is updated multiple times, and then step S203 is executed to obtain the estimated model f ｕ may be output.

[0034] (Example) This embodiment was studied using a simulation. The simulation was conducted in an actual area of ​​Sendai city, where a relatively large intersection was simulated, with multiple MNOs having three base stations in surrounding buildings and roadside units, and five IRSs installed, and where users were present within the intersection and on the sidewalk. The size of the specified area was 48 m ２ is.

[0035] FIG. 6 shows the measured radio wave map X ｕ (b) and Estimated Radio Wave Map X ｕ ~(b) The measured radio wave map is a map transformed into ground truth. ｕ Measurement radio wave map X ｕ (b) We were able to estimate a distribution close to this.

[0036] 7 and 8 show the results of comparing the method of this embodiment with other methods. The method being compared is a time-sharing method of ISR usage time that the inventors have considered as another ISR sharing method. In the time-sharing method, communication time is divided for each MNO 3, and the IRS 4 is controlled so that the total communication capacity of terminals 2 in each MNO 3 is maximized during each time period.

[0037] 7 is a diagram showing the communication capacity of terminal 2 when the number of MNOs 3 is changed. By increasing the number of MNOs 3, the improvement ratio of the communication capacity of the method (prop) of this embodiment to the communication capacity of the time division method (td) increases. Therefore, this embodiment is more advantageous when the number of MNOs 3 is large.

[0038] 8 is a diagram showing the communication capacity of terminal 2 (user) when the number of terminals 2 is changed. Even when the number of terminals 2 is changed, the communication capacity in the method of this embodiment is larger than the communication capacity in the time division method.

[0039] Other Embodiments One embodiment of the present invention has been described in detail above with reference to the drawings, but the specific configuration is not limited to that described above, and various design changes and the like can be made within the scope that does not deviate from the gist of the present invention.

[0040] Estimation model f ｕ However, the present invention is not limited to this. For example, the estimation model f ｕ Alternatively, the estimation unit 52 may store data indicating the relationship between the reflection coefficient pattern of IRS 4 and the radio wave map in MNO 3, and output the corresponding radio wave map in MNO 3 by selecting a reflection coefficient pattern to be stored. Even in this case, the estimation unit 52 can select the optimal reflection coefficient pattern B from the stored reflection coefficient patterns.

[0041] The IRS control device 5 and the estimation model generation device 6 in the above-described embodiment may be partially or entirely implemented by a computer. In this case, a program for implementing this function may be recorded on a computer-readable recording medium, and the program may be loaded into a computer system and executed. Note that the term "computer system" as used herein includes an OS and peripheral hardware. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as recording devices such as hard disks built into a computer system. Furthermore, the term "computer-readable recording medium" may include media that dynamically store programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or telephone lines, or media that store programs for a certain period of time, such as volatile memory within a computer system that serves as a server or client. The program may be designed to implement part of the above-described functions, or may be capable of implementing the above-described functions in combination with a program already stored in the computer system. Furthermore, part or all of the IRS control device 5 and the estimation model generating device 6 may be realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0042] The MNO 3 is not limited to a telecommunications carrier that is a user of the IRS 4, but may also be a user of a terminal 2 in a communication system that uses a different frequency band. Furthermore, each user may have a unique characteristic vector of the radio wave propagation environment. Furthermore, if multiple IRSs 4 exist within a specified area, control may be performed to obtain a reflection coefficient pattern that maximizes the total communication capacity of the multiple users calculated across all the multiple IRSs 4.

[0043] REFERENCE SIGNS LIST 1 Communication system, 2 Terminal, 3 MNO, 4 IRS, 5 IRS control device, 51 Reception unit, 52 Estimation unit, 53 Control unit, 54 Storage unit, 6 Estimation model generation device, 61 Data acquisition unit, 62 Learning unit, 63 Output unit

Claims

1. A storage unit that stores a radio wave map in a predetermined area of a user who uses radio waves when an IRS (Intelligent Reflecting Surface) in the predetermined area is set with a plurality of reflection coefficients; A reception unit that receives a request map when the predetermined area reaches a radio wave state requested by the user; An estimation unit that estimates an optimal reflection coefficient pattern of the IRS that satisfies the request map based on the radio wave map stored in the storage unit; An IRS control device comprising: a reflection coefficient control unit that controls the reflection coefficient of the IRS so as to obtain the optimal reflection coefficient pattern estimated by the estimation unit.

2. The optimal reflection coefficient pattern is a reflection coefficient pattern when the total value of all communication capacities of the user is maximized. The IRS control device according to claim 1.

3. The reflection coefficient is determined by setting the phase of the IRS. The IRS control device according to claim 1.

4. The estimation unit estimates the optimal reflection coefficient pattern using an estimation model learned by machine learning based on a radio wave map when the reflection coefficient stored in the storage unit is set. The IRS control device according to claim 1.

5. There are a plurality of the users. The storage unit stores a radio wave map in the predetermined area for each user. The reception unit receives a plurality of request maps requested by the plurality of users. The estimation unit estimates, as the optimal reflection coefficient pattern, a reflection coefficient pattern that maximizes the communication capacity of all the plurality of users. The IRS control device according to claim 1.

6. The plurality of users use different frequency bands. The IRS control device according to claim 5.

7. Stores a radio wave map in a predetermined area of a user who uses radio waves when an IRS in the predetermined area is set with a plurality of reflection coefficients; Receives a request map when the predetermined area reaches a radio wave state requested by the user; Estimates an optimal reflection coefficient pattern of the IRS that satisfies the request map based on the stored radio wave map; Controls the reflection coefficient of the IRS so as to obtain the estimated optimal reflection coefficient pattern. IRS control method.

8. A program for operating a computer as the IRS control device according to any one of claims 1 to 6. **Claim 9**: In an IRS (Intelligent Reflecting Surface) control device that sets an IRS within a predetermined area with a plurality of reflection coefficients, an estimation unit that estimates a reflection coefficient pattern of the IRS such that the radio wave states of the plurality of users are optimized as a whole from the reflection coefficient patterns that can be set for the IRS, for each optimal radio wave state required by the plurality of users to whom different communication frequencies are assigned within the predetermined area; and a reflection coefficient control unit that controls the plurality of reflection coefficients of the IRS so as to be the reflection coefficient pattern estimated by the estimation unit. An IRS control device comprising the same. **Claim 10**: In an IRS (Intelligent Reflecting Surface) control method for setting an IRS within a predetermined area with a plurality of reflection coefficients, estimating a reflection coefficient pattern of the IRS such that the radio wave states of the plurality of users are optimized as a whole from the reflection coefficient patterns that can be set for the IRS, for each optimal radio wave state required by the plurality of users to whom different communication frequencies are assigned within the predetermined area; and an IRS control method for controlling the plurality of reflection coefficients of the IRS so as to be the estimated reflection coefficient pattern.