Channel estimation device, control method for channel estimation device, and control program for channel estimation device

Abstract

A channel estimation device according to one embodiment of the present invention comprises: an acquisition unit that acquires channel information indicating the communication quality between a base station and a wireless communication terminal; a time-series information processing unit that extracts, from a plurality of pieces of time-series channel information acquired by the acquisition unit and the movement speed of the wireless communication terminal, a predetermined number of channel feature amounts in the time direction of the plurality of pieces of channel information; a generation unit that generates an estimation model obtained by modeling, by machine learning, the relationship between the predetermined number of channel feature amounts and channel information chronologically subsequent to the plurality of pieces of channel information used for extracting the predetermined number of channel feature amounts; and an estimation unit that estimates future channel information of the wireless communication terminal that is the object of estimation, by inputting the predetermined number of channel feature amounts extracted by the time-series information processing unit to the estimation model on the basis of the plurality of pieces of time-series channel information between the wireless communication terminal that is the object of estimation and the base station.

Description

Channel Estimation Device, Control Method for Channel Estimation Device, and Control Program for Channel Estimation Device

[0001] The present invention relates to a channel estimation device, a control method for the channel estimation device, and a control program for the channel estimation device.

[0002] Conventionally, there has been a technique for estimating a channel (communication quality) between a moving wireless communication terminal (UE: User Equipment) and a base station using past channel information based on a predetermined reference signal periodically transmitted from the wireless communication terminal to the base station. In recent years, it has been proposed to use AI (Artificial Intelligence) and ML (Machine Learning) for this channel estimation. For example, Patent Document 1 discloses a communication quality prediction device that acquires environmental information around a wireless communication path as point cloud data in time series and predicts and calculates the future communication quality of the wireless communication path by machine learning.

[0003] Japanese Unexamined Patent Application Publication No. 2023-26789

[0004] A channel estimation device according to an embodiment of the present invention includes an acquisition unit that acquires channel information indicating communication quality between a base station and a wireless communication terminal, a plurality of time-series channel information acquired by the acquisition unit, and a moving speed of the wireless communication terminal. A time-series information processing unit that extracts a predetermined number of channel feature amounts in the time direction of the plurality of channel information, and a generation unit that generates an estimation model that models, by machine learning, the relationship between the predetermined number of channel feature amounts and channel information that is later in time series than the plurality of channel information used for extracting the predetermined number of channel feature amounts. An estimation unit that inputs a predetermined number of channel feature amounts extracted by the time-series information processing unit based on a plurality of time-series channel information between the wireless communication terminal to be estimated and the base station into the estimation model and estimates future channel information of the wireless communication terminal to be estimated.

[0005] In the channel estimation device according to an embodiment of the present invention, the time-series information processing unit may extract a predetermined number of channel feature amounts for each of a plurality of wireless communication terminals having different moving speeds.

[0006] In a channel estimation device according to one embodiment of the present invention, the generation unit may learn the relationships between multiple wireless communication terminals with different moving speeds and generate an estimation model.

[0007] A channel estimation device according to one embodiment of the present invention further comprises an image generation unit that generates channel image information, which is an image of the channel information acquired by the acquisition unit for each acquisition time period, and channel image information may be used as channel information in the extraction of channel features, generation of estimation models, and estimation of future channel information.

[0008] In a channel estimation device according to one embodiment of the present invention, the time-series information processing unit may be an autoencoder.

[0009] A control method for a channel estimation device according to one embodiment of the present invention involves the channel estimation device performing the following steps: acquiring channel information indicating the communication quality between a base station and a wireless communication terminal; extracting a predetermined number of time-direction channel features from the multiple time-series channel information acquired in the acquisition step and the movement speed of the wireless communication terminal; generating an estimation model that models the relationship between the predetermined number of channel features and channel information that is later in the time series than the multiple channel information used to extract the predetermined number of channel features, using machine learning; and inputting the predetermined number of channel features extracted in the extraction step into the estimation model based on the multiple time-series channel information between the wireless communication terminal to be estimated and the base station, in order to estimate the future channel information of the wireless communication terminal to be estimated.

[0010] A control program for a channel estimation device according to one embodiment of the present invention enables the channel estimation device to perform the following functions: a function to acquire channel information indicating the communication quality between a base station and a wireless communication terminal; a function to extract a predetermined number of time-direction channel features from multiple time-series channel information acquired by the acquisition function and the movement speed of the wireless communication terminal; a function to generate an estimation model that models the relationship between a predetermined number of channel features and channel information that is later in the time series than the multiple channel information used to extract the predetermined number of channel features, using machine learning; and a function to input a predetermined number of channel features extracted by the extraction function into the estimation model based on multiple time-series channel information between the wireless communication terminal to be estimated and the base station, in order to estimate the future channel information of the wireless communication terminal to be estimated.

[0011] Figure 1 is a schematic diagram of the channel estimation system configuration according to one embodiment of the present invention. Figure 2 is a diagram illustrating the outline of channel estimation according to one embodiment of the present invention. Figure 3 is an example of a functional block diagram of a channel estimation device according to one embodiment of the present invention. Figure 4 is a flowchart showing an example of operation of a channel estimation device according to one embodiment of the present invention. Figure 5 is an example of a hard block configuration of a computer capable of realizing a channel estimation device according to one embodiment of the present invention. Figure 6 is a schematic diagram illustrating the prior art.

[0012] Hereafter, an embodiment of the invention described herein (also referred to as the present invention) will be explained using the figures. Note that the figures are examples, and the present invention is not limited to those shown in the figures. For example, the number of channel estimation devices, wireless communication terminals, and base stations shown, the functional block diagram, flowchart, and data flow are examples, and the present invention is not limited to these.

[0013] Figure 6 is a schematic diagram illustrating the conventional challenges. In Figure 6, images A1-A4 and B1-B11 are images of channel information between a wireless communication terminal and a base station, and are examples of channel information heatmaps where the horizontal axis represents the number of resource blocks (RS) and the vertical axis represents the number of receiving antennas. Conventional channel estimation assumes that channel fluctuations over short periods are small and obtains channel information by continuously using the last measured channel (sample and hold). Wireless communication terminals are expected to move at various speeds, from low speeds (from walking speed to running speed) to high speeds (from bicycle speed to the speed of cars, trains, etc.). Images A1-A3 and B1-B10 show the channel information over time for wireless communication terminals moving at 3 km / h (low speed range) and 12 km / h (high speed range), respectively. The fluctuation of channel information correlates with the speed of the wireless communication terminal, with faster wireless communication terminals exhibiting greater fluctuations in channel information. Therefore, the faster a wireless communication terminal is moving, the more past channel information is needed to improve the accuracy of channel estimation. In the example in Figure 6, for a wireless communication terminal moving at 3 km / h, the fluctuation of channel information is small, and 3 past channel information points (images A1-A3) are needed to predict the next channel information (image A4). However, for a wireless communication terminal moving at 12 km / h, the fluctuation of channel information is large, and only 10 past channel information points (images B1-B10) are needed to predict the channel information (image B11).

[0014] Thus, when generating a model for channel estimation (estimation model), the number of past channel information variables used as explanatory variables for training will differ depending on the speed of the wireless communication terminal. However, it is not practical to train and generate an estimation model for each speed of the wireless communication terminal. Although AI-based channel estimation has been performed at the research level, due to hardware and specification limitations, a practical method that satisfies the performance requirements (error rate, latency, etc.) required for moving wireless communication terminals has not been established. In other words, it has been difficult to generate a general-purpose estimation model that can be applied to wireless communication terminals with different speeds.

[0015] In contrast, as will be described in detail later, according to one embodiment of the present invention, when training, a time-series information processing unit is provided in front of the input layer. This unit extracts temporal features (such as the period and magnitude of channel information fluctuations and noise distribution) from information regarding the movement speed of the wireless communication terminal and past channel information in the time series, and converts the past channel information into a data format of a predetermined size (a predetermined number). This makes it possible to input the same number of explanatory variables during training, even when the movement speed of the wireless communication terminal is different.

[0016] <System Configuration> Figure 1 shows an example configuration of a channel estimation system according to one embodiment of the present invention. The channel estimation system 600 includes a plurality of wireless communication terminals 300, a radio access network RAN ​​(Radio Access Network) 400, and a core network CN. The RAN 400 includes a base station 200 and a channel estimation device 100. In this embodiment, a 5G (fifth generation) mobile communication system is assumed, and the RAN 400 may be a RAN to which the specifications defined by the industry group O-RAN Alliance (Open Radio Access Network Alliance) are applied. The base station 200 may include a RU (Radio Unit), a DU (Distribution Unit), and a CU (Central Unit). The CU / DU may be implemented by a vRAN (virtual RAN) which implements each function by software on general-purpose hardware.

[0017] The channel estimation device 100 may implement a RIC (RAN Intelligent Controller), which is defined in O-RAN as a logical node for automating and optimizing the parameter design, setting, and operation of base stations. The channel estimation device 100 and the base station 200 are connected to each other via a network for communication. The base station 200 may also have the functions of the channel estimation device 100.

[0018] Figure 2 is a diagram illustrating the outline of channel estimation according to one embodiment of the present invention. According to one embodiment of the present invention, a neural network may be used to generate an AI model (estimation model) for channel estimation. More specifically, in one embodiment of the present invention, channel information may be processed as image information as described above, and a CNN (Convolutional Neural Network) may be used as the machine learning algorithm.

[0019] As shown in Figure 2, according to one embodiment of the present invention, when training an estimation model using a CNN, a time-series information processing unit is provided before the input layer. The time-series information processing unit may be, for example, an encoder that extracts temporal features (period and magnitude of channel information fluctuations, noise distribution, etc.) from information about the movement speed of the wireless communication terminal and past channel information in the time series. This time-series information processing unit converts the past channel information into a data format of a predetermined size (determined number). Therefore, even if the number of data required for training when generating an estimation model would normally differ due to different movement speeds, the number of nodes in the input layer can be made common. The predetermined number of nodes in the input layer can be any number and may be set in advance.

[0020] <Functional Configuration> The functions of each functional unit of the channel estimation device 100 will be explained using Figure 3. The channel estimation device 100 may include an acquisition unit 110, a time-series information processing unit 120, a generation unit 130, an estimation unit 140, and an image generation unit 150.

[0021] The acquisition unit 110 acquires channel information indicating the communication quality between the base station 200 and the wireless communication terminal 300. The channel information may be measured at the base station 200, or it may be measured at the wireless communication terminal 300 and notified to the base station 200. The channel information to be measured is not limited to channel state information (CSI). The acquisition unit 110 may acquire channel information transmitted periodically from multiple wireless communication terminals 300 connected to the base station 200, or at times requested by the base station 200, along with information regarding the mobile speed of the multiple wireless communication terminals 300.

[0022] The time-series information processing unit 120 extracts a predetermined number of time-direction channel features from the time-series channel information acquired by the acquisition unit 110 and the movement speed of the wireless communication terminal 300. The time-series information processing unit 120 may be implemented using an autoencoder. An autoencoder is an unsupervised machine learning method for dimensionality reduction using a neural network. The movement speed of the wireless communication terminal 300 may be measured by the wireless communication terminal 300 from the change in its current position using GPS (Global Positioning System). The time-series information processing unit 120 can convert the multiple channel information into a data format that matches the number of nodes in the input stage of the neural network.

[0023] As described above, the time-series information processing unit 120 extracts a predetermined number of channel features for each of the multiple wireless communication terminals 300 with different movement speeds. In other words, in one embodiment of the present invention, even if the wireless communication terminals 300 have different movement speeds, the number of extracted channel features may be the same.

[0024] The generation unit 130 generates an estimation model that models the relationship between a predetermined number of channel features and channel information that occurs later in the time series than the multiple channel information used to extract the predetermined number of channel features, using machine learning. That is, the estimation model may be generated by machine learning, inputting a predetermined number of channel features extracted by the time-series information processing unit 120 based on multiple time-series channel information acquired in the past and the movement speed of the wireless communication terminal 300, and outputting channel information measured after the multiple time-series channel information used to extract the channel features. In other words, the generation unit 130 may learn the relationship between multiple wireless communication terminals 300 with different movement speeds and generate an estimation model.

[0025] The estimation unit 140 inputs a predetermined number of channel features extracted by the time-series information processing unit 120 based on multiple time-series channel information between the wireless communication terminal 300 and the base station 200 to the estimation model generated by the generation unit 130, and estimates the future channel information of the wireless communication terminal 300 to be estimated. That is, for the wireless communication terminal 300 to be estimated, the time-series information processing unit 120 extracts channel features based on multiple time-series channel information and movement speed, and the estimation unit 140 inputs these channel features into the estimation model and outputs future channel information.

[0026] In one embodiment of the present invention, channel information may be processed as predetermined image information. The image generation unit 150 may generate channel image information by imaging the channel information acquired by the acquisition unit 110 for each acquisition time. The image information may be a heat map of channel information, for example, as described in Figures 2 and 6, with the horizontal axis representing the number of resource blocks and the vertical axis representing the number of receiving antennas. However, the image information is not limited to this. This makes it easier to extract features of the channel information.

[0027] <Control Flow of Channel Estimation Device> The control flow of the channel estimation device 100 described above will be explained using Figure 4. The acquisition unit 110 acquires channel information indicating the communication quality between the base station 200 and the wireless communication terminal 300 (step S11). Next, the time-series information processing unit 120 extracts a predetermined number of channel features in the time direction from the acquired time-series channel information and the movement speed of the wireless communication terminal 300 (step S12). The generation unit 130 generates an estimation model that models the relationship between the predetermined number of channel features and channel information that is later in the time series than the multiple channel information used to extract the predetermined number of channel features, using machine learning (step S13). The estimation unit 140 inputs a predetermined number of channel features extracted by the time-series information processing unit 120 based on the time-series channel information between the wireless communication terminal 300 and the base station 200 to the estimation model and estimates the future channel information of the wireless communication terminal 300 (step S14).

[0028] <Hardware Configuration> The hardware configuration of the channel estimation device 100 will now be described. Figure 6 shows an example of a computer hardware configuration capable of realizing the channel estimation device 100 in this embodiment. The channel estimation device 100 includes a processor 101, storage 102, memory 103, input / output interface (input / output I / F) 104, and communication interface (communication I / F) 105. Each component is interconnected via bus B. The channel estimation device 100 realizes the functions and methods described in this embodiment through the cooperation of these components. For example, when each functional part of the channel estimation device 100 is realized by software, it is realized by the processor 101 executing instructions contained in a program read from the storage 102 into the memory 103. That is, the channel estimation device 100 according to this embodiment functions as an acquisition unit 110, a time-series information processing unit 120, a generation unit 130, an estimation unit 140, and an image generation unit 150 by the processor 101 executing a program read onto the memory 103.

[0029] The processor 101 may include, for example, a central processing unit (CPU), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), a microprocessor, a processor core, a multiprocessor, an ASIC (Application-Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), etc., and may be realized by logic circuits (hardware) or dedicated circuits formed on an integrated circuit (IC (Integrated Circuit) chip, LSI (Large Scale Integration)), etc. It is preferable that the channel estimation device 100 has a processor 101 with high computing power for processing the large amount of data mentioned above.

[0030] The communication interface 105 is implemented as hardware such as a network adapter, communication software, or a combination thereof, and transmits and receives various types of data with external devices. This communication may be performed via wired or wireless connection, and any communication protocol may be used as long as communication between the devices is possible.

[0031] The input / output interface 104 includes an input device for inputting various operations to the channel estimation device 100, and an output device for outputting processing results processed by the channel estimation device 100. The input device includes, for example, hardware keys such as a touch panel, touch display, and keyboard, a pointing device such as a mouse, a camera (operation input via image), and a microphone (operation input via voice). The output device outputs the processing results processed by the processor 101. The output device includes, for example, a touch panel and a speaker.

[0032] The present invention has been described based on various drawings and embodiments, but it should be noted that those skilled in the art will find it easy to make various modifications and alterations based on this disclosure. Therefore, it should be noted that these modifications and alterations are within the scope of the present invention. For example, the functions included in each component, step, etc., can be rearranged in a logically consistent manner, and multiple components or steps, etc., can be combined into one or divided. Furthermore, the configurations shown in the above embodiments may be combined as appropriate. For example, each component described as being provided by the channel estimation device 100 may be implemented by multiple servers in a distributed manner. Also, the processing described above as being performed by the channel estimation device 100 may be performed by the base station 200.

[0033] Furthermore, although the above description used 5G mobile communication systems as an example, the present invention is not limited to this and may also be applied to mobile communication systems such as 4G (fourth generation), 6G (sixth generation), and LTE.

[0034] The programs of each embodiment of this disclosure may be provided stored in a storage medium readable by a channel estimation device. The storage medium is a “non-transient tangible medium” capable of storing the program. The program includes, for example, software programs and control programs. The storage medium may, where appropriate, include one or more semiconductor-based or other integrated circuits (ICs) (e.g., field-programmable gate arrays (FPGAs), application-specific ICs (ASICs), etc.), hard disk drives (HDDs), hybrid hard drives (HHDs), optical disks, optical disk drives (ODDs), magneto-optical disks, magneto-optical drives, floppy diskettes, floppy disk drives (FDDs), magnetic tapes, solid-state drives (SSDs), RAM drives, secure digital cards or drives, any other suitable storage medium, or two or more suitable combinations thereof. The storage medium may, where appropriate, be volatile, non-volatile, or a combination of volatile and non-volatile.

[0035] Furthermore, the program of this disclosure may be provided to the channel estimation device 100 via any transmission medium capable of transmitting the program (such as a communication network or broadcast waves).

[0036] Furthermore, each embodiment of this disclosure can also be realized in the form of data signals embedded in a carrier wave, where the program is embodied by electronic transmission. The programs of this disclosure may be implemented using, for example, scripting languages ​​such as JavaScript® and Python, C language, Go language, Swift®, Koltin®, Java®, etc.

[0037] According to each aspect of this disclosure described above, more accurate channel estimation becomes possible, thereby contributing to the achievement of Sustainable Development Goal (SDG) 9, "Build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation."

[0038] 100 Channel estimation device 110 Acquisition unit 120 Time series information processing unit 130 Generation unit 140 Estimation unit 150 Image generation unit 600 Channel estimation system

Claims

1. A channel estimation device comprising: an acquisition unit that acquires channel information indicating the communication quality between a base station and a wireless communication terminal; a time-series information processing unit that extracts a predetermined number of time-direction channel features from a plurality of time-series channel information acquired by the acquisition unit and the movement speed of the wireless communication terminal; a generation unit that generates an estimation model by machine learning that models the relationship between the predetermined number of channel features and channel information that is later in the time series than the plurality of channel information used to extract the predetermined number of channel features; and an estimation unit that inputs the predetermined number of channel features extracted by the time-series information processing unit to the estimation model based on a plurality of time-series channel information between the wireless communication terminal to be estimated and the base station, and estimates the future channel information of the wireless communication terminal to be estimated.

2. The channel estimation device according to claim 1, wherein the time-series information processing unit extracts a predetermined number of channel features for each of a plurality of wireless communication terminals with different travel speeds.

3. The channel estimation device according to claim 1, wherein the generation unit learns the relationship between a plurality of wireless communication terminals with different moving speeds and generates the estimation model.

4. The channel estimation device according to claim 1, further comprising an image generation unit that generates channel image information, which is an image of the channel information acquired by the acquisition unit for each acquisition time, wherein the channel image information is used as the channel information in the extraction of channel features, the generation of the estimation model, and the estimation of future channel information.

5. The channel estimation device according to claim 1, wherein the time-series information processing unit is an autoencoder.

6. A method for controlling a channel estimation device, which performs the following steps: acquiring channel information indicating the communication quality between a base station and a wireless communication terminal; extracting a predetermined number of time-direction channel features from a plurality of time-series channel information acquired in the acquisition step and the moving speed of the wireless communication terminal; generating an estimation model that models the relationship between the predetermined number of channel features and channel information that is later in the time series than the plurality of channel information used to extract the predetermined number of channel features using machine learning; and inputting the predetermined number of channel features extracted in the extraction step into the estimation model based on a plurality of time-series channel information between the wireless communication terminal to be estimated and the base station, in order to estimate the future channel information of the wireless communication terminal to be estimated.

7. A control program for a channel estimation device that enables the following functions: a function to acquire channel information indicating the communication quality between a base station and a wireless communication terminal; a function to extract a predetermined number of time-direction channel features from a plurality of time-series channel information acquired by the acquisition function and the movement speed of the wireless communication terminal; a function to generate an estimation model that models the relationship between the predetermined number of channel features and channel information that is later in the time series than the plurality of channel information used to extract the predetermined number of channel features using machine learning; and a function to estimate future channel information of the wireless communication terminal to be estimated by inputting the predetermined number of channel features extracted by the extraction function into the estimation model based on a plurality of time-series channel information between the wireless communication terminal to be estimated and the base station.

Images