Information processing infrastructure, program, and information processing method

Abstract

Provided is an information processing infrastructure comprising: a training data storage unit that stores training data including a plurality of pieces of user data and waveform peak data indicating a peak of a signal waveform of a composite signal generated by performing inverse Fourier transform on a plurality of subcarrier signals obtained by performing subcarrier modulation on each of the plurality of pieces of user data; and an execution unit that has a model generation function of generating, by machine learning, an estimation model for estimating, from input data including a plurality of pieces of user data, a peak of a signal waveform of a composite signal generated by performing inverse Fourier transform on a plurality of subcarrier signals obtained by performing subcarrier modulation on each of the plurality of pieces of user data included in the input data, using the plurality of pieces of training data stored in the training data storage unit as teaching data, and a RAN control function of controlling a RAN function by executing AI processing (RAN Intelligent Controller (RIC)) or the like.

Description

Information Processing Infrastructure, Program, and Information Processing Method 【0001】 The present invention relates to an information processing infrastructure, a program, and an information processing method. 【0002】 Patent Document 1 describes a technique related to digital communication involving the transmission of a wireless signal with a limited PAPR (Peak-to-Average Power Ratio). Patent Document 2 describes a technique related to wireless communication performed by wireless communication devices such as a base station and a mobile wireless device. [Prior Art Documents] [Patent Documents] [Patent Document 1] Japanese Patent Application Publication No. 2023-539310 [Patent Document 2] Japanese Patent Application Publication No. 2023-515453 【0003】One type of digital modulation used in mobile communications is OFDM (Orthogonal Frequency Division Multiplexing). OFDM is a digital modulation that is highly efficient in utilizing frequency resources and is suitable for high-speed communication. However, because the shape of the OFDM signal waveform is Gaussian in the time domain, it is difficult to amplify the OFDM signal using linear amplification. Furthermore, primary modulation (sometimes referred to as subcarrier modulation) of digital modulation standardized in mobile communications such as LTE (Long Term Evolution) and NR (New Radio) may include subcarrier modulation with a large number of bits per symbol, resulting in discontinuities in amplitude and phase. In subcarrier modulation with a large number of bits per symbol, the peak of the signal waveform becomes large with a certain probability, making signal clipping more likely. When clipping occurs, the signal waveform becomes distorted. As a result, the characteristics of the signal deteriorate significantly. Therefore, to ensure that an amplifier amplifies a signal reliably in the linear operating region where clipping does not occur, it is necessary to implement a backoff mechanism, that is, to perform amplification with a certain margin above the upper limit of the linear operating region. In particular, amplifiers installed in wireless base stations that can transmit signals with a large PAPR need to ensure a wide dynamic range, which is the range from the minimum signal amplitude to the maximum signal amplitude, by implementing a backoff mechanism to perform amplification without clipping in the high-power operating region. Conventional wireless base stations achieved a wide dynamic range of signal amplitude by installing a Doherty amplifier, which is a combination of a normal amplifier and a peak amplifier with higher gain than the normal amplifier connected in parallel. However, if the backoff mechanism of the amplifier is set to be larger than necessary, it leads to a decrease in the maximum signal amplitude of the signal amplified by the amplifier, thus reducing the power efficiency of the signal transmitted by the wireless base station. Furthermore, amplifiers such as Doherty amplifiers installed in conventional wireless base stations are very high-performance and expensive. In addition, a drawback of installing amplifiers such as Doherty amplifiers in wireless base stations is that the design requirements of the wireless base station become stricter, such as the requirements for high efficiency to reduce power consumption and the requirements for heat dissipation design. 【0004】 In the system according to this embodiment, for example, the functions of RAN (Radio Access Network) can be run on a high-performance GPU (Graphics Processing Unit) server instead of a general-purpose server, thereby allowing the surplus computing resources to be utilized for AI (Artificial Intelligence) processing. Examples of AI processing include AI processing related to RAN control (sometimes referred to as RAN control AI processing) and AI processing unrelated to RAN control (sometimes referred to as non-RAN control AI processing). 【0005】An example of AI-based RAN control processing is the RIC (RAN Intelligent Controller). The RIC is a technology that uses AI to optimize RAN wireless resources and automate RAN operations. The RIC includes Non-RT RIC and Near-RT RIC (Near-Real Time RIC). The Non-RT RIC is sometimes called Centralized RIC. The Non-RT RIC is located within the SMO (Service Management and Orchestration), which manages and orchestrates the RAN. The Non-RT RIC generates and notifies policies related to RAN control and transmits information to the Near-RT RIC. For example, a Non-RT RIC generates a trained model for RAN control by performing machine learning using data collected from the RAN, and sends it to a Near-RT RIC. A Near-RT RIC is sometimes called a Distributed RIC. Compared to a Non-RT RIC, a Near-RT RIC is located closer to the RAN nodes (RU (Radio Unit), DU (Distributed Unit), CU (Central Unit)) and performs control of the RAN nodes and resources. Compared to a Non-RT RIC, a Near-RT RIC performs processing with higher real-time capabilities. For example, a Near-RT RIC performs inference processing related to RAN control using the trained model obtained from a Non-RT RIC. RAN control AI processing is not limited to RICs. 【0006】 Non-RAN-controlled AI processing may correspond to so-called MEC (Multi-access Edge Computing) applications. Examples of non-RAN-controlled AI processing include, but are not limited to, monitoring AI execution processing that determines the situation within the imaging range of an input image, and response AI execution processing that outputs a response to an inquiry made by a user. 【0007】In the system according to this embodiment, for example, an estimation model is generated by machine learning using AI to estimate the peak of the signal waveform of a composite signal produced by subcarrier modulating multiple user data and performing an inverse Fourier transform. Using the generated estimation model, a mechanism is employed to schedule combinations of multiple user data so that the peak of the signal waveform of the composite signal becomes smaller. As a result, the PAPR of the signal transmitted by the wireless base station can be reduced, and clipping of the signal transmitted by the wireless base station can be suppressed even if the maximum signal amplitude of the dynamic range of the amplifier installed in the wireless base station is reduced. Consequently, it is possible to improve the power efficiency of the signal transmitted by the wireless base station while suppressing the degradation of the characteristics of the signal transmitted by the wireless base station, without installing a very high-performance and expensive amplifier such as a Doherty amplifier. 【0008】 According to one embodiment of the present invention, an information processing platform is provided. The information processing platform may include a learning data storage unit that stores learning data including a plurality of user data and waveform peak data indicating the peak of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the plurality of user data. The information processing platform may include an execution unit having a model generation function that uses the plurality of learning data stored in the learning data storage unit as training data to generate an estimation model by machine learning that estimates the peak of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the plurality of user data contained in the input data, from input data including a plurality of user data. 【0009】The information processing infrastructure may further include a user data acquisition unit that acquires a plurality of user data to be transmitted, and a determination unit that uses the estimation model to estimate the peaks of the signal waveforms of a composite signal generated by performing an inverse Fourier transform on a plurality of subcarrier signals obtained by subcarrier modulating each of the at least two user data from input data including the plurality of user data acquired by the user data acquisition unit, for each combination of the plurality of user data included in the input data. Based on these estimation results, the determination unit determines a schedule for allocating wireless resources to the plurality of user data. 【0010】 In any of the above-mentioned information processing infrastructures, the determination unit may determine the schedule such that the peaks of the signal waveforms of the multiple composite signals generated from the multiple user data are reduced. 【0011】 In any of the above-mentioned information processing infrastructures, the decision unit may determine the schedule within the scope of allocating the wireless resources to the plurality of user data according to predetermined allocation conditions. 【0012】 In any of the above-mentioned information processing infrastructures, the decision unit may determine the schedule within the range of allocating the wireless resources to the plurality of user data according to a proportionate fairness scheduling algorithm. 【0013】 In any of the above-mentioned information processing infrastructures, the execution unit may further have a RAN control function for controlling the functions of the RAN (Radio Access Network), and the RAN control function may control the radio base stations constituting the RAN to transmit the plurality of user data using the radio resources allocated according to the schedule determined by the decision unit. 【0014】 In any of the above-mentioned information processing infrastructures, the execution unit may further have a RAN control function for controlling the functions of the RAN. 【0015】In any of the above-mentioned information processing infrastructures, the model generation function may generate the estimation model by machine learning when the computational resources of the execution unit satisfy predetermined resource conditions. 【0016】 In any of the above-mentioned information processing infrastructures, the learning data storage unit may store the learning data which further includes modulation scheme data indicating the modulation scheme of the plurality of user data, and the model generation function may generate the estimation model from the input data which further includes modulation scheme data indicating the modulation scheme of the plurality of user data by machine learning. 【0017】 In any of the above-mentioned information processing infrastructures, the learning data storage unit may store the learning data which further includes user data count data indicating the number of user data, and the model generation function may generate the estimation model from the input data which further includes user data count data indicating the number of user data by machine learning. 【0018】 According to one embodiment of the present invention, a program is provided that, when executed by a computer, causes the computer to function as one of the aforementioned information processing infrastructures. 【0019】 According to one embodiment of the present invention, an information processing method is provided that is performed by a computer that stores training data including a plurality of user data and waveform peak data indicating the peaks of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the plurality of user data. The information processing method may include a model generation step in which, using the plurality of training data stored in the computer as training data, an estimation model is generated by machine learning to estimate the peaks of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the plurality of user data included in the input data, from input data including a plurality of user data to be transmitted. 【0020】It should be noted that the above summary of the invention does not enumerate all the necessary features of the present invention. Furthermore, subcombinations of these features may also constitute an invention. 【0021】 This diagram schematically shows an example of System 10. This is an explanatory diagram illustrating an example of a user data transmission method. This is an explanatory diagram illustrating an example of PAPR. This is an explanatory diagram illustrating an example of the processing flow of System 10. This diagram schematically shows an example of multiple user data to be transmitted by the wireless base station 300. This diagram schematically shows an example of the functional configuration of the information processing infrastructure 200. This diagram schematically shows an example of the functional configuration of the wireless base station 300. This is an explanatory diagram illustrating an example of the processing flow of System 10. This diagram schematically shows an example of the hardware configuration of a computer 1200 that functions as the management infrastructure 100, the information processing infrastructure 200, or the wireless base station 300. 【0022】 The present invention will be described below through embodiments, but these embodiments are not intended to limit the scope of the claimed invention. Furthermore, not all combinations of features described in the embodiments are necessarily essential to the solution of the invention. 【0023】 Figure 1 schematically shows an example of system 10. System 10 may include one or more wireless base stations 300 that constitute the RAN 310. System 10 may include multiple information processing infrastructures 200. System 10 may include a management infrastructure 100 that manages the multiple information processing infrastructures 200. In system 10 according to this embodiment, for example, the management infrastructure 100 and the multiple information processing infrastructures 200 may cooperate to control the RAN 310 and perform AI processing. 【0024】 RAN310 may be a virtualized vRAN (Virtual RAN), and system 10 may perform control of the vRAN. RAN310 may also be a physical RAN, and system 10 may perform control of the physical RAN. 【0025】The AI ​​processing performed by system 10 may include RAN-controlled AI processing (sometimes referred to as RAN_AI). The AI ​​processing performed by system 10 may also include non-RAN-controlled AI processing (sometimes referred to as non-RAN_AI). 【0026】 The information processing infrastructure 200 may be data centers located in various locations. The information processing infrastructure 200 may be composed of multiple devices. The information processing infrastructure 200 may be implemented on a virtualization infrastructure composed of multiple devices. The information processing infrastructure 200 may be implemented by a single device. In other words, the information processing infrastructure 200 may be an information processing device. 【0027】 The information processing infrastructure 200 includes, for example, an execution unit that includes a RAN control function for controlling the functions of the RAN 310 and an application execution function for executing applications. The information processing infrastructure 200 may further include the functions of a wireless base station 300. 【0028】 The RAN control function controls the functions of RAN310, for example, by executing RAN_AI. The RAN control function may also control the functions of RAN310 by executing any other arbitrary processing. 【0029】 The RAN control function controls, for example, the wireless base station 300. The RAN control function controls the wireless base station 300 so that, for example, it uses an antenna to form a wireless communication area and provides mobile communication services to communication terminals 30 within the wireless communication area. 【0030】 The communication terminal 30 is, for example, a mobile phone such as a smartphone. The communication terminal 30 may also be a tablet device or a PC (Personal Computer). The communication terminal 30 may also be a so-called IoT (Internet of Things) device. The communication terminal 30 may include anything that falls under the so-called IoE (Internet of Everything). 【0031】The application execution function may, for example, have the ability to execute AI applications. The application execution function may also have the ability to execute non-RAN_AI applications. The application execution function may execute any other application. 【0032】 The execution unit may contain one or more CPUs (Central Processing Units). The execution unit may contain one or more GPUs. The execution unit may contain multiple superchips, each connected to a CPU and a GPU by an interconnect. This interconnect may be memory consistent and capable of achieving high bandwidth and low latency. Thus, the execution unit may have CPU resources and GPU resources as computing resources. 【0033】 The information processing infrastructure 200 is, for example, located on the core network. The term "on the core network" includes both the area inside and outside the core network. 【0034】 The core network may conform to any mobile communication system. For example, the core network may conform to a 5G (5th Generation) communication system. The core network may conform to a 6G (6th Generation) communication system or later mobile communication systems. The core network may conform to a 3G (3rd Generation) communication system or an LTE communication system. 【0035】 The management infrastructure 100 may be a data center that manages multiple information processing infrastructures 200. The management infrastructure 100 may be composed of multiple devices. The management infrastructure 100 may be implemented on a virtualization infrastructure consisting of multiple devices. The management infrastructure 100 may be implemented by a single device. In other words, the management infrastructure 100 may be a management device. 【0036】The management infrastructure 100 may be called the Core Brain, and the information processing infrastructure 200 may be called the Regional Brain. Note that Figure 1 illustrates a case where a single-layer information processing infrastructure 200 is located below the management infrastructure 100, but this is not the only example. The information processing infrastructure 200 may have multiple layers. For example, if two layers of information processing infrastructure 200 are located below the management infrastructure 100, the management infrastructure 100 may be called the Core Brain, the lower layer of information processing infrastructure 200 may be called the Regional Brain, and the further lower layer of information processing infrastructure 200 may be called the Sub-Regional Brain. 【0037】 Figure 2 is an explanatory diagram illustrating an example of a user data transmission method. Here, we mainly explain an example where the user data transmitted by the wireless base station 300 consists of x user data items: user data 1, user data 2, ..., user data x-1, and user data x. x is an integer greater than or equal to 2. 【0038】 The upper diagram in Figure 2 is an explanatory diagram for illustrating the OFDM (Orthogonal Frequency Division Multiplexing) scheme. As shown in the upper diagram of Figure 2, in the OFDM scheme, one user data occupies all subcarriers contained within the carrier. 【0039】 The lower diagram in Figure 2 is an explanatory diagram for illustrating the OFDMA (Orthogonal Frequency Division Multiple Access) scheme. As shown in the lower diagram of Figure 2, in the OFDMA scheme, multiple user data points are shared among multiple subcarriers within a single carrier. 【0040】 For example, in subcarrier modulation of digital modulation schemes such as OFDM and OFDMA, the bitstream is mapped to signal points defined by the phase and amplitude of the subcarrier. To prevent errors from being concentrated in a single symbol, a bit-level rearrangement process, known as bit interleap, may be performed before the bitstream mapping. 【0041】For example, binary phase-shift keying (BPSK), quadrature PSK (QPSK), eight-phase shift keying (8PSK), sixteen-level quadrature amplitude modulation (16QAM), sixteen-level quadrature amplitude modulation (64QAM), and two-fifteen-level quadrature amplitude modulation (256QAM) are examples of subcarrier modulation for digital modulation schemes such as OFDM and OFDMA. BPSK has 1 bit per symbol, QPSK has 2 bits per symbol, 8PSK has 3 bits per symbol, 16QAM has 4 bits per symbol, 64QAM has 6 bits per symbol, and 256QAM has 8 bits per symbol. 【0042】 Figure 3 is an explanatory diagram illustrating an example of PAPR. P shown in Figure 3 Ａｖｅｒａｇｅ This is the average power of the signal. P is shown in Figure 3. Ｐｅａｋ P is the peak power of the signal. PAPR is the peak power of the signal relative to the average power of the signal. PAPR is P Ｐｅａｋ / P Ａｖｅｒａｇｅ It can be expressed as such. 【0043】 The PAPR of a signal transmitted externally by a wireless base station 300 using its antenna tends to increase as the user data to which the signal is transmitted is subcarrier modulated using a digital modulation scheme with a large number of bits per symbol. For example, among two such signals with the same number of user data to be transmitted, the PAPR of one signal whose user data is subcarrier modulated using the 256QAM scheme is greater than the PAPR of the other signal whose user data is subcarrier modulated using the QPSK scheme. Note that a signal transmitted externally by a wireless base station using its antenna may be referred to as a transmitted signal. 【0044】The PAPR of the transmitted signal from the wireless base station 300 tends to increase as the number of user data items transmitted by the transmitted signal increases. For example, of two transmitted signals in which the user data items to be transmitted are subcarrier modulated using the same digital modulation scheme, the PAPR of one transmitted signal in which the number of user data items to be transmitted is a (where a is a positive integer) consisting of user data 1, user data 2, ..., user data a-1, and user data a is greater than the PAPR of the other transmitted signal in which the number of user data items to be transmitted is b (where b is a positive integer satisfying b < a). 【0045】 Figure 4 is an explanatory diagram illustrating an example of the processing flow of system 10. Here, we mainly explain an example of the processing flow of system 10 when the information processing infrastructure 200 generates an estimation model for estimating the peaks of the signal waveform of a composite signal generated from multiple user data. 【0046】 In Step 1, the information processing infrastructure 200 acquires multiple user data from the wireless base station 300. Here, we will continue the explanation assuming that the information processing infrastructure 200 has acquired N user data, namely user data 1, user data 2, ..., and user data N. Note that N is an integer greater than or equal to 2. 【0047】 In Step 2, the wireless base station 300 generates a composite signal from the N user data that the wireless base station 300 transmitted to the information processing infrastructure 200 in Step 1, and transmits the generated composite signal to the outside using the antenna mounted on the wireless base station 300. Here is an example of the processing flow in which the wireless base station 300 generates a composite signal from N user data and transmits the generated composite signal to the outside as a transmission signal. 【0048】 For example, the wireless base station 300 performs serial / parallel conversion (S / P conversion) on each of the N user data. In the example shown in Figure 4, the wireless base station 300 converts user data 1 to K by performing S / P conversion. １ By performing parallel conversion to individual bitstreams and S / P conversion on user data 2, user data 2 is converted to K２ Convert it into K parallel bit streams, and perform S / P conversion on user data N to convert user data N into K Ｎ parallel bit streams. Here, K １ , K ２ ,..., and K Ｎ are positive integers. 【0049】 Next, the radio base station 300 generates a plurality of subcarrier signals by performing subcarrier modulation on each of the N parallel-converted user data. In an example shown in FIG. 4, the radio base station 300 performs subcarrier modulation on each of the K １ bit streams of the parallel-converted user data 1 to generate L １ subcarrier signals, performs subcarrier modulation on each of the K ２ bit streams of the parallel-converted user data 2 to generate L ２ subcarrier signals,..., and performs subcarrier modulation on each of the K Ｎ bit streams of the parallel-converted user data N to generate L Ｎ subcarrier signals. Therefore, in an example shown in FIG. 4, the radio base station 300 generates L １ +L ２ +...+L Ｎ subcarrier signals from N user data. Here, L １ , L ２ ,..., and L Ｎ are positive integers. 【0050】Next, the wireless base station 300 performs an inverse Fourier transform (IFT) on multiple subcarrier signals generated from N user data. The inverse Fourier transform transforms a signal waveform on the frequency axis into a signal waveform on the time axis. For example, the wireless base station 300 performs a fast inverse Fourier transform (IFT) on multiple subcarrier signals generated from N user data. The fast inverse Fourier transform may be just one example of an inverse Fourier transform. In the example shown in Figure 4, the wireless base station 300 generates L １ +L ２ +...+L Ｎ Perform a fast inverse Fourier transform on each subcarrier signal. 【0051】 Next, the radio base station 300 performs parallel / serial conversion (P / S conversion) on the multiple subcarrier signals that have been converted into signal waveforms on the time axis. The radio base station 300 may generate a composite signal of the multiple subcarrier signals by performing P / S conversion on the multiple subcarrier signals that have been converted into signal waveforms on the time axis. In the example shown in Figure 4, the radio base station 300 performs parallel / serial conversion on the multiple subcarrier signals that have been converted into signal waveforms on the time axis. １ +L ２ +...+L Ｎ By performing P / S conversion on individual subcarrier signals, L １ +L ２ +...+L Ｎ It generates a composite signal of individual subcarrier signals. 【0052】 Next, the radio base station 300 amplifies the signal amplitude of the combined signal of the multiple subcarrier signals. For example, the radio base station 300 performs a digital-to-analog conversion (D / A conversion) on the combined signal of the multiple subcarrier signals, and inputs the analog-converted combined signal of the multiple subcarrier signals into an amplifier, thereby amplifying the signal amplitude of the combined signal of the multiple subcarrier signals. In the example shown in Figure 4, the radio base station 300, L １ +L ２ +...+L ＮThe combined signal of the individual subcarrier signals is D / A converted, and the analog converted L １ +L ２ +...+L Ｎ By inputting the combined signal of these subcarrier signals into the amplifier, L １ +L ２ +...+L Ｎ The signal amplitude of the combined signal of the individual subcarrier signals is amplified. 【0053】 Subsequently, the radio base station 300 uses its onboard antenna to transmit a combined signal of multiple subcarrier signals with amplified signal amplitudes to the outside as a transmission signal. In one example shown in Figure 4, the radio base station 300 transmits a combined signal of multiple subcarrier signals with amplified signal amplitudes. １ +L ２ +...+L Ｎ The combined signal of these subcarrier signals is transmitted externally as the transmission signal. 【0054】 In Step 3, the radio base station 300 identifies the peak of the signal waveform of the composite signal generated from N user data in Step 2. For example, the radio base station 300 identifies the peak of the signal waveform of the composite signal of multiple subcarrier signals whose signal amplitudes have been amplified. In the example shown in Figure 4, the radio base station 300 identifies the peak of the signal waveform of L whose signal amplitude has been amplified. １ +L ２ +...+L Ｎ The peaks of the signal waveform of the combined signal of the individual subcarrier signals are identified. The radio base station 300 may identify the peaks of the signal waveform of the combined signal using any device capable of identifying the peaks of the signal waveform of the combined signal. 【0055】 In Step 4, the information processing infrastructure 200 acquires waveform peak data from the wireless base station 300, which shows the peak of the signal waveform of the composite signal identified by the wireless base station 300 in Step 3. The waveform peak data is, for example, the peak power (P) of the composite signal. Ｐｅａｋ This is peak power data indicating the peak of the signal waveform. The waveform peak data may be any other data that can identify the peak of the signal waveform of the composite signal. 【0056】In Step 5, the information processing infrastructure 200 stores the N user data acquired from the wireless base station 300 in Step 1 and the waveform peak data acquired from the wireless base station 300 in Step 4 as training data. In Step 6, the information processing infrastructure 200 uses the training data, including the training data stored in Step 5, as training data to generate an estimation model using machine learning to estimate the peak of the signal waveform of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals, each of which is obtained by subcarrier modulating multiple user data points contained in the input data, from input data containing multiple user data points. 【0057】 In Step 7, the information processing infrastructure 200 transmits the estimation model generated in Step 6 to the wireless base station 300. Subsequently, when the wireless base station 300 acquires multiple user data to be transmitted, the wireless base station 300 uses the estimation model acquired from the information processing infrastructure 200 in Step 7 to determine a schedule for allocating wireless resources to the multiple user data. For example, the wireless base station 300 uses the estimation model to estimate the peaks of the signal waveforms of a composite signal generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the at least two user data for each combination of the multiple user data, and based on the estimation results, determines a schedule for allocating wireless resources to the multiple user data so that the peaks of the signal waveforms of the composite signal generated from the multiple at least two user data are smaller. The process of determining a schedule for allocating wireless resources to the multiple user data using the estimation model will be described later. Note that wireless resources may include time resources and frequency resources. 【0058】Conventional wireless base stations have been designed using the peak PAPR of the transmitted signal as the basis for their design requirements, in order to prevent clipping even when the PAPR of the transmitted signal is at its peak. As a result, the design requirements for conventional wireless base stations were extremely stringent. Furthermore, even when the PAPR of the transmitted signal was small, such as when the transmitted signal was subcarrier modulated using modulation schemes with a small number of bits per symbol, such as BPSK or QPSK, or when the amount of user data included in the transmitted signal was small, conventional wireless base stations transmitted the transmitted signal in the same way as when the PAPR of the transmitted signal was large. Therefore, conventional wireless base stations sometimes wasted power when transmitting the transmitted signal. 【0059】In contrast, according to the system 10 of this embodiment, the information processing infrastructure 200 generates an estimation model by machine learning using multiple training data based on the transmission patterns when the wireless base station 300 has transmitted multiple user data to the outside in the past, and provides the generated estimation model to the wireless base station 300. The wireless base station 300 uses the estimation model provided by the information processing infrastructure 200 to estimate the peaks of the signal waveform of the composite signal generated from the multiple user data to be transmitted. Based on the estimation result of estimating the peaks of the signal waveform of the composite signal, the wireless base station 300 determines a schedule for allocating wireless resources to the multiple user data to be transmitted. By allocating wireless resources to the multiple user data to be transmitted according to the schedule determined using the estimation model, the wireless base station 300 can transmit the multiple user data to the outside with a smaller overall peak in the signal waveform of the transmitted signal compared to when wireless resources are allocated to the multiple user data to be transmitted according to a schedule determined simply using a scheduling algorithm. As a result, the wireless base station 300 can reduce the PAPR of the transmitted signal as a whole. Therefore, the system 10 of this embodiment can contribute to reducing the PAPR of the transmitted signal of the wireless base station. By reducing the PAPR of the transmitted signal from the wireless base station, the wireless base station no longer needs to perform amplification in the high-power operating range. In this case, the wireless base station can prevent clipping of the transmitted signal by using a lower-performance and less expensive amplifier compared to a very high-performance and expensive amplifier such as a Doherty amplifier. Furthermore, if the information processing infrastructure 200 generates an estimation model with high estimation accuracy, an amplifier with appropriate backoff can be installed in the wireless base station. Therefore, the system 10 according to this embodiment can contribute to both the relaxation of design requirements for wireless base stations compared to conventional wireless base stations and the improvement of power efficiency of wireless base stations compared to conventional wireless base stations. As a result, the system 10 according to this embodiment can contribute to the inexpensive realization of highly power-efficient mobile communication. 【0060】Figure 5 schematically shows an example of multiple user data to be transmitted by the wireless base station 300. Here, the number of multiple user data to be transmitted by the wireless base station 300 is N. １ +N ２ +...+N ｍ It is assumed that there are individual units. Furthermore, N １ , N ２ , ..., and N ｍ is an integer greater than or equal to 2. 【0061】 The wireless base station 300, for example, allocates wireless resources to multiple user data to be transmitted according to predetermined allocation conditions. The wireless base station 300, for example, groups the multiple user data to be transmitted based on the allocation priority for allocating wireless resources, and allocates wireless resources more preferentially to user data belonging to the group with a higher allocation priority among the multiple user data to be transmitted. 【0062】 In one example shown in Figure 5, the wireless base station 300 is N １ +N ２ +...+N ｍ The user data is grouped into m groups: the first group with the highest assignment priority, the second group with the second highest assignment priority, and so on, up to the mth group with the mth highest assignment priority. １ +N ２ +...+N ｍ N of the individual user data １ Each user data belongs to the first group, N １ +N ２ +...+N ｍ N of the individual user data ２ Each user data belongs to the second group, ..., N １ +N ２ +...+N ｍ N of the individual user data ｍEach user data belongs to the m-th group. Note that user data belonging to the first group may be referred to as the first user data, user data belonging to the second group as the second user data, ..., and user data belonging to the m-th group as the m-th user data. Also, m is an integer greater than or equal to 2. 【0063】 The wireless base station 300 determines a schedule for allocating wireless resources to multiple user data to be transmitted, within the scope of allocating wireless resources to multiple user data to be transmitted according to allocation conditions, using, for example, an estimation model provided by the information processing infrastructure 200. The wireless base station 300 determines a schedule for allocating wireless resources to multiple user data to be transmitted, for example, by using the estimation model for each set of user data belonging to the same group among the multiple user data to be transmitted. Here, using an example of the multiple user data to be transmitted by the wireless base station 300 shown in Figure 5, an example of how the wireless base station 300 determines a schedule for allocating wireless resources to multiple user data to be transmitted, within the scope of allocating wireless resources to multiple user data to be transmitted according to allocation conditions, using an estimation model, will be explained. 【0064】 For example, the wireless base station 300 uses an estimation model to determine N １ +N ２ +...+N ｍ N belonging to the first group of individual user data １ From input data containing individual first user data, N included in the input data １ For each combination of at least two first user data from the N first user data, the peak of the signal waveform of the composite signal generated from those at least two first user data is estimated. The wireless base station 300 then considers the input data as N １ Based on the estimation results when including multiple first user data, N such that the peaks of the signal waveforms of multiple composite signals generated from multiple such at least two first user data are smaller. １ Determine the schedule for allocating wireless resources to each individual first user data. 【0065】 Next, the radio base station 300 uses the estimation model to estimate, for each combination of at least two of the N second user data included in the input data among the N + N +... + N user data that belong to the second group, the peak of the signal waveform of the composite signal generated from the at least two second user data. The radio base station 300 determines a schedule for allocating radio resources to the N second user data such that the peaks of the signal waveforms of the plurality of composite signals generated from the plurality of at least two second user data become smaller based on the estimation result when the input data includes the N second user data. １ + N ２ +... + N ｍ Among the N user data, for the N second user data belonging to the second group, ２ from the input data including the N second user data, for each combination of at least two of the N second user data included in the input data, ２ the peak of the signal waveform of the composite signal generated from the at least two second user data is estimated. When the input data includes the N second user data, the radio base station 300, based on the estimation result, ２ for the N second user data, determines a schedule for allocating radio resources such that the peaks of the signal waveforms of the plurality of composite signals generated from the plurality of at least two second user data become smaller. ２ 【0066】 Similarly, the radio base station 300 uses the estimation model to sequentially determine a schedule for allocating radio resources to a plurality of user data belonging to a group with a high allocation priority among the N + N +... + N user data. Then, the radio base station 300 determines a schedule for allocating radio resources to the N second user data belonging to the m-th group among the N + N +... + N user data. By doing so, a schedule for allocating radio resources to the N + N +... + N user data is determined. １ + N ２ +... + N ｍ From the plurality of user data belonging to the group with a high allocation priority among the N + N +... + N user data, a schedule for allocating radio resources to the plurality of user data is determined in order. Then, the radio base station 300 １ + N ２ +... + N ｍ Among the N + N +... + N user data, for the N m-th user data belonging to the m-th group, ｍ by determining a schedule for allocating radio resources up to the N m-th user data, １ + N ２ +... + N ｍ a schedule for allocating radio resources to the N + N +... + N user data is determined. 【0067】According to the system 10 shown in Figure 5, the wireless base station 300 determines a schedule for allocating wireless resources to multiple user data to be transmitted, within the range of allocating wireless resources to multiple user data to be transmitted according to allocation conditions, based on the estimation results obtained by estimating the peaks of the signal waveform of the composite signal generated from multiple user data to be transmitted using an estimation model. As a result, the wireless base station 300 can reduce the overall PAPR of the transmitted signal while allocating wireless resources fairly among the user data. Consequently, the system 10 shown in Figure 5 can contribute to both the realization of fair allocation of wireless resources among user data and the inexpensive realization of highly power-efficient mobile communication. 【0068】 Figure 6 schematically shows an example of the functional configuration of the information processing infrastructure 200. The information processing infrastructure 200 includes a learning data acquisition unit 202, a learning data storage unit 204, an execution unit 206, a model storage unit 212, a model acquisition unit 214, a model transmission unit 215, a user data acquisition unit 216, an input data acquisition unit 218, a determination unit 222, and a schedule information transmission unit 224. However, it is not necessarily required that the information processing infrastructure 200 have all of these components. 【0069】 The training data acquisition unit 202 acquires training data for generating an estimation model. The training data acquisition unit 202 may store the acquired training data in the training data storage unit 204. 【0070】 The information processing infrastructure 200 acquires various types of data, for example, by receiving various types of data such as training data. The information processing infrastructure 200 receives various types of data, for example, via a core network. The information processing infrastructure 200 may also receive various types of data via the Internet. 【0071】 The information processing infrastructure 200 acquires various data from, for example, the wireless base station 300 that constitutes the RAN 310. The information processing infrastructure 200 may also acquire various data from any other device. 【0072】The training data includes, for example, multiple user data and waveform peak data showing the peaks of the signal waveform of a composite signal generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data. The training data further includes, for example, waveform average data showing the average of the signal waveform of the composite signal. The waveform average data includes, for example, the average power (P) of the composite signal. Ａｖｅｒａｇｅ This is average power data indicating the average of the signal waveform of the composite signal. The waveform average data may be any other data that can identify the average of the signal waveform of the composite signal. The training data further includes, for example, modulation scheme data indicating the modulation scheme of the plurality of user data. The training data further includes, for example, user data count data indicating the number of the plurality of user data. 【0073】 The execution unit 206 performs various processes. The execution unit 206 includes, for example, a RAN control function 207 and a model generation function 209. The RAN control function 207 and the model generation function 209 may use the same computing resources. 【0074】 The RAN control function 207 controls the functions of the RAN 310. The RAN control function 207 controls the functions of the RAN 310, for example, by executing RAN_AI. The RAN control function 207 may also control the functions of the RAN 310 by executing any other processing. 【0075】 The model generation function 209 generates an estimated model. For example, the model generation function 209 generates an estimated model using machine learning, using multiple training data stored in the training data storage unit 204 as training data. The model generation function 209 may store the generated estimated model in the model storage unit 212. The application execution function may be an example of the model generation function 209. 【0076】The estimation model is, for example, a model that estimates the peak of the signal waveform of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals, each of which is subcarrier-modulated with multiple user data. The estimation model may also be a model that estimates the PAPR of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals, each of which is subcarrier-modulated with multiple user data. 【0077】 The model generation function 209 generates an estimated model by machine learning when, for example, the computing resources of the execution unit 206 satisfy predetermined resource conditions. The model generation function 209 generates an estimated model by machine learning when, for example, the amount of available resources in the computing resources of the execution unit 206 is greater than a predetermined threshold for available resources. The model generation function 209 generates an estimated model by machine learning when, for example, the amount of available resources of the GPU resources included in the computing resources of the execution unit 206 is greater than a threshold for available resources. The model generation function 209 generates an estimated model by machine learning when, for example, the amount of available resources of the CPU resources included in the computing resources of the execution unit 206 is greater than a threshold for available resources. The model generation function 209 generates an estimated model by machine learning when, for example, the sum of the available resources of the GPU resources and CPU resources included in the computing resources of the execution unit 206 is greater than a threshold for available resources. 【0078】 The model acquisition unit 214 acquires an estimated model. The model acquisition unit 214 acquires an estimated model similar to the estimated model generated by the model generation function 209, for example. The model acquisition unit 214 may store the acquired estimated model in the model storage unit 212. 【0079】 The model acquisition unit 214 acquires an estimated model, for example, from a model generation device that generates an estimated model using machine learning. The model acquisition unit 214 may also acquire an estimated model from a wireless base station 300. 【0080】 The model transmission unit 215 transmits the estimated model stored in the model storage unit 212. The model transmission unit 215 transmits the estimated model to the wireless base station 300, for example. 【0081】 The information processing infrastructure 200 transmits various types of information, such as estimation models, via a core network, for example. The information processing infrastructure 200 may also transmit various types of information via the Internet. 【0082】 The user data acquisition unit 216 acquires multiple user data to be transmitted. For example, the user data acquisition unit 216 acquires multiple user data to be transmitted from the wireless base station 300. 【0083】 The input data acquisition unit 218 acquires input data to be input to the estimation model stored in the model storage unit 212. The input data includes, for example, a plurality of user data acquired by the user data acquisition unit 216. The input data further includes, for example, modulation scheme data indicating the modulation scheme of the plurality of user data. The input data further includes, for example, user data count data indicating the number of the plurality of user data. 【0084】 The decision unit 222 determines a schedule for allocating wireless resources to multiple user data acquired by the user data acquisition unit 216. The decision unit 222 determines this schedule, for example, using an estimated model stored in the model storage unit 212. 【0085】 The decision unit 222 determines the schedule based on estimation results obtained by using an estimation model to estimate the peaks of the signal waveforms of a composite signal generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the at least two user data points included in the input data acquired by the input data acquisition unit 218. The decision unit 222 determines the schedule such that the peaks of the signal waveforms of the multiple composite signals generated from the multiple at least two user data points become smaller. 【0086】For example, if the user data acquisition unit 216 acquires multiple user data, namely user data 1, user data 2, user data 3, and user data 4, and a composite signal is generated from two user data, the determination unit 222 uses an estimation model to estimate the following peaks in the signal waveform of the composite signal generated from user data 1 and user data 2, the peaks in the signal waveform of the composite signal generated from user data 1 and user data 3, the peaks in the signal waveform of the composite signal generated from user data 1 and user data 4, the peaks in the signal waveform of the composite signal generated from user data 2 and user data 3, the peaks in the signal waveform of the composite signal generated from user data 2 and user data 4, and the peaks in the signal waveform of the composite signal generated from user data 3 and user data 4. Next, the determination unit 222 estimates the peak of the signal waveform of the composite signal for each combination of six user data, and based on the estimation results, it determines the pattern with the smallest peak of the signal waveform of the composite signal from among three patterns: a pattern in which a composite signal is generated from user data 1 and user data 2 and also from user data 3 and user data 4; a pattern in which a composite signal is generated from user data 1 and user data 3 and also from user data 2 and user data 4; and a pattern in which a composite signal is generated from user data 1 and user data 4 and also from user data 2 and user data 3. After that, the determination unit 222 determines the schedule so as to allocate wireless resources to user data 1, user data 2, user data 3, and user data 4 using the pattern with the smallest peak of the signal waveform of the composite signal. 【0087】The decision unit 222 determines the schedule based on estimation results obtained by using an estimation model to estimate the PAPR of a composite signal generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the at least two user data points included in the input data acquired by the input data acquisition unit 218. The decision unit 222 determines the schedule such that the PAPR of the multiple composite signals generated from the multiple at least two user data points becomes smaller. 【0088】 The decision unit 222 determines the schedule within the range of allocating wireless resources to a plurality of user data acquired by the user data acquisition unit 216 according to predetermined allocation conditions, for example, using an estimation model. The decision unit 222 groups the plurality of user data based on the allocation priority for allocating wireless resources, for example, using an estimation model, and determines the schedule within the range of allocating wireless resources more preferentially to user data belonging to the group with a higher allocation priority among the plurality of user data. The decision unit 222 determines the schedule within the range of allocating wireless resources more preferentially to user data belonging to the group with a higher allocation priority among the plurality of user data, based on estimation results obtained by estimating the peak of the signal waveform of the composite signal generated from at least two user data belonging to the same group for each combination of at least two user data belonging to the same group among the plurality of user data, for example, using an estimation model. The decision unit 222, for example, uses an estimation model to estimate the PAPR of the composite signal generated from at least two user data belonging to the same group for each combination of at least two user data belonging to the same group. Based on the estimation results, the decision unit determines the schedule within the range in which it allocates wireless resources more preferentially to user data belonging to the group with a higher allocation priority among the plurality of user data. 【0089】The determination unit 222 determines the allocation priority of each of the multiple user data, for example, based on the time during which wireless resources are not allocated. The determination unit 222 determines the allocation priority of each of the multiple user data, for example, such that the allocation priority of one user data for which the time during which wireless resources are not allocated is 1 hour is higher than the allocation priority of other user data for which the time during which wireless resources are not allocated is 2 hours, which is shorter than 1 hour. The determination unit 222 determines the allocation priority of each of the multiple user data, for example, based on QoS (Quality of Service). The determination unit 222 determines the allocation priority of each of the multiple user data, for example, such that the allocation priority of one user data with a QoS at a 1st QoS level is higher than the allocation priority of other user data with a QoS at a 2nd QoS level, which is lower than 1st QoS level. The determination unit 222 determines the allocation priority of each of the multiple user data, for example, based on both the time during which wireless resources are not allocated and QoS. 【0090】 The determination unit 222 determines the schedule within the range of allocating wireless resources to the plurality of user data according to, for example, a Proportional Fairness scheduling algorithm. The determination unit 222 may also determine the schedule within the range of allocating wireless resources to the plurality of user data according to any other scheduling algorithm. 【0091】 The RAN control function 207 controls the radio base station 300 that constitutes the RAN 310 to transmit multiple user data to be transmitted using radio resources allocated according to a schedule determined by the determination unit 222, for example. The radio base station 300 may transmit multiple user data to be transmitted in accordance with the control by the RAN control function 207. 【0092】 The schedule information transmission unit 224 transmits schedule information indicating the schedule determined by the determination unit 222. For example, the schedule information transmission unit 224 transmits the schedule information to the wireless base station 300. 【0093】 Figure 7 schematically shows an example of the functional configuration of a wireless base station 300. The wireless base station 300 includes a user data acquisition unit 302, a signal generation unit 303, a synthesized signal transmission unit 308, a specification unit 312, a learning data acquisition unit 314, a learning data storage unit 316, a learning data transmission unit 318, a model generation unit 322, a model storage unit 324, a model acquisition unit 326, an input data acquisition unit 328, a determination unit 332, and a schedule information acquisition unit 334. However, it is not necessarily required that the wireless base station 300 have all of these configurations. 【0094】 The user data acquisition unit 302 acquires user data to be transmitted. For example, the user data acquisition unit 302 acquires user data from one or more communication terminals 30 as user data to be transmitted. 【0095】 The signal generation unit 303 generates various signals. The signal generation unit 303 includes, for example, a subcarrier signal generation unit 304 and a composite signal generation unit 306. 【0096】 The subcarrier signal generation unit 304 generates a subcarrier signal. The subcarrier signal generation unit 304 generates a subcarrier signal from, for example, user data acquired by the user data acquisition unit 302. The subcarrier signal generation unit 304 generates a subcarrier signal by, for example, performing S / P conversion on the user data and subcarrier modulation on the parallel-converted user data. 【0097】 The combined signal generation unit 306 generates a combined signal. The combined signal generation unit 306 generates a combined signal by, for example, performing an inverse Fourier transform on multiple subcarrier signals generated by the subcarrier signal generation unit 304, and then performing a P / S transform on the multiple subcarrier signals that have been waveform-converted into signal waveforms on the time axis. The combined signal generation unit 306 generates a combined signal by, for example, performing a fast inverse Fourier transform on multiple subcarrier signals generated by the subcarrier signal generation unit 304, and then performing a P / S transform on the multiple subcarrier signals that have been waveform-converted into signal waveforms on the time axis. 【0098】The combined signal transmission unit 308 transmits the combined signal generated by the signal generation unit 303 to the outside using an antenna mounted on the wireless base station 300. The combined signal transmission unit 308 transmits the combined signal to the outside using the antenna by, for example, D / A conversion of the combined signal and amplifying the signal amplitude of the analog-converted combined signal using an amplifier mounted on the wireless base station 300. 【0099】 The identification unit 312 identifies the characteristics of the composite signal. For example, the identification unit 312 identifies the characteristics of the composite signal transmitted externally by the composite signal transmission unit 308. 【0100】 The identification unit 312 identifies, for example, the peak of the signal waveform of the composite signal. The identification unit 312 also identifies, for example, the average of the signal waveform of the composite signal. 【0101】 The training data acquisition unit 314 acquires training data for generating an estimation model. The training data acquisition unit 314 may store the acquired training data in the training data storage unit 316. 【0102】 The learning data includes, for example, a plurality of user data acquired by the user data acquisition unit 302 and waveform peak data indicating the peaks of the signal waveforms of the composite signal generated from the plurality of user data, as identified by the identification unit 312. The learning data further includes, for example, waveform average data indicating the average of the signal waveforms of the composite signal generated from the plurality of user data, as identified by the identification unit 312. The learning data further includes, for example, modulation scheme data indicating the modulation scheme of the plurality of user data. The learning data further includes, for example, user data count data indicating the number of the plurality of user data. 【0103】 The learning data transmission unit 318 transmits the learning data stored in the learning data storage unit 316. The learning data transmission unit 318 transmits the learning data to the information processing infrastructure 200. 【0104】 The wireless base station 300 transmits various types of data, such as training data, via, for example, the core network. The wireless base station 300 may also transmit various types of data via the internet. 【0105】The model generation unit 322 generates an estimation model. For example, the model generation unit 322 generates an estimation model by machine learning using multiple training data stored in the training data storage unit 316 as training data. The model generation unit 322 may generate an estimation model similar to the estimation model generated by the model generation function 209. The model generation unit 322 may store the generated estimation model in the model storage unit 324. 【0106】 The model acquisition unit 326 acquires an estimated model. The model acquisition unit 326 acquires an estimated model similar to the estimated model generated by the model generation function 209, for example. The model acquisition unit 326 may store the acquired estimated model in the model storage unit 324. 【0107】 The model acquisition unit 326 acquires an estimated model from, for example, the information processing infrastructure 200. The model acquisition unit 326 may also acquire an estimated model from a model generation device that generates estimated models using machine learning. 【0108】 The input data acquisition unit 328 acquires input data to be input to the estimation model stored in the model acquisition unit 326. The input data includes, for example, a plurality of user data acquired by the user data acquisition unit 302. The input data further includes, for example, modulation scheme data indicating the modulation scheme of the plurality of user data. The input data further includes, for example, user data count data indicating the number of the plurality of user data. 【0109】 The decision unit 332 determines a schedule for allocating wireless resources to a plurality of user data acquired by the user data acquisition unit 302. The decision unit 332 determines the schedule using, for example, an estimation model stored in the model storage unit 324. The decision unit 332 determines the schedule based on estimation results obtained by using the estimation model to estimate the peaks of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the at least two user data from the input data acquired by the input data acquisition unit 328. 【0110】 The decision unit 332 may determine a schedule for allocating wireless resources to multiple user data in the same manner as the decision unit 222 determines a schedule for allocating wireless resources to multiple user data. The decision unit 332 may have the same functions as the decision unit 222. 【0111】 The schedule information acquisition unit 334 acquires schedule information. For example, the schedule information acquisition unit 334 acquires schedule information from the information processing infrastructure 200. 【0112】 The wireless base station 300 acquires various information, for example, by receiving various information such as schedule information. The wireless base station 300 receives various information, for example, via the core network. The wireless base station 300 may also receive various information via the internet. 【0113】 The signal generation unit 303 generates various signals by, for example, allocating wireless resources to a plurality of user data acquired by the user data acquisition unit 302 according to a schedule determined by the determination unit 332. The signal generation unit 303 may also generate various signals by allocating wireless resources to a plurality of user data acquired by the user data acquisition unit 302 according to a schedule indicated by the schedule information acquired by the schedule information acquisition unit 334. 【0114】 Figure 8 is an explanatory diagram illustrating an example of the processing flow of system 10. Here, the starting state is defined as the state in which the information processing infrastructure 200 has not yet generated an estimation model. 【0115】 In step 102 (steps may be abbreviated as S), the user data acquisition unit 302 acquires user data from each of the multiple communication terminals 30. In S104, the signal generation unit 303 generates a composite signal from the multiple user data acquired by the user data acquisition unit 302 in S102. The composite signal transmission unit 308 transmits the composite signal generated by the signal generation unit 303 to the outside. 【0116】In S106, the identification unit 312 identifies the peak of the signal waveform of the composite signal generated by the signal generation unit 303 in S104 and transmitted to the outside by the composite signal transmission unit 308. The learning data acquisition unit 314 acquires learning data that includes multiple user data acquired by the user data acquisition unit 302 in S102 and waveform peak data indicating the peak of the signal waveform of the composite signal identified by the identification unit 312 in S106. 【0117】 In S108, the learning data transmission unit 318 transmits the learning data acquired by the learning data acquisition unit 314 in S106 to the information processing infrastructure 200. The learning data acquisition unit 202 acquires the learning data transmitted by the wireless base station 300 and stores it in the learning data storage unit 204. 【0118】 In S110, the model generation function 209 generates an estimated model by machine learning, using multiple training data, including the training data stored in the training data storage unit 204 in S108, as training data. In S112, the model transmission unit 215 transmits the estimated model generated by the model generation function 209 in S110 to the wireless base station 300. 【0119】 In S114, the user data acquisition unit 302 acquires user data from each of the multiple communication terminals 30. In S116, the determination unit 332 uses the estimated model acquired by the model acquisition unit 326 from the information processing infrastructure 200 in S112 to determine a schedule for allocating wireless resources to the multiple user data acquired by the user data acquisition unit 302 in S114. In S118, the signal generation unit 303 generates a composite signal from the multiple user data by allocating wireless resources to the multiple user data according to the schedule determined by the determination unit 332 in S116. The composite signal transmission unit 308 transmits the composite signal generated by the signal generation unit 303 to the outside. 【0120】Figure 9 schematically shows an example of the hardware configuration of a computer 1200 that functions as a management infrastructure 100, an information processing infrastructure 200, or a wireless base station 300. A program installed on the computer 1200 can cause the computer 1200 to function as one or more "parts" of the apparatus according to this embodiment, or to cause the computer 1200 to execute operations associated with the apparatus according to this embodiment or such one or more "parts", and / or to cause the computer 1200 to execute a process or a stage of such process according to this embodiment. Such a program may be executed by the CPU 1212 to cause the computer 1200 to execute specific operations associated with some or all of the blocks in the flowcharts and block diagrams described herein. 【0121】 The computer 1200 according to this embodiment includes a CPU 1212, RAM 1214, and a graphics controller 1216, which are interconnected by a host controller 1210. The computer 1200 also includes input / output units such as a communication interface 1222, a storage device 1224, a DVD drive 1226, and an IC card drive, which are connected to the host controller 1210 via an input / output controller 1220. The DVD drive 1226 may be a DVD-ROM drive and a DVD-RAM drive, etc. The storage device 1224 may be a hard disk drive and a solid-state drive, etc. The computer 1200 also includes legacy input / output units such as a ROM 1230 and a keyboard, which are connected to the input / output controller 1220 via an input / output chip 1240. 【0122】 The CPU 1212 operates according to the programs stored in the ROM 1230 and RAM 1214, thereby controlling each unit. The graphics controller 1216 acquires the image data generated by the CPU 1212 and stores it in the frame buffer provided in the RAM 1214 or within itself, so that the image data is displayed on the display device 1218. 【0123】 The communication interface 1222 communicates with other electronic devices via a network. The storage device 1224 stores programs and data used by the CPU 1212 in the computer 1200. The DVD drive 1226 reads programs or data from the DVD-ROM 1227, etc., and provides them to the storage device 1224. The IC card drive reads programs and data from the IC card and / or writes programs and data to the IC card. 【0124】 The ROM 1230 stores boot programs and / or hardware-dependent programs of the computer 1200, which are executed by the computer 1200 when activated. The input / output chip 1240 may also connect various input / output units to the input / output controller 1220 via USB ports, parallel ports, serial ports, keyboard ports, mouse ports, etc. 【0125】 The program is provided on a computer-readable storage medium such as a DVD-ROM 1227 or an IC card. The program is read from the computer-readable storage medium and installed on a storage device 1224, RAM 1214, or ROM 1230, which are examples of computer-readable storage media, and executed by the CPU 1212. The information processing described within these programs is read by the computer 1200, resulting in coordination between the program and the various types of hardware resources described above. The apparatus or method may be configured to realize the operation or processing of information in accordance with the use of the computer 1200. 【0126】For example, when communication is performed between a computer 1200 and an external device, the CPU 1212 may execute a communication program loaded into the RAM 1214 and, based on the processing described in the communication program, instruct the communication interface 1222 to perform communication processing. Under the control of the CPU 1212, the communication interface 1222 reads transmission data stored in a transmission buffer area provided in a recording medium such as the RAM 1214, storage device 1224, DVD-ROM 1227, or IC card, transmits the read transmission data to the network, or writes received data received from the network to a reception buffer area or the like provided on the recording medium. 【0127】 Furthermore, the CPU 1212 may read all or necessary parts of a file or database stored on an external recording medium such as a storage device 1224, a DVD drive 1226 (DVD-ROM 1227), or an IC card into the RAM 1214, and perform various types of processing on the data in the RAM 1214. The CPU 1212 may then write the processed data back to the external recording medium. 【0128】 Various types of information, such as various types of programs, data, tables, and databases, may be stored on the recording medium and subjected to information processing. The CPU 1212 may perform various types of processing on the data read from the RAM 1214, including various types of operations, information processing, conditional judgments, conditional branching, unconditional branching, information retrieval / replacement, etc., as described throughout this disclosure and specified by the program instruction sequence, and write the results back to the RAM 1214. The CPU 1212 may also retrieve information in files, databases, etc., within the recording medium. For example, if a plurality of entries are stored in the recording medium, each having an attribute value of a first attribute associated with an attribute value of a second attribute, the CPU 1212 may search among the plurality of entries for an entry that matches the specified condition for the attribute value of the first attribute, read the attribute value of the second attribute stored in that entry, and thereby obtain the attribute value of the second attribute associated with the first attribute that satisfies a predetermined condition. 【0129】 The program or software module described above may be stored on or near the computer 1200 in a computer-readable storage medium. Alternatively, a recording medium such as a hard disk or RAM provided within a server system connected to a dedicated communication network or the Internet can be used as a computer-readable storage medium, thereby providing the program to the computer 1200 via the network. 【0130】 In this embodiment, blocks in the flowchart and block diagram may represent a stage in a process in which an operation is performed or a "part" of a device that has the role of performing an operation. A particular stage and "part" may be implemented by a dedicated circuit, a programmable circuit supplied with computer-readable instructions stored on a computer-readable storage medium, and / or a processor supplied with computer-readable instructions stored on a computer-readable storage medium. The dedicated circuit may include digital and / or analog hardware circuits, and may include integrated circuits (ICs) and / or discrete circuits. The programmable circuit may include reconfigurable hardware circuits, such as field-programmable gate arrays (FPGAs) and programmable logic arrays (PLAs), which include logical AND, logical OR, exclusive OR, negated AND, negated OR, and other logical operations, flip-flops, registers, and memory elements. 【0131】Computer-readable media may include any tangible device capable of storing instructions to be executed by a suitable device, and as a result, computer-readable media having instructions stored therein will comprise a product that includes instructions that can be executed to create means for performing operations specified in a flowchart or block diagram. Examples of computer-readable media may include electronic storage media, magnetic storage media, optical storage media, electromagnetic storage media, semiconductor storage media, etc. More specific examples of computer-readable media may include floppy disks (registered trademark), diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), electrically erasable programmable read-only memory (EEPROM), static random access memory (SRAM), compact disk read-only memory (CD-ROM), digital multipurpose disc (DVD), Blu-ray (registered trademark) disc, memory stick, integrated circuit card, etc. 【0132】 Computer-readable instructions may include assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk®, Java®, C++, and conventional procedural programming languages ​​such as the C programming language or similar programming languages. 【0133】Computer-readable instructions are provided locally or via a wide area network (WAN) such as a local area network (LAN) or the internet to the processor or programmable circuit of a programmable data processing device such as a computer, and may be executed to create means for performing operations specified in a flowchart or block diagram. Here, the computer may be a PC (personal computer), tablet computer, smartphone, workstation, server computer, general-purpose computer, or special-purpose computer, and may also be a computer system in which multiple computers are connected. Such a computer system in which multiple computers are connected is also called a distributed computing system and is a computer in a broad sense. In a distributed computing system, multiple computers execute a program by having each computer execute a part of the program and by passing data during program execution between computers as needed. 【0134】 Examples of processors include computer processors, central processing units (CPUs), processing units, microprocessors, digital signal processors, controllers, and microcontrollers. A computer may have one or more processors. In a multiprocessor system with multiple processors, each processor executes a portion of the program, and the processors collectively execute the program by passing program execution data between them as needed. For example, in the execution of multitasks, each of the multiple processors may execute a portion of each task in small chunks by switching tasks at each time slice. In this case, which part of a program each processor executes changes dynamically. Which part of a program each of the multiple processors executes may also be statically determined by multiprocessor-aware programming. 【0135】This invention can contribute to the low-cost realization of highly power-efficient mobile communications, and therefore can contribute to achieving Sustainable Development Goal (SDG) 7, "Affordable and Clean Energy," and Goal 9, "Industry, Innovation, and Infrastructure." 【0136】 Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be apparent to those skilled in the art that various modifications or improvements can be made to the above embodiments. It will be clear from the claims that such modified or improved forms may also be included in the technical scope of the present invention. 【0137】 It should be noted that the execution order of operations, procedures, steps, and stages in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not explicitly stated as "before" or "prior to," and that these can be performed in any order unless the output of a previous operation is used in a later operation. Even if the operation flow in the claims, specifications, and drawings is described using phrases such as "first," and "next," for convenience, this does not mean that it is mandatory to perform the operations in that order. 【0138】10 System, 30 Communication terminal, 100 Management infrastructure, 200 Information processing infrastructure, 202 Learning data acquisition unit, 204 Learning data storage unit, 206 Execution unit, 207 RAN control function, 209 Model generation function, 212 Model storage unit, 214 Model acquisition unit, 215 Model transmission unit, 216 User data acquisition unit, 218 Input data acquisition unit, 222 Determination unit, 224 Schedule information transmission unit, 300 Wireless base station, 302 User data acquisition unit, 303 Signal generation unit, 304 Subcarrier signal generation unit, 306 Composite signal generation unit, 308 Composite signal transmission unit, 310 RAN, 312 Identification unit, 314 Learning data acquisition unit, 316 Learning data storage unit, 318 Learning data transmission unit, 322 Model generation unit, 324 Model storage unit, 326 Model acquisition unit, 328 Input data acquisition unit, 332 Determination unit, 334 Schedule information acquisition unit, 1200 Computer, 1210 Host controller, 1212 CPU, 1214 RAM, 1216 Graphics controller, 1218 Display device, 1220 Input / output controller, 1222 Communication interface, 1224 Storage device, 1226 DVD drive, 1227 DVD-ROM, 1230 ROM, 1240 Input / output chip

Claims

A learning data storage unit stores learning data that includes multiple user data and waveform peak data that shows the peak of the signal waveform of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data, An execution unit having a model generation function that uses the learning data storage unit to use the learning data stored in the learning data storage unit as training data to generate an estimation model that estimates the peak of the signal waveform of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data contained in the input data, using the learning data stored in the learning data storage unit as training data, and using machine learning to generate an estimation model that estimates the peak of the signal waveform of a composite signal generated by performing an inverse Fourier transform on multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data contained in the input data, which are all part of an input data including multiple user data. An information processing infrastructure equipped with the necessary components. 　 A user data acquisition unit that acquires multiple user data to be transmitted, A determination unit determines a schedule for allocating wireless resources to the plurality of user data based on the estimation result obtained by using the estimation model described above, and estimating the peak of the signal waveform of a composite signal generated by inverse Fourier transforming a plurality of subcarrier signals obtained by subcarrier modulating each of the at least two user data from the input data including the plurality of user data acquired by the user data acquisition unit, for each combination of at least two user data included in the input data, and The information processing infrastructure according to claim 1, further comprising: 　 The information processing platform according to claim 2, wherein the determination unit determines the schedule such that the peaks of the signal waveforms of the multiple composite signals generated from the multiple at least two user data are smaller. 　 The information processing platform according to claim 2 or 3, wherein the determination unit determines the schedule within the range of allocating the wireless resources to the plurality of user data according to predetermined allocation conditions. 　 The information processing infrastructure according to claim 4, wherein the determination unit determines the schedule within the range of allocating the wireless resources to the plurality of user data according to a proportionate fairness scheduling algorithm. 　 The execution unit further has a RAN control function that controls the functions of RAN (Radio Access Network), The RAN control function controls the radio base stations constituting the RAN to transmit the plurality of user data using the radio resources allocated according to the schedule determined by the determination unit. The information processing infrastructure according to any one of claims 2 to 5. 　 The information processing infrastructure according to any one of claims 1 to 5, wherein the execution unit further has a RAN control function for controlling the functions of RAN. 　 The information processing infrastructure according to claim 7, wherein the model generation function generates the estimation model by machine learning when the computing resources of the execution unit satisfy predetermined resource conditions. 　 The learning data storage unit stores the learning data, which further includes modulation scheme data indicating the modulation scheme of the plurality of user data. The aforementioned model generation function generates the estimation model by machine learning from the input data, which further includes modulation scheme data indicating the modulation scheme of the multiple user data. An information processing platform according to any one of claims 1 to 8. 　 The learning data storage unit stores the learning data which further includes user data count data indicating the number of such user data, The aforementioned model generation function generates the estimation model by machine learning from the input data, which further includes user data count data indicating the number of such user data. An information processing platform according to any one of claims 1 to 9. 　 A program that, when executed by a computer, causes the computer to function as an information processing platform according to any one of claims 1 to 10. 　 An information processing method performed by a computer that stores training data including multiple user data and waveform peak data indicating the peaks of the signal waveform of a composite signal generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data, Model generation step: Using the multiple training data stored in the computer as training data, a model is generated by machine learning to estimate the peak of the signal waveform of a composite signal, which is generated by inverse Fourier transforming multiple subcarrier signals obtained by subcarrier modulating each of the multiple user data included in the input data, from input data including multiple user data to be transmitted. An information processing method comprising:

Images