JP2024008684A - Communication device, control method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data collection method and a usage method of learning data to implement machine learning to estimate optimal MCS information.

SOLUTION: In order to estimate an MCS, some or all of an SNR, a measured PER, and a target PER of the moving speed of a communication device input as input data to a learned model, and information on the MCS calculated by the learned model is notified to other communication devices.

SELECTED DRAWING: Figure 4

COPYRIGHT: (C)2024,JPO&INPIT

Description

The present invention relates to a communication device compliant with the IEEE802.11 standard.

The IEEE 802.11 series standard is known as a communication standard related to wireless LAN (Wireless Local Area Network, hereinafter referred to as WLAN). The latest standard, IEEE802.11be, uses Multi-Link technology to achieve low-latency communication in addition to high peak throughput (Patent Document 1).

JP 2018-50133 Publication

In the successor standard to the IEEE 802.11 standard, the introduction of AI (Artificial Intelligence) and ML (Machine Learning) is being considered.

On the other hand, in the IEEE802.11 series standard, a mechanism called Link Adaptation is used to determine MCS (Modulation and Coding Scheme) information to be applied to a communication frame. In Link Adaptation, the sender transmits a communication frame including an MCS request and requests the receiver for MCS information to be applied. The receiver determines MCS information from the communication quality at the time of receiving the communication frame including the MCS request, and transmits the communication frame including the MCS response storing the MCS information to the sender. The sender that receives the MCS response applies the MCS information to communication frames transmitted thereafter. Furthermore, even without an MCS request from the sender, the receiver can independently return an MCS response depending on the communication quality of the communication frame received from the sender.

It is conceivable to use machine learning to optimize MCS information applied to communication frames, but conventionally, there are data collection methods and learning data usage methods to realize machine learning to estimate MCS information. I didn't.

In view of the above problems, an object of the present invention is to provide a data collection method and a learning data usage method for realizing machine learning in order to estimate optimal MCS information.

In view of the above problems, the communication device of the present invention applies some or all of the SNR, measured PER, target PER, and moving speed of the communication device to a learned model in order to estimate MCS (Modulation and Coding Scheme). an input means for inputting data as input data; a calculation means for calculating the MCS using the trained model by inputting the data through the input means; and notifying another communication device of information on the MCS calculated by the calculation means. and a notification means to do so.

According to the present invention, it is possible to provide a data collection method and a learning data usage method for realizing machine learning in order to estimate optimal MCS information.

It is a diagram showing an example of a network configuration. It is a diagram showing an example of the hardware configuration of an AP/STA. It is a functional block diagram including AP/STA. FIG. 2 is a conceptual diagram of a structure using a trained model consisting of input data, a trained model, and output data. FIG. 3 is a diagram showing the flow of the system in the present invention. FIG. 3 is a diagram showing an example of Control Information of an MCS request/response included in a communication frame in the present invention. It is a figure which shows the flowchart in AP or STA of this invention.

(Embodiment 1)
FIG. 1 shows an example of a network configuration according to the first embodiment. The wireless communication system in FIG. 1 is a wireless network including an AP 101 and an STA 102. AP is also a form of STA because it has the same functions as STA except that it has a relay function.

The AP 101 communicates with each STA 102 according to the wireless communication method of the IEEE802.11 standard. STAs within the circle 100 indicating the reachable range of the signal transmitted by the AP 102 can communicate with the AP 101. In this embodiment, the AP 101 and each STA 102 communicate according to the IEEE802.11 standard. The AP 101 establishes wireless links 103 and 104 with each STA 102 via a predetermined association process or the like. Note that the number of wireless links may be one or three or more. Further, the number of STAs and APs may be two or more.

FIG. 2 shows the hardware configuration of the AP/STA in the present invention. An example of the hardware configuration includes a storage section 201, a control section 202, a functional section 203, a calculation section 204, an input section 205, an output section 206, a communication section 207, and an antenna 208.

The storage unit 201 is constituted by a memory such as a ROM or a RAM, and stores various information such as programs for performing various operations described later and communication parameters for wireless communication. In addition to memories such as ROM and RAM, the storage unit 201 may include storage media such as flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, CD-Rs, magnetic tapes, nonvolatile memory cards, and DVDs. may also be used. Further, the storage unit 201 may include a plurality of memories or the like.

The control unit 202 includes, for example, a processor such as a CPU or an MPU, an ASIC (Application Specific Integrated Circuit), a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), and the like. Here, CPU is an acronym for Central Processing Unit, and MPU is an acronym for Micro Processing Unit. The AP is controlled by executing a program stored in the storage unit 201. Note that the control unit 202 may control the AP through cooperation between a program stored in the storage unit 201 and an OS (Operating System). Further, the control unit 202 may be made up of a plurality of processors such as multi-core processors, and may control the AP.

Further, the control unit 202 controls the functional unit 203 to execute predetermined processing such as AP function, imaging, printing, and projection. The functional unit 203 is hardware for the AP to execute predetermined processing.

The calculation unit 204 includes, for example, a processor such as a GPU or a TPU, an ASIC (application specific integrated circuit), a DSP (digital signal processor), an FPGA (field programmable gate array), or the like.

The calculation unit 204 operates as hardware for performing estimation calculations using machine learning results and calculating machine learning itself. Here, GPU is an acronym for Graphical Processing Unit, and TPU is an acronym for Tensor Processing Unit. TPU is an example of a systolic array type hardware processor specialized for machine learning, and the calculation resources include a product-accumulator, a buffer register installed adjacent to the product-accumulator, and an active processor implemented in hardware. It has a conversion function. It also has an instruction decoder that interprets TPU instructions for instructing the flow of calculations and controls the above-mentioned calculation resources. This TPU functions as a so-called neural processing unit (NPU). Although the example in FIG. 1 shows an example in which the AP 101 and the STA 102 themselves perform machine learning, a data collection server and an estimation server used for machine learning may be prepared separately from the AP 101 and the STA 102.

Since these processors perform calculations together with the control unit 202, some calculations may be shared. GPUs and TPUs can perform efficient calculations by processing more data in parallel, so when learning multiple times using a learning model such as deep learning, processing is performed on GPUs and TPUs. This is effective. Therefore, in this embodiment, in addition to the control unit 202, a GPU or TPU is used for the calculation unit 204 for processing by the learning unit of the estimation server. Specifically, when a learning program including a learning model is executed, learning is performed by the control unit 202 or calculation unit 204 working together to perform calculations. Note that the processing of the learning section may be performed only by the control section 202 or the calculation section 204. Further, the estimation section may also use the calculation section 204 similarly to the learning section.

The input unit 205 accepts various operations from the user. The output unit 206 performs various outputs to the user. Here, the output from the output unit 206 includes at least one of display on the screen, audio output from a speaker, vibration output, and the like. Note that, like a touch panel, both the input section 205 and the output section 206 may be implemented in one module.

The communication unit 207 is configured to be able to perform wireless communication in accordance with the successor standard to the IEEE 802.11 EHT standard (also referred to as the 802.11be standard), which aims for a maximum transmission speed of over 90 Gbps-100 Gbps. . This successor standard to 802.11be sets out new goals to achieve, such as support for highly reliable communications and low-latency communications. Based on the above, in this embodiment, the successor standard of IEEE802.11be, which aims for a maximum transmission speed of over 90 Gbps to 100 Gbps, is tentatively named IEEE802.11HR (High Reliability).

Note that the name IEEE802.11HR was established for convenience based on the goals to be achieved by the successor standard and the main features of the standard, and it may be given a different name once the standard is finalized. On the other hand, it should be noted that this specification and the appended claims are essentially successor standards to the 802.11be standard and are applicable to any successor standard that may support wireless communications. .

The communication unit 207 performs processing of encoding, decoding, and modulation/demodulation of wireless communication data in accordance with the IEEE 902.11 standard series such as the IEEE 802.11 EHT standard and the IEEE 802.11 HR standard. Further, the communication unit 207 controls wireless communication based on Wi-Fi and IP (Internet Protocol) communication. Furthermore, the communication unit 207 controls the antenna 208 to transmit and receive wireless signals for wireless communication.

Furthermore, when a data collection server and an estimation server used for machine learning are prepared separately from the AP 101 and the STA 102, the servers are configured with a so-called Neumann type computer. More specifically, the server includes one or more memories and one or more processors that correspond to the control unit 204, and calculation resources such as a GPU and a TPU that correspond to the calculation unit 204. In this case, the GPU and TPU of the server operate as hardware for performing estimation calculations using machine learning results and for calculating machine learning itself.

FIG. 3 shows functional blocks of a learning system for communication devices 101 and 102 according to the present invention. The communication devices 101 and 102 have a data transmitting/receiving section 303, and the communication section 207 transmits and receives communication frames through the communication section 207 and the antenna 208. The data storage unit 301 uses the storage unit 201. Furthermore, the storage unit 201 and the control unit 202 are expanded to include a communication-related data management unit 302. The communication-related data management unit 302 manages metadata that is the source of learning data, receives estimation results, and communicates requests for the same. The learning data generation unit 332 converts metadata from the communication-related data management unit into input data or teacher data.

The learning unit 333 generates a learning model based on the learning data from the learning data generating unit 332. The generated learning model is stored in the data storage unit 331.

The learning unit 333 may include an error detection unit and an update unit. The error detection unit obtains an error between the output data output from the output layer of the neural network and the teacher data according to the input data input to the input layer. The error detection unit may use a loss function to calculate the error between the output data from the neural network and the teacher data.

The updating unit updates connection weighting coefficients between nodes of the neural network, etc., based on the error obtained by the error detection unit, so that the error becomes smaller. This updating unit updates the connection weighting coefficients and the like using, for example, an error backpropagation method. The error backpropagation method is a method of adjusting connection weighting coefficients between nodes of each neural network so that the above-mentioned error is reduced.

Note that when machine learning is incorporated into the AP 101 or the STA 102 itself, a single device such as the AP 101 or the STA 102 will have all the functions shown in FIG. 3.

When the data collection server and estimation server used for machine learning are provided separately from the AP 101 and the STA 102, functions such as collection and inference will be handled by the separate server as described above. Although FIG. 3 illustrates a case where the AP 101 or the STA 102 performs both learning and inference, the present invention is not limited to this, and the inference processing may be implemented in the AP 101. In this case, the estimation server transmits trained model data generated based on the received input/output data to the AP 101. In this case, the AP 101 may be configured to have the function of the inference unit 334. The AP 101 stores trained model data received from the estimation server 106. Then, the inference unit 334 of the AP 101 may be configured to calculate an estimated value using input data for inference obtained from the surrounding environment and operating conditions collected by itself and learned model data.

FIG. 4 is a conceptual diagram showing the input/output structure using the learning model of this embodiment.

In this embodiment, an example in which the STA 102 itself uses a learning model will be described.

As input data for the learning model, for example, the SNR, measured PER, target PER, and moving speed of the STA 102 when the STA 102 receives a communication frame from the AP 101 are used. Here, SNR stands for Signal to Noise Ration, and PER stands for Packet Error Rate. Table 1 shows an example of a learning data set in which input parameters and correct parameters are associated. Note that the unit of SNR shown in Table 1 is dB, the unit of measured PER is %, and the unit of target PER is %.

The SNR and the measured PER may be matched by values measured from communication frames most recently received by the STA 102 from the AP 101, or may be matched by average values of values measured over a predetermined period. The target PER is a PER value that is a minimum target that should be met when estimating MCS (Modulation and Coding Scheme) information. Furthermore, if the STA 102 is a mobile device, the moving speed may be measured using an acceleration sensor or the like, or may be predicted based on the degree of change in SNR, and may be used as input data.

Based on input data, the learning model outputs MCS information that maximizes communication throughput and minimizes delay time within a range below the target PER.

MCS information includes NSS, which is the number of spatial streams for MIMO communication, MCS Index, which is information on the combination of modulation method and encoding method, and whether to use DCM, which transmits the same information redundantly using multiple different carrier waves, etc. can be mentioned. Here, NSS stands for Number of Spatial Stream, and DCM stands for Dual Carrier Modulation. The MCS information also includes information such as RU Allocation indicating the OFDMA allocated frequency, BW indicating the bandwidth, PPDU Format specifying the format of the PPDU, and Coding Type specifying the error correction coding method such as LDPC or BCC. Here, RU stands for Resource Unit, BW stands for Bandwidth, LDPC stands for Low-density parity-check, and BCC stands for Binary Convolutional Coding.

Learning using learning data may be performed in advance before the wireless communication device operates, or learning may be performed while the wireless communication device transmits and receives communication frames. When a wireless communication device learns while transmitting and receiving communication frames, if the measured PER when receiving the communication frame to which the estimated MCS information is applied is lower than the target PER, it may be adopted as training data.

Note that although this embodiment shows an example in which the STA 102 uses the learning model, the AP 101 may also use the learning model.

Note that specific algorithms for machine learning include the nearest neighbor method, the Naive Bayes method, decision trees, and support vector machines. Another example is deep learning, which uses neural networks to generate feature quantities and connection weighting coefficients for learning by itself. Any available algorithm among the above algorithms can be applied to this embodiment as appropriate.

FIG. 5 is a diagram illustrating the operation of a system to which the present invention can be applied using the structure of the learning model shown in FIG. 4.

First, the AP 101 transmits a communication frame including an MCS request to the STA 102 (S501). The MCS request may specify the RU Allocation and BW for which MCS information is requested. When receiving a communication frame including an MCS request, the STA 102 measures communication quality, which is input data when determining MCS information. Communication quality includes signal quality such as SNR, CNR, SIR, and RSSI when receiving a communication frame including an MCS request. Here, CNR stands for Carrier to Noise Ratio, SIR stands for Signal to Interference Ratio, and RSSI stands for Received Signal Strength Indicator. In addition to SNR, PER, BER (Bit Error Rate), and FER (Frame Error Rate) are measured. Other information may be used as long as it can indicate communication quality. Further, the communication frame used for measuring the communication quality may be a communication frame different from the communication frame including the MCS request, as long as it is a communication frame that the AP 101 transmits to the STA 102. Further, the communication quality may be determined by a value measured from the most recently received communication frame, or may be determined by an average value of values measured over a predetermined period.

The STA 102 inputs the SNR, the measured PER, its own moving speed, and the target PER to a learning model, and obtains MCS information as an output. The STA 102 stores the obtained MCS information in a communication frame including the stored MCS response, and transmits it to the AP 101 (S502).

After receiving the communication frame including the MCS response, the AP 101 transmits a communication frame to which the MCS information included in the MCS response is applied to the STA 102 (S503).

Even if the STA 102 does not receive a communication frame including an MCS request, it can independently transmit a communication frame including an MCS response depending on the communication quality of the communication frame received from the AP 101. For example, a communication frame including an MCS response may be transmitted autonomously in response to a large change in the SNR or PER of a received communication packet.

Although this example shows an example in which the AP 101 transmits a communication frame including an MCS request to the STA 102, the STA 102 may transmit a communication frame including an MCS request to the AP 101.

FIG. 6 shows an example of Control Information of an MCS request/response included in a communication frame according to the present invention. In this example, it is assumed that common Control Information is used for MCS requests and responses. Control Information can be stored in the MAC header of a communication frame.

The Unsolicited MFB subfield (600) is set to 1 when an MCS response is returned without an MCS request, and set to 0 when an MCS response is returned when an MCS request is received.

The MRQ subfield (601) is set to 1 if it is an MCS request and set to 0 if it is an MCS response.

The NSS subfield (602) indicates the number of MIMO spatial streams recommended when returning an MCS response. For example, the number of spatial streams from 1 to 16 can be specified. The NSS subfield (602) is not used for MCS requests.

The MCS Index subfield (603) indicates recommended combination information of modulation method and encoding method. The MCS Index subfield (603) is not used for MCS requests.

The DCM subfield (604) indicates whether application of DCM is recommended. It is set to 1 when DCM is recommended, and set to 0 when it is not recommended. The DCM subfield (604) is not used for MCS requests.

The RU Allocation subfield (605) indicates the OFDMA RU Allocation for which MCS information is requested in the case of an MCS request. Recommended OFDMA RU Allocation in case of MCS response is shown.

The BW subfield (606) indicates the bandwidth for which MCS information is requested in the case of an MCS request. For example, it can be specified in a range of 20 MHz to 640 MHz. Indicates the recommended bandwidth for MCS responses.

The MSI/Partial PPDU Parameters subfield (607) indicates the sequence number of the MCS request when the Unsolicited MFB subfield is 1. Furthermore, when the Unsolicited MFB subfield is 0, it is used as the PPDU Format subfield (611) and Coding Type subfield (612).

The PPDU Format subfield (611) indicates a PPDU format supported by the IEEE802.11 standard, such as SU PPDU or MU PPDU.

The Coding Type subfield (612) indicates the error correction code system, and can select, for example, BCC or LDPC.

The Tx Beamforming subfield (608) is set to 1 when beamforming is applied to the PPDU during communication quality measurement, and is set to 0 when beamforming is not applied.

FIG. 7 is a flowchart showing the flow of processing performed when the control unit 202 executes a program stored in the storage unit 201 of the AP or STA of the present invention. This flowchart process is started in response to the start of data reception operation from the opposing communication device. In order to simplify the explanation, this flowchart only describes the data reception process, which is the essence of the present invention, and omits the data transmission process. Although this example shows an example in which the STA 102 receives data from the AP 101, the same flowchart can be applied even when the AP 101 receives data from the STA 102.

The STA 102 determines whether the communication application has instructed it to stop the data reception operation from the AP 101 (S700). If it is determined in S700 that a stop instruction has been given, the data reception process is stopped, and if it is determined in S700 that a stop instruction has not been given, the SNR and PER, which are the communication quality of the communication frame received from the AP 101, are determined. Measure and record in internal memory (S701). Next, the STA 102 determines whether the measured PER is less than or equal to the target PER (S702). If it is determined in S702 that the value is below the target, the following processing is performed. That is, the SNR, measured PER, target PER, and moving speed of the STA 102 that were recorded in the internal memory when estimating the MCS information are input information, and the applied MCS information is input to the learning model as teacher data, The learning model is updated (S703). Note that if the STA 102 uses only the learned model and does not update the learned model, the processes of S702 and S703 may be omitted. Next, the STA 102 determines whether an MCS request is included in the communication frame received from the opposite device (S704). If it is determined in S704 that the MCS request is not included, the process branches to S705, and if it is determined that the MCS request is included, the process branches to S706.

If it is determined in S704 that an MCS request is not included, the STA 102 determines whether the SNR/PER of the received communication frame has changed significantly or whether its own moving speed has changed significantly (S705). If it is determined in S705 that there has been no significant change, it is determined that updating the MCS information is unnecessary and the process returns to S701. If it is determined in S705 that there has been a large change, it is determined that the MCS information needs to be updated, and the process proceeds to S706. In S706, the measured SNR, PER, target PER, and moving speed of the STA 102 are input to the learning model as input data, and MCS information is obtained as output data. Next, the STA 102 transmits to the AP 101 a communication frame including an MCS response containing the acquired MCS information. The AP 101 updates the MCS information of the communication frame to be transmitted according to this MCS information.

As described above, the communication device of the present invention is capable of determining MCS information to be applied to a communication frame by machine learning, and notifying the opposite communication device of the result.

(Modified example)
In this embodiment, a standard name such as IEEE802.11HR is used as an example of a successor standard to IEEE802.11be, but the name is not limited thereto. For example, the standard name may be HRL (High ReLiability). Further, the standard name may be HRW (High Reliability Wireless). Alternatively, VHT (Very High Reliability) may be used. Further, the standard name may be EHR (Extremely High Reliability). Further, the standard name may be UHR (Ultra High Reliability). Alternatively, it may be LL (Low Latency). Further, the standard name may be VLL (Very Low Latency). Further, the standard name may be ELL (Extremely Low Latency). Alternatively, it may be ULL (Ultra Low Latency). Further, the standard name may be HRLL (High Reliable and Low Latency). Further, the standard name may be URLL (Ultra-Reliable and Low Latency). Further, the standard name may be URLLC (Ultra-Reliable and Low Latency Communications). Moreover, other different names may be used.

Note that a part of the data set for learning (combination of input values and teaching data) generated by the learning data generation unit 322 can be used not only for learning but also for performance evaluation of a trained data model. The inference server intentionally does not use a part of the data set generated by the learning data generation unit 322 for learning, but stores it separately as a data set for evaluation. In terms of trained model data, this evaluation data set is a combination of unknown input values that have not been used for learning in the past and teacher data (correct data).

The inference server calculates an inference result using the learned model data trained by the learning unit 333 and input values of the evaluation data set. Next, the performance of the trained model is evaluated by comparing the inference results with the training data.

Then, as a result of evaluating the performance, if the correct answer rate exceeds a predetermined threshold (for example, 90%), the operation of the inference process can be started.

(Other embodiments)
The present invention can also be realized by a process in which a program that implements one or more functions of the above-described embodiments is supplied to a system or device via a network or a storage medium, and a computer of the system or device reads and executes the program. be. A computer has one or more processors or circuits and may include separate computers or a network of separate processors or circuits for reading and executing computer-executable instructions.

The processor or circuit may include a central processing unit (CPU), microprocessing unit (MPU), graphics processing unit (GPU), application specific integrated circuit (ASIC), or field programmable gateway (FPGA). The processor or circuit may also include a digital signal processor (DSP), a data flow processor (DFP), or a neural processing unit (NPU).

Claims (5)

A communication device,
In order to estimate MCS (Modulation and Coding Scheme), some or all of SNR (Signal to Noise Ration), measured PER (Packet Error Rate), target PER, and moving speed of the communication device are applied to the learned model. an input means for inputting as input data;
Calculation means for calculating the MCS using the learned model by inputting it through the input means;
A communication device comprising: notification means for notifying another communication device of the MCS information calculated by the calculation means. further comprising receiving means for receiving a frame for requesting MCS estimation from the other communication device,
2. The communication device according to claim 1, wherein the MCS is calculated by the calculating means based on information of the frame received by the receiving means. The MCS information includes NSS (Number of Spatial Stream), MCS Index, DCM (Dual Carrier Modulation), RU (Resource Unit) Allocation, Bandwidth, PPDU for It is characterized by including some or all of mat, Coding Type. The communication device according to claim 1. A communication method for a communication device, the communication method comprising:
In order to estimate MCS (Modulation and Coding Scheme), some or all of SNR (Signal to Noise Ration), measured PER (Packet Error Rate), target PER, and moving speed of the communication device are applied to the learned model. an input step of inputting as input data;
a calculation step of calculating the MCS using the learned model by inputting it in the input step;
A communication method for a communication device, comprising: a notification step of notifying another communication device of the MCS information calculated by the calculation means. A program for causing a computer to function as each means of the communication device according to claim 1.

Images