WIFI beam rapid training method and device

Abstract

The application provides a WIFI beam rapid training method and device, the method comprises the following steps: determining the best sector through the STA feedback position information and received signal strength; performing beam training preprocessing to obtain a wide transmitting beam; determining the best wide beam pair through a wide beam training process assisted by sensing information; determining the best narrow beam pair through a narrow beam training process assisted by sensing information based on the best wide beam; in the beam training process, the sensing process is to obtain the position and shape sensing information of the STA by using a pre-trained convolutional neural network (CNN) model. The WIFI beam rapid training method and device provided by the application can utilize the sensing information of the WIFI communication device when performing communication beam training, and can perform targeted training in the direction where the best communication beam may exist, so that the communication beam training cost can be reduced, the device access time can be reduced, the communication delay can be reduced, and the communication quality can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a WIFI millimeter wave beam rapid training method and device. BACKGROUND

[0002] WIFI, as a short-range wireless communication technology, is one of the key technologies commonly used in the Internet of Things, and is particularly suitable for smart home and smart city fields. Millimeter wave WIFI is mainly based on IEEE 802.11ad and 802.11ay standard specifications. Millimeter wave chips developed in accordance with the above standard specifications can provide high-speed wireless transmission and are suitable for high-definition video, AR / VR and industrial automation applications. With the popularity of WIFI devices and the ubiquity of WIFI networks, WIFI signals will produce reflection, diffraction and scattering phenomena when propagating in physical space.

[0003] The 60GHz millimeter wave frequency band has abundant available spectrum resources, but due to the increase in path loss, the attenuation is very serious, so directional communication technology using beamforming (BF) is introduced to provide additional transmit antenna gain or receive antenna gain to overcome signal attenuation. With the beamforming technology, beamforming training (BFT) is first performed. The process of BFT is specifically the process of aligning the transmit and receive beams of the access point (AP) and the station (STA). The communication beam training in the existing IEEE 802.11ay standard is mainly based on the maximum received signal strength criterion of the signal transmitting end and the receiving end to continuously try and error in each direction to find the best communication beam pair. This method is a traversal algorithm, which requires all beams of the STA and the AP to be tried. The STA and the AP find the best transmit beam and the best receive beam, which has a certain blindness, and the process is a traversal process, which takes a long time and causes large communication delay. SUMMARY

[0004] The present application provides a WIFI beam rapid training method and device to solve the technical problem of long time consumption and large communication delay caused by the WIFI beam training method in the existing IEEE 802.11ad / ay standard.

[0005] In a first aspect, the present application provides a WIFI millimeter wave beam rapid training method, comprising:

[0006] Determine the best sector by feeding back the position information and the received signal strength of the STA;

[0007] Perform beam training preprocessing to obtain a wide transmit beam;

[0008] determining the best wide beam pair through a wide beam training process assisted by perception information; determining the best narrow beam pair through a narrow beam training process assisted by perception information based on the best wide beam; and obtaining the perception information of the STA, including the position and shape of the STA, by using a trained convolutional neural network (CNN) model in the beam training process.

[0009] In some embodiments, the determination of the best sector through the feedback of the position information and the received signal strength by the STA includes:

[0010] In the sector scanning phase initiated by the transmitter, the beam request message sent by the STA is obtained, the beam request message carrying the position information and the signal strength of the received signal of the receiver, and the beam of the beam selection interval is sent;

[0011] In the sector scanning phase initiated by the receiver, the beam response message is sent to the STA, the beam response message carrying the position information and the signal strength of the received signal of the receiver, and the beam of the beam selection interval is sent;

[0012] The best transmission sector and the best reception sector are determined according to the position information and the signal strength.

[0013] In some embodiments, the determination of the best wide beam pair through the beam training process assisted by perception information includes: obtaining the best wide beam pair by using a beam training process assisted by perception information in the wide beam training process; and the determination of the best narrow beam pair through the beam training process assisted by perception information based on the best wide beam includes: adjusting the beam width to a narrow beam within the coverage area of the best wide beam, and obtaining the best wide beam pair by using a beam training process assisted by perception information in the narrow beam training process.

[0014] In some embodiments, the frame format in the perception process includes a request and a response message.

[0015] In some embodiments, the beam training process assisted by perception information includes:

[0016] Step 1: The station (STA) sends a beam training request assisted by perception information to an access point (AP), the AP receives the request, performs an STA perception process to obtain STA perception information, adjusts the beam direction based on the STA perception information and the strength of the received signal, calculates the transmission / reception beam pair with a stronger beam signal based on the size of the CSI, the shape and position of the STA, and the beam width, and obtains a list of the best transmission / reception communication beam pairs;

[0017] Step 2: The AP sends a beam training response assisted by perception information to the STA, carrying the list of the best reception communication beam pairs calculated by the AP based on the perception information assistance;

[0018] Step 3, the AP sends training frames to the STA in turn according to the best transmission beam, and the STA receives the training frames according to the best reception beam list received in step 2, and records the signal reception strength corresponding to the beam pair;

[0019] Step 4: the STA sends a feedback frame to the AP, feeding back the best transmission beam obtained in the beam training in step 3 and the CSI of the STA and the like information;

[0020] Step 5: the STA and the AP complete the fast communication beam training assisted by the perception information, and determine the best beam pair.

[0021] In some embodiments, the perception information includes the position information and shape information of the STA, which are obtained through a STA perception process; the STA perception process is an online inference process based on a pre-trained CNN network model; more specifically, the input data of the CNN model inference process is the real-time channel matrix and CSI data reported by the STA, and the type and model number of the STA reported by the STA during initial registration; the output of the CNN model inference process is the perception information of the STA: the position information and shape information of the STA;

[0022] The pre-trained CNN model is obtained by an offline training method, and the offline training includes:

[0023] A training data set is constructed according to the historical data collected by the AP, including the channel matrix reported by the STA, the type of the STA, the model number of the STA, and the CSI data fed back by the STA; the training data in the training data set is input into the CNN model to complete the training of the CNN model, and a usable CNN model is obtained.

[0024] In a second aspect, the application further provides a WIFI beam fast training device, which includes:

[0025] A sector scanning module is configured to determine the best sector by feeding back the position information and the received signal strength of the STA;

[0026] A preprocessing module is configured to perform beam training preprocessing to obtain a wide transmission beam;

[0027] A wide beam scanning module is configured to determine the best wide beam pair through a perception information assisted wide beam training process;

[0028] A narrow beam scanning module is configured to determine the best narrow beam pair through a perception information assisted narrow beam training process based on the best wide beam.

[0029] A perception module is configured to obtain the perception information of the STA, including the position and shape of the STA, by using a trained convolutional neural network (CNN) model during the beam training process.

[0030] In a third aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the WIFI beam rapid training method according to any one of the above aspects when executing the computer program.

[0031] In a fourth aspect, the present application provides a non-transitory computer readable storage medium, having a computer program stored thereon, wherein the computer program is executable on a processor to implement the WIFI beam rapid training method according to any one of the above aspects.

[0032] In a fifth aspect, the present application provides a computer program product, comprising a computer program, wherein the computer program is executable on a processor to implement the WIFI beam rapid training method according to any one of the above aspects.

[0033] The WIFI beam rapid training method and device provided by the present application can utilize the sensing information of the WIFI communication device when performing communication beam training, and perform targeted training in the direction where the optimal communication beam is likely to exist, thereby reducing the overhead of beam training, reducing the device access time, reducing the communication delay, and improving the communication quality. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0035] Figure 1 is a flowchart of the WIFI beam rapid training method provided by the present application.

[0036] Figure 2 is a schematic diagram of the sector-level beam scanning process provided by the present application.

[0037] Figure 3 is a schematic diagram of the beacon period being divided into multiple time periods in the 802.11ad / ay protocol.

[0038] Figure 4 is a schematic diagram of the I-TXSS stage provided by the present application.

[0039] Figure 5 is a schematic diagram of the R-TXSS stage provided by the present application.

[0040] Figure 6 is a schematic diagram of the SSW Feedback stage provided by the present application.

[0041] Figure 7 is a schematic diagram of the SSW ACK stage provided by the application.

[0042] Figure 8 is a schematic diagram of the perception information assisted beam training process provided by the application.

[0043] Figure 9 is a schematic diagram of the off-line training CNN model process provided by the application.

[0044] Figure 10 is a schematic diagram of the CNN model provided by the application.

[0045] Figure 11 is a schematic diagram of the frame format with perception information provided by the application.

[0046] Figure 12 is a schematic diagram of the structure of the WIFI beam rapid training device provided by the application.

[0047] Figure 13 is a schematic diagram of the structure of the electronic device provided by the application. DETAILED DESCRIPTION

[0048] The STA devices in the WIFI communication network are various in shape, such as mobile phones, sweeping robots, notebook computers, etc. The communication node AP can obtain the size, shape and contour of the STA device through perception, and can utilize the perception information to perform targeted training in the direction where the optimal communication beam may exist when performing communication beam training, which can not only reduce the overhead of beam training, but also reduce the device access time, reduce the communication delay and improve the communication quality. The application can assist the AP to reduce the beam training range. The application provides a communication and perception combined rapid beam training method. The WIFI device detects and perceives the WIFI communication environment and the communication node through wireless perception technology, and perceives the position, moving speed and other information of the communication node. The perception information and the perception result can be used to assist the channel estimation, equalization, beam management and other modules of the receiving end and the transmitting end of the communication system to select the algorithm, set the algorithm parameters and optimize the algorithm.

[0049] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0050] Figure 1 is a schematic diagram of the WIFI beam rapid training method provided by the application, asFigure 1 As shown, the method comprises the following:

[0051] Step 101, feedback position information and received signal strength by STA, determine the best sector.

[0052] Specifically, the process of beamforming (BFT) is specified in the 802.11 protocol, which includes sector level sweep (SLS) and beam refinement protocol (BRP). Figure 2 is a sector level beam sweep process schematic diagram provided by the application, as Figure 2 As shown, the SLS process can occur in the "beacon transmission interval (BTI) + adaptive beamforming training (A-BFT)" phase and the data transfer interval (DTI) phase, and the BRP process can only occur in the DTI phase.

[0053] The SLS process is a process of interaction between the initiator and the responder, which usually includes four sub-stages, namely initiator sector sweep (ISS), responder sector sweep (RSS), sector sweep feedback (SSW-Feedback, SSW-FBCK) and sector sweep ACK (SSW-ACK). The SLS stage only trains the sending sectors of the initiator and the responder, and is mandatory, and the BRP is used to train the receiving sector and perform beam pairing. In the existing protocol, a traversal algorithm is used in the BRP process to realize beam pairing, and the technical scheme of the present application optimizes the existing method. In the sector sweep stage, the sector is selected according to the feedback position information and received signal strength of the STA

[0054] Figure 3 is a schematic diagram of dividing the beacon interval (Beacon Interval) in the 802.11ad / ay protocol into multiple time intervals, as Figure 3As shown, one beacon period can be divided into a beacon header interval (BHI) and a data transfer interval (DTI). The BHI includes a beacon transmission interval (BTI), an association beamforming training interval (A-BFT), and an announcement transmission interval (ATI). The DTI contains a scheduling period (SP) and a contention-based access period (CBAP). The sector level sweep (SLS) training phase mainly completes the initiator and responder transmit sector training, and the training process includes four sub-stages of initiator-transmit sector sweep (I-TXSS), responder transmit sector sweep (R-TXSS), sector sweep feedback (SSW-Feedback), and sector sweep ACK (SSW-ACK). In the existing protocol, the beam training process is performed according to the traversal method, and the two-stage beam training method of sensing information cooperation is proposed in the present application.

[0055] Figure 4 is a schematic diagram of the I-TXSS phase provided by the present application, as shown in Figure 4 In the I-TXSS phase: the initiator sends a beacon frame or a special training frame in each sector to discover the responder, and the responder receives in quasi-omnidirectional mode at this time. This phase is used to train the transmit beam of the initiator, and generally occurs during the BTI. Different from the method of sequentially sending messages by the initiator in the existing protocol, the present application proposes that the initiator sends the beam request message of the training process with position information and the signal strength of the signal that can be received by the responder, and the beam of the transmit beam selection interval.

[0056] Figure 5 is a schematic diagram of the R-TXSS phase provided by the present application, as shown in Figure 5As shown in the R-TXSS stage: the responder sends the SSW frame in each sector, and carries the best sending sector information of the initiator in the last stage in the SSW frame, the initiator receives in quasi-omni mode at this time, and this stage is used to train the sending sector of the responder, and generally occurs during the A-BFT. The responder sends the position information and the signal strength capable of receiving the signal of the receiver in the beam request message of the training process, and sends the beam of the beam selection interval.

[0057] The sending party and the receiving party determine a smaller range according to the position information and the signal strength, and repeat the steps of the I-TXSS stage and the R-TXSS stage, so that the best sending sector and the best receiving sector can be quickly determined.

[0058] Figure 6 The SSW Feedback stage provided by the application is shown in the schematic diagram as shown in the figure. Figure 6 As shown in the SSW-Feedback stage: the initiator sends the SSW-Feedback frame in the best sector specified in the last step, and carries the sending sector list of the responder sorted according to the receiving quality and the best sector of the responder in the last stage in the frame; when the sending party is a STA, the SSW-Feedback frame carries the channel state information (Channel State Information, CSI) and the channel matrix of each sector in the sector list, and the AP receives and saves the CSI information and the channel matrix; the responder receives in quasi-omni mode at this time, and this stage generally occurs during the A-BFT.

[0059] Figure 7 The SSW ACK stage provided by the application is shown in the schematic diagram as shown in the figure. Figure 7 As shown in the SSW-ACK stage: the responder sends the SSW-ACK frame in the best sector specified in the last step, and carries the sending sector list of the initiator sorted according to the quality in the frame, and the initiator receives in quasi-omni mode at this time. The training of the SLS stage is completed at this step, and the SLS is generally performed in the BHI, and the subsequent BRP training is generally performed in the SP.

[0060] Step 102, performing beam training preprocessing to obtain a wide transmission beam.

[0061] Specifically, after the sector is selected, the beam-level training is performed, and the beam-level training is divided into two stages of wide beam training and narrow beam training, and a preprocessing process is added in the process. The beam width is adjusted by controlling the antenna switch, so that the first stage beam training process is performed on the wide beam, thereby selecting a wider beam alignment.

[0062] The transmitter and the receiver turn off the antennas outside the best sector by using antenna switch control and power control technology, and make most of the antennas in an inactive state, which is equivalent to using only a small number of antenna elements in the antenna array to transmit the training beam. Since the number of antennas participating in the work is reduced, the width of the transmission beam will be widened, and the coverage of the main lobe will be wider. At this time, only a small number of training beams are needed to search for the approximate interval where the target user is located.

[0063] Step 103: Determine the best wide beam by beam-level beam training scanning.

[0064] Specifically, the present application proposes a perception information assisted communication beam training under the existing IEEE 802.11ad / ay standard system.

[0065] Figure 8 The present application provides a perception information assisted beam training process diagram, as shown in Figure 8 The present application provides a perception information assisted beam training process diagram, as shown in

[0066] Step 1: The station STA sends a perception information assisted beam training request to the access point AP. After receiving the request, the AP performs a STA perception process to obtain STA perception information, and then adjusts the beam direction based on the perception information. The AP calculates the transmission / reception beam pair with strong beam signal according to the CSI size, the shape and position of the STA, and the beam width, and obtains a list of the best transmission / reception communication beam pairs.

[0067] The STA perception process is as follows:

[0068] The AP adopts a convolutional neural network (CNN) model to realize perception of the shape and position of the STA device, and the STA perception process is an online inference process based on a pre-trained CNN network model. The input data of the CNN model inference process includes real-time channel matrices reported by the STA in the SSW-Feedback stage and CSI data fed back by the STA, as well as the type and model of the STA reported when the STA is initially registered; the output of the CNN model inference process is STA perception information: position information and shape information of the STA. The CNN model training is divided into two stages of offline training and online inference according to the generation and use process of the CNN model. In the offline training stage, an offline training method is used to obtain a pre-trained CNN model. In the offline training process, the input data set required for training of the CNN network for STA device shape and position information perception is generated according to historical data collected by the AP. The training data set input data of the CNN model includes but is not limited to channel matrices reported by the STA, CSI data fed back by the STA, the type of the STA device, and the model of the STA device; the training data set output data includes the shape of the STA device and the position of the STA device; then the CNN model training set data set is used for offline training of the model, and finally the trained CNN network model parameters are saved; in the online inference stage, the type and model of the STA device are reported to the AP when the STA device is initially connected and registered and are saved in the AP, the STA device reports real-time channel matrices and CSI data, and the AP transmits the received input data to the trained CNN model after processing, and can directly output the perception result: the shape of the STA device and the position of the STA device.

[0069] Since the shapes of STA devices of the same type and model are the same, the shape of the STA device is quantified and independently measured by the sum of the maximum distances of the three directions x, y and z of the device, and the position of the STA device is measured by a two-element tuple of longitude and latitude information.

[0070] Step 2: The AP sends a beam training response assisted by perception information to the STA, carrying the best receiving communication beam pair list calculated by the AP based on the perception information.

[0071] Step 3: The AP sequentially sends training frames to the STA in the direction according to the best sending beam determined in the previous step, and the STA receives the training frames in the direction according to the best receiving communication beam list received in step 2, and records the signal reception strength corresponding to the beam pair.

[0072] Step 4: The STA sends a feedback frame to the AP, feeding back the best sending beam obtained in the beam training in step 3 and the CSI information of the STA.

[0073] Step 5: The STA and the AP complete the perception-based fast communication beam training to determine the optimal beam.

[0074] In some embodiments, the method further comprises:

[0075] constructing a training data set according to the channel matrix reported by the STA, the CSI data reported by the STA device, the type of the STA device, and the model of the STA device;

[0076] inputting the training data in the training data set into the CNN model to complete the training of the CNN model.

[0077] The process of offline training of the CNN model comprises:

[0078] The offline training process refers to training the CNN model using large-scale data in a non-real-time environment, and deploying the trained model to the AP for inference after the training is completed. Figure 9 is a flowchart of the offline training process of the CNN model provided by the present application, as shown in Figure 9 The offline training process comprises data preparation and preprocessing, forward propagation, loss calculation, backward propagation, and model optimization update to obtain the trained CNN model.

[0079] Data preparation and preprocessing: In this stage, data acquisition, data augmentation, data normalization, and data division of the data required for model training are completed, and the specific process is as follows:

[0080] Data acquisition: training sample data is acquired from a database formed by historical data, and the data of the present embodiment includes but is not limited to channel matrix, CSI data reported by the STA device, type of the STA device, and model of the STA device, and an error threshold T is set.

[0081] Data augmentation: the acquired sample data is augmented by methods such as random cropping, rotation, flipping, and noise addition to increase data diversity and prevent overfitting during model training.

[0082] Data normalization: the enhanced data is standardized, and in the present embodiment, normalization is adopted to normalize the data to the interval [0, 1] to accelerate the convergence speed.

[0083] Data division: the normalized data is divided into a training set (about 70-80%) and a test set (about 20-30%); wherein the training set data is used as input data for executing the CNN model training, the test set is used to evaluate the accuracy of the model and optimize the model; and a group of data is taken as label data.

[0084] Forward propagation: the training set data is input into the model to determine the CNN model.

[0085] Figure 10 This is a schematic diagram of the CNN model provided by the present invention, as shown below. Figure 10 As shown, the CNN model used in this embodiment consists of an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. The forward propagation process is as follows: the input layer receives the model input data; local features are extracted in the convolutional layer, multiple filters are applied, and feature mapping is performed. In this embodiment, a 3×3 convolutional kernel is used for feature extraction; the activation function is the ReLU (Rectified Linear Unit) function: f(x) = max(0, x) to increase non-linearity; the pooling layer uses average pooling to reduce the dimensionality of features and improve computational efficiency; the fully connected layer (FC) maps the extracted high-level features to the final perceptual task output features; the output layer outputs the final output value obtained from model training.

[0086] Loss Calculation: The test set is used as input data to the trained CNN model to obtain the shape and position of the STA device from the training output. The output data is compared with the real value data used as labels to calculate the error. The shape error of the device is the difference between the perceived shape measurement of the device and the real value. The position error of the device is the sum of the perceived position error of the device and the difference between the longitude and latitude of the real position of the device. When the sum of the errors is greater than a preset threshold T, the model input is adjusted and retrained until the error is less than the threshold.

[0087] Backpropagation: The gradient of the error with respect to the weight parameters of each layer of the CNN model is calculated using the chain rule, and the gradient is backpropagated to each layer (convolutional layer, fully connected layer).

[0088] Model optimization and update: Update training data and model parameters (such as weights and biases) and retrain.

[0089] Figure 11 This is a schematic diagram of the frame format with sensing information provided by the present invention, such as... Figure 11 As shown, this application modifies the frame format in the protocol during the sensing process by adding bits to the frame format to represent request and response messages.

[0090] Step 104: Based on the optimal wide beam, determine the optimal narrow beam pair through a narrow beam training process assisted by sensing information.

[0091] Specifically, since some STAs can be located between two wide beam coverage ranges or narrow beam coverage ranges, in order to provide the best beam direction for all STAs, the awareness information training is added in the beam training process, and the volume, position and other information of the STA are obtained through the awareness information training; after the AP determines the wide beam range with strong signal, the narrow beam beam training assisted by the awareness information is performed on the selected coarse-grained space of the wide beam according to the wide beam awareness information; further, the narrow beam width is adjusted according to the more accurate position volume information of the STA, the codebook selection beam direction is transmitted through the adjustment of the narrow beam, and the optimal narrow beam pair is determined through the narrow beam training process. Figure 8

[0092] Step 105, in the beam training process, the awareness information of the STA is obtained by using the trained convolutional neural network (CNN) model, including the position and shape of the STA.

[0093] Specifically, in the narrow beam training process, the AP obtains the channel matrix reported by the STA, the type of the STA, the model of the STA, and the CSI, inputs the channel matrix, the type of the STA, the model of the STA, and the CSI into the CNN network model, and obtains the CSI information fed back by the STA under the condition of coarse beam transmission signal by using the convolution kernel of different scales of the CNN network model. In addition, the accurate position information and volume information of the STA can also be perceived. Through the awareness information, the AP dynamically adjusts the width of the narrow beam and obtains the best narrow beam pair for transmission and reception.

[0094] The WIFI signal is transmitted by relying on a wireless channel, and the CSI is a measure of the properties of the wireless channel, which records the changes of the wireless channel under the influence of multiple factors such as distance attenuation, environmental attenuation and signal scattering. The WIFI system based on the 802.11n and above standards combines MIMO and OFDM technologies, that is, the transmitting end and the receiving end have multiple antennas, and the WIFI signal is modulated by using the OFDM technology. At this time, the CSI measured at the receiving end is a complex matrix with a size of , wherein represents the number of transmitting antennas, represents the number of receiving antennas, represents the number of subcarriers.

[0095] ​In a wireless transmission environment, the transmission of WIFI signals can be affected by obstacles, resulting in phenomena such as reflection, scattering, refraction, and diffraction, thereby causing the receiving end to receive WIFI signals from multiple paths in different directions, which is called multipath effect. A typical WIFI communication system is composed of a WIFI signal transmitting end (such as a wireless router), a receiving end (such as a mobile phone, a computer, or other devices), and a user. WIFI signals propagate in a multipath manner, including a line of sight (LoS) path from the transmitting end to the receiving end, and a path reflected by stationary objects such as walls, which is called a static path. When a person performs a gesture action in a WIFI environment, it causes dynamic changes in some of the propagation paths of WIFI signals, and these paths are called dynamic paths. Due to the influence of multipath effect, the receiving end receives signals on multiple paths. Each path on the signal will be affected by different channel attenuation, phase delay, and scattering. The receiving end can obtain the CSI information of the signal through channel estimation technology.

[0096] When the volume of the obstacle is greater than a certain threshold value, different positions of the obstacle have different effects on the propagation path of WIFI, especially when the obstacle moves or its shape changes, different positions will also have different effects on the propagation of WIFI. Correspondingly, the receiving end will collect different CSI information. The shape and position of the obstacle can be related to the CSI information, and the feature information related to the volume and position of the obstacle can be extracted from the CSI information, and then input into a machine learning algorithm to realize the perception of the position and volume information of the obstacle.

[0097] The present application introduces the perception information of STA in the process of beam training and beam tracking, which can perceive the position, shape and other information of the object, so as to reduce the range of the beam in the process of beam training and beam tracking; further, the present application proposes a multi-stage scanning beam training scheme, which includes sector scanning, coarse-grained beam scanning and fine-grained beam scanning. The sector scanning follows the sector-level scanning of the existing technology, the coarse-grained beam scanning is a beam-level scanning, which uses wide beam training to find the best transmitting beam and receiving wide beam in a shorter time, and then uses narrow beam training within the coverage range of the wide beam, which consumes less time than directly using narrow beam training in a larger range, thereby reducing the communication delay.

[0098] The WIFI beam rapid training device provided by the present application is described below. The WIFI beam rapid training device described below can be referred to in conjunction with the WIFI beam rapid training method described above.

[0099] Figure 12is a structural schematic diagram of the WIFI beam rapid training device provided by the application, as Figure 12 The application provides a WIFI beam rapid training device, which comprises:

[0100] The sector scanning module 1001 is used for determining an optimal sector by feeding back position information and received signal strength of the STA.

[0101] The preprocessing module 1002 is used for performing beam training preprocessing to obtain a wide transmission beam.

[0102] The wide beam scanning module 1003 is used for determining an optimal wide beam pair through a wide beam training process assisted by sensing information.

[0103] The narrow beam scanning module 1004 is used for determining an optimal narrow beam pair through a narrow beam training process assisted by sensing information based on the optimal wide beam.

[0104] The sensing module 1005 is used for obtaining sensing information of the STA, including the position and shape of the STA, by using a trained convolutional neural network (CNN) model during the beam training process.

[0105] In some embodiments, the determination of the optimal sector by feeding back the position information and the received signal strength of the STA comprises:

[0106] In the sector scanning stage of the initiator, a beam request message sent by the STA is acquired, the beam request message comprises position information and signal strength of a signal capable of being received by the receiver, and a beam of a beam selection interval is sent.

[0107] In the sector scanning stage of the responder, a beam response message is sent to the STA, the beam response message comprises position information and signal strength of a signal capable of being received by the receiver, and a beam of a beam selection interval is sent.

[0108] The optimal transmission sector and the optimal reception sector are determined according to the position information and the signal strength.

[0109] In some embodiments, the determination of the optimal wide beam pair through the beam training process assisted by the sensing information comprises: obtaining the optimal wide beam pair by using a sensing information assisted beam training process in the wide beam training process; and the determination of the optimal narrow beam pair through the beam training process assisted by the sensing information based on the optimal wide beam comprises: adjusting the beam width to a narrow beam in the coverage area of the optimal wide beam, and obtaining the optimal narrow beam pair by using a sensing information assisted beam training process in the narrow beam training process.

[0110] In some embodiments, the frame format in the sensing process comprises a request and a response message.

[0111] In some embodiments, the perception information assisted beam training process comprises:

[0112] Step 1, the station STA sends a perception information assisted beam training request to the access point AP, the AP receives the request and performs a STA perception process to obtain STA perception information, adjusts the beam direction based on the STA perception information and the strength of the received signal; the AP calculates the transmission / reception beam pair with stronger beam signal based on the CSI size, the shape and position of the STA, and the beam width, and obtains a list of optimal transmission / reception communication beam pairs;

[0113] Step 2, the AP sends a perception information assisted beam training response to the STA, carrying the list of optimal reception communication beam pairs calculated by the AP based on the perception information assistance;

[0114] Step 3, the AP sequentially sends training frames to the STA according to the optimal transmission beam, and the STA receives the training frames according to the optimal reception communication beam list received in step 2, and records the signal reception strength corresponding to the beam pair;

[0115] Step 4: the STA sends a feedback frame to the AP, feeding back the optimal transmission beam obtained in the beam training in step 3 and the CSI and other information of the STA;

[0116] Step 5: the STA and the AP complete the perception information assisted fast communication beam training and determine the optimal beam pair.

[0117] In some embodiments, the perception information includes STA position information and shape information, which is obtained through a STA perception process; the STA perception process is an online inference process based on a pre-trained CNN network model; more specifically, the input data of the CNN model inference process is the real-time channel matrix and CSI data reported by the STA, as well as the type and model of the STA reported by the STA during initial registration; the output of the CNN model inference process is the STA perception information: the position information and the shape information of the STA;

[0118] The pre-trained CNN model is obtained by an offline training method, which includes:

[0119] A training data set is constructed according to the historical data collected by the AP, including the channel matrix reported by the STA, the type of the STA, the model of the STA, and the CSI data fed back by the STA; the training data in the training data set is input into the CNN model to complete the training of the CNN model, and a usable CNN model is obtained.

[0120] Specifically, the WIFI beam rapid training device provided by the embodiment of the present application can realize all the method steps achieved by the WIFI beam rapid training method embodiment and achieve the same technical effects. Here, the same parts and beneficial effects in the method embodiment will not be described in detail.

[0121] Figure 13 An example of a schematic diagram of the physical structure of an electronic device is shown in Figure 13 The electronic device can include a processor 810, a communications interface 820, a memory 830, and a communications bus 840, wherein the processor 810, the communications interface 820, and the memory 830 can communicate with each other through the communications bus 840. The processor 810 can invoke the logic instructions in the memory 830 to execute the WIFI beam rapid training method, which includes:

[0122] Determine the best sector by feeding back the position information and the received signal strength of the STA;

[0123] Perform beam training preprocessing to obtain a wide transmission beam;

[0124] Determine the best wide beam pair through the wide beam training process assisted by the perception information;

[0125] Based on the best wide beam, determine the best narrow beam pair through the narrow beam training process assisted by the perception information;

[0126] In the beam training process, the perception information of the STA, including the position and shape of the STA, is obtained by using the trained convolutional neural network (CNN) model.

[0127] In addition, the logic instructions in the memory 830 described above can be implemented in the form of a software function unit and sold or used as an independent product. When used, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0128] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program being stored in a non-transitory computer readable storage medium, and the computer program being executable by a processor to enable the computer to perform the WIFI beam fast training method provided by the above-mentioned methods, the method comprising:

[0129] determining the optimal sector by feeding back the position information and the received signal strength by the STA;

[0130] performing the beam training preprocessing to obtain the wide transmission beam;

[0131] determining the optimal wide beam pair through the wide beam training process assisted by the perception information;

[0132] determining the optimal narrow beam pair through the narrow beam training process assisted by the perception information based on the optimal wide beam;

[0133] in the beam training process, the perception information of the STA is obtained by using the trained convolutional neural network (CNN) model, including the position and shape of the STA.

[0134] In another aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, the computer program being executable by a processor to implement the WIFI beam fast training method provided by the above-mentioned methods, the method comprising:

[0135] determining the optimal sector by feeding back the position information and the received signal strength by the STA;

[0136] performing the beam training preprocessing to obtain the wide transmission beam;

[0137] determining the optimal wide beam pair through the wide beam training process assisted by the perception information;

[0138] determining the optimal narrow beam pair through the narrow beam training process assisted by the perception information based on the optimal wide beam;

[0139] in the beam training process, the perception information of the STA is obtained by using the trained convolutional neural network (CNN) model, including the position and shape of the STA.

[0140] The apparatus embodiments described above are only schematic, wherein the units shown as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place or distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment scheme. Those skilled in the art can understand and implement without creative labor.

[0141] Those skilled in the art can clearly understand the implementation of the embodiments by means of software and necessary general hardware platforms through the description of the above embodiments, and of course, the embodiments can also be implemented by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, and the computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0142] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A WIFI beam fast training method, characterized in that, Comprise: Determine the best sector by STA feedback position information and received signal strength; Wide transmit beam is obtained by beam training preprocessing; Determine the best wide beam pair through the wide beam training process assisted by perception information; Determine the best narrow beam pair through the narrow beam training process assisted by perception information based on the best wide beam; In the beam training process, the perception information of the STA is obtained by using the trained convolutional neural network (CNN) model, including the position and shape of the STA; The beam training process assisted by perception information comprises: Step 1: The STA sends a beam training request assisted by perception information to the AP, and the AP receives the request and performs a STA perception process to obtain STA perception information, adjusts the beam direction based on the STA perception information and the received signal strength; the AP calculates the transmission / reception beam pair with stronger beam signal based on the size of the CSI, the shape and position of the STA, and the beam width, and obtains a list of the best transmission / reception communication beam pairs; Step 2: The AP sends a beam training response assisted by perception information to the STA, carrying the list of the best reception communication beam pairs calculated by the AP based on the perception information; Step 3: The AP sends training frames to the STA in turn according to the best transmission beam, and the STA receives the training frames according to the best reception communication beam list received in step 2, and records the signal reception strength corresponding to the beam pair; Step 4: The STA sends a feedback frame to the AP, feeding back the best transmission beam obtained in the beam training in step 3 and the CSI information of the STA; Step 5: The STA and the AP complete the fast communication beam training assisted by perception information, and determine the best beam pair.

2. The method of claim 1, wherein, Determine the best sector by STA feedback position information and received signal strength, comprising: In the initiator transmission sector scanning phase, obtain the beam request message sent by the STA, which contains position information and signal strength that can receive the signal of the receiver, and the beam selection interval of the transmission beam; In the responder transmission sector scanning phase, send a beam response message to the STA, which contains position information and signal strength that can receive the signal of the receiver, and the beam selection interval of the transmission beam; Determine the best transmission sector and the best reception sector according to the position information and the signal strength. 3.The WIFI beam rapid training method of claim 1, wherein, Determine the best wide beam pair through the beam training process assisted by perception information, comprising: In the wide beam training process, the beam training process assisted by perception information is used to obtain the best wide beam pair; Determine the best narrow beam pair through the beam training process assisted by perception information based on the best wide beam, comprising: In the coverage area of the best wide beam, adjust the beam width to a narrow beam, and in the narrow beam training process, use the beam training process assisted by perception information to obtain the best narrow beam pair.

4. The method of claim 1, wherein, The frame format in the perception process contains a request and a response message.

5. The method of claim 1, wherein, The perception information includes the position information and shape information of the STA, which is obtained through the STA perception process; the STA perception process is an online inference process based on a pre-trained CNN network model; More specifically, the input data of the CNN model inference process is the real-time channel matrix and CSI data reported by the STA, as well as the type and model of the STA reported by the STA during initial registration; the output of the CNN model inference process is the STA perception information: the position information and shape information of the STA; The pre-trained CNN model is obtained by an offline training method, and the offline training comprises: constructing a training data set according to historical data collected by the AP, including the channel matrix reported by the STA, the type of the STA, the model of the STA, and the CSI data fed back by the STA; inputting the training data in the training data set into the CNN model to complete the training of the CNN model, and obtaining a usable CNN model. 6.A WIFI beam rapid training apparatus, characterized in that, Comprise: a sector scanning module for determining the best sector by the STA feeding back the position information and the received signal strength; a preprocessing module for preprocessing the beam training to obtain a wide transmission beam; a wide beam scanning module for determining the best wide beam pair through a perception information assisted wide beam training process; a narrow beam scanning module for determining the best narrow beam pair through a perception information assisted narrow beam training process based on the best wide beam; a perception module for obtaining the perception information of the STA, including the position and shape of the STA, by using the trained convolutional neural network (CNN) model during the beam training process; The perception information assisted beam training process comprises: Step 1: the STA sends a perception information assisted beam training request to the AP, the AP receives the request and performs a STA perception process to obtain STA perception information, and adjusts the beam direction based on the STA perception information and the received signal strength; the AP calculates the transmission / reception beam pair with stronger beam signal based on the CSI size, the shape and position of the STA, and the beam width, and obtains a list of best transmission / reception communication beam pairs; Step 2: the AP sends a perception information assisted beam training response to the STA, carrying the list of best reception communication beam pairs calculated by the AP based on the perception information; Step 3: the AP sends training frames to the STA in turn according to the best transmission beam, and the STA receives the training frames in a targeted manner according to the list of best reception communication beams received in step 2, and records the signal reception strength corresponding to the beam pair; Step 4: the STA sends a feedback frame to the AP, feeding back the best transmission beam obtained in the beam training in step 3 and the CSI information of the STA; Step 5: the STA and the AP complete the perception information assisted fast communication beam training to determine the best beam pair.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to realize the WIFI beam fast training method of any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the WIFI beam fast training method of any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the WIFI beam fast training method of any one of claims 1 to 5.

Images