Communication and detection method of dual-frequency communication system

Abstract

According to the communication and detection method of the dual-frequency communication system, the problems of blocking detection and beam forming vector selection are solved by using pilot signals, the process of channel estimation and channel feedback is reduced during the period, and the overhead is greatly reduced; meanwhile, due to the integrated use of sub-6GHz and millimeter waves, the communication system can give full play to the advantages of the two frequency bands, the millimeter wave frequency band can provide high throughput service, the sub-6GHz frequency band can provide wider coverage range and relatively stable communication, it is ensured that user equipment with poor millimeter wave signal quality can perform stable data transmission, and the user experience is improved. The problem of low spectrum efficiency of millimeter waves in a blocking state is solved, and the communication rate is maximized; a feature fusion algorithm is adopted, and blocking state detection and beam forming vector selection can be completed by extracting features of sub-6GHz pilot signals and combining correlation of transmitted and received pilot signals, so that a more accurate classification task is completed.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and more particularly to a communication and detection method for a dual-frequency communication system. Background Art

[0002] The millimeter-wave (mmWave) frequency band, due to its unique advantages, has become a new research hotspot in the communications field. Located between microwaves and infrared, the mmWave band boasts relatively unused spectrum resources and can leverage large bandwidths for high-speed transmission, holding enormous potential for revolutionizing wireless communications. However, its use is subject to several limitations, including high path loss during signal propagation and high susceptibility to link obstruction. These challenges pose significant obstacles to the widespread deployment of mmWave communications. To mitigate severe path loss, millimeter-wave communications use multiple transmit / receive antennas and perform beamforming operations to increase received signal power. Beamforming utilizes multiple antennas to control the direction and shape of radio wave transmission. This technique adjusts the phase and amplitude of the signals transmitted by the antennas to focus the wireless signal in a specific direction, thereby improving signal strength and quality. A readily implemented beamforming design approach is to select the optimal beamforming vector that achieves maximum gain from a given beamforming codebook. The most straightforward approach to selecting the optimal beamforming vector is to exhaustively search all candidate beamforming vectors. However, this approach is computationally expensive and computationally complex when the number of candidate vectors is large. Existing research has proposed several strategies to reduce the cost of beamforming design by leveraging the peculiarities of millimeter-wave channels. Numerous studies have demonstrated similarities between the CSI in the sub-6 GHz band and the millimeter-wave band. Leveraging sub-6 GHz CSI can assist in millimeter-wave beamforming vector selection. In addition, information obtained from the sub-6 GHz band can be used to assist in the establishment of mmWave links.

[0003] However, in real-world situations, there may be link congestion between base stations and users. Millimeter-wave communications have inherent drawbacks such as poor penetration, high propagation loss, and limited coverage. Therefore, it's crucial to select the appropriate frequency band for communication based on the congestion status of the communication link, while also using sub-6 GHz pilot signals for congestion detection and beamforming vector selection to reduce the overhead required for channel estimation and feedback. Summary of the Invention

[0004] In order to solve the problems of possible link blockage between the base station and the user during communication signal transmission, and the inherent defects of millimeter wave communication such as poor penetration ability, large propagation loss and limited coverage, the present invention proposes a communication and detection method for a dual-frequency communication system.

[0005] In order to achieve the above technical effects, the technical solutions of the present invention are as follows: A communication and detection method for a dual-band communication system, the dual-band communication system including a base station and a user, wherein the base station and the user perform dual-band communication via a sub-6 GHz frequency band and a millimeter wave frequency band, and the communication and detection between the base station and the user includes the following steps: Transmitting pilot signals based on the uplink sub-6GHz frequency band and receive pilot signals To detect whether the communication link between the base station and the user is blocked; Depending on whether the communication link between the base station and the user is blocked, the base station will determine the beamforming vector for the downlink transmission; If the communication link is not blocked, the base station sends a signal to the user through the millimeter wave antenna array and selects the optimal beamforming vector from the millimeter wave beamforming codebook; If the communication link is blocked, the base station sends a signal to the user through the sub-6 GHz antenna array and selects the optimal beamforming vector from the sub-6 GHz beamforming codebook.

[0006] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: By using pilot signals to complete blocking detection and beamforming vector selection problems, the channel estimation and channel feedback processes are reduced, and the overhead is greatly reduced. At the same time, the integrated use of sub-6GHz and millimeter waves enables the communication system to give full play to the advantages of the two frequency bands. The millimeter wave band can provide high-throughput services, and the sub-6GHz band can provide a wider coverage range and relatively stable communication, ensuring that user equipment with poor millimeter wave signal quality can perform stable data transmission, overcoming the problem of low millimeter wave spectrum efficiency in the blocked state, and thus maximizing the communication rate. The correlation between the transmitted and received pilot signals is combined to complete blocking state detection and beamforming vector selection. It is suitable for the actual environment of different frequency bands and different numbers of antennas in dual-frequency communication systems and uncertain link communication environments, thereby completing more accurate classification tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 This is a model diagram of a dual-frequency communication system according to an embodiment of the present invention.

[0008] Figure 2 This is a flow chart of a method for joint design of communication and detection according to an embodiment of the present invention.

[0009] Figure 3 This is a structural diagram of the CBD algorithm model shown in an embodiment of the present invention.

[0010] Figure 4 This is a diagram of the basic CNN unit structure shown in an embodiment of the present invention.

[0011] Figure 5 This is a structural diagram of the PSFF algorithm model shown in an embodiment of the present invention.

[0012] Figure 6 This is a bottleneck residual block structure diagram shown in an embodiment of the present invention.

[0013] Figure 7 This is a comparison chart of experimental results of blockage detection accuracy shown in an embodiment of the present invention. DETAILED DESCRIPTION

[0014] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0015] The terms used in this invention are for the purpose of describing specific embodiments only and are not intended to limit the invention. The singular forms "a," "the," and "the" used in this invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0016] It should be understood that although the terms "first," "second," "third," etc. may be used in the present invention to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0017] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Example 1 This embodiment provides a communication and detection method for a dual-frequency communication system. Figure 1As shown, there are base stations and users, and dual-band communication is carried out between the base station and the user through the sub-6GHz frequency band and the millimeter wave frequency band.

[0019] The communication and detection between the base station and the user includes the following steps, the flow chart of which is as follows: Figure 2 As shown: Transmitting pilot signals based on the uplink sub-6GHz frequency band and receive pilot signals To detect whether the communication link between the base station and the user is blocked; Depending on whether the communication link between the base station and the user is blocked, the base station will determine the beamforming vector for the downlink transmission; If the communication link is not blocked, the base station sends a signal to the user through the millimeter wave antenna array and selects the optimal beamforming vector from the millimeter wave beamforming codebook; If the communication link is blocked, the base station sends a signal to the user through the sub-6 GHz antenna array and selects the optimal beamforming vector from the sub-6 GHz beamforming codebook.

[0020] This embodiment uses sub-6 GHz pilot signals to detect obstacles in the communication link system, achieving a sensing function. After sensing, base station transmissions are appropriately scheduled, transmitting millimeter-wave signals to users in an unblocked state and using sub-6 GHz for blocked users. Beamforming design is also implemented using sub-6 GHz pilot signals. This sensing-before-communication approach overcomes the low efficiency of millimeter-wave spectrum in blocked states, thereby maximizing communication rates.

[0021] Example 2 This embodiment is based on the embodiment 1 and changes the transmission pilot signal of the uplink sub-6GHz frequency band to and receive pilot signals The CBD algorithm model and the PSFF algorithm model are input to detect whether the communication link between the base station and the user is blocked.

[0022] CNN, with its unique advantages, demonstrates exceptional capabilities in handling multi-classification problems. It can accurately extract local features between elements and perform structured feature extraction, effectively identifying and classifying complex data patterns. Inspired by this, a CNN-based algorithm was proposed to address the problems faced in dual-band communication systems. Specifically, a deep learning algorithm was used to effectively extract features related to the blocking state and beamforming vector from the pilot signal in the uplink sub-6GHz band, thereby solving the problem under consideration. For the two sub-problems of blocking detection and beamforming vector selection, deep learning algorithms with different structures were designed: the CBD (CNN-based Blockage Detection) algorithm for blocking detection, and the Pilot Signal Feature Fusion (PSFF) algorithm for beamforming vector selection.

[0023] In order to solve the blocking detection sub-problem, it is necessary to establish a mapping relationship between the uplink sub-6GHz received pilot signal and the blocking state. This mapping relationship is implicit and may be highly nonlinear, so a deep learning method is used to learn this mapping relationship. The algorithm for blocking detection is to obtain the received pilot signal from the sub-6GHz Extract features from the link to determine whether the link is in a blocking or non-blocking state.

[0024] Because the dynamic range of the received pilot signal is large, we divide its value by its infinity norm to normalize it.

[0025] In an optional embodiment, the value of the received pilot signal is normalized by dividing it by its infinite norm; the calculation expression is:

[0026] in, represents the normalized received pilot signal, Indicates receiving a pilot signal; Decompose the transmission pilot signal of the sub-6GHz frequency band into and Input the algorithm model and decompose the normalized received pilot signal into and Enter the algorithm model.

[0027] Furthermore, for the input data and , both are complex data, but neural networks cannot extract features from complex data. Decompose into and ,Will Decompose into and .

[0028] In an optional embodiment, the transmission pilot signal and the reception pilot signal based on the uplink sub-6 GHz frequency band are input into a CBD algorithm model to detect whether the communication link between the base station and the user is blocked.

[0029] In an optional embodiment, the CBD algorithm model structure is as follows: Figure 3 Shown, including: The basic CNN module combination, the first maximum pooling layer, the first expansion layer, the first fully connected layer and the first Softmax layer are connected in sequence; wherein the basic CNN module combination includes Basic CNN units; Exemplarily, the basic CNN module combination includes a first CNN unit, using a filter size of 3×3, a channel of 16, a stride of 1, using ReLU as an activation function, using 0 padding, and an output dimension of 4×32×16; The second CNN unit uses a filter size of 3×3, 64 channels, a stride of 1, ReLU as the activation function, 0 padding, and an output dimension of 4×32×64; The third CNN unit uses a filter size of 3×3, a channel of 256, a stride of 1, a ReLU as the activation function, 0 padding, and an output dimension of 4×32×64; Furthermore, the first maximum pooling layer uses maximum pooling, with an input matrix of 4×32×64, a filter of 2×2, and an output matrix of 2×16×64; The first expansion layer: the flattening layer, which converts the multi-dimensional input matrix into one dimension and transitions to the fully connected layer. The input matrix is 2×16×256, which is flattened into a vector of 1×1×2048, and the number of output nodes is 2048; The first fully connected layer has 256 input nodes and 2 output nodes. The first maximum pooling layer and the first fully connected layer use the Dropout mechanism to lose input features; the second fully connected layer is also connected to a softmax layer.

[0030] The basic CNN unit structure is as follows Figure 4 As shown, it includes a first batch normalization layer, a first activation layer, a first convolutional layer, and a first dropout layer connected in sequence; The normalized received pilot signal is decomposed and input into the CBD algorithm model to first extract features to obtain the first feature map with a dimension of 4×32×64. The first feature map is subjected to the maximum pooling operation to obtain the second feature map with a matrix of 2×16×64. Figure 1 After dimensionalization, the blocking state is output through the full connection operation, and the blocking state is converted into the probability of being detected as blocking and non-blocking through the softmax function; Furthermore, after completing data preprocessing, Input into the basic CNN module. The basic CNN module contains multiple basic CNN units, which are mainly used for convolution operations. The features after the convolution operation will be input into the maximum pooling layer (MaxPooling). By reducing the mapping space dimension of the elements, the maximum pooling layer can significantly reduce the number of parameters and the complexity of the model. This operation retains the main features of the input data while filtering out irrelevant information, thereby improving the abstraction and generalization capabilities of the network. The output of the maximum pooling layer is fed into the expansion layer (Flatten) to convert it into a one-dimensional vector. The output is then fed back to the fully connected layer, followed by the softmax layer. This is used to generate a probability distribution of all available categories. Here, and denote the probability of being detected as blocking and non-blocking, respectively.

[0031] The PSFF algorithm model structure is as follows Figure 5 Shown, including: The second batch normalization layer, the second fully connected layer, the second activation layer, the third fully connected layer, and the third activation layer are connected in sequence; The basic CNN module combination, the first bottleneck residual block, the first global average pooling layer, the fourth fully connected layer, and the fourth activation layer are connected in sequence; The first feature fusion module, the third batch normalization layer, the fifth fully connected layer, the fifth activation layer, the second maximum pooling layer, the sixth fully connected layer and the second Softmax layer are connected in sequence; The output end of the third activation layer is connected to the input end of the first feature fusion module, and the output end of the fourth activation layer is connected to the input end of the first feature fusion module; The first bottleneck residual block structure is as follows Figure 6 As shown, it includes a fourth batch normalization layer, a sixth activation layer, a second convolutional layer, a fifth batch normalization layer, a seventh activation layer, a third convolutional layer, a sixth batch normalization layer, and a fourth convolutional layer connected in sequence.

[0032] Exemplarily, the PSFF algorithm model includes an input pilot signal subnetwork model and an output pilot signal subnetwork model; The input pilot signal sub-network model is sequentially connected to the batch normalization layer, the fully connected layer, and the activation layer.

[0033] The second fully connected layer has 1024 input nodes and 1024 output nodes. The third fully connected layer has 1024 input nodes and 512 output nodes. The output pilot signal sub-network model sequentially connects the batch basic CNN module, bottleneck residual block, average pooling layer, fully connected layer, and activation layer.

[0034] The fourth CNN unit uses a filter size of 3×3, 64 channels, a stride of 2, ReLU as the activation function, 0 padding, and an output dimension of 2×16×64.

[0035] The fifth CNN unit uses a filter size of 3×3, 64 channels, a stride of 1, ReLU as the activation function, 0 padding, and an output dimension of 2×16×64; The sixth CNN unit uses a filter size of 3×3, a channel of 32, a stride of 1, a ReLU as the activation function, 0 padding, and an output dimension of 2×16×32; The first bottleneck residual block uses a filter size of 3×3, 16 channels, a dilation of 6 times, a stride of 2, and an output dimension of 1×8×16; The second bottleneck residual block uses a filter size of 3×3, a channel of 64, a dilation of 6 times, a stride of 2, and an output dimension of 1×4×64; The fourth bottleneck residual block uses a filter size of 3×3, a channel of 256, a dilation of 6 times, a stride of 2, and an output dimension of 1×2×256; The fourth fully connected layer has 512 input nodes and 512 output nodes, and uses ReLU as the activation function; The first global average pooling layer input features, the number of output nodes is 512; The input pilot signal sub-network model and the output pilot signal sub-network model are spliced together to complete feature fusion, with the number of output nodes being 1024. The batch normalization layer, the fully connected layer, and the activation layer are sequentially connected. The fifth fully connected layer has 1024 input nodes, uses ReLU as the activation function, and has 512 output nodes. The sixth fully connected layer has 512 input nodes, uses ReLU as the activation function, and has D output nodes, which is the number of beamforming vectors. The Dropout mechanism is used after the pooling layer and the fully connected layer to lose input features; the fifth fully connected layer is also connected to a softmax layer.

[0036] Furthermore, a feature fusion network is proposed using the transmitted and received pilot signals. First, two sub-networks are used to extract features from the transmitted and received signals respectively, and then the features of the two are connected using splicing, and the output features are used for further feature extraction. By combining these two different types of features, a richer and more discerning feature representation is formed. Specifically, through the feature fusion algorithm, features related to the blocking state and beamforming vector are effectively extracted from the pilot signal in the uplink sub-6GHz band, thereby solving the problem under consideration. In the transmitting signal sub-model, a basic neural network is used for feature extraction. The network consists of a multi-layer stack, each layer contains a fully connected layer, an activation layer, a dropout layer, etc., where all fully connected layers have the same width, and each layer has 1024 neurons. In the receiving signal sub-model, the preprocessed data is input into a basic network, which contains a neural network for shallow feature extraction. The output of the base network is then connected in series with the core network. The core network consists of multiple bottleneck residual blocks and adopts the MobileNetV2 structure, which improves the classification performance of the model by mapping low-dimensional features to high-dimensional space. Finally, the feature map undergoes a global average pooling (GAP) operation after the core network. This converts the output of the core network into a fixed-length feature vector and reduces the risk of overfitting. The output is then sent to the fully connected layer, which is used to generate a probability distribution over all available categories, where Indicates the d The probability that the vector is the optimal beamforming vector, where . Perform the softmax operation on the elements within the loss function and then calculate the loss. The Softmax operation formula is as follows:

[0037] in, represents the first vector corresponding to the input to the softmax layer d the value of the element, Indicates the d The probability that an element is correctly detected. The detected blocking state can be determined by identifying the index of the element with the highest probability, which can be calculated using the following equation:

[0038] in, Indicates the blocking state.

[0039] The loss function expression is:

[0040] Among them, if the d beams are actually optimal, then ,otherwise .

[0041] Exemplarily, the optimization algorithm used is Adam.

[0042] In this embodiment, a feature fusion algorithm based on pilot signals is proposed. Different models are used to extract features from the transmitted and received pilot signals respectively. After completing the preliminary feature extraction of the two, the output features are combined through feature splicing to complete a more accurate classification task.

[0043] Example 3 This embodiment further explains the present solution in detail based on Embodiment 1 and Embodiment 2.

[0044] In an optional embodiment, the base station includes two antenna array sets, one antenna array operates in the sub-6GHz frequency band and is equipped with antennas; an antenna array operating in the millimeter wave band, equipped with antennas; the user uses a single antenna in both the sub-6 GHz and millimeter wave bands; the two antenna arrays are located in the same location.

[0045] Furthermore, it is assumed that the two antenna arrays for the sub-6 GHz and millimeter wave bands are located at the same location. Orthogonal Frequency Division Multiplexing (OFDM) transmission is used in both the sub-6 GHz and millimeter wave bands.

[0046] In the uplink, the user sends a pilot signal to the base station for channel estimation. represents the uplink channel coefficient matrix for the sub-6 GHz band, where K is the number of sub-6GHz subcarriers, Indicates in k The uplink channel coefficient vector of the sub-6 GHz array from the user to the base station on the subcarriers.

[0047] In an optional embodiment, a received pilot signal is obtained according to an uplink channel coefficient vector of a sub-6 GHz antenna array; the calculation expression thereof is:

[0048] in, Indicates receiving the pilot signal, Indicates in k The uplink channel coefficient vector of the sub-6 GHz array from the user to the base station on the subcarriers, represents the uplink transmission pilot signal, represents the additive white Gaussian noise in the sub-6GHz band, represents the variance, K Indicates the number of sub-6 GHz subcarriers.

[0049] Furthermore, for sub-6GHz communication links, the uplink CSI on each subcarrier can be obtained using channel estimation. For example, the common least squares (LS) channel estimation method can be used to obtain the channel state information. .

[0050] use represents the downlink channel coefficient matrix of the sub-6GHz frequency band. In addition, the downlink channel coefficient matrix of the millimeter wave frequency band is used Indicates that Q is the number of subcarriers in the millimeter wave band, Indicates in Q The downlink channel coefficient vector from the base station's millimeter wave array to the user on subcarriers.

[0051] In the sub-6GHz frequency band, the base station transmits power In the k Subcarriers transmit signals , the user's received signal can be expressed as:

[0052] in, is the beamforming vector of the sub-6 GHz antenna array of the base station, which is selected from the sub-6 GHz beamforming codebook .

[0053] use Indicates the transmission signal from the user to the base station, Indicates receiving a signal. and They are k The transmitted and received signals of the subcarriers.

[0054] In an optional embodiment, the transmission pilot signal based on the uplink sub-6GHz frequency band is and receive pilot signals Input the CBD algorithm model and the PSFF algorithm model to obtain the detection result of whether the communication link between the base station and the user is blocked, and calculate the probability that the detection result of the blocking state is consistent with the actual blocking state; its mathematical expression is:

[0055] in, Defined as known Events under the conditions The probability of Indicates the actual blocking status between the base station and the user, Indicates the result of detecting the blocking state. Indicates the transmission of pilot signal, Indicates receiving the pilot signal.

[0056] Furthermore, the base station will select an appropriate frequency band for downlink transmission and will determine whether the beamforming vector selection sub-problem is for the sub-6 GHz band or the millimeter wave band. l Indicates the actual blocking status between the base station and the user. If the link is not blocked, then l = 0; otherwise, l = 1. It is represented as the result of the blocking detection sub-problem. If it is detected that the link between the base station and the user equipment is unblocked, ;otherwise, The goal of the blocking detection subproblem is to maximize the detection accuracy of the blocking state.

[0057] Furthermore, the detected blocking state is known. If , the base station sends signals to users through the sub-6GHz antenna array, and the beamforming vector selection subproblem aims to select the sub-6GHz beamforming codebook Otherwise, the base station sends a signal to the user through the millimeter wave antenna array. The beamforming vector selection subproblem aims to select the optimal beamforming vector from the millimeter wave beamforming codebook. Select the optimal beamforming vector. Use the uplink sub-6GHz frequency band to transmit and receive signals. To solve the beamforming vector selection problem, this can reduce the overhead of channel estimation and channel feedback. and The beamforming vector selection subproblem aims to select the optimal sub-6GHz and millimeter wave beamforming vectors based on the received and transmitted pilot signals. , maximizes the probability of obtaining the optimal beamforming vector from a given codebook, which can maximize the downlink communication rate.

[0058] In an optional embodiment, if the communication link is not blocked, the base station sends a signal to the user through the millimeter wave antenna array and selects the optimal beamforming vector from the millimeter wave beamforming codebook; its mathematical expression is:

[0059] in, Defined as known Events under the conditions The probability of Indicates the sub-6GHz band beamforming vector selection result, represents the optimal beamforming vector in the sub-6GHz band, Indicates the transmission of pilot signal, Indicates receiving a pilot signal; If the communication link is blocked, the base station sends a signal to the user through the sub-6 GHz antenna array and selects the optimal beamforming vector from the sub-6 GHz beamforming codebook; its mathematical expression is:

[0060] in, Defined as known Events under the conditions The probability of Indicates the beamforming vector selection result in the millimeter wave band, represents the optimal beamforming vector in the millimeter wave band, Indicates the transmission of pilot signal, Indicates receiving the pilot signal.

[0061] Furthermore, the blocking detection subproblem can be viewed as a binary classification problem with two categories, namely blocking and non-blocking states. The beamforming vector selection subproblem is a multi-classification problem with two cases. As for the beamforming vector selection subproblem, it belongs to a multi-classification problem with two different cases. When the link state is determined to be blocked by the blocking detection subproblem, the sub-GHz band will be used for downlink communication due to its relatively insensitive characteristics to blocking. At this point, a problem with multi-classification problem with multiple categories. categories, which actually correspond to the sub-6GHz band beamforming codebook in When the blocking detection result is non-blocking, the millimeter wave band will be used for downlink communication due to its significant advantages of large bandwidth and high-speed communication. multi-classification problem with multiple categories. The categories also correspond to the millimeter wave band beamforming codebook in beamforming vector options.

[0062] Indicates the icandidate beamforming vectors, and In (2.2), represents the additive white Gaussian noise in the sub-6GHz band, where Denotes the variance. Indicates the k The bandwidth of the sub-6GHz subcarriers.

[0063] In the millimeter wave frequency band, the base station transmits power In the q Subcarriers transmit signals , the user's received signal can be expressed as:

[0064] in, is the beamforming vector of the millimeter wave antenna array of the base station, which selects the millimeter wave beamforming codebook , Represents additive white Gaussian noise in the millimeter wave band.

[0065] Further, Indicates the j candidate beamforming vectors, and .

[0066] In an optional embodiment, the downlink communication rate of the millimeter wave frequency band is obtained according to the optimal beamforming vector in the millimeter wave beamforming codebook; the calculation expression is:

[0067] in, represents the downlink channel coefficient matrix of the millimeter wave frequency band, represents the beamforming vector of the base station’s millimeter-wave antenna array, Indicates the q subcarriers, Indicates the q The bandwidth of millimeter-wave subcarriers, Indicates the transmit power, represents the variance, represents the downlink channel coefficient vector from the base station’s millimeter-wave array to the user, represents the beamforming vector.

[0068] In an optional embodiment, the downlink communication rate in the sub-6 GHz frequency band is obtained according to the optimal beamforming vector in the sub-6 GHz beamforming codebook; the calculation expression is:

[0069] in, Sub-6GHz band downlink channel coefficient matrix, represents the beamforming vector of the sub-6GHz antenna array of the base station, Indicates the k subcarriers, Indicates the k The bandwidth of sub-6GHz subcarriers, Indicates the transmit power, represents the variance, represents the downlink channel coefficient vector from the sub-6GHz array of the base station to the user, represents the beamforming vector.

[0070] In this embodiment, a pilot signal in the sub-6 GHz band is used to detect whether the link between the base station and the user is blocked or unblocked. The downlink beamforming vector is selected based on the result of the blocking detection subproblem. The optimal beamforming vector is selected from the beamforming codebook for the sub-6 GHz band or from the beamforming codebook for the millimeter wave band to maximize the achievable downlink communication rate.

[0071] This embodiment also provides a comparison chart of experimental results on the accuracy of blockage detection, such as Figure 7 Shown, showing The relationship between the blocking detection accuracy and SNR when the SNR is 2, 4, 8, and 16 respectively.

[0072] Transmitting and receiving pilot signals also demonstrate high performance in blocking detection, with detection accuracy increasing with increasing SNR. At an SNR of 20dB, the accuracy reaches over 90%. Results demonstrate that the proposed feature fusion algorithm can achieve excellent blocking detection using pilot signals in the sub-6GHz band. Therefore, using transmitting and receiving pilot signals to detect link blocking is a viable approach.

[0073] Each embodiment of the present invention is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiment. The device embodiment described above is merely exemplary. The modules described as separate components may or may not be physically separated. When implementing the scheme of the present invention, the functions of each module can be implemented in the same one or more software and / or hardware. It is also possible to select some or all of the modules according to actual needs to achieve the purpose of the scheme of this embodiment.

[0074] Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A communication and detection method for a dual-band communication system, wherein the dual-band communication system includes a base station and a user, and the base station and the user perform dual-band communication via a sub-6 GHz frequency band and a millimeter wave frequency band, characterized in that: The communication and detection between the base station and the user includes the following steps: Transmitting pilot signals based on the uplink sub-6GHz frequency band and receive pilot signals To detect whether the communication link between the base station and the user is blocked; Depending on whether the communication link between the base station and the user is blocked, the base station will determine the beamforming vector for the downlink transmission; If the communication link is not blocked, the base station sends a signal to the user through the millimeter wave antenna array and selects the optimal beamforming vector from the millimeter wave beamforming codebook; If the communication link is blocked, the base station sends a signal to the user through the sub-6 GHz antenna array and selects the optimal beamforming vector from the sub-6 GHz beamforming codebook.

2. The communication and detection method of a dual-frequency communication system according to claim 1, characterized in that: The base station includes two antenna array sets, one antenna array operates in the sub-6GHz frequency band and is equipped with antennas; the other antenna array operates in the millimeter wave band and is equipped with antennas; the user uses a single antenna in both the sub-6 GHz and millimeter wave bands; the two antenna arrays are located in the same location.

3. The communication and detection method of a dual-frequency communication system according to claim 1, characterized in that: The received pilot signal is obtained based on the uplink channel coefficient vector of the sub-6GHz antenna array; its calculation expression is: in, Indicates receiving the pilot signal, Indicates in k The uplink channel coefficient vector of the sub-6 GHz array from the user to the base station on the subcarriers, represents the uplink transmission pilot signal, represents the additive white Gaussian noise in the sub-6GHz band, represents the variance, K Indicates the number of sub-6 GHz subcarriers.

4. The communication and detection method of a dual-frequency communication system according to claim 1, wherein: The value of the received pilot signal is divided by its infinite norm for normalization; the calculation expression is: in, represents the normalized received pilot signal, Indicates receiving the pilot signal.

5. The communication and detection method of a dual-frequency communication system according to claim 1, wherein: The transmission pilot signal and the reception pilot signal based on the uplink sub-6 GHz frequency band are input into the CBD algorithm model to detect whether the communication link between the base station and the user is blocked.

6. The communication and detection method of a dual-frequency communication system according to claim 5, characterized in that: The CBD algorithm model includes: The basic CNN module combination, the first maximum pooling layer, the first expansion layer, the first fully connected layer and the first Softmax layer are connected in sequence; wherein the basic CNN module combination includes Basic CNN units; The basic CNN unit includes a first batch normalization layer, a first activation layer, a first convolutional layer and a first dropout layer connected in sequence; The normalized received pilot signal is decomposed and input into the CBD algorithm model. Feature extraction is first performed to obtain a first feature map with a dimension of 4×32×64. The first feature map is subjected to a maximum pooling operation to obtain a second feature map with a matrix of 2×16×64. The second feature map is then converted to a one-dimensional state through a full connection operation to output the blocking state. The blocking state is converted into the probability of being detected as blocked or non-blocked through the softmax function. The PSFF algorithm model includes: The second batch normalization layer, the second fully connected layer, the second activation layer, the third fully connected layer, and the third activation layer are connected in sequence; The basic CNN module combination, the first bottleneck residual block, the first global average pooling layer, the fourth fully connected layer, and the fourth activation layer are connected in sequence; The first feature fusion module, the third batch normalization layer, the fifth fully connected layer, the fifth activation layer, the second maximum pooling layer, the sixth fully connected layer and the second Softmax layer are connected in sequence; The output end of the third activation layer is connected to the input end of the first feature fusion module, and the output end of the fourth activation layer is connected to the input end of the first feature fusion module; The first bottleneck residual block includes a fourth batch normalization layer, a sixth activation layer, a second convolutional layer, a fifth batch normalization layer, a seventh activation layer, a third convolutional layer, a sixth batch normalization layer, and a fourth convolutional layer, which are connected in sequence.

7. The communication and detection method of a dual-frequency communication system according to claim 6, characterized in that: The transmission pilot signal based on the uplink sub-6GHz frequency band and receive pilot signals The CBD algorithm model is input to detect whether the communication link between the base station and the user is blocked. The probability that the detection result of the blocking state is consistent with the actual blocking state is calculated. The mathematical expression is: in, Defined as known Events under the conditions The probability of Indicates the actual blocking status between the base station and the user, Indicates the detection result of the blocking state, Indicates the transmission of pilot signal, Indicates receiving the pilot signal.

8. The communication and detection method of a dual-frequency communication system according to claim 1, characterized in that: If the communication link is not blocked, the base station sends a signal to the user through the millimeter wave antenna array and selects the optimal beamforming vector from the millimeter wave beamforming codebook; its mathematical expression is: in, Defined as known Events under the conditions The probability of Indicates the sub-6GHz band beamforming vector selection result, represents the optimal beamforming vector in the sub-6GHz band, Indicates the transmission of pilot signal, Indicates receiving a pilot signal; If the communication link is blocked, the base station sends a signal to the user through the sub-6 GHz antenna array and selects the optimal beamforming vector from the sub-6 GHz beamforming codebook; its mathematical expression is: in, Defined as known Events under the conditions The probability of Indicates the beamforming vector selection result in the millimeter wave band, represents the optimal beamforming vector in the millimeter wave band, Indicates the transmission of pilot signal, Indicates receiving the pilot signal.

9. The communication and detection method of a dual-frequency communication system according to any one of claims 1 to 8, characterized in that: The downlink communication rate in the millimeter wave band is obtained based on the optimal beamforming vector in the millimeter wave beamforming codebook; its calculation expression is: in, represents the downlink channel coefficient matrix of the millimeter wave frequency band, represents the beamforming vector of the base station’s millimeter-wave antenna array, Indicates the q subcarriers, Indicates the q The bandwidth of millimeter-wave subcarriers, Indicates the transmit power, represents the variance, represents the downlink channel coefficient vector from the base station’s millimeter-wave array to the user, represents the beamforming vector.

10. The communication and detection method of a dual-frequency communication system according to any one of claims 1 to 9, characterized in that: The downlink communication rate in the sub-6 GHz band is obtained based on the optimal beamforming vector in the sub-6 GHz beamforming codebook; its calculation expression is: in, Sub-6GHz band downlink channel coefficient matrix, represents the beamforming vector of the sub-6GHz antenna array of the base station, Indicates the k subcarriers, Indicates the k The bandwidth of sub-6GHz subcarriers, Indicates the transmit power, represents the variance, represents the downlink channel coefficient vector from the sub-6GHz array of the base station to the user, represents the beamforming vector.

Images