Wireless communication link quality confidence interval prediction method and system based on Transform and multi-head map attention network

Abstract

The invention discloses a wireless communication link quality confidence interval prediction method and system based on Transform and a multi-head graph attention network. The method comprises the following steps: constructing a normalized quality evaluation data matrix sequence of a wireless communication link and an adjacent matrix of a network topology structure; decomposing the normalized quality evaluation data matrix sequence into a stationary data matrix sequence and a noise data matrix sequence by adopting wavelet decomposition; calculating a standard deviation data matrix sequence of the noise data matrix sequence; capturing stationary data matrix sequence residual structure output and standard deviation data matrix sequence residual structure output with spatial features by using a residual structure multi-head image attention network; and predicting the output of the next moment and calculating the quality confidence interval of the wireless communication link. According to the method, the wavelet decomposition, the residual structure multi-head map attention network and the Transform model are combined, the spatial-temporal characteristics of the link quality data are effectively extracted, and a more accurate basis is provided for prediction of the wireless communication link quality confidence interval.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless communication link quality confidence interval prediction, and in particular to a method and system for wireless communication link quality confidence interval prediction based on Transformer and multi-head graph attention network. Background Art

[0002] In wireless communication systems, link quality assessment and prediction are critical to ensuring network performance and assisting high-level protocol design. Traditional methods such as hard metrics based on physical layer parameters and soft metrics based on packet reception statistics have many defects. Hard metrics are limited in measurement and susceptible to interference, while soft metrics are insensitive to short-term link changes. Methods based on statistics and curve fitting to construct mapping relationships rely on large amounts of data and fixed time windows, making it difficult to adapt to the time-varying characteristics of links, and the burst characteristics of links further highlight their limitations.

[0003] At the same time, the characteristics of wireless links are complex. There are connected, transitional, and non-connected areas in space. The transitional areas fluctuate greatly and the identification of intermediate links in different device scenarios varies greatly. In terms of time, intermediate links are sudden and traditional methods are difficult to cope with their rapid changes. All of these require the ability to accurately capture the spatiotemporal changes of links, comprehensively weigh multiple factors such as accuracy, speed, and stability, and flexibly adapt to different scenarios. However, existing methods are difficult to meet these requirements and new solutions are urgently needed. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network. The method combines wavelet decomposition, graph attention network and Transformer model to effectively extract the spatiotemporal characteristics of link quality data, providing a more accurate basis for the prediction of wireless communication link quality confidence interval.

[0005] In order to achieve the above object, the present invention provides the following technical solutions: The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network provided by the present invention comprises the following steps: S1, constructs a normalized quality assessment data matrix sequence of wireless communication links and an adjacency matrix of the network topology structure; S2, using wavelet decomposition to decompose the normalized quality assessment data matrix sequence into a stationary data matrix sequence and a noise data matrix sequence; S3, calculate the standard deviation data matrix sequence of the noise data matrix sequence; S4, using the residual structure multi-head graph attention network to capture the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence; S5, the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence are respectively input into the Transformer to predict the output at the next moment and calculate the confidence interval of the wireless communication link quality.

[0006] Furthermore, in step S1, constructing a normalized quality assessment data matrix sequence of a wireless communication link includes the following steps: The original data is modeled as ; Among them, the matrix represents the kth wireless communication link quality assessment data matrix recorded within the time range t, K represents the number of all data matrices recorded within the time range t, N represents the total number of nodes in the wireless communication link network topology structure, and C represents the type of parameters for evaluating the quality of the wireless communication link.

[0007] The quality assessment data matrix sequence is normalized and mapped to [0,1]. The normalization process is: ; in, is the original data under the i-th quality assessment parameter, and are the maximum and minimum values ​​in the data under the i-th quality assessment parameter, respectively. is the normalized data under the i-th quality assessment parameter.

[0008] Furthermore, in step S1, constructing the adjacency matrix A of the network topology structure includes the following steps: Numbering N nodes in the wireless communication link network from 0 to N-1; create The all-zero matrix A is used as the adjacency matrix; Add weight information to edges ,make ,in Represents a node, represents the neighboring nodes of the i-th node, Represents the weight value of the edge between node i and node j.

[0009] Further, in step S2, the wavelet decomposition comprises the following steps: Select the wavelet basis function as db series wavelet; Determine the threshold , where t is the number of elements in the normalized quality assessment data matrix sequence; Determine the number of decomposition levels L=2; Determine the threshold function ; Where b is the wavelet coefficient of the original sequence on the second layer during the wavelet decomposition process; Decompose the normalized quality assessment data matrix sequence with wavelet to obtain a stationary data matrix sequence and the noise data matrix sequence ; in, represents the kth stationary data matrix within the time range t, represents the kth noise data matrix within the time range t.

[0010] Further, in step S3, calculating the standard deviation data matrix sequence of the noise data matrix sequence includes the following steps: The standard deviation of the noise data matrix sequence is expressed as: ; in, is the standard deviation data matrix, N is the total number of nodes, C is the type of quality assessment parameters, and t is the time step; Calculate the standard deviation using the formula: ; in, , is the standard deviation value of the i-th node under the j-th quality evaluation parameter at the K-th time step, G is the length of the standard deviation window used to calculate, is the noise data value under the jth quality evaluation parameter of the ith node in m time steps.

[0011] Further, in step S4, the residual structure multi-head graph attention network captures the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence, including the following steps: The stationary data matrix sequence and the adjacency matrix A of the network topology as input; The stationary data matrix sequence The linear transformation is , the linear transformation process is: ; in, represents the quality assessment data matrix recorded at the kth time step within the time range t, is the transformed feature matrix, is the linear transformation matrix; The attention mechanism of the graph attention network is used to calculate the correlation coefficient between nodes. The correlation coefficient calculation formula is: ; in, represents the attention coefficient between node i and its neighbor node j at the kth time step, is a nonlinear activation function, is the initialized learnable matrix, and Respectively represent the feature vectors of node i and its neighbor node j at the kth time step; Normalize the attention coefficient: ; in, is the set of neighbor nodes of node i; The multi-head attention mechanism is used to obtain a stable data space feature matrix sequence: ; in, is the spatial feature matrix extracted at the kth time step, N is the total number of nodes, and M is the spatial feature dimension of a single node at a certain time step.

[0012] The standard deviation data matrix sequence and the adjacency matrix A of the network topology as input; The standard deviation data space feature matrix sequence is obtained by using the same stationary data matrix sequence steps: ; Furthermore, the multi-head attention mechanism is used to obtain a stationary data space feature matrix sequence, which specifically includes the following steps: The weighted sum of the features of neighboring nodes is used to obtain a single-head output: in, is a nonlinear activation function, is the normalized attention coefficient; Concatenate the multi-head attention outputs. The concatenation process is: ; in, represents the output feature of the pth attention head of the i-th node, represents the output feature of the p-th attention head of neighbor node j, represents the feature vector of the i-th node at the k-th time step, Used to adjust the concatenated feature dimension to the target dimension; Further, in step S4, the design steps of the residual structure are: The stationary data space feature matrix sequence extracted by the multi-head graph attention network And standard deviation data space characteristic matrix sequence Perform batch normalization, the calculation formula is: ; ; in and They are respectively the batch normalized stationary data space feature matrix sequence and the standard deviation data space feature matrix sequence; The batch-normalized stationary data space feature matrix sequence And standard deviation data space characteristic matrix sequence They are connected to the previous layer input residual respectively, and the process of the first layer residual connection is: ; ; in is the residual structure output of the stationary data matrix sequence, The residual structure output for the and standard deviation data matrix sequence; Will and Then input the above residual structure to obtain the residual structure output of the final stationary data matrix sequence and Residual structure output for standard deviation data matrix sequence ; Furthermore, the step S5, inputting the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence into the Transformer to predict the output at the next moment and integrating them to obtain the confidence interval of the wireless communication link quality includes the following steps: Output the residual structure of the stationary data matrix sequence and Residual structure output for standard deviation data matrix sequence Expand by time step; in represents the standard deviation spatial feature matrix of the kth time step, N is the total number of nodes, and C is the type of quality assessment parameters; The matrix for each time step is and Add dynamic position coding respectively, the formula is: in and is the node-level position encoding matrix generated based on time step k, which is used to preserve temporal and spatial associations; The matrix sequence after position encoding and The Transformer encoder is input separately, and the temporal dependency between time steps is captured through the self-attention mechanism, and the predicted stable data feature matrix for the next moment is output. and standard deviation feature matrix ; right and Perform node-level denormalization to restore the original data scale. The formula is: in , and , are the original data and standard deviation extreme value of the i-th node respectively; Calculate the node-level denormalized and Get the wireless communication link quality confidence lower limit matrix of C quality assessment parameters and the upper confidence limit matrix ; in, is a standard normal distribution Quantile, represents the significance level; The confidence level of prediction under different quality assessment parameters is weighted averaged to obtain the reliability confidence interval matrix containing all nodes. ; in is the weight matrix of different quality assessment parameter types, satisfying , C is the type of quality assessment parameter; The wireless communication link quality confidence interval prediction system based on Transformer and graph attention network provided by the present invention includes a memory, a processor and a computer program stored in the memory and executable on the processor, and the above method is implemented when the processor executes the program.

[0013] The beneficial effects of the present invention are: The invention provides a wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network, and the invention relates to the field of wireless communication link quality confidence interval prediction, including constructing a normalized quality assessment data matrix sequence of a wireless communication link and an adjacency matrix of a network topology structure; using wavelet decomposition to decompose the normalized quality assessment data matrix sequence into a stationary data matrix sequence and a noise data matrix sequence; calculating the standard deviation data matrix sequence of the noise data matrix sequence; using a residual structure multi-head graph attention network to capture the stationary data matrix sequence residual structure output with spatial characteristics and the standard deviation data matrix sequence residual structure output; inputting the stationary data matrix sequence residual structure output with spatial characteristics and the standard deviation data matrix sequence residual structure output into the Transformer to predict the output at the next moment and calculate the wireless communication link quality confidence interval. The method of the invention combines wavelet decomposition, residual structure multi-head graph attention network and Transformer model, effectively extracts the spatiotemporal characteristics of link quality data, and provides a more accurate basis for the prediction of wireless communication link quality confidence intervals.

[0014] In this method, the wireless communication link quality assessment data is a multivariate time series, which is modeled as a matrix sequence. The link quality assessment parameters such as the signal-to-noise ratio, signal strength, and packet reception and retransmission of each node can be predicted simultaneously. The confidence interval of a single parameter is easily affected by noise and outliers, while combining multiple parameters can reduce the impact of these factors and make the final reliability confidence interval more robust.

[0015] This method uses the wavelet decomposition preprocessing mechanism to decompose the wireless communication link quality assessment data matrix sequence into a stationary matrix sequence and a non-stationary matrix sequence, which is conducive to quickly identifying and predicting the short-term changes in the intermediate link quality and effectively capturing the burst characteristics of the link. This decomposition method can also reduce data complexity, allowing subsequent models to focus more on data with different characteristics and improve the model's ability to process complex data.

[0016] This method uses a multi-head graph attention network. The multi-head attention mechanism can simultaneously focus on different feature subspaces of nodes and capture the complex relationships between nodes from multiple angles. In the wireless communication link quality prediction scenario, different attention heads can focus on different link quality evaluation parameters such as received signal strength, signal-to-noise ratio, packet reception and retransmission, so as to fully explore the spatial correlation between nodes. Compared with single-head attention, multi-head attention provides richer information, which helps the model to more carefully characterize the node relationship.

[0017] This method models the wireless communication link data sequence and network topology structure. It first extracts spatial features through the residual structure of the graph attention network, comprehensively considers the spatial correlation between nodes with similar geographical locations and the temporal characteristics between sequences, and can be applied to the scenario of offload transmission, providing a more fine-grained and flexible option for wireless communication link quality assessment and prediction. The results are then input into the Transformer, which can adaptively capture the long-distance time dependency of link quality data and mine complex time associations. The combination of the two can fully capture the spatiotemporal characteristics of wireless communication link quality data.

[0018] This method combines wavelet decomposition, Transformer and multi-head graph attention network technology to predict the confidence interval of wireless communication link quality, providing network managers with key information about the reliability of predictions, helping them make more scientific and reasonable decisions.

[0019] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for explanation.

[0021] Figure 1 This is a simplified schematic diagram of the process.

[0022] Figure 2 Graph of the multi-head graph attention network of our method.

[0023] Figure 3 This is the principle framework diagram of this method.

[0024] Figure 4 This is a comparison chart of the prediction results of various models for the wireless communication link quality of node 1 under three parameters: RSSI, SNR and PRT.

[0025] Figure 5 is the average error of each model under different parameters. DETAILED DESCRIPTION

[0026] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0027] refer to Figure 1 As shown, Figure 1is a simplified flow chart of the method. This embodiment discloses a wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network, including the following steps: S1, constructs a normalized quality assessment data matrix sequence of wireless communication links and an adjacency matrix of the network topology structure; It should be noted that the quality assessment parameters of the wireless communication link include hard metrics (received signal strength, signal-to-noise ratio, link quality indication, etc.), soft metrics (data link layer packet reception or retransmission time) and hybrid metrics. In this embodiment, three wireless communication link quality assessment parameters are used, namely, the signal-to-noise ratio (SNR) and signal strength (RSSI) in the hard metrics and the data link layer packet reception or retransmission time statistics (PRT) in the soft metrics.

[0028] It should be noted that during data collection, a wireless communication network consisting of 5 nodes was constructed. The nodes were randomly distributed within the range of 0 to 100 meters, and the positions were located by GPS. Each node was equipped with an IEEE 802.11 wireless communication module that supports 2.4GHz and 5GHz dual-bands. The data collection frequency was once every 10 minutes for 24 hours. The collection parameters included received signal strength (RSSI, in dBm), signal-to-noise ratio (SNR, in dB), and packet reception time (PRT, in ms). During the collection process, random noise and periodic interference were introduced, and burst interference was simulated during the 8:00-10:00 and 18:00-20:00 periods every day, resulting in a decrease in RSSI, a decrease in SNR, and an increase in PRT. The collected data is stored in the form of a matrix, and the matrix dimensions correspond to time points, nodes, and quality assessment parameters, respectively.

[0029] It should be noted that constructing a normalized quality assessment data matrix sequence of a wireless communication link in S1 includes the following steps: S101, the original data is modeled as ; Among them, the matrix represents the kth wireless communication link quality assessment data matrix recorded within the time range t, K represents the number of all data matrices recorded within the time range t, N represents the total number of nodes in the wireless communication link network topology structure, and C represents the type of parameters for evaluating the quality of the wireless communication link.

[0030] S102, normalize and map the quality assessment data matrix sequence to [0,1], the normalization process is: in, is the original data under the i-th quality assessment parameter, and are the maximum and minimum values ​​in the data under the i-th quality assessment parameter, respectively. is the normalized data under the i-th quality assessment parameter.

[0031] It should be noted that the adjacency matrix A for constructing the network topology structure in S1 includes the following steps: S111, numbering N nodes in the wireless communication link network from 0 to N-1; S112, Create The all-zero matrix A is used as the adjacency matrix; S113, add weight information to the edge ,make ,in Represents a node, Represents the neighbor node of the i-th node; Represents the weight value of the edge between node i and node j.

[0032] S2, using wavelet decomposition to decompose the normalized quality assessment data matrix sequence into a stationary data matrix sequence and a noise data matrix sequence; It should be noted that the wavelet decomposition in S2 includes the following steps: S201, selecting the wavelet basis function as the db series wavelet; S202, determine threshold , where t is the number of elements in the normalized quality assessment data matrix sequence; S203, determine the decomposition level L=2; S204, determine the threshold function ; Where b is the wavelet coefficient of the original sequence on the second layer during the wavelet decomposition process; S205, decomposing the normalized quality assessment data matrix sequence by wavelet to obtain a stationary data matrix sequence and the noise data matrix sequence ; in, represents the kth stationary data matrix within the time range t, represents the kth noise data matrix within the time range t.

[0033] S3, calculate the standard deviation data matrix sequence of the noise data matrix sequence; It should be noted that the calculation of the standard deviation data matrix sequence of the noise data matrix sequence in S3 includes the following steps: S301, the standard deviation data matrix sequence of the noise data matrix sequence is expressed as: ; in, is the standard deviation data matrix, N is the total number of nodes, C is the type of quality assessment parameters, and t is the time step; S302, calculate the standard deviation, the formula is: ; in, , is the standard deviation value of the i-th node under the j-th quality evaluation parameter at the K-th time step, G is the length of the standard deviation window used to calculate, is the noise data value under the jth quality evaluation parameter of the ith node in m time steps.

[0034] S4, using the residual structure multi-head graph attention network to capture the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence; It should be noted that in S4, the residual structure multi-head graph attention network captures the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence, which includes the following steps: S401, the stationary data matrix sequence and the adjacency matrix A of the network topology as input; S402, the stationary data matrix sequence The linear transformation is , the linear transformation process is: in, represents the quality assessment data matrix recorded at the kth time step within the time range t, is the transformed feature matrix, is the linear transformation matrix; S403, using the attention mechanism of the graph attention network to calculate the correlation coefficient between nodes, the correlation coefficient calculation formula is: ; in, represents the attention coefficient between node i and its neighbor node j at the kth time step, is a nonlinear activation function, is the initialized learnable matrix, and Respectively represent the feature vectors of node i and its neighbor node j at the kth time step; S404, normalize the attention coefficient: ; in, is the set of neighbor nodes of node i; S405, use the multi-head attention mechanism to obtain a stable data space feature matrix sequence: ; in, is the spatial feature matrix extracted at the kth time step, N is the total number of nodes, and M is the spatial feature dimension of a single node at a certain time step.

[0035] S406, the standard deviation data matrix sequence and the adjacency matrix A of the network topology as input; S407, using the same stationary data matrix sequence step to obtain the standard deviation data space feature matrix sequence: ; It should be noted that obtaining a stationary data space feature matrix sequence by using a multi-head attention mechanism in S405 includes the following steps: S411, weighted sum of features of neighboring nodes to obtain single-head output: ,in, is a nonlinear activation function, is the normalized attention coefficient; S412, concatenate the multi-head attention outputs. The concatenation process is: ; in, represents the output feature of the p-th attention head of the i-th node, represents the output feature of the p-th attention head of neighbor node j, represents the feature vector of the i-th node at the k-th time step, It is used to adjust the concatenated feature dimension to the target dimension. Figure 2 As shown, Figure 2 The three single-head outputs are shown in the figure merged together through the concat function.

[0036] It should be noted that, in S4, the design steps of the residual structure are: S421, extracting the stationary data space feature matrix sequence of the multi-head graph attention network And standard deviation data space characteristic matrix sequence Perform batch normalization, the calculation formula is: ; ; in and They are respectively the batch normalized stationary data space feature matrix sequence and the standard deviation data space feature matrix sequence; S422, batch normalized stationary data space feature matrix sequence And standard deviation data space characteristic matrix sequence They are connected to the previous layer input residual respectively, and the process of the first layer residual connection is: ; ; in is the residual structure output of the stationary data matrix sequence, The residual structure output for the and standard deviation data matrix sequence; S423, will and Then input the above residual structure to obtain the residual structure output of the final stationary data matrix sequence and Residual structure output for standard deviation data matrix sequence ; S5, the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence are respectively input into the Transformer to predict the output at the next moment and calculate the confidence interval of the wireless communication link quality.

[0037] In summary, the quality assessment data matrix sequence and the adjacency matrix are analyzed and processed, and the stationary data prediction results and the standard deviation prediction results are obtained through the residual block and the transformer respectively, and finally the confidence interval prediction results are output; Figure 3 As shown, Figure 3 This is the principle framework diagram of this method. Figure 3 The working principle of each module in this method is shown in FIG. It should be noted that in S5, the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence are respectively input into the Transformer to predict the output at the next moment and integrate to obtain the confidence interval of the wireless communication link quality, which specifically includes the following steps: S501, output the residual structure of the stationary data matrix sequence as well as Residual structure output for standard deviation data matrix sequence Expand by time step; in represents the standard deviation spatial feature matrix of the kth time step, N is the total number of nodes, and C is the type of quality assessment parameters; S502, for each time step matrix and Add dynamic position coding respectively, the formula is: , ; in and is the node-level position encoding matrix generated based on time step k, which is used to preserve temporal and spatial associations; S503, the matrix sequence after position encoding and The Transformer encoder is input separately, and the temporal dependency between time steps is captured through the self-attention mechanism, and the predicted stable data feature matrix for the next moment is output. and standard deviation feature matrix ; S504, calculate the node-level denormalized and Get the wireless communication link quality confidence lower limit matrix of C quality assessment parameters and the upper confidence limit matrix ; in, is a standard normal distribution Quantile, represents the significance level; S505: weighted average of the predicted confidence levels under different quality assessment parameter types to obtain a reliability confidence interval matrix containing all nodes ; in is the weight matrix of different quality assessment parameter types, satisfying , C is the type of quality assessment parameter.

[0038] like Figure 4 As shown, Figure 4 This is a comparison chart of the prediction results of each model for the three parameters of RSSI, SNR and PRT for the wireless communication link quality of node 1. It can be seen from the figure that this method combines wavelet decomposition, multi-head graph attention network and Transformer model, and shows higher fitting accuracy in the prediction of the three parameters of RSSI, SNR and PRT. The degree of fit between its prediction curve and the true value is significantly better than that of other models, especially during the period of sudden interference. The model provided by this embodiment can adapt to the drastic changes in link quality more quickly, and the prediction results fluctuate less, showing higher robustness and stability. In contrast, the prediction errors of other models during the period of sudden interference are large, and the curves deviate significantly from the true value.

[0039] like Figure 5 As shown, Figure 5It is the average error of each model under different parameters. It can be seen from the figure that the average error comparison of this method and other models in three parameters of received signal strength, signal-to-noise ratio and packet reception time. The average error of the model provided by this method in all parameters is significantly lower than that of other models, indicating that its prediction accuracy is higher. In contrast, the average error of the LSTM model is the highest, and the errors of the GAT-LSTM and GAT-Transformer models are between the two. This shows that the model provided in this embodiment can effectively extract the spatiotemporal characteristics of link quality data, reduce prediction errors, and provide a reliable basis for wireless communication link quality assessment and prediction by combining wavelet decomposition, multi-head graph attention network and Transformer model. This embodiment also provides a wireless communication link quality confidence interval prediction system based on Transformer and multi-head graph attention network, including a memory, a processor and a computer program stored in the memory and run on the processor, and the processor implements the above method when executing the program.

[0040] The above-described embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or changes made by those skilled in the art based on the present invention are within the protection scope of the present invention. The protection scope of the present invention shall be subject to the claims.

Claims

1. A wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network, characterized by: The following steps are involved: S1, constructs a normalized quality assessment data matrix sequence of wireless communication links and an adjacency matrix of the network topology structure; S2, using wavelet decomposition to decompose the normalized quality assessment data matrix sequence into a stationary data matrix sequence and a noise data matrix sequence; S3, calculate the standard deviation data matrix sequence of the noise data matrix sequence; S4, using the residual structure multi-head graph attention network to capture the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence; S5, the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence are respectively input into the Transformer to predict the output at the next moment and calculate the confidence interval of the wireless communication link quality.

2. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S1, constructing a normalized quality assessment data matrix sequence of a wireless communication link includes the following steps: The original data is modeled as ; Among them, the matrix represents the kth wireless communication link quality evaluation data matrix recorded within the time range t, K represents the number of all data matrices recorded within the time range t, N represents the total number of nodes in the wireless communication link network topology structure, and C represents the type of parameters for evaluating the quality of the wireless communication link; The quality assessment data matrix sequence is normalized and mapped to [0,1]. The normalization process is: ; in, is the original data under the i-th quality assessment parameter, and are the maximum and minimum values ​​in the data under the i-th quality assessment parameter, respectively. is the normalized data under the i-th quality assessment parameter.

3. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S1, constructing the adjacency matrix A of the network topology structure includes the following steps: Numbering N nodes in the wireless communication link network from 0 to N-1; create The all-zero matrix A is used as the adjacency matrix; Add weight information to edges ,make ,in Represents a node, represents the neighboring nodes of the i-th node, Represents the weight value of the edge between node i and node j.

4. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S2, wavelet decomposition comprises the following steps: Select the wavelet basis function as db series wavelet; Determine the threshold , where t is the number of elements in the normalized quality assessment data matrix sequence; Determine the number of decomposition levels L=2; Determine the threshold function ; Where b is the wavelet coefficient of the original sequence on the second layer during the wavelet decomposition process; Decompose the normalized quality assessment data matrix sequence with wavelet to obtain a stationary data matrix sequence and the noise data matrix sequence ; in, represents the kth stationary data matrix within the time range t, represents the kth noise data matrix within the time range t.

5. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S3, calculating the standard deviation data matrix sequence of the noise data matrix sequence includes the following steps: The standard deviation of the noise data matrix sequence is expressed as: ; in, is the standard deviation data matrix, N is the total number of nodes, C is the type of quality assessment parameters, and t is the time step; Calculate the standard deviation using the formula: ; in, , is the standard deviation value of the i-th node under the j-th quality evaluation parameter at the K-th time step, G is the length of the standard deviation window used to calculate, is the noise data value under the jth quality evaluation parameter of the ith node in m time steps.

6. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S4, the residual structure multi-head graph attention network captures the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence, including the following steps: The stationary data matrix sequence and the adjacency matrix A of the network topology as input; The stationary data matrix sequence The linear transformation is , the linear transformation process is: ； in, represents the quality assessment data matrix recorded at the kth time step within the time range t, is the transformed feature matrix, is the linear transformation matrix; The attention mechanism of the graph attention network is used to calculate the correlation coefficient between nodes. The correlation coefficient calculation formula is: ； in, represents the attention coefficient between node i and its neighbor node j at the kth time step, is a nonlinear activation function, is the initialized learnable matrix, and Respectively represent the feature vectors of node i and its neighbor node j at the kth time step; Normalize the attention coefficient: ; in, is the set of neighbor nodes of node i; The multi-head attention mechanism is used to obtain a stable data space feature matrix sequence: ; in, is the spatial feature matrix extracted at the kth time step, N is the total number of nodes, and M is the spatial feature dimension of a single node at a certain time step; The standard deviation data matrix sequence and the adjacency matrix A of the network topology as input; The standard deviation data space feature matrix sequence is obtained by using the same stationary data matrix sequence steps: .

7. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 6 is characterized in that: The multi-head attention mechanism is used to obtain a stable data space feature matrix sequence, which specifically includes the following steps: The weighted sum of the features of neighboring nodes is used to obtain a single-head output: in, is a nonlinear activation function, is the normalized attention coefficient; Concatenate the multi-head attention outputs. The concatenation process is: ; in, represents the output feature of the pth attention head of the i-th node, represents the output feature of the p-th attention head of neighbor node j, represents the feature vector of the i-th node at the k-th time step, Used to adjust the concatenated feature dimension to the target dimension.

8. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1 is characterized in that: In step S4, the design steps of the residual structure are: The stationary data space feature matrix sequence extracted by the multi-head graph attention network And standard deviation data space characteristic matrix sequence Perform batch normalization, the calculation formula is: ； ； in and They are respectively the batch normalized stationary data space feature matrix sequence and the standard deviation data space feature matrix sequence; The batch-normalized stationary data space feature matrix sequence And standard deviation data space characteristic matrix sequence They are connected to the previous layer input residual respectively, and the process of the first layer residual connection is: ； ； in is the residual structure output of the stationary data matrix sequence, The residual structure output of the standard deviation data matrix sequence; Will and Then input the above residual structure to obtain the residual structure output of the final stationary data matrix sequence The residual structure output of the standard deviation data matrix sequence .

9. The wireless communication link quality confidence interval prediction method based on Transformer and multi-head graph attention network according to claim 1, characterized in that: The step S5, inputting the residual structure output of the stationary data matrix sequence with spatial characteristics and the residual structure output of the standard deviation data matrix sequence into the Transformer to predict the output at the next moment and calculate the confidence interval of the wireless communication link quality, comprises the following steps: Output the residual structure of the stationary data matrix sequence as well as Residual structure output for standard deviation data matrix sequence Expand by time step; in represents the standard deviation spatial feature matrix of the kth time step, N is the total number of nodes, and C is the type of quality assessment parameters; The matrix for each time step is and Add dynamic position coding respectively, the formula is: in and is the node-level position encoding matrix generated based on time step k, which is used to preserve temporal and spatial associations; The matrix sequence after position encoding and The Transformer encoder is input separately, and the temporal dependency between time steps is captured through the self-attention mechanism, and the predicted stable data feature matrix for the next moment is output. and standard deviation feature matrix ; right and Perform node-level denormalization to restore the original data scale. The formula is: in , and , are the original data and standard deviation extreme value of the i-th node respectively; Calculate the node-level denormalized and Get the wireless communication link quality confidence lower limit matrix of C quality assessment parameters and the upper confidence limit matrix ; in, is a standard normal distribution Quantile, represents the significance level; The confidence level of prediction under different quality assessment parameters is weighted averaged to obtain the reliability confidence interval matrix containing all nodes. ; in is the weight matrix of different quality assessment parameter types, satisfying , C is the type of quality assessment parameter.

10. A wireless communication link quality confidence interval prediction system based on Transformer and multi-head graph attention network, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images