Network link optimization method and device, equipment, storage medium and product

Abstract

The invention discloses a network link optimization method and device, equipment, a storage medium and a product, and relates to the technical field of wireless communication, and the network link optimization method comprises the steps: obtaining network link operation data; the network link operation data is input to a preset optimization model, prediction processing is carried out on the network link operation data based on the preset optimization model, a link optimization scheme is obtained, and the link optimization scheme comprises at least one of transmission parameter adjustment, optimal transmission path selection and interference channel avoidance. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data and historical link stability data of a network link; and dynamically adjusting the network link according to the link optimization scheme. According to the invention, the transmission efficiency and stability of a network link are improved.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, and in particular to a network link optimization method, device, equipment, storage medium and computer program product. Background Art

[0002] In 5G networks, microwave backhaul links serve as core transmission channels and undertake the important task of transmitting base station data back to the core network. The performance of microwave backhaul links directly affects the transmission efficiency and stability of the entire network.

[0003] At present, traditional microwave backhaul link optimization methods usually rely on pre-set rules and experience. However, as the scale of the network expands and the complexity increases, these fixed rules and parameters may gradually become invalid when traditional methods cope with dynamically changing network environments, and they cannot continuously provide optimal transmission performance, resulting in insufficient link transmission efficiency and stability.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide a network link optimization method, device, equipment, storage medium and computer program product, aiming to solve the technical problem in the related technology that network link optimization relies on pre-set rules and experience, cannot continuously provide optimal transmission performance, and leads to insufficient link transmission efficiency and stability.

[0006] To achieve the above objectives, the present application proposes a network link optimization method, which includes:

[0007] Obtain network link operation data;

[0008] The network link operation data is input into a preset optimization model, and based on the preset optimization model, the network link operation data is predicted and processed to obtain a link optimization scheme, wherein the link optimization scheme includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0009] The network links are dynamically adjusted using the link optimization solution.

[0010] In one embodiment, the preset optimization model includes a graph neural network and a temporal convolutional network, and the step of performing prediction processing on the network link operation data based on the preset optimization model to obtain a link optimization solution includes:

[0011] Based on the preset optimization model, extracting the network feature vector and the timing feature vector of the network link operation data respectively;

[0012] Processing the network feature vector through a graph neural network to obtain a first output solution, and processing the time series feature vector through a time convolutional network to obtain a second output solution;

[0013] The first output solution and the second output solution are integrated to obtain a link optimization solution.

[0014] In one embodiment, before the step of inputting the network link operation data into a preset optimization model, the method further includes:

[0015] Acquire a historical operation sample data set, wherein the historical operation sample data set includes historical signal strength data, historical transmission rate data, and historical link stability data;

[0016] Preprocessing the historical operation sample data set to obtain preprocessed data;

[0017] Based on the preprocessed data, the model to be trained is iteratively trained until the preset loss function converges, thereby obtaining a trained preset optimization model.

[0018] In one embodiment, the step of iteratively training the model to be trained based on the preprocessed data until the preset loss function converges to obtain a trained preset optimization model includes:

[0019] Inputting the preprocessed data into a model to be trained, wherein the model to be trained includes a hybrid network architecture combining a graph neural network and a temporal convolutional network;

[0020] Initialize the weight parameters of the hybrid network architecture, and based on the hybrid network architecture, perform forward propagation calculation on the preprocessed data to obtain a first output result of the graph neural network and a second output result of the temporal convolutional network;

[0021] Merging the first output result and the second output result to obtain a final output result;

[0022] Determine the loss value of the preset loss function corresponding to the final output result, and generate a preset optimization model after the training is completed after the loss value reaches the minimum value.

[0023] In one embodiment, the step of dynamically adjusting the network link using the link optimization solution includes at least one of the following:

[0024] Adjusting the transmission parameters of each network node in the network link;

[0025] According to the current network status and transmission requirements, select the best transmission path corresponding to the link optimization solution;

[0026] The transmission channels of each network node in the network link are adjusted to avoid interference channels.

[0027] In one embodiment, after the step of dynamically adjusting the network link using the link optimization solution, the method further includes:

[0028] Continuously monitoring network data of the network link, wherein the network data includes environmental interference data, signal attenuation data, and network traffic data;

[0029] Inputting the network data into a preset optimization model, and based on the preset optimization model, performing propagation calculation on the comprehensive input vector corresponding to the network data to obtain a model optimization solution;

[0030] The preset optimization model is continuously optimized using the model optimization scheme.

[0031] In addition, to achieve the above purpose, the present application also proposes a network link optimization device, the network link optimization device comprising:

[0032] A first acquisition module, used to acquire network link operation data;

[0033] a processing module, configured to input the network link operation data into a preset optimization model, and perform prediction processing on the network link operation data based on the preset optimization model to obtain a link optimization solution, wherein the link optimization solution includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels, and the preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0034] The adjustment module is used to dynamically adjust the network link using the link optimization solution.

[0035] In addition, to achieve the above-mentioned purpose, the present application also proposes a network link optimization device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the network link optimization method described above.

[0036] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the network link optimization method described above are implemented.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the network link optimization method described above are implemented.

[0038] The present application proposes a network link optimization method, device, equipment, storage medium and computer program product. The present application obtains network link operation data, inputs the network link operation data into a preset optimization model, and predicts and processes the network link operation data through the preset optimization model to obtain a link optimization plan. The preset optimization model is obtained by iterative training based on the historical signal strength data, historical transmission rate data and historical link stability data of the network link, so that the model can take into account the complex and changeable network environment in which the network link is located from many aspects, and after comprehensive prediction processing, a more accurate optimization plan is generated. In addition, the link optimization plan obtained by model prediction can be used to adjust the transmission parameters of the current network link, the optimal transmission path, or avoid interference channels, etc., so as to ensure the link transmission efficiency and stability of the link in a complex and changeable network environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0041] Figure 1 A flowchart of the first embodiment of the network link optimization method of the present application is provided;

[0042] Figure 2 A schematic diagram of the model processing flow involved in the network link optimization method of this application;

[0043] Figure 3 A flowchart of the second embodiment of the network link optimization method of the present application is provided;

[0044] Figure 4 A schematic diagram of the overall implementation process involved in the network link optimization method of this application;

[0045] Figure 5 This is a schematic diagram of the module structure of the network link optimization device according to an embodiment of the present application;

[0046] Figure 6This is a schematic diagram of the device structure of the hardware operating environment involved in the network link optimization method in the embodiment of the present application.

[0047] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0048] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0049] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0050] The main solutions of the embodiments of this application are:

[0051] Obtain network link operation data;

[0052] The network link operation data is input into a preset optimization model, and based on the preset optimization model, the network link operation data is predicted and processed to obtain a link optimization scheme, wherein the link optimization scheme includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0053] The network links are dynamically adjusted using the link optimization solution.

[0054] Because the existing microwave backhaul link optimization methods in related technologies mainly rely on pre-set rules and experience, they show obvious deficiencies when facing complex and changing network environments. Traditional methods are difficult to adapt to factors such as environmental interference, signal attenuation, and network traffic fluctuations in real time, resulting in insufficient link transmission efficiency and stability. In addition, traditional solutions are usually based on fixed rules and preset parameters, lack adaptive learning capabilities, and cannot provide optimal transmission performance when the number of users increases and data traffic changes, resulting in network congestion and increased transmission delays. Related technologies also often only consider a single or a few factors, lack the ability to comprehensively analyze multiple factors, cannot fully optimize link performance, and are difficult to cope with complex network environments and emergencies.

[0055] The present application proposes a network link optimization method, device, equipment, storage medium and computer program product. The present application obtains network link operation data, inputs the network link operation data into a preset optimization model, and predicts and processes the network link operation data through the preset optimization model to obtain a link optimization plan. The preset optimization model is obtained by iterative training based on the historical signal strength data, historical transmission rate data and historical link stability data of the network link, so that the model can take into account the complex and changeable network environment in which the network link is located from many aspects, and after comprehensive prediction processing, a more accurate optimization plan is generated. In addition, the link optimization plan obtained by model prediction can be used to adjust the transmission parameters of the current network link, the optimal transmission path, or avoid interference channels, etc., so as to ensure the link transmission efficiency and stability of the link in a complex and changeable network environment.

[0056] It should be noted that the execution subject of this embodiment may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a network link optimization device, a base station, etc. The following takes a base station as an example to illustrate this embodiment and the following embodiments.

[0057] Based on this, the embodiment of the present application provides a network link optimization method, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the network link optimization method of the present application.

[0058] In this embodiment, the network link optimization method includes steps S10 to S30:

[0059] Step S10, obtaining network link operation data;

[0060] It should be noted that the network link operation data may be the operation data of the network link collected on the base station and the network node. The network link includes but is not limited to the microwave backhaul link. The network link operation data includes signal strength data, transmission rate data and link stability data. By collecting these data in real time, a microwave backhaul link operation data set is constructed.

[0061] Specifically, the process of obtaining network link operation data is as follows:

[0062] 1. Install monitoring equipment on 5G base stations and network nodes. The monitoring equipment includes signal receivers, data processors and data transmission modules;

[0063] 2. The monitoring equipment collects the signal strength data S of the microwave return link in real time i , transmission rate data R i and link stability data Li ,in:

[0064] S i Represents the signal strength data at time point i, in dBm;

[0065] R i Represents the transmission rate data at time point i, in Mbps;

[0066] L i Represents the link stability data at time point i, in percentage;

[0067] 3. Collect the signal strength data S i , transmission rate data R i and link stability data L i , stored in the data storage module, and the microwave backhaul link operation data set D is constructed:

[0068]

[0069] 4. The data storage module periodically transmits the microwave backhaul link operation data set D to the data processor.

[0070] Step S20, inputting the network link operation data into a preset optimization model, and performing prediction processing on the network link operation data based on the preset optimization model to obtain a link optimization solution, wherein the link optimization solution includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0071] It should be noted that the preset optimization model is a neural network model that combines a graph neural network and a temporal convolutional network. By combining the two models to process the input network link operation data, the network characteristics and timing characteristics of the same data can be predicted and processed, thereby generating more accurate prediction results.

[0072] It should be noted that the microwave backhaul link is dynamically adjusted through a preset optimization model based on real-time network link operation data and optimization solutions, including adjusting transmission parameters, selecting the best transmission path and avoiding interference channels.

[0073] Among them, adjusting the transmission parameters means adjusting the transmission power, encoding method, etc. in the network link, selecting the best transmission path means selecting the network link that is most suitable for transmitting the current data, and avoiding interference channels means selecting a suitable transmission channel to prevent other channels from interfering with the current channel.

[0074] It should be noted that the preset optimization model takes into account multiple factors such as environmental interference, signal attenuation and network traffic, and generates an optimization plan through comprehensive analysis. The model generates a more comprehensive and accurate optimization plan by comprehensively processing various data such as signal strength, transmission rate, link stability, interference intensity, path loss, traffic load and throughput. The multi-factor comprehensive analysis capability enables the optimization plan to better cope with complex and changeable network environments in practical applications, which has obvious advantages over traditional methods.

[0075] In a feasible implementation, the preset optimization model includes a graph neural network and a time convolutional network, and the step S20 of predicting and processing the network link operation data based on the preset optimization model to obtain a link optimization solution includes:

[0076] Based on the preset optimization model, extracting the network feature vector and the timing feature vector of the network link operation data respectively;

[0077] It should be noted that the preset optimization model includes two types of neural networks, namely graph neural network and temporal convolutional network. Graph neural network is a neural network model for processing graph structure data, and temporal convolutional network is a neural network model for processing time series data to capture long-term dependencies. After the preset optimization model receives the network link operation data, it extracts the network feature vector and time series feature vector of the data respectively, and inputs them into the two neural networks for processing.

[0078] Processing the network feature vector through a graph neural network to obtain a first output solution, and processing the time series feature vector through a time convolutional network to obtain a second output solution;

[0079] It should be noted that in the graph neural network part, the embedding vector of each node is calculated to obtain the first output scheme, which is the prediction result obtained after the graph neural network processes the network feature vector.

[0080] It should be noted that, in the temporal convolutional network part, the convolution output of the current time step is calculated to obtain the second output scheme, which is the prediction result obtained after the temporal convolutional network processes the time series feature vector.

[0081] The first output solution and the second output solution are integrated to obtain a link optimization solution.

[0082] It should be noted that the first output solution and the second output solution are integrated to output the final link optimization solution.

[0083] Specifically, a hybrid neural network architecture combined with 5G network is used to generate link optimization solutions. The hybrid neural network architecture combines adaptive graph neural network and spatiotemporal transformation network to process network topology, time series and environmental change data:

[0084]

[0085] in, is the link optimization solution, W agnn is the weight matrix of the adaptive graph neural network, x graph is the graph structure feature vector, W stn is the weight matrix of the spatiotemporal transformation network, x time is the time series feature vector, and b1 is the bias vector.

[0086] It should be noted that the model processing flow chart is as follows Figure 2 As shown, the preset optimization model includes an adaptive graph neural network and a space-time transformation network (temporal convolutional network). The graph neural network processes the network feature vector and outputs a node embedding vector. The space-time transformation network processes the time series feature vector to obtain a convolution output. Finally, the node embedding vector is fused with the convolution output through a fusion layer to obtain the final output link optimization solution.

[0087] Step S30: dynamically adjust the network link using the link optimization solution.

[0088] It should be noted that the preset optimization model dynamically adjusts the microwave backhaul link based on the real-time microwave backhaul link operation data set and optimization scheme, including adjusting at least one of the transmission parameters, selecting the best transmission path, and avoiding interference channels, or all of them.

[0089] In a feasible implementation manner, the step S30 of dynamically adjusting the network link using the link optimization solution includes at least one of the following:

[0090] Adjusting the transmission parameters of each network node in the network link;

[0091] It should be noted that the transmission parameters θ are adjusted, including the transmission power, modulation mode M and coding rate C:

[0092]

[0093] Where α and β are adjustment coefficients, N is the noise power, and g(·) and h(·) are functions for adjusting the modulation mode and coding rate according to the transmission rate and link stability, respectively.

[0094] According to the current network status and transmission requirements, select the best transmission path corresponding to the link optimization solution;

[0095] It should be noted that the best transmission path is selected, and the link selection P is optimized according to the current network status and transmission requirements:

[0096]

[0097] Among them, d i is the path length, R i,r is the transmission rate on the path, L i,r is the link stability of the path, P opt Represents a set of optimal transmission paths.

[0098] The transmission channels of each network node in the network link are adjusted to avoid interference channels.

[0099] It should be noted that avoiding interference channels and adjusting channel selection C can ensure stable link transmission:

[0100]

[0101] Among them, C opt represents the set of optimal channel selection, I j is the interference intensity of the interference channel, W j is the channel bandwidth, S i,r is the real-time signal strength on channel j.

[0102] It should be noted that the generated dynamic adjustment scheme is fed back to 5G base stations and network nodes and applied in real time to the actual operation of the microwave backhaul link:

[0103]

[0104] in, It is a new dynamic adjustment scheme, δ(t) is a time function, which is used for real-time adjustment.

[0105] In a feasible implementation manner, after the step S30 of dynamically adjusting the network link using the link optimization solution, the method further includes:

[0106] Continuously monitoring network data of the network link, wherein the network data includes environmental interference data, signal attenuation data, and network traffic data;

[0107] It should be noted that after applying the dynamically adjusted link optimization scheme, the preset optimization model continues to monitor the operating status of the microwave backhaul link and performs further optimization and adjustment based on new real-time data, forming a closed-loop optimization process.

[0108] It should be noted that the preset optimization model takes into account multiple factors such as environmental interference, signal attenuation and network traffic, and generates an optimization plan through comprehensive analysis, so as to continuously iterate and update the model.

[0109] It should be noted that the preset optimization model takes into account multiple factors such as environmental interference, signal attenuation, and network traffic, and uses the data of these factors as input features to form a comprehensive input vector x total :

[0110] x total =[x env ,x decay ,x flow ];

[0111] Among them, x env represents the environmental interference feature vector, x decay represents the signal attenuation feature vector, x flow Represents the network traffic feature vector;

[0112] Environmental interference feature vector x env Including interference intensity I i and interference frequency F i , where I i represents the interference intensity at time point i, F i represents the interference frequency at time point i;

[0113] Signal attenuation feature vector x decay Including path loss L p and multipath attenuation M d , where L p represents the path loss, M d Indicates multipath attenuation;

[0114] Network traffic feature vector x flow Including traffic load T l and throughput T p , where T l Indicates the traffic load, T p Indicates throughput.

[0115] Inputting the network data into a preset optimization model, and based on the preset optimization model, performing propagation calculation on the comprehensive input vector corresponding to the network data to obtain a model optimization solution;

[0116] It should be noted that the comprehensive input vector x total Input into the improved deep neural network model, and obtain the model optimization solution y through forward propagation and back propagation calculation. opt :

[0117] y opt=σ(W total ·x total +b total );

[0118] Among them, W total is the weight matrix, b total is the bias parameter.

[0119] The preset optimization model is continuously optimized using the model optimization scheme.

[0120] It should be noted that the generated optimization solution y opt Feedback to the system, real-time adjustment of microwave backhaul link parameters to achieve optimization effect, the preset optimization model continuously accumulates new data during system operation, automatically updates and optimizes its own algorithms and parameters, the specific update process is as follows:

[0121]

[0122] Among them, η1 is the learning rate, is the loss function, which is used to evaluate the current optimization effect.

[0123] In this embodiment, the hybrid neural network architecture includes an adaptive graph neural network and a hybrid neural network architecture. The graph neural network dynamically adjusts the weights of nodes and edges to adapt to changes in network topology. The feature vector of each node is updated by the following formula:

[0124]

[0125] in, represents the set of neighbor nodes of node i, e ij is the edge weight, c ij is the normalization factor, W (t) is the weight matrix at time step t, b (t) is the bias vector, σ is the activation function;

[0126] The spatiotemporal transformer network captures complex patterns in time series data through a multi-head attention mechanism, and the output z at each time step is t Calculated by the following formula:

[0127] z t =Concat(head1,head2,…,head h )W O ;

[0128] Among them, the output of each attention head is:

[0129]

[0130] Among them, Q i,K i ,V i are query, key and value matrices respectively, W O is the output matrix, d k The dimension of the key.

[0131] Finally, the optimization effect is evaluated regularly, performance reports are generated, and suggestions for further improvement are provided.

[0132] It should be noted that the preset optimization model has adaptive learning capabilities, can continuously accumulate new data during system operation, and automatically update and optimize its own algorithms and parameters. This self-learning and improvement capability enables the system to maintain optimization effects in long-term operation and gradually adapt to the ever-changing network environment and usage requirements. The system also regularly evaluates the optimization effects, generates performance reports, and provides suggestions for further improvement to ensure that the optimization plan is always in the best state.

[0133] The present application proposes a network link optimization method, device, equipment, storage medium and computer program product. The present application obtains network link operation data, inputs the network link operation data into a preset optimization model, and predicts and processes the network link operation data through the preset optimization model to obtain a link optimization plan. The preset optimization model is obtained by iterative training based on the historical signal strength data, historical transmission rate data and historical link stability data of the network link, so that the model can take into account the complex and changeable network environment in which the network link is located from many aspects, and after comprehensive prediction processing, a more accurate optimization plan is generated. In addition, the link optimization plan obtained by model prediction can be used to adjust the transmission parameters of the current network link, the optimal transmission path, or avoid interference channels, etc., so as to ensure the link transmission efficiency and stability of the link in a complex and changeable network environment.

[0134] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 3 , before the step S20 of inputting the network link operation data into the preset optimization model, the method further includes steps S100 to S300:

[0135] Step S100, obtaining a historical operation sample data set, wherein the historical operation sample data set includes historical signal strength data, historical transmission rate data, and historical link stability data;

[0136] It should be noted that the historical running sample data set D h :

[0137] D h ={(S i,h ,Ri,h ,L i,h )|i∈N};

[0138] Among them, S i,h represents the historical signal strength data at time point i, R i,h represents the historical transmission rate data at time point i, L i,h represents the historical link stability data at time point i;

[0139] Step S200, preprocessing the historical operation sample data set to obtain preprocessed data;

[0140] It should be noted that the preprocessing process can be:

[0141] 1. Run the sample data set D on history h Delete missing values, correct outliers and fill in gaps;

[0142] 2. Normalize the cleaned historical running sample data set and convert the signal strength data S i,h , transmission rate data R i,h and link stability data L i,h Convert to a uniform dimension range to obtain a normalized data set D norm :

[0143] D norm ={(S′ i,h ,R′ i,h ,L′ i,h )|i∈N};

[0144] Among them, S′ i,h Represents the normalized historical signal strength data, R′ i,h Represents the normalized historical transmission rate data, L′ i,h represents the normalized historical link stability data;

[0145] 3. Normalize the dataset D norm Stored in the data storage module.

[0146] Step S300, based on the preprocessed data, iteratively train the model to be trained until the preset loss function converges, thereby obtaining a trained preset optimization model.

[0147] It should be noted that the preprocessed data is input into the model to be trained, the two neural networks in the model are iteratively trained, the prediction results are output, and the weight parameters of the hybrid network architecture are updated using the back propagation algorithm. (weight parameters of graph neural network), (weight parameter of the temporal convolutional network) and bias parameter bi , to minimize the loss function. When the loss function reaches the minimum value, that is, the loss function converges, the iterative training of the model is stopped and the preset optimization model is output.

[0148] In a feasible implementation manner, the step S300 of iteratively training the model to be trained based on the preprocessed data until the preset loss function converges to obtain a trained preset optimization model includes:

[0149] Inputting the preprocessed data into a model to be trained, wherein the model to be trained includes a hybrid network architecture combining a graph neural network and a temporal convolutional network;

[0150] It should be noted that the preprocessed data D norm Input into the model to be trained, which combines a hybrid network architecture of graph neural network and temporal convolutional network to process network topology and time series data;

[0151] Initialize the weight parameters of the hybrid network architecture, and based on the hybrid network architecture, perform forward propagation calculation on the preprocessed data to obtain a first output result of the graph neural network and a second output result of the temporal convolutional network;

[0152] It should be noted that the hybrid network architecture is initialized, including randomly initializing the node weight parameters of the graph neural network. and the convolution kernel weight parameters of the temporal convolutional network And the corresponding bias parameter b i ,in, represents the weight from the i-th node to the j-th node in GNN (Graph Neural Network, a neural network model used to process graph structured data). represents the weights from the i-th neuron to the j-th neuron in TCN (TCN (Temporal Convolutional Network) is a neural network model for processing sequence data and capturing long-term dependencies), b i represents the bias of the i-th neuron;

[0153] S33, using the preprocessed historical running sample data set D norm Perform forward propagation calculation:

[0154] In the graph neural network part, the embedding vector h of each node is calculated i :

[0155]

[0156] Among them, σ represents the activation function, represents the set of neighbor nodes of node i, x j represents the input feature vector of node j;

[0157] In the temporal convolutional network part, the convolution output of the current time step is calculated

[0158]

[0159] Among them, f represents the activation function, x t+k represents the input data at time step t+k, K is the convolution kernel size, and b t are the convolution kernel weight and bias, ∈ t represents the noise term at time step t;

[0160] Merging the first output result and the second output result to obtain a final output result;

[0161] It should be noted that the output of the fusion graph neural network and the temporal convolutional network is used to obtain the final output

[0162]

[0163] Among them, α, β and γ are the fusion weight coefficients, φ j is the connection coefficient, and M is the total number of input data.

[0164] Determine the loss value of the preset loss function corresponding to the final output result, and generate a preset optimization model after the training is completed after the loss value reaches the minimum value.

[0165] It should be noted that the design improves the loss function Optimize specific link performance indicators, including transmission delay, bandwidth utilization, and signal interference, to obtain loss values

[0166]

[0167]

[0168] Where m is the number of data samples, is the predicted value of the i-th sample, y i is the actual value of the i-th sample, T i is the transmission delay of the ith sample, is the expected transmission delay, U i is the bandwidth utilization of the i-th sample, is the expected bandwidth utilization, I iis the signal interference of the i-th sample, is the desired signal interference, λ1, λ2 and λ3 are the corresponding weight parameters;

[0169] Using the back-propagation algorithm to update the weight parameters of the hybrid network architecture and the bias parameter b i , to minimize the loss function

[0170]

[0171] Where η is the learning rate;

[0172] Repeat the iterative training steps until the loss function After convergence, a trained improved deep neural network model is obtained to identify and predict the key factors affecting the performance of microwave backhaul links and optimize microwave backhaul links in 5G networks.

[0173] In this embodiment, a hybrid neural network model is trained through various types of network data sets. The system can monitor the status of microwave backhaul links in 5G networks in real time and make dynamic adjustments based on real-time data. The model can automatically adjust transmission parameters, select the best transmission path, and avoid interference channels. The dynamic adjustment capability enables the link to respond quickly to different loads and interference conditions, maintain efficient and stable transmission, and significantly improve transmission efficiency and link stability.

[0174] For example, to help understand the implementation process of the network link optimization method obtained by combining this embodiment with the above-mentioned embodiment 1, please refer to Figure 4 , Figure 4 A brief flowchart of a network link optimization method is provided, specifically:

[0175] First, monitoring equipment is installed on the base station or each network node to collect data, and then the data is pre-processed to train the neural network model. After the model training is completed, the operation data on the network link is collected to dynamically adjust the network link to adapt to the complex and changing environment. Then, through multi-factor comprehensive analysis, adaptive learning, performance evaluation and reporting, it is determined whether the model needs to be further optimized. If the model needs to be optimized and adjusted, the updated model is used to dynamically adjust the microwave backhaul link again.

[0176] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the network link optimization method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0177] This application also provides a network link optimization device, please refer to Figure 5, the network link optimization device comprises:

[0178] A first acquisition module 10 is used to acquire network link operation data;

[0179] A processing module 20 is used to input the network link operation data into a preset optimization model, and based on the preset optimization model, perform prediction processing on the network link operation data to obtain a link optimization solution, wherein the link optimization solution includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0180] The adjustment module 30 is used to dynamically adjust the network link using the link optimization solution.

[0181] Optionally, the processing module includes:

[0182] An extraction unit, configured to extract a network feature vector and a timing feature vector of the network link operation data based on the preset optimization model;

[0183] A processing unit, configured to process the network feature vector through a graph neural network to obtain a first output solution, and process the time series feature vector through a temporal convolutional network to obtain a second output solution;

[0184] The first fusion unit is used to fuse the first output solution and the second output solution to obtain a link optimization solution.

[0185] Optionally, the device further comprises:

[0186] A second acquisition module is used to acquire a historical operation sample data set, wherein the historical operation sample data set includes historical signal strength data, historical transmission rate data, and historical link stability data;

[0187] A preprocessing module, used for preprocessing the historical operation sample data set to obtain preprocessed data;

[0188] The training module is used to iteratively train the model to be trained based on the preprocessed data until the preset loss function converges, thereby obtaining a preset optimization model that has been trained.

[0189] Optionally, the training module includes:

[0190] An input unit, used to input the preprocessed data into a model to be trained, wherein the model to be trained includes a hybrid network architecture combining a graph neural network and a temporal convolutional network;

[0191] A computing unit, used to initialize weight parameters of the hybrid network architecture, and based on the hybrid network architecture, perform forward propagation calculation on the preprocessed data to obtain a first output result of the graph neural network and a second output result of the temporal convolutional network;

[0192] A second fusion unit, used to fuse the first output result and the second output result to obtain a final output result;

[0193] A generation unit is used to determine the loss value of the preset loss function corresponding to the final output result, and generate a preset optimization model after the training is completed after the loss value reaches a minimum value.

[0194] Optionally, the adjustment module at least includes:

[0195] A first adjustment unit, configured to adjust transmission parameters of each network node in the network link;

[0196] A selection unit, configured to select an optimal transmission path corresponding to the link optimization solution according to the current network status and transmission requirements;

[0197] The second adjustment unit is used to adjust the transmission channel of each network node in the network link to avoid the interference channel.

[0198] Optionally, the device further comprises:

[0199] A monitoring module, used to continuously monitor network data of the network link, wherein the network data includes environmental interference data, signal attenuation data, and network traffic data;

[0200] A calculation module, used for inputting the network data into a preset optimization model, and performing propagation calculation on the comprehensive input vector corresponding to the network data based on the preset optimization model to obtain a model optimization solution;

[0201] The optimization module is used to continuously optimize the preset optimization model using the model optimization scheme.

[0202] The network link optimization device provided by the present application adopts the network link optimization method in the above embodiment to solve the technical problem of network link optimization. Compared with the prior art, the beneficial effects of the network link optimization device provided by the present application are the same as the beneficial effects of the network link optimization method provided by the above embodiment, and the other technical features in the network link optimization device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0203] The present application provides a network link optimization device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the network link optimization method in the above-mentioned embodiment one.

[0204] Reference below Figure 6 , which shows a schematic diagram of the structure of a network link optimization device suitable for implementing the embodiment of the present application. The network link optimization device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The network link optimization device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0205] like Figure 6As shown, the network link optimization device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the network link optimization device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the network link optimization device to communicate with other devices wirelessly or wired to exchange data. Although the figure shows a network link optimization device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0206] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0207] The network link optimization device provided by the present application adopts the network link optimization method in the above embodiment to solve the technical problem of network link optimization. Compared with the prior art, the beneficial effects of the network link optimization device provided by the present application are the same as the beneficial effects of the network link optimization method provided by the above embodiment, and the other technical features in the network link optimization device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0208] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0209] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0210] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the network link optimization method in the above-mentioned embodiment.

[0211] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0212] The above-mentioned computer-readable storage medium may be included in the network link optimization device; or it may exist independently without being assembled into the network link optimization device.

[0213] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the network link optimization device, the network link optimization device:

[0214] Obtain network link operation data;

[0215] The network link operation data is input into a preset optimization model, and based on the preset optimization model, the network link operation data is predicted and processed to obtain a link optimization scheme, wherein the link optimization scheme includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link;

[0216] The network links are dynamically adjusted using the link optimization solution.

[0217] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0218] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0219] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0220] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned network link optimization method, and can solve the technical problems of network link optimization. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the network link optimization method provided in the above-mentioned embodiment, and will not be repeated here.

[0221] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned network link optimization method when executed by a processor.

[0222] The computer program product provided by this application can solve the technical problem of network link optimization. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the network link optimization method provided by the above embodiment, which will not be repeated here.

[0223] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A network link optimization method, characterized in that: The method comprises: Obtain network link operation data; The network link operation data is input into a preset optimization model, and based on the preset optimization model, the network link operation data is predicted and processed to obtain a link optimization scheme, wherein the link optimization scheme includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels. The preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link; The network links are dynamically adjusted using the link optimization solution.

2. The method according to claim 1, characterized in that The preset optimization model includes a graph neural network and a time convolutional network. The step of performing prediction processing on the network link operation data based on the preset optimization model to obtain a link optimization solution includes: Based on the preset optimization model, extracting the network feature vector and the timing feature vector of the network link operation data respectively; Processing the network feature vector through a graph neural network to obtain a first output solution, and processing the time series feature vector through a time convolutional network to obtain a second output solution; The first output solution and the second output solution are integrated to obtain a link optimization solution.

3. The method according to claim 1, characterized in that Before the step of inputting the network link operation data into the preset optimization model, the step further includes: Acquire a historical operation sample data set, wherein the historical operation sample data set includes historical signal strength data, historical transmission rate data, and historical link stability data; Preprocessing the historical operation sample data set to obtain preprocessed data; Based on the preprocessed data, the model to be trained is iteratively trained until the preset loss function converges, thereby obtaining a trained preset optimization model.

4. The method according to any one of claims 3, characterized in that The step of iteratively training the model to be trained based on the preprocessed data until the preset loss function converges to obtain a trained preset optimization model comprises: Inputting the preprocessed data into a model to be trained, wherein the model to be trained includes a hybrid network architecture combining a graph neural network and a temporal convolutional network; Initialize the weight parameters of the hybrid network architecture, and based on the hybrid network architecture, perform forward propagation calculation on the preprocessed data to obtain a first output result of the graph neural network and a second output result of the temporal convolutional network; Merging the first output result and the second output result to obtain a final output result; Determine the loss value of the preset loss function corresponding to the final output result, and generate a preset optimization model after the training is completed after the loss value reaches the minimum value.

5. The method according to claim 1, characterized in that The step of dynamically adjusting the network link using the link optimization solution includes at least one of the following: Adjusting the transmission parameters of each network node in the network link; According to the current network status and transmission requirements, select the best transmission path corresponding to the link optimization solution; The transmission channels of each network node in the network link are adjusted to avoid interference channels.

6. The method according to claim 1, characterized in that After the step of dynamically adjusting the network link using the link optimization solution, the method further includes: Continuously monitoring network data of the network link, wherein the network data includes environmental interference data, signal attenuation data, and network traffic data; Inputting the network data into a preset optimization model, and based on the preset optimization model, performing propagation calculation on the comprehensive input vector corresponding to the network data to obtain a model optimization solution; The preset optimization model is continuously optimized using the model optimization scheme.

7. A network link optimization device, characterized in that: The device comprises: A first acquisition module, used to acquire network link operation data; a processing module, configured to input the network link operation data into a preset optimization model, and perform prediction processing on the network link operation data based on the preset optimization model to obtain a link optimization solution, wherein the link optimization solution includes adjusting transmission parameters, selecting an optimal transmission path, and avoiding interference channels, and the preset optimization model is obtained by iterative training based on historical signal strength data, historical transmission rate data, and historical link stability data of the network link; The adjustment module is used to dynamically adjust the network link using the link optimization solution.

8. A network link optimization device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the network link optimization method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the network link optimization method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the network link optimization method according to any one of claims 1 to 6 are implemented.

Images