Network slicing resource optimization methods, devices, systems and products

Abstract

The application relates to the communication technology field and discloses a network slice resource optimization method, device, system and product. The method comprises the following steps: collecting the traffic data of each edge node in a target network in a current time step according to the size of the time step; the traffic data comprises short-time traffic data and long-time traffic data of all network slices in the edge node; predicting the predicted traffic data of each edge node in a next time step based on the traffic data of each edge node at present; determining a target allocation strategy based on the predicted traffic data of each edge node in the next time step; determining the target resource configuration in each edge node in the next time step according to the target allocation strategy and the predicted traffic data of each edge node in the next time step; and allocating network resources to each edge node according to the target resource configuration. By using the method, accurate and efficient dynamic network resource allocation of the target network as a whole can be realized, and the overall network performance is improved.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method, apparatus, system and product for optimizing network slicing resources. Background Technology

[0002] In recent years, machine learning-based traffic prediction and resource optimization schemes have emerged, attempting to overcome the limitations of traditional methods. These schemes analyze historical traffic data to try to predict future traffic patterns, thereby guiding the dynamic adjustment of resources. Although these methods have improved resource utilization and network performance to some extent, several challenges remain. For example, traffic prediction on a single time scale may not accurately capture the complex patterns of traffic changes, while centralized resource management mechanisms may lead to high computational burdens and latency issues.

[0003] Network slicing is a key technology in 5G communication, allowing the creation of multiple virtual networks on the same physical network infrastructure to meet the needs of different business scenarios. In a dynamically changing traffic environment, how to efficiently manage and optimize network slice resources has become a pressing issue. Current dynamic optimization solutions typically rely on predicting future traffic changes based on current traffic data. However, different business scenarios exhibit varying traffic patterns, and predicting using only a uniform time period cannot accurately capture the traffic patterns of all scenarios. This leads to inaccurate traffic predictions, low resource utilization, and difficulty in ensuring the efficient operation of business systems. Summary of the Invention

[0004] In view of this, this application aims to propose a network slicing resource optimization method, apparatus, system and product to achieve accurate dynamic network resource allocation and improve the overall performance of the network.

[0005] To achieve the above objectives, the technical solution of this application is as follows:

[0006] A first aspect of this application provides a method for optimizing network slicing resources, the method comprising:

[0007] According to the size of the time step, collect the traffic data of each edge node in the target network within the current time step; the traffic data includes: short-term traffic data and long-term traffic data of all network slices in the edge node;

[0008] Based on the current traffic data of each edge node, predict the predicted traffic data of each edge node in the next time step;

[0009] Based on the predicted traffic data of each edge node in the next time step, a target allocation strategy is determined; based on the target allocation strategy and the predicted traffic data of each edge node in the next time step, the target resource configuration in each edge node in the next time step is determined.

[0010] Assign network resources to each edge node according to the target resource configuration.

[0011] Optionally, based on the current traffic data of each edge node, predict the predicted traffic data of each edge node in the next time step, including:

[0012] Construct a first prediction model based on reinforcement learning, and train the first prediction model using the Q-learning algorithm;

[0013] Using the first prediction model that has been trained, the traffic data of each edge node is processed to obtain the predicted traffic data of each edge node in the next time step.

[0014] Optionally, based on the current traffic data of each edge node, predict the predicted traffic data of each edge node in the next time step, including:

[0015] A second prediction model based on a neural network is constructed and trained. The second prediction model includes sub-neural networks deployed at each edge node. In the trained second prediction model, the parameters of each sub-neural network are the same.

[0016] The local traffic data is processed using the sub-neural networks of the second prediction model to obtain the local predicted traffic data for the next time step.

[0017] Optionally, training a second prediction model based on a neural network includes:

[0018] The local sub-neural network is iteratively trained using traffic data from each edge node. In each training round, the loss value is calculated and the parameters of the sub-neural network are updated based on the actual local traffic data and the predicted traffic data of the sub-neural network.

[0019] Calculate the comprehensive parameters based on the parameters of all sub-neural networks;

[0020] Based on the comprehensive parameters, the parameters of all sub-neural networks are updated, and the next round of training is performed.

[0021] Optionally, predict the predicted traffic data for each edge node in the next time step, including:

[0022] Obtain a first predicted value for the edge node based on long-term traffic data, and a second predicted value based on short-term traffic data;

[0023] Long-term and short-term weights are set based on the device type of the edge node; the sum of the long-term and short-term weights is 1;

[0024] The predicted traffic data of the edge nodes are calculated based on the first predicted value, the second predicted value, the long-term weight, and the short-term weight of each edge node.

[0025] Optionally, based on the target allocation strategy and the predicted traffic data of each edge node in the next time step, the target resource configuration in each edge node in the next time step is determined, including:

[0026] Obtain the initial allocation strategy of the target network;

[0027] Based on the initial allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration in the next time step is calculated as the target resource configuration.

[0028] Optionally, after calculating the initial resource configuration for the next time step, the following is also included:

[0029] Obtain the current resource allocation of each network slice in each edge node, and calculate the overall utility and fairness of the target network based on the current resource allocation;

[0030] Based on the predicted traffic data of each edge node, the initial allocation strategy is adjusted using the gradient descent method to maximize the overall utility and optimize the fairness, thus obtaining the first allocation strategy.

[0031] Based on the first allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration of each edge node in the next time step is calculated, and the target resource configuration is updated.

[0032] Optionally, after collecting traffic data from each edge node in the target network within the current time step, the method further includes:

[0033] The traffic data is preprocessed, specifically including: removing invalid and erroneous data from the traffic data; detecting abnormal data in the traffic data and replacing the abnormal data with valid data from the previous time step; and compressing the traffic data using principal component analysis to obtain compressed data.

[0034] Optionally, the network slicing resource optimization method further includes:

[0035] Obtain the current network resource configuration of multiple target networks, and calculate the overall utility and overall fairness;

[0036] Based on optimization preferences, utility weights and fairness weights are set; the sum of the utility weights and the fairness weights is 1; the optimization preference is either utility priority or fairness priority.

[0037] Calculate the overall performance based on the utility weight and the fairness weight;

[0038] A comprehensive allocation strategy for the multiple target networks is obtained, and the comprehensive allocation strategy is adjusted using gradient descent to maximize the comprehensive performance, thus obtaining a second allocation strategy.

[0039] Based on the second allocation strategy, the current network resource configuration of each target network and the predicted traffic data, calculate the network resource configuration of each target network in the next time step as the comprehensive resource configuration.

[0040] Based on the comprehensive resource allocation, network resources are allocated to each target network.

[0041] According to a second aspect of the present application, a network slicing resource optimization apparatus is provided, used to implement the steps of the method provided in the first aspect of the present application, the apparatus comprising:

[0042] The perception module is configured to collect traffic data of each edge node in the target network within the current time step according to the size of the time step; the traffic data includes: short-term traffic data and long-term traffic data of all network slices in the edge node;

[0043] The prediction module is configured to predict the traffic data of each edge node in the next time step based on the current traffic data of each edge node.

[0044] The optimization module is configured to determine the target allocation strategy based on the predicted traffic data of each edge node in the next time step; and to determine the target resource configuration in each edge node in the next time step according to the target allocation strategy and the predicted traffic data of each edge node in the next time step.

[0045] The allocation module is configured to allocate network resources to each edge node according to the target resource configuration.

[0046] According to a third aspect of the embodiments of this application, a network slicing resource optimization system is provided, comprising:

[0047] Multiple target networks, including: an access network, a transmission network, and a core network; wherein, the access network is deployed with:

[0048] Multiple edge nodes in the method provided in the first aspect of the embodiments of this application; wherein each edge node is deployed with a network slicing resource optimization device provided in the second aspect of the embodiments of this application;

[0049] The data processing center is used to aggregate traffic data or parameters of sub-neural networks sent by various edge nodes.

[0050] According to a fourth aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the method provided in the first aspect of the embodiments of this application.

[0051] According to a fifth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the method provided in the first aspect of the present application.

[0052] According to a sixth aspect of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the steps in the method provided in the first aspect of the present application.

[0053] The network slicing resource optimization method provided in this application collects current traffic data for each edge node at different time scales. Based on the current multi-scale traffic data, it predicts the traffic data of the edge nodes in the next time step, thereby determining the target allocation strategy for each edge node in the next time step. The specific network resource configuration in the next time step is determined according to the target allocation strategy, and dynamically adjusted based on the current network resource configuration.

[0054] Given the significant differences in traffic variation patterns across different network slices and their associated business scenarios, traffic data based on a single time scale cannot accurately capture the traffic variation patterns across all business scenarios. Therefore, this application introduces multi-time-scale traffic data when collecting traffic data from edge nodes. By collecting short-term traffic data to capture instantaneous changes and short-term trends in node traffic, and by collecting long-term traffic data to capture long-term trends and periodic changes in node traffic, this approach comprehensively covers all business scenarios, enabling more complete acquisition of node traffic variation patterns and accurate prediction of node traffic in the next time step. Furthermore, based on the prediction results, accurate dynamic network resource allocation is achieved, improving overall network resource utilization and overall network performance. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is a flowchart of a network slicing resource optimization method proposed in an embodiment of this application;

[0057] Figure 2 This is a flowchart of adjusting the target allocation strategy in one embodiment of this application;

[0058] Figure 3 This is a flowchart of adjusting the target allocation strategy across layers in one embodiment of this application;

[0059] Figure 4 This is a schematic diagram of a network slicing resource optimization device according to an embodiment of this application;

[0060] Figure 5 This is a schematic diagram of a network slicing resource optimization system proposed in an embodiment of this application;

[0061] Figure 6 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0062] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0063] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0064] In the various embodiments of this application, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0065] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects as detailed in this application.

[0066] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0067] The present application will now be described in detail with reference to the accompanying drawings and embodiments.

[0068] Figure 1 This is a flowchart of a network slicing resource optimization method proposed in an embodiment of this application. For example... Figure 1 As shown, the method includes:

[0069] S1: Collect traffic data of each edge node in the target network within the current time step according to the size of the time step; the traffic data includes: short-term traffic data and long-term traffic data of all network slices in the edge node;

[0070] S2: Based on the current traffic data of each edge node, predict the predicted traffic data of each edge node in the next time step;

[0071] S3: Based on the predicted traffic data of each edge node in the next time step, determine the target allocation strategy; according to the target allocation strategy and the predicted traffic data of each edge node in the next time step, determine the target resource configuration in each edge node in the next time step.

[0072] S4: Allocate network resources to each edge node according to the target resource configuration.

[0073] In this embodiment, the target network is a 5G network. Sensing modules are deployed at each edge node in the target network to monitor and record traffic-related information for each node in real time, including inbound traffic, outbound traffic, traffic rate, and traffic direction. The sensing modules collect multi-timescale traffic data from the local edge nodes according to a pre-set time step size. This traffic data includes short-term and long-term traffic data. Based on the traffic data collected from each edge node, predictions are made to obtain the predicted traffic data for the edge nodes in the next time step. Then, based on the predicted traffic data for each edge node, a target allocation strategy for network resources at the edge nodes is determined. The target resource configuration for each edge node in the next time step is calculated according to the target allocation strategy. Based on this target resource configuration, the current network resource configuration (e.g., bandwidth allocation) is dynamically adjusted.

[0074] In this embodiment, by deploying sensing modules at edge nodes, the timeliness of traffic data collection is ensured. By collecting traffic data from edge nodes at multiple time scales in real time, a comprehensive capture of traffic change patterns for all service types on the edge nodes is achieved. Specifically, short-term traffic data is collected to capture recent traffic change trends, while long-term traffic data is collected to capture long-term trends and historical patterns. This allows for more accurate prediction of edge node traffic data in the next time step. Based on the predicted traffic data, dynamic network resource allocation for network slices in each edge node is implemented, improving overall network resource utilization and performance.

[0075] As one embodiment of this application, after collecting the traffic data of each edge node in the target network within the current time step, the method further includes:

[0076] The traffic data is preprocessed, specifically including: removing invalid and erroneous data from the traffic data; detecting abnormal data in the traffic data and replacing the abnormal data with valid data from the previous time step; and compressing the traffic data using principal component analysis to obtain compressed data.

[0077] In one embodiment, each edge node collects traffic data through a sensing module and then sends it to a central data processing center for preprocessing. Finally, the short-term and long-term traffic data are processed into compressed data for subsequent traffic data prediction.

[0078] Specifically, the acquisition period is defined as Δt, and the traffic data x of the i-th edge node recorded in each acquisition period is... i (t) can be represented in vector form:

[0079] x i (t)=[x i1 (t),x i2 (t),…,x im (t)];

[0080] Where, x ij (t) represents the j-th type of traffic data of the i-th edge node at the current time step.

[0081] Furthermore, the perception module calculates the current total traffic data X of the local node. i (t):

[0082]

[0083] A data processing center is also deployed within the target network, connected to each edge node. In this embodiment, each edge node sends its total local traffic data to the data processing center, where preprocessing operations are performed.

[0084] Specifically, the preprocessing includes the following steps:

[0085] (1) Data cleaning, including removing invalid and erroneous traffic data. Invalid data is defined as data points with a traffic value of 0 or a negative number. The following expression is used to remove invalid data, yielding valid traffic data x. i ′ (t):

[0086]

[0087] Erroneous data typically includes abnormal peaks and troughs. Abnormal peaks and troughs are defined as data points that exceed the historical data mean by 3-5 times the standard deviation. The following expression is then used to remove erroneous data, yielding the traffic data x″. i (t):

[0088]

[0089] Where, μ j Let σ be the mean of the j-th type of traffic data. j Let be the standard deviation of the j-th type of traffic data.

[0090] (2) Detect outliers in the traffic data and replace them with normal values. Specifically, use the Z-score method to detect outliers and calculate x. ij (t) The Z-score of traffic data, defined as outliers as data points exceeding 3-5 times the standard deviation, is detected using the following expression:

[0091]

[0092] Where, μ j Let σ be the mean of the j-th type of traffic data. j Let x' be the standard deviation of the j-th type of traffic data. Outliers are handled by using the valid traffic data x″ from the previous time step (t-1). i (t) is replaced to obtain the traffic data x″′ i (t):

[0093]

[0094] (3) Data Compression. In this embodiment, principal component analysis is used to reduce the dimensionality of the traffic data to reduce data transmission overhead. The covariance matrix C of the traffic data is calculated, and the eigenvalues ​​and eigenvectors are solved. The eigenvector corresponding to the largest eigenvalue in the previous time step (t-1) is selected as the principal component, and data compression is performed using the following expression:

[0095] Y i(t)=W T x′′ i (t)

[0096] Where W is a matrix composed of the feature vectors of the previous traffic data, Y i (t) is the compressed traffic data of the i-th edge node at the current time step.

[0097] In this embodiment, by removing invalid and erroneous data and processing outliers, unnecessary data computation is reduced for subsequent prediction stages, thereby improving prediction efficiency. Furthermore, compressing the collected traffic data reduces the transmission overhead of communication between edge nodes and the data processing center, thus saving network resources and improving the overall efficiency and accuracy of the network.

[0098] As one implementation of this application, based on the current traffic data of each edge node, the predicted traffic data of each edge node in the next time step is predicted, including:

[0099] Construct a first prediction model based on reinforcement learning, and train the first prediction model using the Q-learning algorithm;

[0100] Using the first prediction model that has been trained, the traffic data of each edge node is processed to obtain the predicted traffic data of each edge node in the next time step.

[0101] In one embodiment, a first prediction model based on reinforcement learning is constructed, the model is trained using the Q-learning algorithm, and the trained first prediction model is used to optimize network resources for each edge node.

[0102] Specifically, we define the state space S: states s t Including compressed traffic data Y at the current time step i (t) and network resource allocation R t ;

[0103] Define action space A: action a t This includes adjusting network resource allocation strategies (g).

[0104] Based on the actual and predicted traffic data in the next time step, calculate the error reward r(s). t ,a t ):

[0105]

[0106] in, It is the predicted traffic data for the next time step, X i (t+1) is the actual traffic data in the next time step.

[0107] Update the Q-value table using the following expression:

[0108]

[0109] Where α is the learning rate and γ is the discount factor; ′ Indicates the state s at the next time step. t+1 To obtain the maximum Q value, specifically:

[0110]

[0111] During the training of the first prediction model, the Q-value is continuously updated by defining a suitable state space s, action space a, and reward function r, and continuously interacting with the environment to minimize the prediction error.

[0112] In this embodiment, leveraging the characteristic of reinforcement learning that it does not require prior knowledge when facing unknown and complex environments and learns solely through interaction with the environment, a first prediction model based on reinforcement learning is deployed in each edge node. The first prediction model autonomously learns the dynamic traffic change patterns at multiple time scales in the edge nodes, thereby providing a more accurate basis for the dynamic adjustment of network resources.

[0113] As one implementation of this application, based on the current traffic data of each edge node, the predicted traffic data of each edge node in the next time step is predicted, including:

[0114] A second prediction model based on a neural network is constructed and trained. The second prediction model includes sub-neural networks deployed at each edge node. In the trained second prediction model, the parameters of each sub-neural network are the same.

[0115] The local traffic data is processed using the sub-neural networks of the second prediction model to obtain the local predicted traffic data for the next time step.

[0116] In one embodiment, a second prediction model is constructed using federated learning technology to achieve comprehensive optimization of all edge nodes in the network. The second prediction model includes multiple sub-neural networks, with the same sub-neural network deployed on each edge node. The model uses deep learning for resource allocation, and global model parameters are aggregated through interaction between each edge node and the data processing center. After the second prediction model is trained, the parameters of the sub-neural networks in all edge nodes are identical. By using each sub-neural network to predict local multi-timescale traffic data, network resource optimization is achieved for each edge node.

[0117] In this embodiment, the traffic data collected by each edge node does not need to be uploaded to the data processing center, thus protecting user privacy. Furthermore, compared to traditional centralized computing methods, this solution distributes model computation across various edge nodes, and the edge nodes only need to transmit a small amount of model parameter data to the data processing center, reducing the computational burden on individual nodes and improving resource allocation efficiency and overall network performance.

[0118] In addition, each edge node sends the sub-neural network parameters to the data processing center, which then aggregates the parameters based on considerations of the global nodes, achieving efficient global network resource optimization and making the dynamic adjustment of network resources of each edge node in the network more flexible and efficient.

[0119] As one embodiment of this application, training a second prediction model based on a neural network includes:

[0120] The local sub-neural network is iteratively trained using traffic data from each edge node. In each training round, the loss value is calculated and the parameters of the sub-neural network are updated based on the actual local traffic data and the predicted traffic data of the sub-neural network.

[0121] Calculate the comprehensive parameters based on the parameters of all sub-neural networks;

[0122] Based on the comprehensive parameters, the parameters of all sub-neural networks are updated, and the next round of training is performed.

[0123] In one embodiment, the local sub-neural network is iteratively trained using traffic data from each edge node, and parameter aggregation is performed in each training round.

[0124] Specifically, in each training round, the predicted traffic data of the sub-neural network is calculated using the prediction function, and the loss value is calculated based on the local actual traffic data and the predicted traffic data to update the parameters θ of the sub-neural network. i Among them, the loss value L i (θ i The calculation method is as follows:

[0125]

[0126] Where, N i Let f(Y) be the data volume of the i-th edge node, l be the loss function, and f(Y) be the loss function. i (t j );θ i Y is the prediction function of the sub-neural network. i (t j ) is the j-th type of compressed traffic data of the i-th edge node in the current time step.

[0127] For the parameters θ of each sub-neural network i Perform aggregation and calculate the federated parameter θglobal:

[0128]

[0129] Where M is the total number of edge nodes, and N is the total amount of compressed traffic data across all edge nodes.

[0130] Based on the aggregated federated parameters θglobal, the parameters θ of each sub-neural network are updated. i Then, continue with the next round of training until the model converges.

[0131] This embodiment uses federated learning technology to construct a second prediction model, enabling each edge node to jointly train the model while protecting local data privacy. This improves the generalization ability and prediction accuracy of the second prediction model and achieves efficient global network slice resource optimization in large-scale, distributed 5G networks.

[0132] As one implementation of this application, predicting the predicted traffic data of each edge node in the next time step includes:

[0133] Obtain a first predicted value for the edge node based on long-term traffic data, and a second predicted value based on short-term traffic data;

[0134] Long-term and short-term weights are set based on the device type of the edge node; the sum of the long-term and short-term weights is 1;

[0135] The predicted traffic data of the edge nodes are calculated based on the first predicted value, the second predicted value, the long-term weight, and the short-term weight of each edge node.

[0136] In one embodiment, short-term and long-term traffic data of each edge node are predicted separately using a prediction model (first prediction model or second prediction model), and the predicted traffic data of the edge node in the next time step is obtained by combining the prediction results of multiple time scales and conducting comprehensive analysis, thereby improving the accuracy of the prediction results.

[0137] Short-term traffic data reflects the trend of traffic changes over a recent period of time. For capturing short-term traffic fluctuations, it is defined as... Specifically, the most recent T s Traffic data at each time step is represented as follows:

[0138]

[0139] Among them, Y i (t) is the compressed traffic data of the i-th edge node at the current time step.

[0140] Long-term traffic data reflects long-term traffic trends and periodic characteristics, and is defined as follows: Long-term traffic data is from the past T l Traffic data at each time step is represented as follows:

[0141]

[0142] Among them, Y i (t) is the compressed traffic data of the i-th edge node at the current time step.

[0143] Combining short-term and long-term traffic data allows for a more comprehensive understanding of network traffic patterns and enables the construction of more accurate multi-time-scale prediction models, thereby improving prediction accuracy.

[0144] Specifically, long-term traffic data Input the prediction model to obtain the first predicted flow data: Short-term traffic data Input the prediction model to obtain the second predicted flow data:

[0145] Based on the device configuration type or service deployment of the edge nodes, set the corresponding long-term weight w. l and short-term weight w s This indicates a node's preference for long-term and short-term traffic. The long-term and short-term weights satisfy the following relationship: w s +w l =1.

[0146] Based on the first and second predicted values, calculate the predicted flow data.

[0147]

[0148] In this embodiment, the prediction results of short-term and long-term traffic data are weighted and fused to generate the final predicted load value. This comprehensive consideration of traffic characteristics at different time scales leads to more accurate and stable prediction results. This multi-time-scale prediction method not only improves prediction accuracy but also better adapts to different dynamic changes in traffic. Compared with single-time-scale prediction methods in related technologies, the method of this application can more comprehensively capture traffic change patterns, thus providing a more accurate basis for the dynamic adjustment of network resources.

[0149] As one implementation of this application, the target resource configuration in each edge node in the next time step is determined based on the target allocation strategy and the predicted traffic data of each edge node in the next time step, including:

[0150] Obtain the initial allocation strategy of the target network;

[0151] Based on the initial allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration in the next time step is calculated as the target resource configuration.

[0152] In one embodiment, the initial allocation strategy in the target network (i.e., the allocation strategy at the current time step) is used as the target allocation strategy, and the target resource configuration for the next time step is calculated based on the predicted traffic data of each edge node. Specifically, the local target resource configuration R in the next time step... slice,t+1 for:

[0153]

[0154] in, Here, R represents the predicted traffic data for the i-th edge node, g is the initial allocation strategy, and R is the predicted traffic data for the i-th edge node. slice,t This refers to the current network resource configuration of the target network.

[0155] As one embodiment of this application, after calculating the initial resource configuration in the next time step, the method further includes:

[0156] Obtain the current resource allocation of each network slice in each edge node, and calculate the overall utility and fairness of the target network based on the current resource allocation;

[0157] Based on the predicted traffic data of each edge node, the initial allocation strategy is adjusted using the gradient descent method to maximize the overall utility and optimize the fairness, thus obtaining the first allocation strategy.

[0158] Based on the first allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration of each edge node in the next time step is calculated, and the target resource configuration is updated.

[0159] In one embodiment, when dynamically optimizing network resources, it is also necessary to consider the overall utility of the network and the fairness of each edge node. Therefore, in this embodiment, the initial allocation strategy of the network is dynamically optimized based on overall utility and fairness to further improve the overall resource utilization and performance of the network.

[0160] Figure 2 This is a flowchart illustrating the adjustment of the target allocation strategy in one embodiment of this application. For example... Figure 2As shown, based on the current network resource configuration of each node in the target network, the initial target resource configuration is calculated in conjunction with the initial allocation strategy. Then, based on the current overall network resource configuration, the overall utility and fairness are calculated. Taking both factors into account, the gradient descent method is used to adjust the initial allocation strategy to maximize overall utility while achieving optimal fairness, thus obtaining the first allocation strategy.

[0161] First, the network resources of each edge node are configured, and the overall utility U(R) and fairness F(R) of the target network are calculated:

[0162]

[0163] Where N is the number of edge nodes, R i U represents the network resource configuration of the i-th edge node. i It is the utility function of the i-th edge node.

[0164] Then, the initial allocation strategy g is adjusted using gradient descent to maximize overall utility and achieve optimal fairness, yielding the optimized result, i.e., the first allocation strategy g. opt The specific optimization strategies are as follows:

[0165] (Maximizing overall utility)

[0166] F(R)≥θ; (Fairness is not less than the fairness threshold)

[0167] (The allocated network resources cannot exceed the total network resources.)

[0168] (The allocated network resources are not negative)

[0169] Among them, R i Let R represent the network resource configuration of the i-th edge node, θ be the fairness threshold, and R be the network resource allocation. total It represents the total allocatable network resources.

[0170] Furthermore, based on the optimization results (first allocation strategy), the network resource configuration of the target network is dynamically adjusted, and a new target resource configuration R is generated based on the following expression. slice-opt,t+1 :

[0171]

[0172] Among them, g opt It is the first allocation strategy. R is the predicted traffic data for the i-th edge node. slice,t This refers to the current network resource configuration of the target network.

[0173] Furthermore, based on the target resource allocation Rslice-opt,t+1 This allocates corresponding network resources to the network slices of each edge node.

[0174] In this embodiment, the overall utility of the target network and the fairness of each edge node are considered during the optimization of resource allocation. This ensures the fairness of resource allocation for each edge node while maximizing the overall utility of the network, thereby ensuring the service quality of each edge node.

[0175] As one embodiment of this application, the network slicing resource optimization method further includes:

[0176] Obtain the current network resource configuration of multiple target networks, and calculate the overall utility and overall fairness;

[0177] Based on optimization preferences, utility weights and fairness weights are set; the sum of the utility weights and the fairness weights is 1; the optimization preference is either utility priority or fairness priority.

[0178] Calculate the overall performance based on the utility weight and the fairness weight;

[0179] A comprehensive allocation strategy for the multiple target networks is obtained, and the comprehensive allocation strategy is adjusted using gradient descent to maximize the comprehensive performance, thus obtaining a second allocation strategy.

[0180] Based on the second allocation strategy, the current network resource configuration of each target network and the predicted traffic data, calculate the network resource configuration of each target network in the next time step as the comprehensive resource configuration.

[0181] Based on the comprehensive resource allocation, network resources are allocated to each target network.

[0182] In this embodiment, network resource configurations of multiple target networks are combined to dynamically allocate network resources across network layers, thereby achieving end-to-end resource optimization across layers. Figure 3 This is a flowchart illustrating the adjustment of a cross-layer target allocation strategy in one embodiment of this application. For example... Figure 3 As shown, the network resource configurations of each target network are first obtained, and the overall utility and fairness are calculated. Utility weights and fairness weights are set according to the optimization preference of dynamically adjusting network resources, indicating the preference for overall utility and fairness in this optimization. The overall performance is calculated based on the set weights. Based on the overall performance, the gradient descent method is used to adjust the overall allocation strategy of all target networks to maximize the overall performance, resulting in the optimized second allocation strategy.

[0183] In one embodiment, taking an access network, a transmission network, and a core network as examples, multiple target networks first obtain the current network resource status of each target network, and then determine the current network resource configuration. Specifically, the access network resource status is A(t), the transmission network resource status is T(t), and the core network resource status is C(t). The network resource status includes bandwidth, latency, packet loss rate, available resource quantity, etc.

[0184] Define resource state vector R layer :

[0185] R layer = [a(t); T(t); c(t)].

[0186] Based on the network resource allocation of each target network, calculate the overall comprehensive utility U(R(t)) and comprehensive fairness F(R(t)) of the network:

[0187]

[0188] Where N is the number of target networks, U(R) is the overall utility of the i-th network, and R(t) represents the network resource configuration of the i-th target network.

[0189] The goal of cross-network layer optimization is to maximize the overall performance of the network and ensure resource utilization and service quality at each network layer. Based on this, the comprehensive performance J(R(t)) of cross-layer optimization is defined as follows:

[0190] J(R(t))=αU(R(t))+βF(R(t));

[0191] Where U(R(t)) is the overall utility function across layers, F(R(t)) is the fairness function across layers, and α and β are weight coefficients, representing the relative importance of utility and fairness, respectively, and satisfying α+β=1.

[0192] A comprehensive allocation strategy for all target networks is obtained, and gradient descent is used as the optimization algorithm to adjust the comprehensive allocation strategy to maximize the overall utility J(R(t)), thus obtaining the second allocation strategy. The specific optimization rules are as follows:

[0193] (Maximizing overall utility)

[0194] (The allocated network resources cannot exceed the total network resources.)

[0195] (The allocated network resources are not negative)

[0196] Among them, R i R is the network resource configuration for the i-th target network. totalIt represents the total allocatable network resources.

[0197] Furthermore, based on the optimization results (second allocation strategy), a comprehensive resource allocation is calculated, and the network resource allocation of each target network is dynamically adjusted based on this comprehensive resource allocation. Specifically, a new comprehensive resource allocation R is generated based on the following expression. layer,t+1 :

[0198] R layer,t+1 =h(R(t),R layer );

[0199] Where h is the optimization result (second allocation strategy), R layer It is the current resource state vector.

[0200] In this embodiment, considering the multi-layered characteristics of the network, the network resource configuration of multiple target networks (network layers) is obtained and cross-layer optimization is performed to achieve dynamic resource allocation optimization across network layers based on overall utility and fairness. This enables more refined and efficient network resource management for multi-layered networks, improves the overall performance of the network, and ensures the stability and reliability of services in each network layer.

[0201] Based on the same inventive concept, one embodiment of this application provides a network slicing resource optimization device. (Reference) Figure 4 , Figure 4 This is a schematic diagram of a network slicing resource optimization device 100 according to an embodiment of this application. Figure 4 As shown, the device includes:

[0202] The perception module 101 is configured to collect traffic data of each edge node in the target network within the current time step according to the size of the time step; the traffic data includes: short-term traffic data and long-term traffic data of all network slices in the edge node;

[0203] The prediction module 102 is configured to predict the predicted traffic data of each edge node in the next time step based on the current traffic data of each edge node.

[0204] The optimization module 103 is configured to determine the target allocation strategy based on the predicted traffic data of each edge node in the next time step; and to determine the target resource configuration in each edge node in the next time step according to the target allocation strategy and the predicted traffic data of each edge node in the next time step.

[0205] The allocation module 104 is configured to allocate network resources to each edge node according to the target resource configuration.

[0206] As one embodiment of this application, the apparatus further includes a first training module configured to construct a first prediction model based on reinforcement learning and train the first prediction model using a Q-learning algorithm.

[0207] The prediction module 102 is specifically configured to use the trained first prediction model to process the traffic data of each edge node to obtain the predicted traffic data of each edge node in the next time step.

[0208] As one embodiment of this application, the apparatus further includes a second training module configured to construct and train a second prediction model based on a neural network, the second prediction model including sub-neural networks deployed at each edge node; in the trained second prediction model, the parameters of each sub-neural network are the same;

[0209] The prediction module 102 is specifically configured to use each sub-neural network of the second prediction model to process the local traffic data to obtain the local predicted traffic data for the next time step.

[0210] In one embodiment of this application, the second training module is configured to construct and train a second prediction model based on a neural network, specifically including:

[0211] The local sub-neural network is iteratively trained using traffic data from each edge node. In each training round, the loss value is calculated and the parameters of the sub-neural network are updated based on the actual local traffic data and the predicted traffic data of the sub-neural network.

[0212] Calculate the comprehensive parameters based on the parameters of all sub-neural networks;

[0213] Based on the comprehensive parameters, the parameters of all sub-neural networks are updated, and the next round of training is performed.

[0214] In one embodiment of this application, the prediction module 102 is configured to predict the predicted traffic data of each edge node in the next time step, specifically including:

[0215] Obtain a first predicted value for the edge node based on long-term traffic data, and a second predicted value based on short-term traffic data;

[0216] Long-term and short-term weights are set based on the device type of the edge node; the sum of the long-term and short-term weights is 1;

[0217] The predicted traffic data of the edge nodes are calculated based on the first predicted value, the second predicted value, the long-term weight, and the short-term weight of each edge node.

[0218] As one embodiment of this application, the optimization module 103 is configured to determine the target resource configuration in each edge node in the next time step based on the target allocation strategy and the predicted traffic data of each edge node in the next time step, specifically including:

[0219] Obtain the initial allocation strategy of the target network;

[0220] Based on the initial allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration in the next time step is calculated as the target resource configuration.

[0221] In one embodiment of this application, the optimization module 103 is further configured to perform the following operations after calculating the initial resource configuration for the next time step:

[0222] Obtain the current resource allocation of each network slice in each edge node, and calculate the overall utility and fairness of the target network based on the current resource allocation;

[0223] Based on the predicted traffic data of each edge node, the initial allocation strategy is adjusted using the gradient descent method to maximize the overall utility and optimize the fairness, thus obtaining the first allocation strategy.

[0224] Based on the first allocation strategy, the current network resource configuration of each edge node, and the predicted traffic data, the network resource configuration of each edge node in the next time step is calculated, and the target resource configuration is updated.

[0225] In one embodiment of this application, the sensing module 101 is configured to perform the following steps after collecting traffic data of each edge node in the target network within the current time step:

[0226] The traffic data is preprocessed, specifically including: removing invalid and erroneous data from the traffic data; detecting abnormal data in the traffic data and replacing the abnormal data with valid data from the previous time step; and compressing the traffic data using principal component analysis to obtain compressed data.

[0227] In one embodiment of this application, the optimization module 103 is further configured to perform the following steps:

[0228] Obtain the current network resource configuration of multiple target networks, and calculate the overall utility and overall fairness;

[0229] Based on optimization preferences, utility weights and fairness weights are set; the sum of the utility weights and the fairness weights is 1; the optimization preference is either utility priority or fairness priority.

[0230] Calculate the overall performance based on the utility weight and the fairness weight;

[0231] A comprehensive allocation strategy for the multiple target networks is obtained, and the comprehensive allocation strategy is adjusted using gradient descent to maximize the comprehensive performance, thus obtaining a second allocation strategy.

[0232] Based on the second allocation strategy, the current network resource configuration of each target network and the predicted traffic data, calculate the network resource configuration of each target network in the next time step as the comprehensive resource configuration.

[0233] The allocation module 104 is also configured to allocate network resources to each target network based on the comprehensive resource configuration.

[0234] Based on the same inventive concept, one embodiment of this application provides a network slicing resource optimization system. Figure 5 This is a schematic diagram of a network slicing resource optimization system proposed in an embodiment of this application. For example... Figure 5 As shown, the system includes:

[0235] Multiple target networks, including: an access network, a transmission network, and a core network; wherein, the access network is deployed with:

[0236] Multiple edge nodes as described in any of the above embodiments; wherein each edge node is equipped with a network slicing resource optimization device as described in any of the above embodiments;

[0237] The data processing center is used to aggregate traffic data or parameters of sub-neural networks sent by various edge nodes.

[0238] Based on the same inventive concept, one embodiment of this application provides a computer program product. This computer program product includes a computer program that, when executed by a processor, implements the steps of the network slicing resource optimization method as described in any of the above embodiments of this application.

[0239] Based on the same inventive concept, one embodiment of this application provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the network slicing resource optimization method as described in any of the above embodiments of this application.

[0240] Based on the same inventive concept, one embodiment of this application provides an electronic device, with reference to... Figure 6 . Figure 6 This is a schematic diagram of an electronic device according to an embodiment of this application. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the steps in the network slicing resource optimization method as described in any of the above embodiments of this application.

[0241] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0242] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0243] For the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and components involved are not necessarily essential to this application.

[0244] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0245] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0246] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0247] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0248] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, this application is to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of this application.

[0249] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0250] The above provides a detailed description of the network slicing resource optimization method, apparatus, system, and product provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A network slice resource optimization method, characterized in that, The method comprises the following steps: According to the size of the time step, collect the traffic data of each edge node in the target network in the current time step; The traffic data includes short-term traffic data and long-term traffic data of all network slices in the edge node; Based on the traffic data of each edge node at present, the predicted traffic data of each edge node in the next time step is predicted, which specifically includes: obtaining the first prediction value of the edge node based on the long-term traffic data, and the second prediction value based on the short-term traffic data; Set the long-term weight and short-term weight based on the device type of the edge node; The sum of the long-term weight and short-term weight is 1; According to the first prediction value, the second prediction value, the long-term weight and the short-term weight of each edge node, the predicted traffic data of the edge node is calculated; Based on the predicted traffic data of each edge node in the next time step, determine the target allocation strategy; According to the target allocation strategy and the predicted traffic data of each edge node in the next time step, determine the target resource configuration of each edge node in the next time step; According to the target resource configuration, allocate network resources to each edge node.

2. The network slice resource optimization method of claim 1, wherein, Based on the traffic data of each edge node at present, the predicted traffic data of each edge node in the next time step is predicted, which includes: Construct a first prediction model based on reinforcement learning, and train the first prediction model using Q learning algorithm; Using the trained first prediction model, the traffic data of each edge node is processed to obtain the predicted traffic data of each edge node in the next time step. 3.The network slice resource optimization method of claim 1, wherein, Based on the traffic data of each edge node at present, the predicted traffic data of each edge node in the next time step is predicted, which includes: Construct and train a second prediction model based on neural network, which includes a sub neural network deployed in each edge node; The parameters of each sub neural network in the trained second prediction model are the same; Using each sub neural network of the second prediction model, the local traffic data is processed to obtain the predicted traffic data of the local in the next time step.

4. The network slice resource optimization method of claim 3, wherein, Training the second prediction model based on neural network includes: Iteratively train the local sub neural network using the traffic data of each edge node. In each round of training, based on the actual traffic data and the predicted traffic data of the sub neural network, the loss value is calculated, and the parameters of the sub neural network are updated; According to the parameters of all sub neural networks, calculate the comprehensive parameters; Based on the comprehensive parameters, update the parameters of all sub neural networks and perform the next round of training.

5. The network slice resource optimization method of claim 1, wherein, According to the target allocation strategy and the predicted traffic data of each edge node in the next time step, the target resource configuration of each edge node in the next time step is determined, which includes: Obtain the initial allocation strategy of the target network; According to the initial allocation strategy, the current network resource configuration and the predicted traffic data of each edge node, the network resource configuration in the next time step is calculated as the target resource configuration.

6. The network slice resource optimization method of claim 5, wherein, After calculating the initial resource configuration in the next time step, it also includes: Obtaining current resource allocation of each network slice in each edge node, calculating overall utility and fairness of the target network based on the current resource allocation; Based on the predicted traffic data of each edge node, the initial allocation strategy is adjusted using gradient descent method to maximize the overall utility and optimize the fairness, obtaining a first allocation strategy; Based on the first allocation strategy, current network resource configuration of each edge node and predicted traffic data, network resource configuration in the next time step in each edge node is calculated, and the target resource configuration is updated.

7. The network slice resource optimization method of claim 1, wherein, After collecting the traffic data of each edge node in the target network in the current time step, it further includes: The traffic data is preprocessed, specifically including: removing invalid data and error data in the traffic data; detecting abnormal data in the traffic data, replacing the abnormal data with valid data in the previous time step; using principal component analysis method to compress the traffic data to obtain compressed data. 8.The method of claim 1, wherein, Further comprising: Obtaining current network resource configuration of multiple target networks, calculating comprehensive utility and comprehensive fairness; According to the optimization preference, setting utility weight and fairness weight; the sum of the utility weight and the fairness weight is 1; the optimization preference is utility priority or fairness priority; Based on the utility weight and the fairness weight, the comprehensive performance is calculated; Obtaining the comprehensive allocation strategy of the multiple target networks, adjusting the comprehensive allocation strategy using gradient descent method to maximize the comprehensive performance, obtaining a second allocation strategy; According to the second allocation strategy, current network resource configuration of each target network and predicted traffic data, calculating network resource configuration of each target network in the next time step as comprehensive resource configuration; Based on the comprehensive resource configuration, network resources are allocated to each target network. 9.A network slice resource optimization apparatus, characterized in that, For implementing the method of any one of claims 1-8, comprising: A perception module configured to collect traffic data of each edge node in the target network in the current time step according to the size of the time step; the traffic data includes short-time traffic data and long-time traffic data of all network slices in the edge node; A prediction module configured to predict predicted traffic data of each edge node in the next time step based on the traffic data of each edge node at present; An optimization module configured to determine a target allocation strategy based on the predicted traffic data of each edge node in the next time step; determine target resource configuration in each edge node in the next time step according to the target allocation strategy and the predicted traffic data of each edge node in the next time step; An allocation module configured to allocate network resources to each edge node according to the target resource configuration. 10.A network slice resource optimization system, characterized in that, Comprising: Multiple target networks, the multiple target networks include: access network, transmission network and core network; wherein, multiple edge nodes are deployed in the access network; wherein, each edge node is deployed with a network slice resource optimization device as claimed in claim 9; A data processing center for aggregating traffic data or parameters of sub neural networks sent by each edge node. ​ 11. A computer program product comprising a computer program, characterized in that, The computer program, which when executed by a processor, implements the steps of the method according to any one of claims 1-8.

12. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program, which when executed by a processor, implements the steps of the method according to any one of claims 1-8.

13. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor, when executing the computer program, implements the steps of the method according to any one of claims 1-8.

Images