Network resource configuration method, device, electronic device and medium based on federated learning

Abstract

The present application discloses a network resource configuration method, device, electronic device and medium based on federated learning. By applying the technical solution of the present application, the macro base station can receive the model parameters of multiple micro base stations under the federated learning architecture and summarize them to obtain the global model parameters and then transmit them back to the micro base station. Then, the micro base station trains the prediction model based on the global model parameters to obtain a trained target prediction model. On the one hand, the present application can achieve a method of using a federated learning framework to realize multi-party collaborative training of the load prediction model of base stations between slices while ensuring user privacy and data non-sharing, and can achieve the purpose of distributed resource prediction of base stations at the slice level. At the same time, it also avoids the problem existing in the related technology that the insufficient resource prediction performance of the base station in the network slice easily affects the progress of user services.

Description

Technical Field

[0001] The present application relates to data processing technology, and in particular to a network resource configuration method, device, electronic device and medium based on federated learning. Background Art

[0002] Network Slicing (NS) is one of the core technologies of 5G / B5G communication systems to empower vertical industries. It refers to the virtualization of unified physical network infrastructure resources and abstraction into multiple end-to-end (E2E) logical networks.

[0003] In related technologies, in order to realize the vision of automated management and orchestration of network slices, the running network slices need to have the ability to perceive, predict and actively respond to user needs in real time. In wireless networks, network loads often have temporal and spatial correlations and can be predicted and perceived in advance. Among them, existing load prediction technologies are usually implemented through centralized controllers. For example, the controller collects global user data and network status information, centrally trains global ML models and performs predictive analysis to achieve real-time perception of the future load status of the slice.

[0004] However, in the network slicing scenario, a slice network may be deployed on multiple base stations, and its historical load data is geographically dispersed. In addition, network slices are isolated from each other, and user data privacy is sensitive, making it difficult to establish a centralized machine learning model to analyze and optimize the resource prediction performance of network slices. This leads to problems that affect business progress. Summary of the invention

[0005] The embodiments of the present application provide a method, device, electronic device and medium for configuring network resources based on federated learning, which are used to solve the problem in the related art that the insufficient resource prediction performance of base stations in network slicing easily affects the progress of user services.

[0006] According to one aspect of an embodiment of the present application, a network resource configuration method based on federated learning is provided, which is applied to a micro base station, including:

[0007] Using a local target resource prediction model, determining the traffic information to be matched in a future time period, wherein the target resource prediction model is a model obtained by collaborative training of the micro base station under a federated learning framework;

[0008] If it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, where the resource acquisition request carries the traffic information to be matched;

[0009] Receive the slice resource configuration signaling allocated by the macro base station, and complete the resource configuration of the network slice based on the slice resource configuration signaling.

[0010] Optionally, in another embodiment of the method of the present application, before using the local target resource prediction model to determine the traffic information to be matched consumed in the future time period, it also includes:

[0011] Obtaining an initial resource prediction model sent by the macro base station;

[0012] Using local data to train the initial resource prediction model to obtain initial model parameters;

[0013] Sending the initial model parameters to the macro base station, so that the macro base station performs federation aggregation on the received initial model parameters to obtain global model parameters;

[0014] The global model parameters sent by the macro base station are received, and the target resource prediction model is obtained based on the global model parameters.

[0015] Optionally, in another embodiment of the method of the present application, obtaining the target resource prediction model based on the global model parameters includes:

[0016] Using the global model parameters, performing a preset model accuracy test on the initial resource prediction model;

[0017] If it is determined that the initial resource prediction model meets the preset model accuracy, the initial resource prediction model is used as the target resource prediction model; or,

[0018] If it is determined that the initial resource prediction model does not meet the preset model accuracy, the initial resource prediction model is trained using the global model parameters to obtain the target resource prediction model.

[0019] Optionally, in another embodiment of the above method of the present application, after determining the traffic information to be matched consumed in the future time period, the method further includes:

[0020] If it is determined that the network resources provided by the network slice can cover the traffic information to be matched, the resource configuration of the network slice is completed in a preset manner.

[0021] According to another aspect of an embodiment of the present application, a network resource configuration method based on federated learning is provided, which is applied to a macro base station, including:

[0022] Receive a resource acquisition request sent by a micro base station and carrying traffic information to be matched, where the resource acquisition request is used to allocate preset resources for a network slice associated with the micro base station;

[0023] Determine the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice;

[0024] Sending slice resource configuration signaling to the micro base station for allocating the preset resources to the network slice.

[0025] Optionally, in another embodiment of the method of the present application, determining the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice includes:

[0026] Extracting the resource demand contained in the traffic information to be matched; and obtaining the network priority and network deployment location corresponding to the network slice;

[0027] Using an optimization algorithm, a resource allocation decision is made based on the resource demand, the network priority, and the network deployment location to determine the preset resources allocated to the network slice.

[0028] Optionally, in another embodiment of the method of the present application, before receiving the resource acquisition request carrying the traffic information to be matched sent by the micro base station, it also includes:

[0029] Sending an initial resource prediction model to the plurality of micro base stations, and receiving initial model parameters sent by the plurality of micro base stations, wherein the initial model parameters are obtained by the micro base station training the initial resource prediction model using local data;

[0030] Federate and aggregate the received initial model parameters to obtain global model parameters;

[0031] The global model parameters are sent to the micro base station.

[0032] According to another aspect of the embodiment of the present application, a network resource configuration device based on federated learning is provided, which is applied to a micro base station, including:

[0033] A determination module is configured to determine the traffic information to be matched consumed in a future time period by using a local target resource prediction model, wherein the target resource prediction model is a model obtained by collaborative training of the micro base station under a federated learning framework;

[0034] A sending module is configured to send a resource acquisition request to a macro base station associated with itself if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, wherein the resource acquisition request carries the traffic information to be matched;

[0035] The configuration module is configured to receive the slice resource configuration signaling allocated by the macro base station and complete the resource configuration of the network slice based on the slice resource configuration signaling.

[0036] According to another aspect of the embodiments of the present application, an electronic device is provided, including:

[0037] A memory for storing executable instructions; and

[0038] A display is used to execute the executable instructions together with the memory to complete the operation of any of the above-mentioned network resource configuration methods based on federated learning.

[0039] According to another aspect of an embodiment of the present application, a computer-readable storage medium is provided for storing computer-readable instructions, which, when executed, perform the operations of any of the above-mentioned network resource configuration methods based on federated learning.

[0040] In the present application, the micro base station can use the local target resource prediction model to determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by the micro base station through collaborative training under the federated learning framework; if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, and the resource acquisition request carries the traffic information to be matched; the slice resource configuration signaling allocated by the macro base station is received, and the resource configuration of the network slice is completed based on the slice resource configuration signaling. The macro base station can receive the model parameters in multiple micro base stations under the federated learning architecture and summarize them to obtain the global model parameters and then transmit them back to the micro base station. Then the micro base station trains the prediction model based on the global model parameters to obtain the trained target prediction model. On the one hand, the present application can achieve a method of using the federated learning framework to realize the multi-party collaborative training of the load prediction model of the base station between slices under the premise of ensuring user privacy and data non-sharing, and can achieve the purpose of distributed resource prediction of base stations at the slice level. At the same time, it also avoids the problem that the resource prediction performance of the base station in the network slice is insufficient and easily affects the progress of user services in the related technology.

[0041] The technical solution of the present application is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present application and, together with the description, serve to explain the principles of the present application.

[0043] The present application can be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0044] Figure 1 A schematic diagram of a network resource configuration method based on federated learning provided in an embodiment of the present application is shown;

[0045] Figure 2 A schematic diagram of the system architecture of a network resource configuration method based on federated learning provided in an embodiment of the present application is shown;

[0046] Figure 3 A schematic diagram of a process of a network resource configuration method based on federated learning provided in an embodiment of the present application is shown;

[0047] Figure 4 A schematic diagram showing another method for configuring network resources based on federated learning provided in an embodiment of the present application is shown;

[0048] Figure 5 A schematic diagram of a process for a macro base station to determine resource allocation in another network resource configuration based on federated learning provided by an embodiment of the present application is shown;

[0049] Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present application is shown;

[0050] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present application is shown;

[0051] Figure 8 A schematic diagram of a storage medium provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0052] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0053] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0054] The following description of at least one exemplary embodiment is merely illustrative in nature and is not intended to limit the present application, its application, or uses.

[0055] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0056] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0057] In addition, the technical solutions between the various embodiments of the present application can be combined with each other, but it must be based on the fact that ordinary technicians in the field can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0058] It should be noted that all directional indications in the embodiments of the present application (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0059] Combine the following Figure 1-Figure 3 To describe a method for performing network resource configuration based on federated learning according to an exemplary embodiment of the present application. It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this regard. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0060] The present application also proposes a network resource configuration method, device, electronic device and medium based on federated learning.

[0061] Figure 1 The flowchart of a method for configuring network resources based on federated learning according to an embodiment of the present application is schematically shown. Figure 1 As shown, the method is applied to a micro base station, comprising:

[0062] S101, using the local target resource prediction model, determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by collaborative training of micro base stations under the federated learning framework.

[0063] S102: If it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, and the resource acquisition request carries the traffic information to be matched.

[0064] S103, receive the slice resource configuration signaling allocated by the macro base station, and complete the resource configuration of the network slice based on the slice resource configuration signaling.

[0065] Among the related technologies, Network Slicing (NS) is one of the core technologies of 5G / B5G communication systems to empower vertical industries. It refers to the virtualization of unified physical network infrastructure resources and abstraction into multiple end-to-end (End to End, E2E) logical networks.

[0066] Furthermore, these logical networks are isolated from each other and each serves a specific business scenario independently, thus meeting the customized and diversified quality of service (QoS) requirements of users. Network slicing is based on SDN / NFV technology and is a set of virtual network functions (VNFs) and their resources. Different VNFs can implement a specific network function.

[0067] Among them, a network slice is composed of multiple VNFs, thus forming a complete logical network to realize customized services. In a network slice, it can be divided into three parts: radio network (AN) sub-slice, bearer network (TN) sub-slice and core (CN) network sub-slice. In the 5G era, ITU formally defines three typical application scenarios of network slicing: enhanced mobile broadband (eMBB), massive machine type communication (mMTC), and ultra-reliable and low latency communication (URLLC).

[0068] At present, network slicing can basically realize the cross-domain integration of end-to-end business processes and provide customized services to users. The technical goal of the next stage is to realize cross-domain automated management, orchestration and configuration of network slicing and enhance the end-to-end automation capability of network slicing under the premise of guaranteeing the service level agreement (SLA). This is inseparable from the support of load sensing technology driven by artificial intelligence (AI) and big data.

[0069] Furthermore, for the use of network slicing to implement AI-based load perception and resource prediction technology, in order to realize the vision of automated management and orchestration of network slicing, the running slices need to have the ability to perceive and predict in real time and actively respond to user needs. In wireless networks, network loads often have temporal and spatial correlations and can be predicted and perceived in advance. Due to differences in business types and user groups, the traffic patterns of different slices are heterogeneous (for example, different business types have different traffic peak periods, and some scenarios have stronger traffic bursts, etc.).

[0070] The essence of network resource prediction is time series prediction. Traditional resource prediction is mainly based on time series models. The specific parameters of the model are solved according to actual data, and finally the time series model with known parameters is used for time series prediction. Common time series models include Moving Average (MA), Auto Regressive (AR), Auto Regressive Moving Average (ARMA), and Auto Regressive Integrated Moving Average (ARIMA).

[0071] In response to the above problems, this application proposes a network resource configuration method based on federated learning. The idea is that the micro base station uses the local target resource prediction model to determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by the micro base station through collaborative training under the federated learning framework; if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, and the resource acquisition request carries the traffic information to be matched; the slice resource configuration signaling allocated by the macro base station is received, and the resource configuration of the network slice is completed based on the slice resource configuration signaling.

[0072] Furthermore, in the 6G era, a large amount of user data will be distributed at the edge of the network, which requires wireless networks to have more ubiquitous and native intelligence capabilities. Specifically, AI is not only deployed in the centralized SDN controller that manages slices, but also exists in network nodes and user equipment (UE) in a distributed paradigm. Among various distributed network architectures, federated learning (FL) is considered to be a potential important solution for realizing 6G ubiquitous intelligence due to its advantages such as protecting privacy and overcoming data silos.

[0073] Federated learning is essentially a distributed machine learning framework. The goal is to achieve joint modeling and improve the effectiveness of AI models while ensuring data privacy, security, and legal compliance. Under the FL framework, local clients use a large amount of local user data to train local models, and upload local model parameters to centralized base stations / edge servers for aggregation to obtain a global model. This can achieve multi-party collaborative training of ML models while protecting user privacy and not sharing data.

[0074] In one embodiment, the prediction model in the embodiment of the present application can be a linear regression model (LinearRegression, LR), a support vector regression model (Support Vector Regression, SVR), a long short-term memory model (Long Short Time Memory, LSTM) neural network, etc.

[0075] As an example, the prediction model can be a long short-term memory model. It can be understood that the LSTM network is an improved recurrent neural network (RNN) that can solve the problem that RNN cannot handle long-distance dependencies. It adds a state (called a cell state) to the hidden layer of the original RNN to save the long-term state, and introduces a gate structure to control the retention and discarding of historical memory information, control the retention of highly relevant information, and delete less relevant content.

[0076] Therefore, LSTM has good memory performance and can effectively solve the gradient explosion and gradient vanishing problems of RNN. ML-based network resource prediction is mainly implemented through centralized network elements such as data centers, centralized controllers (such as SDN controllers) or intelligent network functions (such as network data intelligent analysis functions (NWDAF)). The centralized control unit uniformly collects global user data and network status information, performs intelligent analysis, prediction and decision-making on the data, and realizes the perception of the future load status of the slice.

[0077] In one way, Figure 2 The figure shows the system architecture diagram of the network resource configuration method based on federated learning. In the two-layer heterogeneous network scenario composed of macro and micro base stations, different network slices (composed of different service flows) share the same physical network infrastructure in one area. That is, one micro base station can carry multiple slices, or one slice can be deployed on multiple micro base stations.

[0078] Among them, users requesting the same service (slice) may access different micro base stations. Therefore, the sample data in each slice is distributed on multiple micro base stations and is geographically dispersed. In order to collect distributed slice data and provide relatively reliable slice-level predictions with lower communication overhead and latency, in the solution of this application, the load prediction between base stations in a slice is implemented based on horizontal federated learning.

[0079] Further, such as Figure 3 As shown, it is a flowchart of the network resource configuration method based on federated learning proposed in this application, which includes:

[0080] Step 1: Each micro base station participating in federated learning receives the initial resource prediction model sent by the macro base station.

[0081] It should be noted that the various micro base stations may belong to the same slice network or to different slice networks.

[0082] Step 2: The micro base station trains the initial resource prediction model using local data to obtain initial model parameters.

[0083] Step 3: The micro base station sends the initial model parameters to the macro base station.

[0084] In this step, after receiving the initial model parameters sent by multiple micro base stations, the macro base station can perform federation aggregation on the received initial model parameters to obtain global model parameters.

[0085] In one manner, the aggregation mode of the macro base station may be FedAvg, FedProx, SCAFFOLD, etc. This application does not limit this.

[0086] Step 4: The macro base station sends the global model parameters of the current slice to the micro base station.

[0087] Step 5: The micro base station uses the global model parameters to perform a preset model accuracy test on the initial resource prediction model to determine whether the current model has converged. There are two cases:

[0088] The first case: if it is determined that the initial resource prediction model meets the preset model accuracy, the initial resource prediction model is used as the target resource prediction model.

[0089] The second case: if it is determined that the initial resource prediction model does not meet the preset model accuracy, the initial resource prediction model is trained using the global model parameters until the target resource prediction model is obtained.

[0090] At this point, the intra-slice load prediction based on horizontal federated learning is completed, and each base station executes the next decision-making process according to the prediction results.

[0091] Step 6: The micro base station uses a local target resource prediction model to determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by collaborative training of the micro base station under the federated learning framework.

[0092] Step 7: The micro base station calculates the network resources provided by the network slice associated with itself based on the prediction results (i.e., the traffic information to be matched).

[0093] Step 8: The micro base station determines whether the network resources provided by the network slice can cover the traffic information to be matched. If sufficient, the slice resources are directly allocated according to the requested amount. Otherwise, execute step 9.

[0094] Step 9: The micro base station sends a resource acquisition request carrying traffic information to be matched to the macro base station.

[0095] Step 10: The macro base station performs traffic segmentation and makes resource allocation decisions based on information such as resource demand, network priority corresponding to the network slice, and network deployment location through optimization algorithms (linear optimization, game theory, reinforcement learning, etc.), thereby determining the preset resources allocated to the network slice.

[0096] In one embodiment, the macro base station in the embodiment of the present application stores attribute information of all network slices, and the attribute information includes slice information of all network slices associated with it, including slice type, priority, deployment status, etc.).

[0097] Step 11. The macro base station sends a slice resource configuration signaling to the micro base station for allocating preset resources for the network slice.

[0098] Step 12: The micro base station receives the slice resource configuration signaling allocated by the macro base station, and completes the resource configuration of the network slice based on the slice resource configuration signaling.

[0099] In the present application, the micro base station can use the local target resource prediction model to determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by the micro base station through collaborative training under the federated learning framework; if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, and the resource acquisition request carries the traffic information to be matched; the slice resource configuration signaling allocated by the macro base station is received, and the resource configuration of the network slice is completed based on the slice resource configuration signaling. The macro base station can receive the model parameters in multiple micro base stations under the federated learning architecture and summarize them to obtain the global model parameters and then transmit them back to the micro base station. Then the micro base station trains the prediction model based on the global model parameters to obtain the trained target prediction model. On the one hand, the present application can achieve a method of using the federated learning framework to realize the multi-party collaborative training of the load prediction model of the base station between slices under the premise of ensuring user privacy and data non-sharing, and can achieve the purpose of distributed resource prediction of base stations at the slice level. At the same time, it also avoids the problem that the resource prediction performance of the base station in the network slice is insufficient and easily affects the progress of user services in the related technology.

[0100] Optionally, in another embodiment of the method of the present application, before using the local target resource prediction model to determine the traffic information to be matched consumed in the future time period, the method further includes:

[0101] Obtaining the initial resource prediction model sent by the macro base station;

[0102] Use local data to train the initial resource prediction model to obtain initial model parameters;

[0103] Sending the initial model parameters to the macro base station so that the macro base station aggregates the received initial model parameters to obtain the global model parameters;

[0104] The global model parameters sent by the macro base station are received, and a target resource prediction model is obtained based on the global model parameters.

[0105] Optionally, in another embodiment of the method of the present application, a target resource prediction model is obtained based on the global model parameters, including:

[0106] Use global model parameters to test the accuracy of the initial resource prediction model;

[0107] If it is determined that the initial resource prediction model meets the preset model accuracy, the initial resource prediction model is used as the target resource prediction model; or,

[0108] If it is determined that the initial resource prediction model does not meet the preset model accuracy, the initial resource prediction model is trained using the global model parameters to obtain the target resource prediction model.

[0109] Optionally, in another embodiment of the method of the present application, after determining the traffic information to be matched consumed in the future time period, the method further includes:

[0110] If it is determined that the network resources provided by the network slice can cover the traffic information to be matched, the resource configuration of the network slice is completed in the preset manner.

[0111] In one embodiment, the technical solution proposed in this application can provide a slice-level heterogeneous load prediction method based on federated learning. This achieves the purpose of using the federated learning framework to realize multi-party collaborative training of the slice load prediction model while ensuring user privacy and data non-sharing, and can realize slice-level distributed heterogeneous load prediction and improve prediction performance.

[0112] Among them, since each micro base station adopts the solution of only uploading model parameters instead of uploading original data, it can effectively reduce the communication overhead and delay caused by data exchange, and at the same time effectively avoid the privacy sensitivity and data security issues caused by data sharing.

[0113] In one approach, multiple micro base stations use a horizontal federated learning framework to collaboratively train the current service slice load prediction model. The micro base station records the historical load data of users in the cell, and deploys a local intelligent training unit on it as a local client to train a local load prediction model (for example, an LSTM traffic prediction model).

[0114] In another approach, the macro base station acts as the central server of federated learning, responsible for aggregating model parameters and issuing global models for each slice. Each micro base station participating in the horizontal federation carries the same service (slice) type, with similar features, but different user groups. At the same time, each slice with heterogeneous service types has a parallel horizontal federation traffic load prediction framework and different global models.

[0115] Figure 4 The flowchart of a method for configuring network resources based on federated learning according to an embodiment of the present application is schematically shown. Figure 4 As shown, the method is applied to a macro base station, comprising:

[0116] S201, receiving a resource acquisition request sent by the micro base station and carrying traffic information to be matched, where the resource acquisition request is used to allocate preset resources for a network slice associated with the micro base station.

[0117] S202, based on the traffic information to be matched and the attribute information of the network slice, determine the preset resources allocated to the network slice.

[0118] S203, sending slice resource configuration signaling to the micro base station for allocating preset resources to the network slice.

[0119] Optionally, in another embodiment of the method of the present application, based on the traffic information to be matched and the attribute information of the network slice, determining the preset resources allocated to the network slice includes:

[0120] Extract the resource demand contained in the traffic information to be matched; and obtain the network priority and network deployment location corresponding to the network slice;

[0121] Using optimization algorithms, resource allocation decisions are made based on resource demand, network priority, and network deployment location to determine the preset resources allocated to network slices.

[0122] Optionally, in another embodiment of the method of the present application, before receiving the resource acquisition request carrying the traffic information to be matched sent by the micro base station, it also includes:

[0123] Sending an initial resource prediction model to multiple micro base stations, and receiving initial model parameters sent by the multiple micro base stations, where the initial model parameters are obtained by the micro base stations training the initial resource prediction model using local data;

[0124] Federate and aggregate the received initial model parameters to obtain global model parameters;

[0125] Send the global model parameters to the micro base station.

[0126] In one way, continue with Figure 3 The network resource configuration method based on federated learning proposed in this application is illustrated by taking an example:

[0127] Step 1: Each micro base station participating in federated learning receives the initial resource prediction model sent by the macro base station.

[0128] Step 2: The micro base station trains the initial resource prediction model using local data to obtain initial model parameters.

[0129] Step 3: The micro base station sends the initial model parameters to the macro base station.

[0130] Step 4: The macro base station sends the global model parameters of the current slice to the micro base station.

[0131] Step 5: The micro base station uses the global model parameters to perform a preset model accuracy test on the initial resource prediction model to determine whether the current model has converged. There are two cases:

[0132] Step 6: The micro base station uses a local target resource prediction model to determine the traffic information to be matched in the future time period. The target resource prediction model is a model obtained by collaborative training of the micro base station under the federated learning framework.

[0133] Step 7: The micro base station calculates the network resources provided by the network slice associated with itself based on the prediction results (i.e., the traffic information to be matched).

[0134] Step 8: The micro base station determines whether the network resources provided by the network slice can cover the traffic information to be matched. If sufficient, the slice resources are directly allocated according to the requested amount. Otherwise, execute step 9.

[0135] Step 9: The micro base station sends a resource acquisition request carrying traffic information to be matched to the macro base station.

[0136] Step 10: The macro base station determines the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice.

[0137] In one approach, the macro base station needs to comprehensively consider slice priority, resource demand, and slice deployment, and make the best resource allocation decision through optimization algorithms (linear optimization, game theory, reinforcement learning, etc.).

[0138] As an example, the embodiments of the present application can abstract the multi-slice resource configuration problem into a two-dimensional knapsack problem. The length and width of the knapsack represent the time that the slice occupies resources and the amount of allocated resources, respectively. The optimization target is the weight and (slice SLA satisfaction rate and slice provider revenue) of the items in the knapsack. By optimizing the selection scheme and placement order of items (different slice resource allocations) in the knapsack, the optimal time-frequency resource configuration scheme is obtained.

[0139] It should be noted that since network traffic fluctuates in real time, in order to achieve more fine-grained resource configuration, the embodiment of the present application divides the predicted traffic in the time domain before macro base station resource configuration (within the segmentation granularity, the slice resource allocation amount remains unchanged).

[0140] Further, such as Figure 5 As shown. Where T p is the flow prediction window, T s is the resource scheduling window. After traffic prediction and segmentation, the resource allocation problem of the slice is transformed from the original rectangular knapsack problem to a polygonal knapsack problem (such as Figure 5 The time-frequency resource blocks occupied by slice NS1 are shaped into polygons by the predicted traffic. In this way, the free time-frequency resource blocks can be allocated to other slices to improve resource utilization.

[0141] Taking wireless bandwidth resources as an example: According to Shannon's formula, the mapping relationship between traffic and bandwidth resource demand can be obtained as shown in the following formula:

[0142]

[0143] Where k and i are the numbers of the micro base station and network slice respectively, W k,i represents the amount of bandwidth resources allocated by micro base station k to network slice i, Represents the traffic value of network slice i on micro base station k at time t.

[0144] It is understandable that when there are sufficient remaining resources, wireless bandwidth resources can be directly allocated in a preset manner to meet the slice load requirements. When bandwidth resources are insufficient (i.e., there is resource competition between multiple slices or base stations), since the resource allocation of a single network slice requires the coordination of multiple micro base stations, resource allocation decisions need to be centrally controlled by the macro base station.

[0145] Step 11. The macro base station sends a slice resource configuration signaling to the micro base station for allocating preset resources for the network slice.

[0146] Step 12: The micro base station receives the slice resource configuration signaling allocated by the macro base station, and completes the resource configuration of the network slice based on the slice resource configuration signaling.

[0147] The macro base station can receive the model parameters of multiple micro base stations under the federated learning architecture and summarize them to obtain the global model parameters and then transmit them back to the micro base station. The micro base station then trains the prediction model based on the global model parameters to obtain a trained target prediction model. On the one hand, the present application can achieve a method of using a federated learning framework to realize multi-party collaborative training of load prediction models of base stations between slices while ensuring user privacy and non-sharing of data, and can achieve the purpose of distributed resource prediction of base stations at the slice level. At the same time, it also avoids the problem existing in the related art that the insufficient resource prediction performance of base stations in network slices easily affects the progress of user services.

[0148] Optionally, in another embodiment of the present application, Figure 6 As shown, the present application also provides a network resource configuration device based on federated learning. It is applied to a micro base station and includes:

[0149] The determination module 301 is configured to determine the traffic information to be matched consumed in the future time period by using a local target resource prediction model, wherein the target resource prediction model is a model obtained by collaborative training of the micro base station under a federated learning framework;

[0150] The sending module 302 is configured to send a resource acquisition request to a macro base station associated with itself if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, wherein the resource acquisition request carries the traffic information to be matched;

[0151] The configuration module 303 is configured to receive the slice resource configuration signaling allocated by the macro base station, and complete the resource configuration of the network slice based on the slice resource configuration signaling.

[0152] The macro base station can receive the model parameters of multiple micro base stations under the federated learning architecture and summarize them to obtain the global model parameters and then transmit them back to the micro base station. The micro base station then trains the prediction model based on the global model parameters to obtain a trained target prediction model. On the one hand, the present application can achieve a method of using a federated learning framework to realize multi-party collaborative training of load prediction models of base stations between slices while ensuring user privacy and non-sharing of data, and can achieve the purpose of distributed resource prediction of base stations at the slice level. At the same time, it also avoids the problem existing in the related art that the insufficient resource prediction performance of base stations in network slices easily affects the progress of user services.

[0153] In another embodiment of the present application, the determining module 301 is configured to perform the following steps:

[0154] Obtaining an initial resource prediction model sent by the macro base station;

[0155] Using local data to train the initial resource prediction model to obtain initial model parameters;

[0156] Sending the initial model parameters to the macro base station, so that the macro base station performs federation aggregation on the received initial model parameters to obtain global model parameters;

[0157] The global model parameters sent by the macro base station are received, and the target resource prediction model is obtained based on the global model parameters.

[0158] In another embodiment of the present application, the determining module 301 is configured to perform the following steps:

[0159] Using the global model parameters, performing a preset model accuracy test on the initial resource prediction model;

[0160] If it is determined that the initial resource prediction model meets the preset model accuracy, the initial resource prediction model is used as the target resource prediction model; or,

[0161] If it is determined that the initial resource prediction model does not meet the preset model accuracy, the initial resource prediction model is trained using the global model parameters to obtain the target resource prediction model.

[0162] In another embodiment of the present application, the determining module 301 is configured to perform the following steps:

[0163] If it is determined that the network resources provided by the network slice can cover the traffic information to be matched, the resource configuration of the network slice is completed in a preset manner.

[0164] The present application also provides an electronic device to execute the above-mentioned network resource configuration method based on federated learning. Figure 7 , which shows a schematic diagram of an electronic device provided by some embodiments of the present application. Figure 7 As shown, the electronic device 4 includes: a processor 400, a memory 401, a bus 402 and a communication interface 403, and the processor 400, the communication interface 403 and the memory 401 are connected via the bus 402; the memory 401 stores a computer program that can be run on the processor 400, and when the processor 400 runs the computer program, it executes the network resource configuration method based on federated learning provided in any of the aforementioned embodiments of the present application.

[0165] The memory 401 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage. The communication connection between the device network element and at least one other network element is realized through at least one communication interface 403 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0166] The bus 402 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 401 is used to store programs, and the processor 400 executes the programs after receiving execution instructions. The data identification method disclosed in any implementation of the above-mentioned embodiment of the present application may be applied to the processor 400, or implemented by the processor 400.

[0167] The processor 400 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 400. The above processor 400 can be a general-purpose processor, including a processor (Central Processing Unit, referred to as CPU), a network processor (Network Processor, referred to as NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a readily available programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to be executed, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 401, and the processor 400 reads the information in the memory 401 and completes the steps of the above method in combination with its hardware.

[0168] The electronic device provided in the embodiment of the present application and the network resource configuration method based on federated learning provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented therein.

[0169] The present application also provides a computer-readable storage medium corresponding to the network resource configuration method based on federated learning provided in the above embodiment. Figure 8 The computer-readable storage medium shown is a CD 50 on which a computer program (ie, a program product) is stored. When the computer program is executed by a processor, the network resource configuration method based on federated learning provided in any of the aforementioned embodiments will be executed.

[0170] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0171] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the method for data identification provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0172] It should be noted that:

[0173] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known structures and technologies are not shown in detail so as not to obscure the understanding of this description.

[0174] Similarly, it should be understood that in order to streamline the present application and help understand one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be interpreted as reflecting the following schematic diagram: the claimed application requires more features than the features clearly stated in each claim. More specifically, as reflected in the claims below, the inventive aspects are less than all the features of the single embodiment disclosed above. Therefore, the claims following the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself serves as a separate embodiment of the present application.

[0175] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present application and form different embodiments. For example, in the claims below, any one of the claimed embodiments may be used in any combination.

[0176] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A network resource configuration method based on federated learning, characterized in that: Applied to micro base stations, including: Using a local target resource prediction model, determining the traffic information to be matched in a future time period, wherein the target resource prediction model is a model obtained by collaborative training of the micro base station under a federated learning framework; If it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, a resource acquisition request is sent to the macro base station associated with itself, so that the macro base station receives the resource acquisition request, and determines the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice; a slice resource configuration signaling for allocating the preset resources to the network slice is sent to the micro base station; the resource acquisition request carries the traffic information to be matched; Receive slice resource configuration signaling allocated by the macro base station, and complete resource configuration of the network slice based on the slice resource configuration signaling; The determining, based on the to-be-matched traffic information and the attribute information of the network slice, the preset resources allocated to the network slice includes: Extracting the resource demand contained in the traffic information to be matched; and obtaining the network priority and network deployment location corresponding to the network slice; Using an optimization algorithm, a resource allocation decision is made based on the resource demand, the network priority, and the network deployment location to determine the preset resources allocated to the network slice.

2. The method according to claim 1, characterized in that Before using the local target resource prediction model to determine the traffic information to be matched in the future time period, the method further includes: Obtaining an initial resource prediction model sent by the macro base station; Using local data to train the initial resource prediction model to obtain initial model parameters; Sending the initial model parameters to the macro base station, so that the macro base station performs federation aggregation on the received initial model parameters to obtain global model parameters; The global model parameters sent by the macro base station are received, and the target resource prediction model is obtained based on the global model parameters.

3. The method according to claim 2, characterized in that The step of obtaining the target resource prediction model based on the global model parameters includes: Using the global model parameters, performing a preset model accuracy test on the initial resource prediction model; If it is determined that the initial resource prediction model meets the preset model accuracy, the initial resource prediction model is used as the target resource prediction model; or, If it is determined that the initial resource prediction model does not meet the preset model accuracy, the initial resource prediction model is trained using the global model parameters to obtain the target resource prediction model.

4. The method according to claim 1, characterized in that After determining the to-be-matched traffic information consumed in the future time period, the method further includes: If it is determined that the network resources provided by the network slice can cover the traffic information to be matched, the resource configuration of the network slice is completed in a preset manner.

5. A network resource configuration method based on federated learning, characterized in that: Applied to macro base stations, including: receiving a resource acquisition request carrying traffic information to be matched sent by a micro base station, where the resource acquisition request is used to allocate preset resources for a network slice associated with the micro base station; the micro base station determines the traffic information to be matched consumed in a future time period by using a local target resource prediction model, and if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, the micro base station sends the resource acquisition request to the macro base station associated with itself; Determine the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice; Sending a slice resource configuration signaling for allocating the preset resource to the network slice to the micro base station, so that the micro base station completes the resource configuration of the network slice based on the slice resource configuration signaling; The determining, based on the to-be-matched traffic information and the attribute information of the network slice, the preset resources allocated to the network slice includes: Extracting the resource demand contained in the traffic information to be matched; and obtaining the network priority and network deployment location corresponding to the network slice; Using an optimization algorithm, a resource allocation decision is made based on the resource demand, the network priority, and the network deployment location to determine the preset resources allocated to the network slice.

6. The method according to claim 5, characterized in that Before receiving the resource acquisition request carrying the traffic information to be matched sent by the micro base station, the method further includes: Sending an initial resource prediction model to the plurality of micro base stations, and receiving initial model parameters sent by the plurality of micro base stations, wherein the initial model parameters are obtained by the micro base station training the initial resource prediction model using local data; Federate and aggregate the received initial model parameters to obtain global model parameters; The global model parameters are sent to the micro base station.

7. A network resource configuration device based on federated learning, characterized in that: Applied to micro base stations, including: A determination module is configured to determine the traffic information to be matched consumed in a future time period by using a local target resource prediction model, wherein the target resource prediction model is a model obtained by collaborative training of the micro base station under a federated learning framework; The sending module is configured to send a resource acquisition request to a macro base station associated with itself if it is determined that the network resources provided by the network slice associated with itself cannot cover the traffic information to be matched, so that the macro base station receives the resource acquisition request, and determines the preset resources allocated to the network slice based on the traffic information to be matched and the attribute information of the network slice; and sends a slice resource configuration signaling for allocating the preset resources to the network slice to the micro base station; the resource acquisition request carries the traffic information to be matched; A configuration module is configured to receive a slice resource configuration signaling allocated by the macro base station, and complete resource configuration of the network slice based on the slice resource configuration signaling; The determining, based on the to-be-matched traffic information and the attribute information of the network slice, the preset resources allocated to the network slice includes: Extracting the resource demand contained in the traffic information to be matched; and obtaining the network priority and network deployment location corresponding to the network slice; Using an optimization algorithm, a resource allocation decision is made based on the resource demand, the network priority, and the network deployment location to determine the preset resources allocated to the network slice.

8. An electronic device, characterized in that: include: A memory for storing executable instructions; as well as, A processor is used to execute the executable instructions with the memory to complete the operation of the network resource configuration method based on federated learning in any one of claims 1-6.

9. A computer-readable storage medium for storing computer-readable instructions, characterized in that: When the instruction is executed, the operation of the network resource configuration method based on federated learning in any one of claims 1-6 is performed.

Images