A network optimization method, device, equipment, storage medium and computer program product

Abstract

The application discloses a network optimization method, and the method is applied to a first node. The method comprises the following steps: receiving at least one corrected twin atomic model sent by a second node; wherein the model precision of the twin atomic model is greater than or equal to a first precision threshold value, and the twin atomic model corresponds to a communication network application; performing combination processing on the at least one twin atomic model to obtain a first twin network; and performing model optimization processing on the first twin network to obtain a second twin network, wherein the network precision of the second twin network is greater than or equal to a second precision threshold value, and the stability of the second twin network is greater than or equal to a stability threshold value. The application also discloses a network optimization device, a system, a storage medium and a computer program product.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a network optimization method, apparatus, device, storage medium, and computer program product. Background Technology

[0002] With the development of communication technology, how to pre-verify network strategies in communication systems has become a popular research topic. Currently, a digital twin network technology has been proposed, which constructs twin models based on multiple physical communication entities and orchestrates them to obtain a corresponding simulation environment. This environment enables real-time interaction and mapping with the physical entities. In this way, pre-verification operations of network strategies can be performed within the constructed communication simulation environment, i.e., the digital twin network, to determine the feasibility and reliability of the network strategies. Therefore, ensuring the accuracy of the constructed digital twin network is a pressing technical problem that needs to be solved.

[0003] Application content

[0004] To address the aforementioned technical problems, this application aims to provide a network optimization method, apparatus, device, storage medium, and computer program product. This addresses the issue of poor accuracy in current digital twin network models by proposing a method that optimizes the accuracy of individual digital twin network models, followed by optimizing the accuracy of a digital twin network model obtained by arranging multiple individual models. This improves the accuracy of the digital twin network model and ensures its reliability.

[0005] The technical solution of this application is implemented as follows:

[0006] This application provides a network optimization method, which is applied to a first node, and the method includes:

[0007] Receive a corrected twin atomic model sent by at least one second node; wherein the model accuracy of the twin atomic model is greater than or equal to a first accuracy threshold, and the twin atomic model corresponds to a communication network application;

[0008] At least one of the twin atom models is combined to obtain a first twin network;

[0009] The first twin network is optimized to obtain a second twin network with a network accuracy greater than or equal to a second accuracy threshold and a stability greater than or equal to a stability threshold.

[0010] In the above scheme, the step of performing model optimization on the first twin network to obtain a second twin network with a network accuracy greater than or equal to a second accuracy threshold and a stability greater than or equal to a stability threshold includes:

[0011] Predicting the first set of predicted state parameters of at least one of the twin atom models included in the first twin network at a first time point;

[0012] At the first moment, the first actual state parameter set of at least one of the twin atomic models included in the first twin network is obtained from at least one of the second nodes;

[0013] Based on the first predicted state parameter set and the first actual state parameter set, the first twin network model is optimized to obtain the second twin network.

[0014] In the above scheme, the step of optimizing the first twin network model based on the first predicted state parameter set and the first actual state parameter set to obtain the second twin network includes:

[0015] The network accuracy of the first twin network is calculated based on the first predicted state parameter set and the first actual state parameter set.

[0016] If the network precision is greater than or equal to the second precision threshold, the first twin network is simulated to obtain the simulation results;

[0017] Based on the simulation results, the probability that the network precision of the first twin network is greater than or equal to the second precision threshold is calculated.

[0018] If the first probability is greater than or equal to the stability threshold, the second twin network is determined to be the first twin network.

[0019] In the above scheme, calculating the network accuracy of the first twin network based on the first predicted state parameter set and the first actual state parameter set includes:

[0020] Based on the first predicted state parameter set and the first actual state parameter set, static evaluation parameters, dynamic evaluation parameters and related evaluation parameters are determined.

[0021] The network accuracy of the first twin network is obtained based on the static evaluation parameters, the dynamic evaluation parameters, and the correlation evaluation parameters.

[0022] In the above scheme, determining the static evaluation parameters, dynamic evaluation parameters, and correlation evaluation parameters based on the first predicted state parameter set and the first actual state parameter set includes:

[0023] The first set of predicted state parameters is grouped to obtain a first set of static parameters and a first set of dynamic parameters.

[0024] The first set of actual state parameters is grouped to obtain a second set of static parameters and a second set of dynamic parameters.

[0025] The static evaluation parameters are obtained by performing similarity calculations on the first static parameter set and the second static parameter set.

[0026] The dynamic evaluation parameters are determined based on the first set of dynamic parameters and the second set of dynamic parameters;

[0027] Based on the first set of predicted state parameters, determine the parameter correlation relationship of at least one of the twin atomic models to obtain the first correlation relationship parameter;

[0028] Based on the first set of actual state parameters, determine the parameter association relationship between at least one physical entity node corresponding to the twin atomic model, and obtain the second association relationship parameter;

[0029] The association evaluation parameters are determined based on the first association parameter and the second association parameter.

[0030] In the above scheme, determining the dynamic evaluation parameters based on the first dynamic parameter set and the second dynamic parameter set includes:

[0031] Based on the first set of dynamic parameters, the changing trend of each dynamic parameter is determined to obtain the first set of dynamic changing parameters;

[0032] Based on the second set of dynamic parameters, the changing trend of each dynamic parameter is determined to obtain the second set of dynamic changing parameters;

[0033] Consistency calculations are performed on the first set of dynamic parameters, the second set of dynamic parameters, the first set of dynamically changing parameters, and the second set of dynamically changing parameters to determine the dynamic evaluation parameters.

[0034] In the above scheme, the step of performing consistency calculations on the first dynamic parameter set, the second dynamic parameter set, the first dynamic change parameter set, and the second dynamic change parameter set to determine the dynamic evaluation parameters includes:

[0035] A consistency calculation is performed on the first dynamic parameter set and the second dynamic parameter set to obtain the first consistency parameter;

[0036] A consistency calculation is performed on the first set of dynamically changing parameters and the second set of dynamically changing parameters to obtain a second consistency parameter;

[0037] Calculate the first product of the first weighting coefficient and the first consistency parameter;

[0038] Calculate the second product of the second weighting coefficient and the second consistency parameter;

[0039] The sum of the first product and the second product is calculated to obtain the dynamic evaluation parameters.

[0040] In the above scheme, obtaining the network accuracy of the first Siamese network based on the static evaluation parameters, the dynamic evaluation parameters, and the correlation evaluation parameters includes:

[0041] Determine the static weighting coefficients, dynamic weighting coefficients, and correlation evaluation weighting coefficients;

[0042] Calculate the product of the static weight coefficient and the static evaluation parameter to obtain the weighted static evaluation parameter;

[0043] Calculate the product of the dynamic weight coefficient and the dynamic evaluation parameter to obtain the weighted dynamic evaluation parameter;

[0044] Calculate the product of the correlation evaluation weight coefficient and the correlation evaluation parameter to obtain the weighted correlation evaluation parameter;

[0045] The cumulative sum of the weighted static evaluation parameters, the weighted dynamic evaluation parameters, and the weighted correlation evaluation parameters is calculated to obtain the comprehensive parameters;

[0046] Calculate the first value of the static evaluation parameter;

[0047] Calculate the second value of the dynamic evaluation parameter;

[0048] Calculate the third value of the associated evaluation parameter;

[0049] Calculate the first value, the second value, and the third value to obtain a comprehensive value;

[0050] Calculate a fourth value representing the ratio of the comprehensive parameter to the comprehensive value;

[0051] The network accuracy is obtained by calculating the ratio of the fourth value to the square root of the number of parameters included in the comprehensive parameters.

[0052] The method in the above scheme further includes:

[0053] If the network precision is less than the second precision threshold, or the first probability is less than the stability threshold, a first precision threshold is determined for each of the twin atom models based on the overall precision threshold of the twin network and the probability threshold of the twin network precision reaching the threshold.

[0054] The first precision threshold of the twin atomic model corresponding to each second node is sent to the corresponding second node, so that the second node corrects the corresponding twin atomic model based on the first precision threshold.

[0055] In the above scheme, if the network accuracy is less than the second accuracy threshold, or the first probability is less than the stability threshold, determining the first accuracy threshold for each of the twin atomic models based on the overall accuracy threshold of the twin network and the probability threshold of the twin network accuracy includes:

[0056] If the network accuracy is less than the second accuracy threshold, or the first probability is less than the stability threshold, the second probability is calculated that the model accuracy obtained from the simulation calculation of the first twin network is greater than or equal to the overall accuracy threshold of the twin network within a preset time period.

[0057] If the second probability is greater than or equal to the probability threshold reached by the Siamese network accuracy, the first accuracy threshold of each Siamese atomic model is determined as the current accuracy threshold of the corresponding Siamese atomic model.

[0058] If the second probability is less than the probability threshold for the accuracy of the twin network, a first accuracy threshold for each twin atom model is determined based on the overall accuracy threshold of the twin network, the probability threshold for the accuracy of the twin network, and the first twin network.

[0059] In the above scheme, if the second probability is less than the probability threshold for the Siamese network accuracy, determining the first accuracy threshold for each Siamese atom model based on the overall accuracy threshold of the Siamese network, the probability threshold for the Siamese network accuracy, and the first Siamese network includes:

[0060] If the second probability is less than the probability threshold of the twin network accuracy, calculate the conditional probability that the model accuracy of the first twin network is greater than or equal to the overall accuracy threshold of the twin network when the model accuracy is greater than or equal to the current accuracy threshold of each twin atom model.

[0061] Calculate the prior probability when the accuracy of the model is greater than or equal to the current accuracy threshold of each twin atom model;

[0062] Based on the conditional probability and the prior probability, determine the posterior probability when the model accuracy is greater than or equal to the current accuracy threshold;

[0063] The current accuracy threshold of each twin atom model is optimized to obtain the optimized accuracy threshold of each twin atom model;

[0064] After updating the current precision threshold of each twin atomic model to the corresponding optimized precision threshold, the step of calculating the conditional probability that the model precision of the first twin network is greater than or equal to the overall precision threshold of the twin network when the model precision is greater than or equal to the current precision threshold of each twin atomic model is repeated, until the maximum posterior probability is determined.

[0065] The first precision threshold for each of the twin atomic models is determined to be the current precision threshold corresponding to the maximum posterior probability.

[0066] This application provides a network optimization method, which is applied to a second node, and the method includes:

[0067] Obtain the first precision threshold of the atomic model to be corrected;

[0068] Determine the model accuracy of the atomic model to be corrected; wherein the atomic model to be corrected corresponds to the communication network application;

[0069] Based on the model accuracy and the first accuracy threshold, the atomic model to be corrected is modified to obtain a twin atomic model; wherein the model accuracy of the twin atomic model is greater than or equal to the first accuracy threshold;

[0070] The twin atomic model is sent to the first node so that the first node can build a corresponding first twin network based on the twin atomic model.

[0071] In the above scheme, obtaining the first precision threshold of the atomic model to be corrected includes:

[0072] The first precision threshold is determined to be the initial precision threshold.

[0073] In the above scheme, obtaining the first precision threshold of the atomic model to be corrected includes:

[0074] Receive the first precision threshold sent by the first node; wherein, the first precision threshold is the precision threshold corresponding to the twin atom model;

[0075] Correspondingly, after obtaining the first precision threshold of the atomic model to be corrected, the method further includes:

[0076] After updating the atomic model to be corrected to the twin atomic model, the steps to determine the model accuracy of the atomic model to be corrected are repeated until the first accuracy threshold sent by the first node is no longer received.

[0077] In the above scheme, determining the model accuracy of the atomic model to be corrected includes:

[0078] Predict the second set of predicted state parameters for the atomic model to be corrected at the second time point;

[0079] At the second moment, the second set of actual state parameters of the atomic model to be corrected is obtained;

[0080] The model accuracy of the atomic model to be corrected is calculated based on the second predicted state parameter set and the second actual state parameter set.

[0081] In the above scheme, the step of correcting the atomic model to be corrected based on the model accuracy and the first accuracy threshold to obtain a twin atomic model includes:

[0082] If the model accuracy is greater than or equal to the first accuracy threshold, the twin atom model is determined to be the atom model to be corrected;

[0083] If the model accuracy is less than the first accuracy threshold, calculate the difference between the first accuracy threshold and the model accuracy;

[0084] Calculate the third product of the difference and the regularization coefficient;

[0085] The sum of the loss value of the atomic model to be corrected and the third product is calculated to obtain the adjustment parameters;

[0086] Based on the adjustment parameters, the adjustable parameters in the atomic model to be corrected are adjusted in reverse to obtain an intermediate atomic model;

[0087] The atomic model to be corrected is updated to the intermediate atomic model, and the steps are repeated to determine the model accuracy of the atomic model to be corrected until the twin atomic model is obtained.

[0088] The method in the above scheme further includes:

[0089] Receive terminal communication indicator data sent by the terminal device;

[0090] Receive a twin task request sent by the first node; wherein, the twin task request includes at least: model type, the overall accuracy threshold of the twin network, and the probability threshold of the twin network accuracy reaching;

[0091] Based on the terminal communication index data and the model type, at least one atomic model to be corrected is generated.

[0092] This application provides a first network optimization device, which is applied to a first node. The device includes: a receiving unit, a first processing unit, and a second processing unit; wherein:

[0093] The receiving unit is configured to receive a corrected twin atomic model sent by at least one second node; wherein the model accuracy of the twin atomic model is greater than or equal to a first accuracy threshold, and the twin atomic model corresponds to a communication network application;

[0094] The first processing unit is used to perform combined processing on at least one of the twin atom models to obtain a first twin network;

[0095] The second processing unit is used to perform model optimization processing on the first twin network to obtain a second twin network with network accuracy greater than or equal to a second accuracy threshold and stability greater than or equal to a stability threshold.

[0096] This application provides a second network optimization device, which is applied to a second node. The device includes: an acquisition unit, a determination unit, a correction unit, and a transmission unit; wherein:

[0097] The acquisition unit is used to acquire the first precision threshold of the atomic model to be corrected;

[0098] The determining unit is used to determine the model accuracy of the atomic model to be corrected; wherein the atomic model to be corrected corresponds to a communication network application;

[0099] The correction unit is used to correct the atomic model to be corrected based on the model accuracy and the first accuracy threshold to obtain a twin atomic model.

[0100] The sending unit is used to send the twin atomic model to the first node.

[0101] This application provides a network optimization system, the system comprising at least: a first node and at least one second node; wherein:

[0102] The first node is used to implement the steps of the network optimization method as described in any of the above items;

[0103] The second node is used to implement the steps of the network optimization method as described in any of the above.

[0104] This application provides a storage medium storing a resource conflict indicator program, which, when executed, is used to implement the steps of the network optimization method as described in any of the preceding claims.

[0105] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the network optimization method as described in any of the preceding claims.

[0106] This application provides a network optimization method, apparatus, device, storage medium, and computer program product. After obtaining a first precision threshold of the atomic model to be corrected through a second node, the model precision of the atomic model to be corrected is determined. Based on the model precision and the first precision threshold, the atomic model to be corrected is modified to obtain a twin atomic model. The twin atomic model is then sent to a first node. The first node receives at least one corrected twin atomic model sent by a second node, performs combination processing on at least one twin atomic model to obtain a first twin network, and performs model optimization processing on the first twin network to obtain a second twin network with a network precision greater than or equal to the second precision threshold and a stability greater than or equal to the stability threshold. In this way, the second node corrects the atomic model to be corrected according to the first precision threshold, and sends the corrected twin atomic model to the first node. The first node then combines and processes the corrected twin atomic models sent by at least one second node according to network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision in current digital twin network models. This proposes a method to optimize the precision of a single model in a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability. Attached Figure Description

[0107] Figure 1 A flowchart illustrating a network optimization method provided in this application embodiment. Figure 1 ;

[0108] Figure 2 A flowchart illustrating a network optimization method provided in this application embodiment. Figure 2 ;

[0109] Figure 3 A flowchart illustrating a network optimization method provided in this application embodiment. Figure 3 ;

[0110] Figure 4 A flowchart illustrating a network optimization method provided in this application embodiment. Figure 4 ;

[0111] Figure 5 This is a schematic diagram illustrating an application embodiment provided in this application.

[0112] Figure 6 This is a schematic diagram of the structure of a first network optimization device provided in an embodiment of this application;

[0113] Figure 7 This is a schematic diagram of the structure of a second network optimization device provided in an embodiment of this application;

[0114] Figure 8 This is a schematic diagram of the structure of a network optimization system provided in an embodiment of this application. Detailed Implementation

[0115] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0116] Embodiments of this application provide a network optimization method, referring to... Figure 1 As shown, the method is applied to the first node, and the method includes the following steps:

[0117] Step 101: Receive the corrected twin atomic model sent by at least one second node.

[0118] Among them, the model accuracy of the twin atom model is greater than or equal to the first accuracy threshold, and the twin atom model corresponds to the communication network application.

[0119] In this embodiment, the first node is a management node with service management and orchestration functions, such as a server node or a computing node. The second node is a network service node with communication network services, such as a base station (g-NB) node. The twin-atom model corresponds to each communication network application in the communication network. The relationship between at least one second node is that at least one second node can provide twin-atom modules related to the first twin network. The first accuracy threshold is an empirical value of model accuracy obtained from a large number of experiments, or it can be an empirical value of model accuracy set according to the actual requirements of the twin-atom model.

[0120] Step 102: Combine at least one twin atom model to obtain the first twin network.

[0121] In this embodiment of the application, the first node performs model combination and arrangement processing on the received at least one twin atomic model to achieve model fusion, thereby obtaining the first twin network.

[0122] Step 103: Perform model optimization on the first twin network to obtain a second twin network with a network accuracy greater than or equal to the second accuracy threshold.

[0123] In this embodiment, the first node performs model optimization on the first twin network obtained from the combined processing according to model optimization requirements, resulting in a second twin network with a network accuracy greater than or equal to a second accuracy threshold. That is, the second twin network is a twin network obtained by model optimization of the first twin network. The second accuracy threshold is an empirical value for network accuracy requirements obtained from numerous experiments, or it can be an empirical value for network accuracy requirements set according to actual application needs.

[0124] The network optimization method provided in this application involves a first node receiving a twin atomic model sent by at least one second node, which is corrected according to a first precision threshold dynamically determined by the second node. The first node then performs a combination processing on the at least one twin atomic model to obtain a first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to a second precision threshold and a stability greater than or equal to a stability threshold. In this way, the second node dynamically determines the first precision threshold of the atomic model to be corrected based on the overall precision threshold and the probability threshold of the twin network corresponding to the atomic model to be corrected. Then, the atomic model to be corrected is modified according to the first precision threshold, and the corrected twin atomic model is sent to the first node. The first node then processes the corrected twin atomic models sent by at least one second node according to the network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision of current digital twin network models. This proposes a method to optimize the precision of a single model of a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability.

[0125] Based on the foregoing embodiments, embodiments of this application provide a network optimization method, referring to... Figure 2 As shown, the method is applied to the second node, and the method includes the following steps:

[0126] Step 201: Obtain the first precision threshold of the atomic model to be corrected.

[0127] In the embodiments of this application, the first precision threshold of the atomic model to be corrected can be preset or determined according to the actual application requirements.

[0128] Step 202: Determine the model accuracy of the atomic model to be corrected.

[0129] Among them, the atomic model to be corrected corresponds to the application of communication networks.

[0130] In this embodiment, the second node determines the atomic models to be corrected that it includes, and determines the model accuracy of the atomic models to be corrected based on the relevant characteristics of the second node. Model accuracy is used to represent the reliability of the results of the model to be corrected in the simulation application process.

[0131] Step 203: Based on the model accuracy and the first accuracy threshold, perform model correction on the atomic model to be corrected to obtain the twin atomic model.

[0132] Among them, the model accuracy of the twin atom model is greater than or equal to the first accuracy threshold.

[0133] In this embodiment of the application, the model accuracy of the atomic model to be modified is compared and analyzed with a first accuracy threshold to obtain the analysis results. Based on the obtained analysis results, the atomic model to be modified is modified until a twin atomic model with a model accuracy greater than or equal to the first accuracy threshold is obtained.

[0134] Step 204: Send the twin atom model to the first node.

[0135] Specifically, a twin atomic model is sent to the first node so that the first node can build the corresponding first twin network based on the twin atomic model.

[0136] In this embodiment of the application, the second node sends the corrected twin atomic model to the first node through a communication method agreed upon with the first node, so that the first node can perform subsequent related analysis operations based on the received twin atomic model.

[0137] The network optimization method provided in this application involves obtaining a first precision threshold of the atomic model to be corrected through a second node, determining the model precision of the atomic model to be corrected, performing model correction on the atomic model to be corrected based on the model precision and the first precision threshold, obtaining a twin atomic model, and then sending the twin atomic model to the first node so that the first node receives at least one corrected twin atomic model sent by the second node, performing combination processing on at least one twin atomic model to obtain a first twin network, and performing model optimization processing on the first twin network to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, the second node corrects the atomic model to be corrected according to the first precision threshold, and sends the corrected twin atomic model to the first node. The first node then combines and processes the corrected twin atomic models sent by at least one second node according to network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision in current digital twin network models. This proposes a method to optimize the precision of a single model in a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability.

[0138] Based on the foregoing embodiments, embodiments of this application provide a network optimization method, referring to... Figure 3 As shown, the method includes the following steps:

[0139] Step 301: The second node determines the first precision threshold of the atomic model to be corrected.

[0140] Step 302: The second node determines the model accuracy of the atomic model to be corrected.

[0141] Among them, the atomic model to be corrected corresponds to the application of communication networks.

[0142] In this embodiment of the application, the second node is taken as a base station for illustration. The base station performs accuracy measurement on the atomic model to be corrected that needs model correction, and obtains the model accuracy of the atomic model to be corrected.

[0143] Step 303: The second node performs model correction on the atomic model to be corrected based on the model accuracy and the first accuracy threshold, to obtain the twin atomic model.

[0144] In this embodiment, the base station uses a model correction method to correct the atomic model to be corrected based on the relationship between the model accuracy of the atomic model to be corrected and the first accuracy threshold, thereby obtaining a twin atomic model.

[0145] Step 304: The second node sends the twin atom model to the first node.

[0146] Specifically, a twin atomic model is sent to the first node so that the first node can build the corresponding first twin network based on the twin atomic model.

[0147] Step 305: The first node receives a modified twin atomic model sent by at least one second node.

[0148] Among them, the model accuracy of the twin atom model is greater than or equal to the first accuracy threshold, and the twin atom model corresponds to the communication network application.

[0149] Step 306: The first node performs combination processing on at least one twin atom model to obtain the first twin network.

[0150] In this embodiment of the application, the first node is a service management server node with management and orchestration functions. The service management server node receives a modified twin atomic model sent by at least one second node to obtain at least one twin atomic model. Then, it performs model orchestration and combination processing on the at least one twin atomic model to obtain the first twin network.

[0151] In some application scenarios, the first node can first perform similar model aggregation on at least one twin atomic model to obtain a new model with strong generalizability belonging to the same class. Then, the new model obtained after aggregating these similar models is arranged and combined to obtain the first twin network.

[0152] Step 307: The first node performs model optimization on the first twin network to obtain a second twin network with network accuracy greater than or equal to the second accuracy threshold and stability greater than or equal to the stability threshold.

[0153] In this embodiment, the service management server node performs model optimization on the first twin network obtained by the combined processing until a second twin network is obtained with network accuracy greater than or equal to a second accuracy threshold and stability greater than or equal to a stability threshold.

[0154] Based on the foregoing embodiments, in other embodiments of this application, step 301 can be implemented by step 301a or step 301b:

[0155] Step 301a: The second node determines the first precision threshold as the initial precision threshold.

[0156] In this embodiment, when the second node first obtains the atomic model to be corrected, it determines the first precision threshold of the atomic model to be corrected as the initialization precision threshold. The initialization precision threshold is an empirical value obtained from a large number of experiments, and in some application scenarios, it can be set according to the actual situation.

[0157] Step 301b: The second node receives the first precision threshold sent by the first node.

[0158] The first precision threshold is the precision threshold corresponding to the twin atom model.

[0159] In this embodiment, the first precision threshold sent by the first node can be dynamically set and allocated by the first node based on the twin network including the atomic model to be corrected. In this way, the precision requirements of the twin atomic model can be guaranteed, thereby guaranteeing the precision requirements of the corresponding twin network.

[0160] Correspondingly, after the second node executes step 301b, it is also used to execute the following steps: after updating the atomic model to be corrected to a twin atomic model, it executes the step "determine the model accuracy of the atomic model to be corrected" until the first accuracy threshold sent by the first node is not received.

[0161] In this embodiment of the application, when the second twin network obtained after the first twin network optimization does not meet the requirements, the model accuracy of at least one twin atom model is readjusted according to the model accuracy requirements of the twin network model, so that the model accuracy of each twin atom model is dynamically changed in different analysis processes. By controlling the model accuracy of each twin atom model, the network accuracy requirements of the corresponding twin network can be quickly guaranteed.

[0162] Based on the foregoing embodiments, in other embodiments of this application, step 302 can be implemented by steps 302a to 302c:

[0163] Step 302a: The second node predicts the set of second predicted state parameters of the atomic model to be corrected at the second time point.

[0164] In this embodiment, the second node performs prediction processing based on the atomic model to be corrected, determines the relevant state parameters of the atomic model to be corrected at a future second time, and obtains a second set of predicted state parameters.

[0165] The second time point could be a time interval corresponding to the current time, such as half an hour, an hour, or a day from the current time. The specific time interval can be set according to the actual situation and needs, and no specific limitation is made here.

[0166] Step 302b: At the second time point, the second node obtains the second set of actual state parameters of the atomic model to be corrected.

[0167] In this embodiment of the application, when the second time point arrives, the second node obtains the actual state parameters of the model corresponding to the atomic model to be corrected during actual operation, and obtains the second set of actual state parameters.

[0168] When the second moment arrives, the base station obtains the actual state parameters of the communication network application corresponding to the atomic model to be corrected during actual operation, thereby obtaining the second set of actual state parameters of the atomic model to be corrected.

[0169] Step 302c: The second node calculates the model accuracy of the atomic model to be corrected based on the second predicted state parameter set and the second actual state parameter set.

[0170] In this embodiment, the second node performs an accuracy analysis of the atomic model to be corrected based on the second preset state parameter set and the second actual state parameter set of the atomic model to be corrected, and obtains the model accuracy of the atomic model to be corrected.

[0171] It should be noted that the specific implementation process of step 302c can refer to the subsequent process of determining the network accuracy of the first twin network based on the first preset state parameter set and the first actual state parameter set, which will not be elaborated here.

[0172] Based on the foregoing embodiments, in other embodiments of this application, step 303 can be implemented by step 303a, or steps 303b to 303f:

[0173] Step 303a: If the model accuracy is greater than or equal to the first accuracy threshold, the second node determines the twin atomic model as the atomic model to be corrected.

[0174] In the embodiments of this application, when the model accuracy of the atomic model to be corrected is greater than or equal to the first accuracy threshold, it indicates that the atomic model to be corrected meets the current requirements. Therefore, it can be determined that the twin atomic model is the atomic model to be corrected itself.

[0175] Step 303b: If the model accuracy is less than the first accuracy threshold, the second node calculates the difference between the first accuracy threshold and the model accuracy.

[0176] In this embodiment of the application, when the model accuracy of the atomic model to be corrected is less than the first accuracy threshold, it indicates that the atomic model to be corrected does not meet the requirements and needs to be corrected. At this time, the second node calculates the first accuracy threshold and the model accuracy, for example, by calculating the difference between the first accuracy threshold and the model accuracy.

[0177] Step 303c: The second node calculates the third product of the difference and the regularization coefficient.

[0178] In this embodiment, the regularization coefficient is an empirical value obtained from a large number of experiments, or an empirical value set according to actual needs. The third product of the difference and the regularization coefficient can be calculated using the following formula: Third product = Difference * Regularization coefficient.

[0179] Step 303d: The second node calculates the sum of the loss value of the atomic model to be corrected and the third product to obtain the adjustment parameters.

[0180] In this embodiment, after the second node calculates the loss value of the atomic model to be corrected using the loss function corresponding to the atomic model to be corrected, it continues to calculate the sum of the loss value of the atomic model to be corrected and the third product, thereby obtaining the adjustment parameters for model correction of the atomic model to be corrected.

[0181] Step 303e: The second node adjusts the adjustable parameters in the atomic model to be corrected in reverse based on the adjustment parameters to obtain the intermediate atomic model.

[0182] In this embodiment, the second node uses the calculated adjustment parameters to reverse the adjustable parameters in the atomic model to be corrected, thereby obtaining an intermediate atomic model.

[0183] Step 303f: Update the atomic model to be corrected to an intermediate atomic model. Repeat the steps to determine the model accuracy of the atomic model to be corrected until a twin atomic model is obtained.

[0184] In this embodiment of the application, after updating the atomic model to be corrected to an intermediate atomic model, the process of steps 302 (including 302a to 302c) and 303 (including steps 303a to 303f) is repeated until a twin atomic model with a model accuracy greater than the first accuracy threshold is obtained.

[0185] Based on the foregoing embodiments, in other embodiments of this application, reference is made to... Figure 4 As shown, the second node is also used to execute steps 308 to 310:

[0186] Step 308: The second node receives terminal communication indicator data sent by the terminal device.

[0187] In this embodiment, the terminal device is a device that uses the communication network service provided by the second node. For example, it can be a smart mobile terminal device using the communication network service, such as a smartphone or a smart car. The terminal communication index data refers to parameters obtained by measuring the communication network service during the terminal device's use of the service, such as channel quality parameters.

[0188] Step 309: The second node receives the twin task request sent by the first node.

[0189] The twin task request includes at least: model type, overall accuracy threshold of the twin network, and probability threshold for achieving twin network accuracy.

[0190] In this embodiment of the application, the first node sends a twin task request to the second node to request the second node to provide the corresponding modified twin atomic model.

[0191] Step 310: The second node generates at least one atomic model to be corrected based on the terminal communication indicator data and model type.

[0192] In this embodiment, the second node creates at least one atomic model to be corrected that matches the twin task request, based on the terminal communication indicator data and data type reported by the terminal device. Further, when creating at least one atomic model to be corrected, the second node can also refer to its node feature data to perform the operation. At this point, this is the second node's first time obtaining the atomic model to be corrected. In some application scenarios, after executing step 310, the second node executes step 301a to obtain the twin atomic model, which is then sent to the first node. After the first node processes the received twin atomic model to obtain the first twin network, if a second twin network is not obtained, the first node dynamically determines a first precision threshold for the twin atomic model and sends it to the second node. The second node updates the corresponding atomic model to be corrected to the corresponding twin atomic model and updates its precision threshold to the first precision threshold. Then, based on the first precision threshold sent by the first node, the updated atomic model to be corrected is updated to obtain the twin atomic model. This process continues until the first node finally constructs a second twin network that meets the requirements based on the twin atomic model provided by the second node.

[0193] Based on the foregoing embodiments, in other embodiments of this application, step 307 can be implemented by steps 307a to 307c:

[0194] Step 307a: The first node predicts the first set of predicted state parameters of at least one twin atom model included in the first twin network at the first time point.

[0195] In this embodiment of the application, the first node predicts the first set of predicted state parameters constituting at least one twin atom model of the first twin network at a future first moment.

[0196] Step 307b: At the first moment, the first node obtains the first actual state parameter set of at least one twin atom model included in the first twin network from at least one second node.

[0197] In this embodiment of the application, when the first time is reached, the actual state parameters of at least one twin atom model included in each second node are obtained from each second node, thereby obtaining the first set of actual state parameters.

[0198] Step 307c: The first node performs model optimization processing on the first twin network model based on the first predicted state parameter set and the first actual state parameter set to obtain the second twin network.

[0199] In this embodiment, the first node performs calculation and analysis on the first set of predicted state parameters and the first set of actual state parameters to obtain the calculation and analysis results. Based on the calculation and analysis results, the first twin network model is optimized until a second twin network that meets the requirements is obtained.

[0200] Based on the foregoing embodiments, in other embodiments of this application, step 307c can be implemented by steps a11 to a15:

[0201] Step a11: The first node calculates the network accuracy of the first twin network based on the first predicted state parameter set and the first actual state parameter set.

[0202] In this embodiment, the first node calculates the network accuracy of the first twin network by using the first preset state parameter set and the first actual state parameter set.

[0203] After the first node executes step a11, it can choose to execute steps a12-a13 or steps a14-a15. If the network precision is greater than or equal to the second precision threshold, steps a12-a13 are executed; if the network precision is less than the second precision threshold, or the first probability is less than the stability threshold, step a15 is executed.

[0204] Step a12: If the network precision is greater than or equal to the second precision threshold, perform simulation calculations on the first twin network to obtain the simulation results.

[0205] Step a13: Based on the simulation results, calculate the first probability that the network precision of the first twin network is greater than or equal to the second precision threshold.

[0206] In this embodiment of the application, when the network precision of the first twin network is greater than or equal to the second precision threshold, the stability of the first twin network is statistically analyzed, that is, the probability value of the network precision being greater than or equal to the second precision threshold is statistically analyzed to obtain the first probability.

[0207] Step a14: If the first probability is greater than or equal to the stability threshold, determine the second twin network as the first twin network.

[0208] In this embodiment of the application, when the first probability is greater than or equal to the stability threshold, it indicates that the stability of the first twin network is high. At this time, the network precision of the first twin network is greater than or equal to the second precision threshold, and the first probability is greater than or equal to the stability threshold, indicating that the network precision and stability of the first twin network meet the requirements. At this time, the second twin network can be determined as the first twin network.

[0209] Step a15: If the network precision is less than the second precision threshold, or the first probability is less than the stability threshold, the first node determines the first precision threshold of each twin atom model based on the overall precision threshold of the twin network and the probability threshold of the twin network precision reaching the threshold.

[0210] In this embodiment, when the network accuracy is less than the second accuracy threshold, or the first accuracy threshold is less than the stability threshold, the first node determines the overall accuracy threshold and the probability threshold for achieving the accuracy of the twin network, and then determines the first accuracy threshold for each twin atomic model based on the overall accuracy threshold and the probability threshold for achieving the accuracy of the twin network. This achieves dynamic determination of the accuracy threshold of the twin atomic model.

[0211] Based on the foregoing embodiments, in other embodiments of this application, step a11 can be implemented by steps a111 to a112:

[0212] Step a111: The first node determines the static evaluation parameters, dynamic evaluation parameters, and associated evaluation parameters based on the first predicted state parameter set and the first actual state parameter set.

[0213] In this embodiment, the static evaluation parameters of the first twin network are obtained by analyzing and calculating static parameters representing the static attributes and indicators of the first twin network. These static parameters are inherent attributes of the first twin network and do not change with the network state. The dynamic evaluation parameters of the first twin network are obtained by analyzing dynamic parameters representing the dynamic attributes and indicators of the first twin network. These dynamic parameters change with the network state. The correlation evaluation parameters of the first twin network are obtained by analyzing and calculating parameters describing the explicit or implicit relationships between the parameters of the first twin network and the parameters of associated network nodes, i.e., correlation parameters. The first node classifies the parameters in the first predicted state parameter set and the first actual state parameter set to obtain the corresponding static parameters, dynamic parameters, and correlation parameters. It calculates the static parameters to obtain static evaluation parameters and the dynamic parameters to obtain dynamic evaluation parameters. The first node analyzes the first predicted state parameter set and the second actual state parameter set to determine the corresponding correlation parameters and calculates them to obtain correlation evaluation parameters.

[0214] Step a112: The first node obtains the network accuracy of the first twin network based on the static evaluation parameters, dynamic evaluation parameters, and correlation evaluation parameters.

[0215] In this embodiment of the application, the first node performs calculation and analysis on the calculated static evaluation parameters, dynamic evaluation parameters and correlation evaluation parameters using a preset calculation and analysis method to obtain the network accuracy of the first twin network.

[0216] Based on the foregoing embodiments, in other embodiments of this application, step a111 can be implemented by steps b11 to b17:

[0217] Step b11: The first node groups the first predicted state parameter set to obtain the first static parameter set and the first dynamic parameter set.

[0218] In this embodiment, the first node performs dynamic and static parameter grouping processing on the first predicted state parameter set to obtain a first static parameter set and a first dynamic parameter set.

[0219] Step b12: The first node groups the first actual state parameter set to obtain the second static parameter set and the second dynamic parameter set.

[0220] In this embodiment of the application, the first node performs dynamic parameter and static parameter grouping processing on the first actual state parameter set to obtain a second static parameter set and a second dynamic parameter set.

[0221] Step b13: The first node performs similarity calculation on the first static parameter set and the second static parameter set to obtain static evaluation parameters.

[0222] In this embodiment of the application, the first node uses a similarity calculation method to perform similarity calculation on the first static parameter set and the second static parameter set to obtain static evaluation parameters.

[0223] Step b14: The first node determines the dynamic evaluation parameters based on the first dynamic parameter set and the second dynamic parameter set.

[0224] In this embodiment of the application, there are two analytical indicators in the process of dynamic parameter change: parameters and parameter change. Therefore, the first dynamic parameter set and the second dynamic parameter set are calculated and analyzed from the perspective of these two analytical indicators to obtain dynamic evaluation parameters.

[0225] Step b15: Based on the first set of predicted state parameters, the first node determines the parameter association relationship of at least one twin atomic model and obtains the first association relationship parameter.

[0226] In this embodiment, the first node analyzes the first predicted state parameter set and obtains the first association parameter of the first twin network based on the parameter association relationship between at least one twin atomic model, such as the influence relationship between previous and subsequent operations, the influence relationship between business operations, etc.

[0227] Step b16: Based on the first set of actual state parameters, the first node determines the parameter association relationship between at least one physical entity node corresponding to a twin atomic model, and obtains the second association relationship parameter.

[0228] In this embodiment of the application, the first node analyzes the first set of actual state parameters to obtain a second set of association parameters that match the first association parameters.

[0229] Step b17: The first node determines the association evaluation parameters based on the first association parameter and the second association parameter.

[0230] In this embodiment of the application, the first node performs parameter calculation and analysis on the first association parameter and the second association parameter to obtain the association evaluation parameter.

[0231] Based on the foregoing embodiments, in other embodiments of this application, step b14 can be implemented by steps b141 to b143:

[0232] Step b141: The first node determines the changing trend of each dynamic parameter based on the first dynamic parameter set, thus obtaining the first dynamic changing parameter set.

[0233] In this embodiment of the application, the first node analyzes the parameters in the first dynamic parameter set to determine the trend of each dynamic parameter in the first dynamic parameter set changing over time, thereby obtaining the first dynamic parameter set.

[0234] Step b142: The first node determines the changing trend of each dynamic parameter based on the second dynamic parameter set, thus obtaining the second dynamic changing parameter set.

[0235] In this embodiment of the application, the first node analyzes the parameters in the second dynamic parameter set to determine the trend of each dynamic parameter in the second dynamic parameter set changing over time, thereby obtaining the second dynamic parameter set.

[0236] Step b143: The first node performs consistency calculations on the first dynamic parameter set, the second dynamic parameter set, the first dynamic change parameter set, and the second dynamic change parameter set to determine the dynamic evaluation parameters.

[0237] In this embodiment, the first node uses a consensus calculation method to calculate the first dynamic parameter set, the second dynamic parameter set, the first dynamic change parameter, and the second dynamic change parameter to determine the dynamic evaluation parameters.

[0238] Based on the foregoing embodiments, in other embodiments of this application, step b143 can be implemented by steps c11 to c15:

[0239] Step c11: The first node performs a consistency calculation on the first dynamic parameter set and the second dynamic parameter set to obtain the first consistency parameter.

[0240] Step c12: The first node performs a consistency calculation on the first set of dynamic change parameters and the second set of dynamic change parameters to obtain the second consistency parameter.

[0241] Step c13: The first node calculates the first product of the first weight coefficient and the first consistency parameter.

[0242] In the embodiments of this application, the first weighting coefficient is an empirical value obtained in advance based on a large number of experiments. In some application scenarios, it can also be set according to the actual situation.

[0243] Step c14: The first node calculates the second product of the second weight coefficient and the second consistency parameter.

[0244] In the embodiments of this application, the second weighting coefficient is an empirical value obtained in advance based on a large number of experiments. In some application scenarios, it can also be set according to the actual situation.

[0245] Step c15: The first node calculates the sum of the first product and the second product to obtain the dynamic evaluation parameters.

[0246] Based on the foregoing embodiments, in other embodiments of this application, step a112 can be implemented by steps d11 to d21:

[0247] Step d11: The first node determines the static weight coefficient, dynamic weight coefficient, and correlation evaluation weight coefficient.

[0248] In this embodiment, the static weighting coefficient, dynamic weighting coefficient, and correlation evaluation weighting coefficient are empirical values ​​obtained in advance based on a large number of experiments. In some application scenarios, they can also be set according to the actual situation.

[0249] Step d12: The first node calculates the product of the static weight coefficient and the static evaluation parameter to obtain the weighted static evaluation parameter.

[0250] Step d13: The first node calculates the product of the dynamic weight coefficient and the dynamic evaluation parameter to obtain the weighted dynamic evaluation parameter.

[0251] Step d14: The first node calculates the product of the correlation evaluation weight coefficient and the correlation evaluation parameter to obtain the weighted correlation evaluation parameter.

[0252] Step d15: The first node calculates the sum of the weighted static evaluation parameters, weighted dynamic evaluation parameters, and weighted correlation evaluation parameters to obtain the comprehensive parameters.

[0253] Step d16: The first node calculates the first value of the static evaluation parameter.

[0254] Step d17: The first node calculates the second value of the dynamic evaluation parameter.

[0255] Step d18: The first node calculates the third value of the associated evaluation parameter.

[0256] Step d19: The first node calculates the first, second, and third values ​​to obtain the combined value.

[0257] Step d20: The first node calculates the fourth value of the ratio of the comprehensive parameter to the comprehensive value.

[0258] Step d21: The first node calculates the ratio of the fourth value to the square root of the number of parameters included in the comprehensive parameters, and obtains the network accuracy.

[0259] Based on the foregoing embodiments, in other embodiments of this application, step a15 can be implemented by steps a151 to a153:

[0260] Step a151: Within a preset time period, the first node calculates the second probability that the model accuracy obtained from the simulation calculation of the first twin network is greater than or equal to the overall accuracy threshold of the twin network.

[0261] In this embodiment, the preset duration is an empirical duration obtained from a large number of experiments, or it can be an empirical duration set according to actual needs. Within the preset duration, the first node runs the first twin network, performs simulation calculations on the first twin network, and then calculates the second probability that the model accuracy obtained during the simulation calculation of the first twin network is greater than or equal to the overall accuracy threshold of the twin network within the preset duration.

[0262] After the first node executes step a151, it can choose to execute either step a152 or step a153.

[0263] Step a152: If the second probability is greater than or equal to the probability threshold reached by the Siamese network precision, the first node determines the first precision threshold of each Siamese atomic model as the current precision threshold of the corresponding Siamese atomic model.

[0264] Step a153: If the second probability is less than the probability threshold of the twin network accuracy, the first node determines the first accuracy threshold of each twin atom model based on the overall accuracy threshold of the twin network, the probability threshold of the twin network accuracy, and the first twin network.

[0265] In this embodiment, the first node determines the overall accuracy threshold and the probability threshold for achieving the accuracy of the twin network corresponding to the first twin network. The overall accuracy threshold and the probability threshold for achieving the accuracy of the twin network can be preset accuracy thresholds based on the requirements for the corresponding first twin network. The overall accuracy threshold represents the accuracy of the result when the corresponding twin network is applied, and the probability threshold for achieving the accuracy of the twin network is a certain percentage of empirical values ​​that the number of times the result of the corresponding twin network reaches the overall accuracy threshold within a certain number of iterations.

[0266] The first node adjusts and analyzes the accuracy threshold of each twin atom model based on the overall accuracy threshold and the probability threshold of the twin network accuracy corresponding to the first twin network. It calculates the first accuracy threshold of each twin atom model, so that the first accuracy threshold of the twin atom model matches the requirements of the corresponding twin network. This makes the final twin atom model more in line with the actual requirements, enabling the final first twin network to quickly meet the requirements, thereby reducing the consumption of computing resources.

[0267] Based on the foregoing embodiments, in other embodiments of this application, step a153 can be implemented by steps e11 to e16:

[0268] Step e11: If the second probability is less than the probability threshold when the accuracy of the twin network reaches the probability threshold, the first node calculates the conditional probability that the model accuracy of the first twin network is greater than or equal to the current accuracy threshold of each twin atom model.

[0269] In this embodiment of the application, when the second probability is less than the probability threshold of the twin network accuracy, the first node uses the conditional probability calculation method to calculate the condition that the model accuracy of the first twin network is greater than or equal to the current accuracy threshold of each twin atom model.

[0270] Step e12: The first node calculates the prior probability when the model accuracy is greater than or equal to the current accuracy threshold of each twin atom model.

[0271] In this embodiment, the first node uses a priori algorithm to calculate the prior probability that the model accuracy is greater than or equal to the current accuracy threshold of each twin atom model.

[0272] Step e13: The first node determines the posterior probability when the model accuracy is greater than or equal to the current accuracy threshold of each twin atomic model, based on the conditional probability and the prior probability.

[0273] In this embodiment, the second node uses a posterior probability algorithm to calculate the calculated conditional probability and prior probability to determine the posterior probability corresponding to the current accuracy threshold of each twin atom model when the model accuracy is greater than or equal to the current accuracy threshold.

[0274] Step e14: The first node optimizes the current accuracy threshold of each twin atom model to obtain the optimized accuracy threshold of each twin atom model.

[0275] In this embodiment, the first node can use methods such as gradient ascent or stochastic gradient ascent to adjust the current accuracy threshold and obtain an optimized accuracy threshold.

[0276] Step e15: After the first node updates the current precision threshold of each twin atom model to the corresponding optimized precision threshold, the step is repeated to calculate the conditional probability that the model precision of the first twin network is greater than or equal to the overall precision threshold of the twin network when the model precision is greater than or equal to the current precision threshold of each twin atom model, until the maximum posterior probability is determined.

[0277] In this embodiment of the application, after the first node updates the current precision threshold of each twin atomic model to the optimized precision threshold, it repeats steps a151 and e11 to e15, and iterates the current precision threshold of each twin atomic model repeatedly until the current precision threshold of each twin atomic model maximizes the posterior probability or the gradient approaches zero.

[0278] Step e16: The first node determines the first precision threshold of each twin atomic model as the current precision threshold corresponding to the maximum a posteriori probability.

[0279] In this embodiment, the current precision threshold corresponding to the maximum delay probability is used as the first precision threshold of each atomic model in the twin network. Thus, the precision threshold of each atomic model is dynamically set according to the task's precision requirements for the overall digital twin network, decomposing the overall digital twin network's precision requirements onto each atomic model, thereby ensuring the final overall digital twin network's precision requirements.

[0280] Based on the foregoing embodiments, this application provides a network optimization method for improving the overall accuracy of a digital twin network through hierarchical measurement of accuracy and closed-loop correction. The corresponding application scenario is a terminal device (User, Equipment, UE), a centralized twin task node with operation and maintenance management (OAM) functions, corresponding to the aforementioned first node, and a digital twin network application scenario between at least one base station g-NB. The implementation process of this method can be referred to... Figure 5 As shown, the specific steps include:

[0281] Step f101: The UE reports terminal communication indicator data to the g-NB.

[0282] Among them, terminal communication indicator data may include UE data such as Channel State Information (CSI), Load Update Interval (LUI), Measurement Report (MR), and UE capability.

[0283] Step f102: Construct a twin scenario by centralizing twin task nodes, and specify the digital twin network type (Date Models type) involved in the twin environment, the overall accuracy Min(P) and stability requirements Min(S) of the digital twin network.

[0284] Step f103: The centralized twin task node sends a task request, including DT Modelstype, Min(P) and Min(S), to the g-NB involved in the digital twin network.

[0285] In this context, the OAM in the centralized twin task node can send task requests to the g-NB involved in the task through the E2 interface configured by the OAM for communication with the g-NB.

[0286] In step f104, g-NB constructs a twin model based on the received task request, terminal communication index data, and g-NB base station related data, to obtain the atomic model to be corrected.

[0287] Among them, the atomic model to be corrected can be, for example, an application model corresponding to the communication application, such as a UE behavior model, a channel twin model, a network topology twin model, or a resource allocation model.

[0288] Step f105, g-NB, performs model correction on the atomic model to be corrected, and obtains the corrected twin atomic model.

[0289] Among them, g-NB is equipped with a precision measurement module for measuring the precision of each atomic model to be corrected, a threshold setting module for determining the model threshold of the atomic model to be corrected, and an intelligent verification model for model correction and rectification.

[0290] In some application scenarios, step f105 can be achieved by the following steps:

[0291] Step f1051: g-NB starts the precision measurement module, performs precision measurement on the atomic model to be corrected, and obtains the model precision c of the atomic model to be corrected.

[0292] In this embodiment, the accuracy value of the atomic model to be corrected can be calculated by comprehensively considering the accuracy measures of three indicators: static parameter accuracy (corresponding to the aforementioned static evaluation parameters), dynamic parameter accuracy (corresponding to the aforementioned dynamic evaluation parameters), and parameter correlation accuracy (corresponding to the aforementioned correlation evaluation parameters), thereby enabling a more accurate assessment of the model accuracy of the atomic model to be corrected.

[0293] Specifically, static parameters are various static properties and indices of the atomic model to be modified. Static parameters are related to the inherent properties of the atomic model and do not change with its state. Static parameters can be represented by defining a static parameter eigenvector, defined as sp. m sp represents the static parameters of the atomic model to be modified. r This represents the static parameters of the entity corresponding to the atomic model to be corrected. The accuracy of the static parameters can be measured by calculating the similarity between the eigenvectors of the two models. One calculation method is as follows: Where "·" represents the dot product between vectors, ‖‖ represents the second norm, and c s The closer it is to 1, the higher the accuracy.

[0294] Dynamic parameters are various dynamic attributes and indicators of the atomic model to be corrected. They are related to the state of the atomic model and can change with its state. They include both parameter values ​​and trends in parameter change. Dynamic parameters can be represented by defining a dynamic parameter eigenvector, defined as dp. m dp represents the dynamic parameters of the atomic model to be corrected. r This represents the dynamic parameters of the entity corresponding to the atomic model to be modified. The trend of parameter change can be characterized by the first and second time-domain derivatives of the dynamic parameters, representing the trend and future direction of the dynamic parameters, respectively, i.e., dpt is defined. m dpt represents the trend of dynamic parameter changes in the atomic model to be corrected. r This indicates the trend of dynamic parameter changes for the entity corresponding to the atomic model to be corrected. and These represent the state change trend and future direction of the nth dynamic parameter of the atomic model to be corrected, respectively. and These represent the state change trend and future direction of the nth dynamic parameter of the entity corresponding to the atomic model to be corrected. If the model to be corrected is a channel model, its dynamic parameters can be, for example, signal strength, signal-to-noise ratio, path loss, etc.; the first-order time-domain derivative of the path loss represents the change trend of the loss, and the second-order time-domain derivative represents the future direction of the loss.

[0295] Therefore, the accuracy of dynamic parameters can be represented by the consistency between the dynamic parameter feature vector of the atomic model to be corrected and the dynamic parameter feature vector of the corresponding entity of the atomic model to be corrected, as well as the consistency of the dynamic parameter change trend. One implementation algorithm can be denoted as: Among them, c d The closer the value is to 1, the higher the accuracy. a and b are weighting coefficients, which can be predetermined empirical values.

[0296] Parameter association refers to the explicit or implicit relationships between the parameters of the atomic model to be modified and the parameters of the associated models. Parameter association can be represented by defining an association adjacency vector, for example, it can be denoted as... This represents the parameter relationship between the parameters in the atomic model u to be modified and the parameters in the atomic model v to be modified. This indicates the parameter relationship between the parameters in the entity corresponding to the atomic model u to be corrected and the parameters in the entity corresponding to the atomic model v to be corrected.

[0297] Therefore, the accuracy of parameter association can be represented by the similarity between the association adjacency vector of the atomic model to be corrected and the association adjacency vector of the corresponding entity of the atomic model to be corrected. One calculation algorithm can be denoted as: Among them, c pl The closer it is to 1, the higher the accuracy.

[0298] Finally, the accuracy of the atomic model to be corrected can be measured by the consistency between the results of the atomic model to be corrected and the corresponding entities under the same configuration strategy. One calculation method can be denoted as: Among them, w s w d w pl The values ​​range from 0 to 1, including two endpoints: the weighting coefficients corresponding to the accuracy of static parameters, the accuracy of dynamic parameters, and the accuracy of parameter correlation. These can be predetermined empirical values. The closer the c-value is to 1, the stronger the consistency and the higher the accuracy of the atomic model to be corrected.

[0299] The atomic model to be corrected is the first one constructed; therefore, after g-NB executes f1051, it executes step f1052:

[0300] Step f1052: The g-NB precision measurement module assigns an initial model precision threshold to the atomic model to be corrected.

[0301] Step f1053: If the model accuracy of the atomic model to be corrected is greater than or equal to the initial model accuracy threshold, the twin atomic model is determined to be the atomic model to be corrected.

[0302] Step f1054: If the model accuracy of the atomic model to be corrected is less than the initial model accuracy threshold, g-NB performs model correction on the atomic model to be corrected until a twin atomic model is obtained.

[0303] The model correction process is as follows: The intelligent verification module starts the atomic model adaptive learning mechanism. For the atomic model i to be corrected, it determines |min(c-θ,0)|, that is, after calculating the difference between the model accuracy c of the atomic model to be corrected and the initial model accuracy threshold θ, it determines the minimum value between the difference between the model accuracy c of the atomic model to be corrected and the initial model accuracy threshold θ and 0. The absolute value of this minimum value is determined as the atomic model regularization factor and added to the loss function to correct the atomic model to be corrected. It can be written as: Lcorrected=Loriginal+λ(|min(c-θ,0)|), where λ is the regularization coefficient, which is used to control the influence of the regularization term on the atomic model to be corrected, thereby optimizing the model parameters or model structure until the accuracy reaches the threshold. Lcorrected represents the loss function after model correction, and Loriginal represents the loss function before model correction. In this way, the Lcorrected value can be calculated to optimize the model parameters or model structure of the atomic model to be corrected, and an intermediate atomic model can be obtained. Then, after updating the atomic model to be corrected to the intermediate atomic model, steps f1051 to f1054 are repeated until a twin atomic model with a model accuracy greater than or equal to the initial model accuracy threshold is obtained.

[0304] Step f106: g-NB sends the twin atomic model to the centralized twin task node.

[0305] Step f107: The centralized twin task node receives at least one twin atomic model sent by g-NB.

[0306] Step f108: The centralized twin task nodes are arranged and combined to form at least one twin atomic model to obtain the first twin network.

[0307] Among them, the OAM in the centralized twin task node can arrange at least one twin atomic model received by the UE behavior digital model, channel digital model, resource allocation digital model, etc. based on the topology model to construct an overall twin network and obtain the first twin network.

[0308] Step f109: The centralized twin task node starts its intelligent verification module to perform intelligent verification on the first twin network and obtain the predicted state of the first twin network.

[0309] Specifically, constructing a twin snapshot of the first twin network refers to, based on the first twin network, pre-deducing the predicted states of each twin atomic model at time t+Δt, thus obtaining the predicted state of the first twin network, where time t represents the current time. The pre-deduced predicted state of the first twin network includes the predicted states of each twin atomic model at time t+Δt, denoted as... n represents the total number of twin atomic models received by the centralized twin task node.

[0310] Step f110: At time t+Δt, the centralized twin task nodes collect the actual running state of the corresponding twin atomic model from each g-NB to obtain the actual running state of the first twin network.

[0311] The actual operating state corresponding to each twin atom model can be denoted as:

[0312] Step f111: Based on the predicted state and actual operating state of the first twin network, the twin network accuracy of the centralized twin task nodes is determined.

[0313] The accuracy of the first twin network can be identified by the consistency accuracy of the overall digital twin network, and can be denoted as: C DT =||C(P) m ,P r )||, where ‖‖ represents the calculation of the second-order norm. The specific process can be referred to in step f1051 to determine the model accuracy of the atomic model to be corrected, which will not be elaborated here.

[0314] Step f112: If the accuracy of the twin network is greater than or equal to Min(P), determine the probability that the accuracy of the twin network is greater than or equal to Min(P) by focusing the twin task nodes, and obtain the first probability.

[0315] Step f113: If the first probability is greater than or equal to Min(S), the twin task nodes are concentrated to determine the first twin network as the final twin network, and the twin network construction ends.

[0316] Step f114: If the accuracy of the twin network is less than Min(P), or the first probability is less than Min(S), the twin task nodes are centralized to determine the accuracy threshold of each twin atomic model based on Min(P) and Min(S).

[0317] First, calculate the conditional probability: combining the first twin network, calculate the conditional probability when the model accuracy of twin atomic model i meets the threshold requirement (c i ≥θ i The first twin network meets the threshold requirement (c). total The probability f(c) of ≥Min(P)) total ≥Min(P)c i ≥θ i Second, calculate the prior probability: combining the simulation results of the twin atomic model i, the calculation accuracy satisfies c. i ≥θ i The probability p(c) i ≥θ i Third, calculate the posterior probability: use conditional probability and prior probability to calculate whether the atomic model i to be modified satisfies the accuracy threshold θ. i The posterior probability f(c) i ≥θ i |c total ≥Min(P))=f(c total ≥Min(P)|c i ≥θ i )·p(c i ≥θ i ) / p(c total ≥Min(P)); Fourth, maximum a posteriori probability estimation: optimizing θ i Iterate through the first to third steps to maximize the posterior probability and determine the final accuracy threshold θ of the atomic model to be corrected. i In the fourth case, θ is optimized. i One implementation process can be: (1) Using methods such as gradient ascent or stochastic gradient ascent, for the current precision threshold θ i Adjust θ to optimize the posterior probability towards a higher value. (2) Repeat the iteration of the current precision threshold until the posterior probability is maximized or the gradient approaches zero. At this point, the optimized precision threshold is obtained. Specifically, adjust θ i Make the maximum a posteriori probability estimate calculation formula f(c) i ≥θ i |c total The iteration of f(c) in the direction of increasing (≥Min(P)) can be specifically denoted as: for f(c) i ≥θ i |c total The function ≥Min(P) is used with respect to θ iThe gradient is calculated as follows: Determine the current gradient ascent direction (i.e., the positive gradient direction), and adjust θ with an adjustment step size of α. i This causes the maximum a posteriori probability formula to iterate in the direction of increasing probability: The first two iterations, when the gradient When the value approaches 0, or falls below a very small threshold ε, the iteration ends. At this point, the posterior probability is considered to have been maximized, and the optimized precision threshold θ is obtained. i .

[0318] Step f115: The centralized twin task nodes send the accuracy threshold of each twin atomic model to the corresponding g-NB.

[0319] In step f116, after g-NB receives the precision threshold of the twin atomic model, it updates the atomic model to be corrected to a twin atomic model and updates the corresponding precision threshold to the precision threshold sent by the centralized twin task node. Then, steps f105 to f116 are executed until the first twin network that meets the requirements is obtained.

[0320] In this way, the twin task is analyzed, and the participating models are organized and corrected based on the analysis results. The overall twin environment, composed of multiple models, is then measured and verified, and joint optimization of the multiple models is performed to ensure that the overall twin environment meets user requests. Thus, the overall accuracy of the twin network is improved hierarchically through two closed-loop processes. Before orchestration, atomic models are measured and corrected to ensure their accuracy. After orchestration, the overall twin network formed by multiple models is measured and corrected to ensure its overall accuracy. Furthermore, thresholds are dynamically set; based on the desired twin network accuracy, the threshold for triggering correction for each atomic model is calculated, enabling rapid achievement of the desired twin network accuracy.

[0321] It should be noted that the descriptions of the same steps and contents as in other embodiments in this embodiment can be found in the descriptions in other embodiments, and will not be repeated here.

[0322] The network optimization method provided in this application, after determining the model accuracy of the atomic model to be corrected through the second node, determines the first accuracy threshold of the atomic model to be corrected based on the overall accuracy threshold of the twin network and the probability threshold of the twin network accuracy. Based on the model accuracy and the first accuracy threshold, the atomic model to be corrected is modified to obtain the twin atomic model. After obtaining the twin atomic model, the twin atomic model is sent to the first node so that the first node receives at least one corrected twin atomic model sent by the second node. The at least one twin atomic model is combined to obtain the first twin network. The first twin network is then optimized to obtain the second twin network with a network accuracy greater than or equal to the second accuracy threshold. In this way, the second node dynamically determines the first precision threshold of the atomic model to be corrected based on the overall precision threshold and the probability threshold of the twin network corresponding to the atomic model to be corrected. Then, the atomic model to be corrected is modified according to the first precision threshold, and the modified twin atomic model is sent to the first node. The first node then processes the modified twin atomic models sent by at least one second node according to the network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold and a stability greater than or equal to the stability threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision of current digital twin network models. This proposes a method to optimize the precision of a single model of a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability.

[0323] Based on the foregoing embodiments, embodiments of this application provide a first network optimization device, which can be applied to... Figure 1 and Figures 3-4 In the network optimization method provided in the corresponding embodiment, refer to Figure 6 As shown, the first network optimization device 4 may include: a receiving unit 41, a first processing unit 42, and a second processing unit 43; wherein:

[0324] The receiving unit 41 is used to receive a corrected twin atomic model sent by at least one second node; wherein the model accuracy of the twin atomic model is greater than or equal to a first accuracy threshold, and the twin atomic model corresponds to the communication network application;

[0325] The first processing unit 42 is used to perform combined processing on at least one twin atom model to obtain a first twin network;

[0326] The second processing unit 43 is used to perform model optimization processing on the first twin network to obtain a second twin network with network accuracy greater than or equal to a second accuracy threshold and stability greater than or equal to a stability threshold.

[0327] In other embodiments of this application, the second processing unit includes: a first prediction module, a first acquisition module, and a processing module; wherein:

[0328] The first prediction module is used to predict the first set of predicted state parameters of at least one twin atom model included in the first twin network at the first time step.

[0329] The first acquisition module is used to acquire, at a first moment, a first set of actual state parameters of at least one twin atom model included in the first twin network from at least one second node;

[0330] The processing module is used to perform model optimization processing on the first twin network model based on the first predicted state parameter set and the first actual state parameter set to obtain the second twin network.

[0331] In other embodiments of this application, the processing module is used to implement the following steps:

[0332] The network accuracy of the first twin network is calculated based on the first predicted state parameter set and the first actual state parameter set.

[0333] If the network precision is greater than or equal to the second precision threshold, perform simulation calculations on the first twin network to obtain the simulation results;

[0334] Based on the simulation results, the probability that the network precision of the first twin network is greater than or equal to the second precision threshold is calculated.

[0335] If the first probability is greater than or equal to the stability threshold, the second twin network is determined to be the first twin network.

[0336] In other embodiments of this application, when the processing module performs the step of calculating the network accuracy of the first twin network based on the first predicted state parameter set and the first actual state parameter set, it can be achieved through the following steps:

[0337] Based on the first set of predicted state parameters and the first set of actual state parameters, static evaluation parameters, dynamic evaluation parameters, and associated evaluation parameters are determined.

[0338] The network accuracy of the first twin network is obtained based on static evaluation parameters, dynamic evaluation parameters, and correlation evaluation parameters.

[0339] In other embodiments of this application, when the processing module performs the steps of determining static evaluation parameters, dynamic evaluation parameters, and associated evaluation parameters based on the first predicted state parameter set and the first actual state parameter set, it can be achieved through the following steps:

[0340] The first set of predicted state parameters is grouped to obtain a first set of static parameters and a first set of dynamic parameters.

[0341] The first set of actual state parameters is grouped to obtain the second set of static parameters and the second set of dynamic parameters.

[0342] The similarity between the first set of static parameters and the second set of static parameters is calculated to obtain the static evaluation parameters;

[0343] Based on the first set of dynamic parameters and the second set of dynamic parameters, determine the dynamic evaluation parameters;

[0344] Based on the first set of predicted state parameters, determine the parameter correlation of at least one twin atomic model to obtain the first correlation parameter;

[0345] Based on the first set of actual state parameters, determine the parameter association relationship between physical entity nodes corresponding to at least one twin atomic model, and obtain the second association relationship parameters;

[0346] Based on the first and second correlation parameters, the correlation evaluation parameters are determined.

[0347] In other embodiments of this application, when the processing module performs steps to determine the dynamic evaluation parameters based on the first dynamic parameter set and the second dynamic parameter set, it can be achieved through the following steps:

[0348] Based on the first set of dynamic parameters, the changing trend of each dynamic parameter is determined to obtain the first set of dynamic changing parameters;

[0349] Based on the second set of dynamic parameters, the changing trend of each dynamic parameter is determined to obtain the second set of dynamic changing parameters.

[0350] Consistency calculations are performed on the first set of dynamic parameters, the second set of dynamic parameters, the first set of dynamically changing parameters, and the second set of dynamically changing parameters to determine the dynamic evaluation parameters.

[0351] In other embodiments of this application, when the processing module performs consistency calculations on the first dynamic parameter set, the second dynamic parameter set, the first dynamically changing parameter set, and the second dynamically changing parameter set to determine the dynamic evaluation parameters, it can be achieved through the following steps:

[0352] A consistency calculation is performed on the first dynamic parameter set and the second dynamic parameter set to obtain the first consistency parameter;

[0353] A consistency calculation is performed on the first set of dynamic parameters and the second set of dynamic parameters to obtain the second consistency parameter.

[0354] Calculate the first product of the first weighting coefficient and the first consistency parameter;

[0355] Calculate the second product of the second weighting coefficient and the second consistency parameter;

[0356] Calculate the sum of the first and second products to obtain the dynamic evaluation parameters.

[0357] In other embodiments of this application, when the processing module performs steps to obtain the network accuracy of the first Siamese network based on static evaluation parameters, dynamic evaluation parameters, and correlation evaluation parameters, it can be achieved through the following steps:

[0358] Determine the static weighting coefficients, dynamic weighting coefficients, and correlation evaluation weighting coefficients;

[0359] The weighted static evaluation parameters are obtained by multiplying the static weight coefficients by the static evaluation parameters.

[0360] The weighted dynamic evaluation parameters are obtained by multiplying the dynamic weight coefficients by the dynamic evaluation parameters.

[0361] The weighted correlation evaluation parameters are obtained by multiplying the correlation evaluation weight coefficients and the correlation evaluation parameters.

[0362] The cumulative sum of the weighted static evaluation parameters, weighted dynamic evaluation parameters, and weighted correlation evaluation parameters is calculated to obtain the comprehensive parameters;

[0363] Calculate the first value of the static evaluation parameter;

[0364] Calculate the second value of the dynamic evaluation parameter;

[0365] Calculate the third value of the correlation evaluation parameter;

[0366] Calculate the first, second, and third values ​​to obtain the combined value;

[0367] The fourth value used to calculate the ratio of the comprehensive parameter to the comprehensive value;

[0368] The network accuracy is obtained by calculating the square root of the fourth numerical value to the number of parameters included in the comprehensive parameters.

[0369] In other embodiments of this application, the processing module is further configured to implement the following steps:

[0370] If the network accuracy is less than the second accuracy threshold, or the first probability is less than the stability threshold, the first accuracy threshold of each twin atom model is determined based on the overall accuracy threshold of the twin network and the probability threshold of the twin network accuracy.

[0371] Send the first precision threshold of the twin atomic model corresponding to each second node to the corresponding second node, so that the second node corrects the corresponding twin atomic model based on the first precision threshold.

[0372] In other embodiments of this application, the processing module is used to implement the step of determining the first precision threshold of each twin atom model based on the overall precision threshold of the twin network and the probability threshold of the twin network precision reaching a stability threshold. This can be achieved through the following steps:

[0373] If the network accuracy is less than the second accuracy threshold, or the first probability is less than the stability threshold, the second probability is that the model accuracy obtained from the simulation calculation of the first twin network is greater than or equal to the overall accuracy threshold of the twin network within a preset time period.

[0374] If the second probability is greater than or equal to the probability threshold reached by the Siamese network accuracy, the first accuracy threshold of each Siamese atom model is determined as the current accuracy threshold of the corresponding Siamese atom model.

[0375] If the second probability is less than the probability threshold for the twin network accuracy, the first accuracy threshold for each twin atom model is determined based on the overall accuracy threshold of the twin network, the probability threshold for the twin network accuracy, and the first twin network.

[0376] In other embodiments of this application, when the processing module performs the step of determining the first precision threshold for each twin atom model based on the overall precision threshold of the twin network, the probability threshold of the twin network precision reaching the first precision threshold, and the first twin network, it can be achieved through the following steps:

[0377] If the second probability is less than the probability threshold when the Siamese network accuracy reaches the threshold, calculate the conditional probability that the model accuracy of the first Siamese network is greater than or equal to the overall accuracy threshold of the Siamese network when the model accuracy is greater than or equal to the current accuracy threshold of each Siamese atom model.

[0378] Calculate the prior probability when the model accuracy is greater than or equal to the current accuracy threshold of each twin atom model;

[0379] Based on conditional probability and prior probability, determine the posterior probability when the model accuracy is greater than or equal to the current accuracy threshold of each twin atom model.

[0380] The current accuracy threshold for each twin atom model is optimized to obtain the optimized accuracy threshold;

[0381] After updating the current precision threshold of each twin atom model to the corresponding optimized precision threshold, repeat the steps to calculate the conditional probability that the model precision of the first twin network is greater than or equal to the overall precision threshold of the twin network when the model precision is greater than or equal to the current precision threshold of each twin atom model, until the maximum posterior probability is determined.

[0382] The first precision threshold for each twin atomic model is determined as the current precision threshold corresponding to the maximum a posteriori probability.

[0383] It should be noted that the specific implementation process of the interaction between units and modules in the embodiments of this application can be referred to Figure 1 and Figures 3-4 The implementation process of the network optimization method provided in the corresponding embodiments will not be described in detail here.

[0384] The first network optimization device provided in this application embodiment receives twin atomic models sent by at least one second node and corrected according to a first precision threshold dynamically determined by the second node through a first node, and performs combination processing on at least one twin atomic model to obtain a first twin network, and then performs model optimization processing on the first twin network to obtain a second twin network with network precision greater than or equal to a second precision threshold. In this way, the second node dynamically determines the first precision threshold of the atomic model to be corrected based on the overall precision threshold and the probability threshold of the twin network corresponding to the atomic model to be corrected. Then, the atomic model to be corrected is modified according to the first precision threshold, and the corrected twin atomic model is sent to the first node. The first node then processes the corrected twin atomic models sent by at least one second node according to the network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision of current digital twin network models. This proposes a method to optimize the precision of a single model of a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability.

[0385] Based on the foregoing embodiments, embodiments of this application provide a second network optimization device, which can be applied to... Figures 2-4 In the network optimization method provided in the corresponding embodiment, refer to Figure 7 As shown, the second network optimization device 5 may include: an acquisition unit 51, a determination unit 52, a correction unit 53, and a transmission unit 54; wherein:

[0386] Acquisition unit 51 is used to acquire the first precision threshold of the atomic model to be corrected;

[0387] The determining unit 52 is used to determine the model accuracy of the atomic model to be corrected; wherein the atomic model to be corrected corresponds to the communication network application;

[0388] The correction unit 53 is used to correct the atomic model to be corrected based on the model accuracy and a first accuracy threshold to obtain a twin atomic model; wherein the model accuracy of the twin atomic model is greater than or equal to the first accuracy threshold.

[0389] Sending unit 54 is used to send the twin atomic model to the first node.

[0390] In other embodiments of this application, the acquisition unit is used to implement the following steps:

[0391] The first precision threshold is determined as the initial precision threshold.

[0392] In other embodiments of this application, the acquisition unit is further configured to perform the following steps:

[0393] Receive the first precision threshold sent by the first node; wherein, the first precision threshold is the precision threshold corresponding to the twin atom model;

[0394] Correspondingly, after the acquisition unit, the second network optimization device further includes: a processing unit; wherein:

[0395] The processing unit is used to update the atomic model to be corrected to a twin atomic model, and then repeatedly execute the steps to determine the model accuracy of the atomic model to be corrected until the first accuracy threshold sent by the first node is not received.

[0396] In other embodiments of this application, the determining unit includes: a second prediction module, a second acquisition module, and a first calculation module; wherein:

[0397] The second prediction module is used to predict the second set of predicted state parameters of the atomic model to be corrected at the second time point.

[0398] The second acquisition module is used to acquire the second actual state parameter set of the atomic model to be corrected at the second time.

[0399] The first calculation module is used to calculate the model accuracy of the atomic model to be corrected based on the second set of predicted state parameters and the second set of actual state parameters.

[0400] In other embodiments of this application, the correction unit includes: a first determining module, a second calculating module, an adjusting module, and an updating module; wherein:

[0401] The first determining module is used to determine the twin atomic model as the atomic model to be corrected if the model accuracy is greater than or equal to the first accuracy threshold.

[0402] The second calculation module is used to calculate the difference between the first precision threshold and the model precision if the model precision is less than the first precision threshold.

[0403] The second calculation module is used to calculate the third product of the difference and the regularization coefficient;

[0404] The second calculation module is used to calculate the sum of the loss value of the atomic model to be corrected and the third product to obtain the adjustment parameters;

[0405] The adjustment module is used to reverse the adjustment of the adjustable parameters in the atomic model to be corrected based on the adjustment parameters, so as to obtain an intermediate atomic model.

[0406] The update module updates the atomic model to be corrected to an intermediate atomic model, and repeats the steps to determine the model accuracy of the atomic model to be corrected until a twin atomic model is obtained.

[0407] In other embodiments of this application, the second network optimization device further includes: a receiving unit and a generating unit; wherein:

[0408] The receiving unit is used to receive terminal communication indicator data sent by the terminal device.

[0409] The receiving unit is also used to receive the twin task request sent by the first node; wherein the twin task request includes at least: model type, overall accuracy threshold of the twin network, and probability threshold of the twin network accuracy reaching.

[0410] The generation unit is also used to generate at least one atomic model to be corrected based on terminal communication indicator data and model type.

[0411] It should be noted that the specific implementation process of the interaction between units and modules in the embodiments of this application can be referred to Figures 2-4 The implementation process of the network optimization method provided in the corresponding embodiments will not be described in detail here.

[0412] The second network optimization apparatus provided in this application embodiment determines the model accuracy of the atomic model to be corrected through at least one second node, determines a first accuracy threshold of the atomic model to be corrected based on the overall accuracy threshold of the twin network and the probability threshold of the twin network accuracy, and performs model correction on the atomic model to be corrected based on the model accuracy and the first accuracy threshold to obtain the twin atomic model. After obtaining the twin atomic model, the twin atomic model is sent to the first node so that the first node receives the corrected twin atomic model sent by at least one second node, performs combination processing on at least one twin atomic model to obtain a first twin network, and performs model optimization processing on the first twin network to obtain a second twin network with a network accuracy greater than or equal to the second accuracy threshold. In this way, the second node dynamically determines the first precision threshold of the atomic model to be corrected based on the overall precision threshold and the probability threshold of the twin network corresponding to the atomic model to be corrected. Then, the atomic model to be corrected is modified according to the first precision threshold, and the corrected twin atomic model is sent to the first node. The first node then processes the corrected twin atomic models sent by at least one second node according to the network requirements to obtain the first twin network. Finally, the first twin network is optimized to obtain a second twin network with a network precision greater than or equal to the second precision threshold. In this way, through two-level model optimization, and with different levels of model optimization at different nodes, the computational requirements of the nodes are reduced, solving the problem of poor precision of current digital twin network models. This proposes a method to optimize the precision of a single model of a digital twin network model, and then optimize the precision of a digital twin network model obtained by arranging multiple single models, thereby improving the precision of the digital twin network model and ensuring its reliability.

[0413] Based on the foregoing embodiments, embodiments of this application provide a network optimization system that can be applied to... Figures 1-4 In the network optimization method provided in the corresponding embodiment, refer to Figure 8 As shown, the network optimization system 6 may include: a first node 61 and at least one second node 62; wherein:

[0414] The first node, 61, is used to implement, for example... Figure 1 and Figures 3-4 The implementation process of the network optimization method provided in the corresponding embodiments will not be described in detail here;

[0415] The second node, 62, is used to implement, for example... Figures 2-4 The implementation process of the network optimization method provided in the corresponding embodiments will not be described in detail here.

[0416] Based on the foregoing embodiments, embodiments of this application provide a computer-readable storage medium, simply referred to as a storage medium, which stores one or more programs that can be executed by one or more processors to implement the reference. Figure 1 and Figures 3-4 ,or Figures 2-4 The implementation process of the network optimization method provided in the corresponding embodiments will not be described in detail here.

[0417] Based on the foregoing embodiments, this application also provides a computer program product, including a computer program that can be executed by a first node or at least one second node to complete the steps described in any of the foregoing methods.

[0418] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0419] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will 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 apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0420] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function 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 1 The function specified in one or more boxes.

[0421] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable 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.

[0422] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A network optimization method, characterized by, The method is applied to a first node, and the method comprises: receiving at least one second node sending a corrected processing twin atomic model; wherein the model accuracy of the twin atomic model is greater than or equal to a first accuracy threshold, and the twin atomic model corresponds to an application of a communication network; combining processing of at least one of the twin atomic models to obtain a first twin network; model optimization processing of the first twin network to obtain a second twin network with network accuracy greater than or equal to a second accuracy threshold and stability greater than or equal to a stability threshold.

2. The method of claim 1, wherein, The model optimization processing of the first twin network to obtain the second twin network with network accuracy greater than or equal to the second accuracy threshold and stability greater than or equal to the stability threshold comprises: predicting a first predicted state parameter set of at least one of the twin atomic models included in the first twin network at a first time; acquiring a first actual state parameter set of at least one of the twin atomic models included in the first twin network from at least one of the second nodes at the first time; based on the first predicted state parameter set and the first actual state parameter set, model optimization processing of the first twin network model to obtain the second twin network.

3. The method of claim 2, wherein, The model optimization processing of the first twin network model based on the first predicted state parameter set and the first actual state parameter set to obtain the second twin network comprises: based on the first predicted state parameter set and the first actual state parameter set, calculating the network accuracy of the first twin network; if the network accuracy is greater than or equal to the second accuracy threshold, performing simulation operation on the first twin network to obtain a simulation operation result; based on the simulation operation result, calculating a first probability that the network accuracy of the first twin network is greater than or equal to the second accuracy threshold; if the first probability is greater than or equal to the stability threshold, determining the second twin network as the first twin network.

4. The method of claim 3, wherein, The calculation of the network accuracy of the first twin network based on the first predicted state parameter set and the first actual state parameter set comprises: based on the first predicted state parameter set and the first actual state parameter set, determining static evaluation parameters, dynamic evaluation parameters and correlation evaluation parameters; based on the static evaluation parameters, the dynamic evaluation parameters and the correlation evaluation parameters, obtaining the network accuracy of the first twin network.

5. The method of claim 4, wherein, The determination of the static evaluation parameters, the dynamic evaluation parameters and the correlation evaluation parameters based on the first predicted state parameter set and the first actual state parameter set comprises: grouping the first predicted state parameter set to obtain a first static parameter set and a first dynamic parameter set; grouping the first actual state parameter set to obtain a second static parameter set and a second dynamic parameter set; performing similarity calculation on the first static parameter set and the second static parameter set to obtain the static evaluation parameters; based on the first dynamic parameter set and the second dynamic parameter set, determining the dynamic evaluation parameters; determine a parameter correlation of at least one of the twin atomic models based on the first set of predicted state parameters, to obtain a first correlation parameter; determine a parameter correlation between physical entity nodes corresponding to at least one of the twin atomic models based on the first set of actual state parameters, to obtain a second correlation parameter; determine the correlation evaluation parameter based on the first correlation parameter and the second correlation parameter.

6. The method of claim 5, wherein, The determination of the dynamic evaluation parameter based on the first set of dynamic parameters and the second set of dynamic parameters includes: determine a change trend of each dynamic parameter based on the first set of dynamic parameters, to obtain a first set of dynamic change parameters; determine a change trend of each dynamic parameter based on the second set of dynamic parameters, to obtain a second set of dynamic change parameters; perform consistency calculation on the first set of dynamic parameters, the second set of dynamic parameters, the first set of dynamic change parameters, and the second set of dynamic change parameters, to determine the dynamic evaluation parameter.

7. The method of claim 6, wherein, The consistency calculation on the first set of dynamic parameters, the second set of dynamic parameters, the first set of dynamic change parameters, and the second set of dynamic change parameters to determine the dynamic evaluation parameter includes: perform consistency calculation on the first set of dynamic parameters and the second set of dynamic parameters, to obtain a first consistency parameter; perform consistency calculation on the first set of dynamic change parameters and the second set of dynamic change parameters, to obtain a second consistency parameter; calculate a first product of a first weight coefficient and the first consistency parameter; calculate a second product of a second weight coefficient and the second consistency parameter; calculate a sum of the first product and the second product, to obtain the dynamic evaluation parameter.

8. The method of claim 4, wherein, The determination of the network precision of the first twin network based on the static evaluation parameter, the dynamic evaluation parameter, and the correlation evaluation parameter includes: determine a static weight coefficient, a dynamic weight coefficient, and a correlation evaluation weight coefficient; calculate a product of the static weight coefficient and the static evaluation parameter, to obtain a weighted static evaluation parameter; calculate a product of the dynamic weight coefficient and the dynamic evaluation parameter, to obtain a weighted dynamic evaluation parameter; calculate a product of the correlation evaluation weight coefficient and the correlation evaluation parameter, to obtain a weighted correlation evaluation parameter; calculate a sum of the weighted static evaluation parameter, the weighted dynamic evaluation parameter, and the weighted correlation evaluation parameter, to obtain a comprehensive parameter; calculate a first numerical value of the static evaluation parameter; calculate a second numerical value of the dynamic evaluation parameter; calculate a third numerical value of the correlation evaluation parameter; calculate the first numerical value, the second numerical value, and the third numerical value, to obtain a comprehensive numerical value; calculate a fourth numerical value of a ratio of the comprehensive parameter to the comprehensive numerical value; calculate a ratio of the fourth numerical value to a square root value of a number of parameters included in the comprehensive parameter, to obtain the network precision.

9. The method of claim 3, wherein, The method further includes: if the network accuracy is less than the second accuracy threshold or the first probability is less than the stability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold and a twin network accuracy reaching probability threshold; sending the first accuracy threshold of the twin atomic model corresponding to each of the second nodes to the corresponding second node, so that the second node corrects the corresponding twin atomic model based on the first accuracy threshold.

10. The method of claim 9, wherein, The method further includes: if the network accuracy is less than the second accuracy threshold or the first probability is less than the stability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold and a twin network accuracy reaching probability threshold; if the network accuracy is less than the second accuracy threshold or the first probability is less than the stability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold and a twin network accuracy reaching probability threshold; if the network accuracy is less than the second accuracy threshold or the first probability is less than the stability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold and a twin network accuracy reaching probability threshold; 11. The method of claim 10, wherein, if the network accuracy is less than the second accuracy threshold or the first probability is less than the stability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold and a twin network accuracy reaching probability threshold; if the second probability is greater than or equal to the twin network accuracy reaching probability threshold, determining the first accuracy threshold of each of the twin atomic models as the current accuracy threshold of the corresponding twin atomic model; if the second probability is less than the twin network accuracy reaching probability threshold, determining a first accuracy threshold of each of the twin atomic models based on a twin network overall accuracy threshold, a twin network accuracy reaching probability threshold and the first twin network. The method further includes: if the second probability is less than the twin network accuracy reaching probability threshold, calculating a conditional probability that the model accuracy is greater than or equal to the current accuracy threshold of each of the twin atomic models, and the model accuracy of the first twin network is greater than or equal to the twin network overall accuracy threshold; calculating a prior probability that the model accuracy is greater than or equal to the current accuracy threshold of each of the twin atomic models; based on the conditional probability and the prior probability, determining a posterior probability that the model accuracy is greater than or equal to the current accuracy threshold of each of the twin atomic models; 12. A network optimization method, characterized by, optimizing the current accuracy threshold of each of the twin atomic models to obtain an optimized accuracy threshold of each of the twin atomic models; after updating the current accuracy threshold of each of the twin atomic models to the corresponding optimized accuracy threshold, repeating the step of calculating the conditional probability that the model accuracy is greater than or equal to the current accuracy threshold of each of the twin atomic models, and the model accuracy of the first twin network is greater than or equal to the twin network overall accuracy threshold until a maximum posterior probability is determined; determining the first accuracy threshold of each of the twin atomic models as the current accuracy threshold corresponding to the maximum posterior probability. The method is applied to a second node, and the method includes: obtaining a first accuracy threshold of a to-be-corrected atomic model; determining a model precision of the to-be-corrected atomic model; wherein the to-be-corrected atomic model corresponds to a communication network application; based on the model precision and the first precision threshold, performing model correction on the to-be-corrected atomic model to obtain a twin atomic model; wherein the model precision of the twin atomic model is greater than or equal to the first precision threshold; sending the twin atomic model to the first node, so that the first node assembles a corresponding first twin network based on the twin atomic model.

13. The method of claim 12, wherein, The method further comprises: determining the first precision threshold as an initialization precision threshold.

14. The method of claim 12, wherein, The method further comprises: receiving the first precision threshold sent by the first node; wherein the first precision threshold is a precision threshold corresponding to the twin atomic model; correspondingly, after the step of obtaining the first precision threshold of the to-be-corrected atomic model, the method further comprises: after updating the to-be-corrected atomic model to the twin atomic model, repeatedly performing the step of determining the model precision of the to-be-corrected atomic model until no first precision threshold sent by the first node is received.

15. The method of claim 12, wherein, The method further comprises: predicting a second predicted state parameter set of the to-be-corrected atomic model at a second time; at the second time, obtaining a second actual state parameter set of the to-be-corrected atomic model; based on the second predicted state parameter set and the second actual state parameter set, calculating the model precision of the to-be-corrected atomic model.

16. The method of claim 12, wherein, The method further comprises: if the model precision is greater than or equal to the first precision threshold, determining the twin atomic model as the to-be-corrected atomic model; if the model precision is less than the first precision threshold, calculating a difference between the first precision threshold and the model precision; calculating a third product of the difference and a regularization coefficient; calculating a sum of a loss value of the to-be-corrected atomic model and the third product to obtain an adjustment parameter; based on the adjustment parameter, inversely adjusting an adjustable parameter in the to-be-corrected atomic model to obtain an intermediate atomic model; updating the to-be-corrected atomic model to the intermediate atomic model, and repeatedly performing the step of determining the model precision of the to-be-corrected atomic model until the twin atomic model is obtained.

17. The method of claim 13, wherein, The method further comprises: receiving terminal communication index data sent by a terminal device; receiving a twin task request sent by the first node; wherein the twin task request at least includes: a model type, a whole precision threshold of the twin network, and a twin network precision reaching probability threshold; based on the terminal communication index data and the model type, generating at least one to-be-corrected atomic model.

18. A first network optimization apparatus, comprising: The device is applied to a first node, and the device comprises a receiving unit, a first processing unit and a second processing unit; wherein: The receiving unit is configured to receive at least one corrected twin atomic model sent by a second node; wherein the model precision of the twin atomic model is greater than or equal to a first precision threshold, and the twin atomic model corresponds to a communication network application; The first processing unit is configured to perform combination processing on the at least one twin atomic model to obtain a first twin network; The second processing unit is configured to perform model optimization processing on the first twin network to obtain a second twin network, wherein the network precision of the second twin network is greater than or equal to a second precision threshold, and the stability of the second twin network is greater than or equal to a stability threshold.

19. A second network optimization apparatus, comprising: The device is applied to a second node, and the device comprises an obtaining unit, a determining unit, a correcting unit and a sending unit; wherein: The obtaining unit is configured to obtain a first precision threshold of a to-be-corrected atomic model; The determining unit is configured to determine the model precision of the to-be-corrected atomic model; wherein the to-be-corrected atomic model corresponds to a communication network application; The correcting unit is configured to perform model correction on the to-be-corrected atomic model based on the model precision and the first precision threshold to obtain a twin atomic model; The sending unit is configured to send the twin atomic model to a first node.

20. A network optimization system, comprising: The system comprises at least a first node and at least one second node; wherein: The first node is configured to implement the steps of the network optimization method according to any one of claims 1 to 11; The second node is configured to implement the steps of the network optimization method according to any one of claims 12 to 17.

21. A storage medium, characterized by The storage medium has a resource conflict indication program stored thereon, and the resource conflict indication program, when executed, is configured to implement the steps of the network optimization method according to any one of claims 1 to 11 or claims 12 to 17.

22. A computer program product comprising a computer program, characterized in that, The computer program, when executed by a processor, implements the steps of the network optimization method according to any one of claims 1 to 11 or claims 12 to 17.

Images