Device parameter configuration method and device, electronic device and storage medium

Abstract

The embodiment of the present application provides a device parameter configuration method and apparatus, an electronic device and a storage medium, which belongs to the field of machine learning technology. The method includes: obtaining the number of devices in the target network, device associations, device parameters and network performance indicators of the target network; constructing a target Markov random field based on the number of devices, device associations and device parameters; performing spanning tree sampling on the target Markov random field to obtain multiple uniform random spanning trees; performing belief propagation on each uniform random spanning tree to obtain a parameter edge probability distribution; performing probability merging on the parameter edge probability distribution to obtain a target edge probability distribution; based on the target edge probability distribution, network performance indicators and a preset network performance evaluation function, performing parameter configuration on the device parameters to obtain a target parameter value combination. The embodiment of the present application can improve the efficiency of device parameter configuration in large-scale network optimization problems.

Description

Technical Field

[0001] The present application relates to the field of machine learning technology, and in particular to a device parameter configuration method and apparatus, an electronic device, and a storage medium. Background Art

[0002] With the development of science and technology, the efficient operation of modern production and life is inseparable from various sophisticated and efficient networks, such as transportation networks and communication networks.

[0003] At present, there are two methods for configuring device parameters in the network to improve the performance of the network. One method is to model the network to be optimized as a model for solving a mixed integer linear programming problem, and use the mixed integer linear programming method to solve the model. However, when the scale of the network to be optimized is large, this method is prone to increase the number of selectable device parameters of the network to be optimized, causing the scale of the mixed integer linear programming problem to increase sharply, resulting in increased computational difficulty and time for device parameter configuration. Another method is to model the network to be optimized as a Markov random field, but when the network to be optimized has a ring structure, it is easy to cause the convergence effect of the device parameter configuration to be poor or slow, resulting in the Markov random field model being unable to obtain the optimal device parameter configuration of the network to be optimized. Therefore, how to improve the efficiency of device parameter configuration in large-scale networks in network optimization problems has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] The main purpose of the embodiments of the present application is to propose a device parameter configuration method and apparatus, an electronic device and a storage medium, aiming to improve the efficiency of device parameter configuration in large-scale network optimization problems.

[0005] To achieve the above object, a first aspect of an embodiment of the present application provides a device parameter configuration method, the method comprising:

[0006] Obtaining the number of devices in the target network, device association relationships, device parameters, and network performance indicators of the target network;

[0007] Based on the number of devices, the device association relationship, and the device parameters, a target Markov random field is constructed, wherein the target Markov random field includes a plurality of device nodes;

[0008] Performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees, wherein the uniform random spanning trees include spanning tree nodes;

[0009] Perform belief propagation on each of the uniform random spanning trees to obtain a parameter marginal probability distribution of the spanning tree nodes on each of the uniform random spanning trees;

[0010] Probabilistically merging the parameter edge probability distributions to obtain a target edge probability distribution for each of the device nodes;

[0011] Based on the target edge probability distribution, the network performance index and a preset network performance evaluation function, the device parameters are configured to obtain a target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

[0012] In some embodiments, the target Markov random field includes device association edges, and the step of performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees includes:

[0013] Calculating the equivalent resistance value of the device-associated edge to obtain an estimated probability of the device-associated edge, wherein the estimated probability represents the probability of the device-associated edge participating in constructing a uniform random spanning tree;

[0014] Based on the estimated probability and the device node, uniformly and randomly sample the device-related edges to obtain target relationship edges;

[0015] Based on the device nodes and the target relationship edges, a plurality of the uniformly randomly generated trees are constructed.

[0016] In some embodiments, the uniform random spanning tree includes a spanning tree edge, and the belief propagation is performed on each of the uniform random spanning trees to obtain the parameter edge probability distribution of the spanning tree nodes on each of the uniform random spanning trees, including:

[0017] Obtaining a first mapping relationship and a second mapping relationship, wherein the first mapping relationship represents a mapping relationship between a value combination of the device parameter and a state of the target network, and the second mapping relationship is used to represent a mapping relationship between a parameter value of the device parameter and a cost consumption of a device in the target network;

[0018] Based on the first mapping relationship and the second mapping relationship, calculating the node joint probability distribution of the spanning tree nodes on each of the uniform random spanning trees;

[0019] Calculating a connection edge belief of each spanning tree edge based on the node joint probability distribution of two spanning tree nodes on the spanning tree edge;

[0020] Based on the estimated probability, modifying the connection edge belief to obtain a modified belief;

[0021] The modified belief is propagated to obtain the marginal probability distribution of the parameters of the spanning tree nodes.

[0022] In some embodiments, the modifying the connection edge belief based on the estimated probability to obtain the modified belief includes:

[0023] The device nodes are numbered to obtain device node numbers, and the tree node numbers of the spanning tree nodes on each of the uniform random spanning trees are determined according to the device node numbers;

[0024] Based on the tree node number, the revised belief is calculated using the following formula:

[0025] ;

[0026] in, represents the modified belief of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j, Indicates that the tree node is numbered i, Indicates that the tree node number is j, represents the estimated probability of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j, Represents the connecting edge belief of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j.

[0027] In some embodiments, the probabilistic merging of the parameter edge probability distributions to obtain the target edge probability distribution of each of the device nodes includes:

[0028] Searching for the spanning tree node of each of the device nodes in each of the uniform random spanning trees to obtain a tree node set, wherein the tree node set includes one or more of the spanning tree nodes;

[0029] The parameter edge probability distributions of the spanning tree nodes in the tree node set are multiplied and merged to obtain the target edge probability distribution.

[0030] In some embodiments, the configuring of the device parameters based on the target edge probability distribution, the network performance index and a preset network performance evaluation function to obtain a target parameter value combination includes:

[0031] Based on the target edge probability distribution, performing Gibbs sampling on the device parameters to obtain a first parameter combination;

[0032] Based on the target edge probability distribution, greedily sample the device parameters to obtain a second parameter combination;

[0033] Based on the network performance evaluation function and the network performance indicator, screening out a candidate parameter combination from the first parameter combination and the second parameter combination;

[0034] The target parameter value combination is screened out from the candidate parameter combination and the preset alternative parameter combination, wherein the alternative parameter combination refers to the parameter combination selected by the target network in the previous round of parameter configuration.

[0035] In some embodiments, the step of selecting a candidate parameter combination from the first parameter combination and the second parameter combination based on the network performance evaluation function and the network performance indicator includes:

[0036] Substituting the first parameter combination into the network performance evaluation function to obtain a first network performance;

[0037] Substituting the second parameter combination into the network performance evaluation function to obtain a second network performance;

[0038] Based on the network performance indicator, the first network performance and the second network performance, the candidate parameter combination is determined from the first parameter combination and the second parameter combination, wherein the network performance of the candidate parameter combination is closest to the network performance indicator.

[0039] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application provides a device parameter configuration apparatus, the apparatus comprising:

[0040] A network device information acquisition module is used to obtain the number of devices in the target network, device association relationships, device parameters and network performance indicators of the target network;

[0041] A random field construction module, used to construct a target Markov random field based on the number of devices, the device association relationship, and the device parameters, wherein the target Markov random field includes a plurality of device nodes;

[0042] A spanning tree sampling module, used for performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees, wherein the uniform random spanning trees include spanning tree nodes;

[0043] A spanning tree belief propagation module, used to perform belief propagation on each of the uniform random spanning trees to obtain the parameter marginal probability distribution of the spanning tree nodes on each of the uniform random spanning trees;

[0044] An edge probability calculation module, used for probabilistically merging the parameter edge probability distributions to obtain a target edge probability distribution for each of the device nodes;

[0045] The device parameter configuration module is used to configure the device parameters based on the target edge probability distribution, the network performance indicators and a preset network performance evaluation function to obtain a target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

[0046] To achieve the above objectives, a third aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0047] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0048] The device parameter configuration method and apparatus, electronic device and storage medium proposed in the present application can improve the parameter configuration efficiency of large-scale networks by constructing a target Markov random field using the acquired device quantity, device association relationship and device parameters of the devices in the target network. Furthermore, spanning tree sampling is performed on the target Markov random field to obtain multiple uniform random spanning trees, and belief propagation is performed on the multiple uniform random spanning trees, breaking the closed-loop state of the target Markov random field, allowing belief propagation to converge, thereby obtaining accurate device parameter configuration, and improving the efficiency of device parameter configuration in large-scale networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a flow chart of a device parameter configuration method provided in an embodiment of the present application;

[0050] Figure 2 yes Figure 1 Flow chart of step S103 in FIG.

[0051] Figure 3 yes Figure 1 Flow chart of step S104 in FIG.

[0052] Figure 4 yes Figure 1 Flow chart of step S105 in FIG.

[0053] Figure 5 yes Figure 1 Flow chart of step S106 in FIG.

[0054] Figure 6 yes Figure 5 Flow chart of step S503 in FIG.

[0055] Figure 7 It is a structural diagram of a device parameter configuration apparatus provided in an embodiment of the present application;

[0056] Figure 8 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0058] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0060] First, some nouns involved in this application are analyzed:

[0061] Markov Random Field: Markov Random Field (MRF) is an undirected probabilistic graphical model used to describe the interdependence between a set of random variables. Its undirected graph characteristics allow it to effectively express the conditional independence between variables, making it widely used in various fields, including computer vision, natural language processing, statistical physics, etc. Let's introduce the basic concepts, properties and applications of Markov random fields in detail.

[0062] Uniform random spanning tree: A uniform random spanning tree is a tree randomly generated from an undirected graph, so that each possible spanning tree has the same probability of being selected. In other words, it is a uniform random sampling of all spanning trees in the undirected graph.

[0063] With the development of science and technology, the efficient operation of modern production and life is inseparable from various sophisticated and efficient networks, such as transportation networks and communication networks.

[0064] At present, there are two methods for configuring device parameters in the network to improve the performance of the network. One method is to model the network to be optimized as a model for solving a mixed integer linear programming problem, and use the mixed integer linear programming method to solve the model. However, when the scale of the network to be optimized is large, this method is prone to increase the number of selectable device parameters of the network to be optimized, causing the scale of the mixed integer linear programming problem to increase sharply, resulting in increased computational difficulty and time for device parameter configuration. Another method is to model the network to be optimized as a Markov random field, but when the network to be optimized has a ring structure, it is easy to cause the convergence effect of the device parameter configuration to be poor or slow, resulting in the Markov random field model being unable to obtain the optimal device parameter configuration of the network to be optimized. Therefore, how to improve the efficiency of device parameter configuration in large-scale networks in network optimization problems has become a technical problem that needs to be solved urgently.

[0065] Based on this, the embodiments of the present application provide a device parameter configuration method and apparatus, an electronic device and a storage medium, aiming to improve the efficiency of device parameter configuration in large-scale network optimization problems.

[0066] The recommendation method and device, electronic device and storage medium provided in the embodiments of the present application are specifically described through the following embodiments. First, the recommendation method in the embodiments of the present application is described.

[0067] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0068] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0069] The device parameter configuration method provided in the embodiment of the present application relates to the field of machine learning technology. The device parameter configuration method provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or can be configured as a server cluster or a distributed system composed of multiple physical servers, or can be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the device parameter configuration method, etc., but is not limited to the above forms.

[0070] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0071] Figure 1 is an optional flowchart of the device parameter configuration method provided in the embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S106.

[0072] Step S101, obtaining the number of devices in the target network, device association relationships, device parameters, and network performance indicators of the target network;

[0073] Step S102, constructing a target Markov random field based on the number of devices, device association relationships, and device parameters, wherein the target Markov random field includes a plurality of device nodes;

[0074] Step S103, performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees, wherein the uniform random spanning trees include spanning tree nodes;

[0075] Step S104, performing belief propagation on each uniform random spanning tree to obtain parameter marginal probability distribution of spanning tree nodes on each uniform random spanning tree;

[0076] Step S105, performing probability merging on the parameter edge probability distribution to obtain the target edge probability distribution of each device node;

[0077] Step S106, based on the target edge probability distribution, the network performance index and the preset network performance evaluation function, the device parameters are configured to obtain the target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

[0078] Steps S101 to S106 shown in the embodiment of the present application are to obtain the number of devices, device associations, device parameters and network performance indicators of the target network of the devices in the target network, and then use the obtained number of devices, device associations and device parameters in the target network to perform modeling to obtain a target Markov random field. Further, the target Markov random field is sampled to obtain multiple uniform random spanning trees. Secondly, belief propagation is performed on the sampled multiple uniform random spanning trees to obtain the parameter edge probability distribution of the spanning tree nodes on each uniform random spanning tree. Then, the target edge probability distribution of each device node is obtained by merging the parameter edge probability distributions of the same spanning tree nodes on each uniform random spanning tree. Finally, according to the target edge probability distribution of the device nodes, the network performance indicators of the target network and the pre-fitted network performance evaluation function for characterizing the mapping relationship between the value combination of the device parameters and the performance of the target network, the device parameters of the devices in the target network are configured to obtain a target parameter value combination that is close to or equal to the network performance indicators. Therefore, the present application uses the acquired device quantity, device association relationship and device parameters of the devices in the target network to construct a target Markov random field, which can improve the parameter configuration efficiency of large-scale networks. Furthermore, the target Markov random field is sampled by spanning tree to obtain multiple uniform random spanning trees, and belief propagation is performed on the multiple uniform random spanning trees, breaking the closed-loop state of the target Markov random field, allowing belief propagation to converge, thereby obtaining accurate device parameter configuration, and improving the efficiency of device parameter configuration in large-scale networks.

[0079] In one application scenario, taking the network delay optimization of a communication network as an example, a target Markov random field about the communication network is constructed by obtaining the number of communication devices in the communication network, the correlation relationship between the communication devices and the parameter configuration of the communication devices. After the target Markov random field is constructed, a set number of open-loop uniform random generating trees are sampled from the target Markov random field. Secondly, a set number of uniform random generating trees are used for belief propagation to obtain the target edge probability distribution of each communication device in the target Markov random field. Finally, according to the target edge probability distribution, the communication network performance evaluation function obtained by pre-function fitting and the pre-set communication network delay optimization target, the parameter values ​​of the device parameters of the communication devices in the communication network are selected and configured to obtain a target parameter value combination that can optimize the network delay of the communication network.

[0080] In step S101 of some embodiments, the target network refers to a large-scale network composed of interconnected devices, such as a transportation network and a communication network. The network performance index refers to the state or target value that the target network is expected to achieve. The number of devices refers to the number of devices that are interconnected. Device associations may include information transmission, data exchange, or interactive operations. The present application can obtain the number of devices, device associations, and device parameters of the devices in the target network by querying the configuration document that describes the parameter information in the target network, and then determine the network performance index of the target network according to the set network optimization goals.

[0081] It should be noted that a device may include one or more device parameters, and a device parameter may include one or more parameter values.

[0082] In step S102 of some embodiments, after obtaining the number of devices, device associations and device parameters of the devices in the target network, a Markov random field is constructed using the devices as nodes and the device associations as edges. Based on the number of devices, the number of nodes in the Markov random field is determined. Based on the device parameters, the state space of the nodes in the Markov random field is determined. The Markov random field is converted into a target Markov random field, thereby achieving modeling of the target network.

[0083] It should be noted that the target Markov random field includes device nodes and device-associated edges, wherein the device nodes refer to nodes formed by devices, and the device-associated edges refer to edges formed by device-associated relationships.

[0084] In step S103 of some embodiments, the uniform random spanning tree refers to a spanning tree obtained by randomly and uniformly sampling device nodes and device associated edges in the target Markov random field. After the target Markov random field is constructed, the target Markov random field is sampled by spanning trees using the Wilson's maze algorithm to obtain multiple uniform random spanning trees.

[0085] It should be noted that the uniform random spanning tree includes spanning tree nodes and spanning tree associated edges, wherein the spanning tree nodes refer to nodes of the uniform random spanning tree, and the device associated edges refer to edges of the uniform random spanning tree.

[0086] It is also necessary to know that the number of uniform random spanning trees can be set before spanning tree sampling is performed on the Markov random field, so that when spanning tree sampling is performed on the Markov random field, sampling can be stopped when a preset number of spanning trees are generated.

[0087] See also Figure 2 In some embodiments, step S103 may include but is not limited to steps S201 to S203:

[0088] Step S201, calculating the equivalent resistance value of the device-associated edge to obtain the estimated probability of the device-associated edge, wherein the estimated probability represents the probability of the device-associated edge participating in the construction of the uniform random spanning tree;

[0089] Step S202, based on the estimated probability and the device nodes, uniformly and randomly sample the device-related edges to obtain the target relationship edges;

[0090] Step S203, constructing multiple uniform random spanning trees based on the device nodes and target relationship edges.

[0091] In step S201 of some embodiments, the equivalent resistance value refers to the resistance value between any two nodes in the circuit. It should be noted that in the Markov random field, the equivalent resistance value between any two nodes can represent the probability that the connection edge between the two nodes participates in the construction of any uniform random spanning tree. By calculating the equivalent resistance value of each device-associated edge in the target Markov random field, the estimated probability that each device-associated edge in the target Markov random field participates in the construction of a uniform random spanning tree can be obtained.

[0092] In step S202 and step S203 of some embodiments, after the estimated probability of each edge in the target Markov random field is obtained, the Wilson Labyrinth algorithm is used to sample spanning trees from the target Markov random field according to the estimated probability of each edge to ensure that the sampled spanning trees are uniformly randomly spanning trees.

[0093] In detail, in an embodiment of the present application, any device node in the target Markov random field is used as the starting node of the uniform random spanning tree, and according to the estimated probability of all device-associated edges connected to the starting node, one edge is selected from all device-associated edges connected to the starting node as the target relationship edge connected to the starting node, and the above steps are repeated until a complete spanning tree is constructed. The resulting spanning tree is a uniform random spanning tree. For example, device node 1 in the Markov random field is used as the starting node, and the starting node has two device-associated edges, namely device-associated edge 1 and device-associated edge 2, wherein the estimated probability of device-associated edge 1 is 70%, and the estimated probability of device-associated edge 2 is 30%. Then, the Wilson maze algorithm can select the target relationship edge from device-associated edge 1 and device-associated edge 2 by randomly generating any number from 1 to 10 (1 to 7 indicates selecting device-associated edge 1 as the target relationship edge, and 8 to 10 indicates selecting device-associated edge 2 as the target relationship edge).

[0094] In steps S201 to S203 shown in the present embodiment, the estimated probability of each device-associated edge is obtained by calculating the equivalent resistance value of each device-associated edge in the target Markov random field, and then the device-associated edge is selected using the Wilson maze algorithm according to the estimated probability of each edge to obtain the target relationship edge. Furthermore, a uniform random spanning tree is generated according to the target relationship edge and the device nodes on the target relationship edge, so that open-loop processing of the target Markov random field can be achieved, thereby avoiding falling into a closed loop when belief propagation is performed on the target Markov random field, thereby causing the belief convergence to be unable to converge. In addition, uniform random sampling of the target Markov random field can ensure that the probability of the spanning tree generated in the target Markov random field is uniform, and there will be no bias or bias towards a certain type of spanning tree, thereby improving the accuracy of device parameter configuration.

[0095] In step S104 of some embodiments, after a preset number of uniform random spanning trees are sampled from the target Markov random field, belief propagation is performed on each uniform random spanning tree to obtain the marginal probability distribution of each spanning tree node on each uniform random spanning tree, that is, the parameter marginal probability distribution. It should be noted that the uniform random spanning tree also includes spanning tree edges.

[0096] See also Figure 3 In some embodiments, step S104 may include but is not limited to steps S301 to S305:

[0097] Step S301, obtaining a first mapping relationship and a second mapping relationship, wherein the first mapping relationship represents a mapping relationship between a value combination of a device parameter and a state of a target network, and the second mapping relationship is used to represent a mapping relationship between a parameter value of a device parameter and a cost consumption of a device in the target network;

[0098] Step S302, calculating the node joint probability distribution of the spanning tree nodes on each uniform random spanning tree based on the first mapping relationship and the second mapping relationship;

[0099] Step S303, calculating the connection edge belief of each spanning tree edge based on the node joint probability distribution of two spanning tree nodes on the spanning tree edge;

[0100] Step S304, based on the estimated probability, modify the connection edge belief to obtain a modified belief;

[0101] Step S305, propagating the modified belief to obtain the marginal probability distribution of the parameters of the spanning tree nodes.

[0102] In step S301 of some embodiments, the first mapping relationship is used to characterize the mapping relationship between the combination of parameter values ​​of device parameters and the state of the target network. For example, when the parameter value of device parameter 1 of device 1 is 10, the parameter value of device parameter 2 is 11, the parameter value of device parameter 1 of device 2 is 9, and the parameter value of device parameter 2 is 12, according to the first mapping relationship, it can be obtained that the state of the target network composed of device 1 and device 2 can be 3. The second mapping relationship represents the cost consumption of the target network when the device operates under the selected device parameters, for example, when the communication device in the communication network operates under the specified device operating parameters, the operating resources consumed by the communication network. The first mapping relationship and the second mapping relationship can be obtained by querying the configuration document of the target network.

[0103] In step S302 of some embodiments, after the first mapping relationship and the second mapping relationship are obtained, the node joint probability distribution of each spanning tree node in each uniform random spanning tree based on the other spanning tree nodes can be determined according to the first mapping relationship and the second mapping relationship.

[0104] It should be noted that in some embodiments of the present application, the sum-product algorithm can be used to perform belief propagation on the uniform random spanning tree to obtain the node joint probability distribution of each spanning tree node in the uniform random spanning tree based on the remaining spanning tree nodes.

[0105] In step S303 of some embodiments, a single uniform random spanning tree only contains local information in the target Markov random field, resulting in that when belief propagation is performed on each uniform random spanning tree, the belief on each spanning tree node is different from the belief on each device node when belief propagation is performed on the target Markov random field. Therefore, it is also necessary to calculate the connecting edge belief of each spanning tree edge based on the node joint probability distribution of the two spanning tree nodes on each spanning tree edge in the uniform random spanning tree.

[0106] In step S304 of some embodiments, after obtaining the connection edge belief of each spanning tree edge in the uniform random spanning tree, the connection edge belief is corrected to ensure that the corrected belief is close to the belief of each edge in the target Markov random field when belief propagation is performed on the target Markov random field, thereby reducing the computational error caused by the difference between the uniform random spanning tree and the target Markov random field.

[0107] In detail, the present application can obtain the device node number by marking the device node with a number, and then determine the tree node number of the spanning tree node on each uniform random spanning tree according to the device node number. Further, based on the tree node number, the revised belief is calculated using the following formula:

[0108] ;

[0109] in, represents the revised belief of the spanning tree edge between the spanning tree node with tree node number i and the spanning tree node with tree node number j, Indicates that the tree node number is i, Indicates that the tree node number is j, represents the estimated probability of the spanning tree edge between the spanning tree node with tree node number i and the spanning tree node with tree node number j, Represents the connection edge belief of the spanning tree edge between the spanning tree node with tree node number i and the spanning tree node with tree node number j.

[0110] In step S305 of some embodiments, after obtaining the revised belief of each spanning tree edge, the revised belief is propagated to the spanning tree nodes through the belief propagation algorithm, and the edge probability distribution of each spanning tree node, that is, the parameter edge probability distribution, can be obtained. It should be noted that the parameter edge probability distribution of each spanning tree node is the parameter edge probability distribution of each device parameter of the spanning tree node under the condition of specified parameter values ​​based on the neighboring spanning tree nodes, wherein the neighboring spanning tree nodes of the spanning tree node are nodes that have edges connected to the spanning tree node.

[0111] In step S105 of some embodiments, the spanning tree nodes with the same tree node number on each uniform random spanning tree are probabilistically merged to obtain the edge probability distribution of the device nodes with the same device node number as the tree node number, that is, the target edge probability distribution.

[0112] See also Figure 4 In some embodiments, step S105 may include but is not limited to steps S401 to S402:

[0113] Step S401, searching for a spanning tree node of each device node in each uniform random spanning tree to obtain a tree node set, wherein the tree node set includes one or more spanning tree nodes;

[0114] Step S402, multiplying and merging the parameter edge probability distributions of the spanning tree nodes in the tree node set to obtain a target edge probability distribution.

[0115] In step S401 of some embodiments, the spanning tree node corresponding to each device node in the target Markov random field in each uniform random spanning tree is queried, and one or more spanning tree nodes with the same tree node number can be obtained. The spanning tree nodes with the same tree node number are grouped together to obtain a tree node set.

[0116] In step S402 of some embodiments, the parameter edge probability distribution of each spanning tree node in the tree node set is multiplied and combined to obtain the target edge probability distribution of the device node having the same number as the spanning tree node.

[0117] It should be noted that by repeating the above step S402 for each device node in the target Markov random field, the target edge probability distribution of each device node in the target Markov random field can be obtained.

[0118] In detail, the target edge probability distribution of each device node in the Markov random field can be calculated using the following formula:

[0119] ;

[0120] in, Indicates that the device node number is i. The parameter value of the device parameter of the device node with device node number i is x. Represents a set of parameter values ​​for device parameters of all device nodes. represents the marginal probability distribution of equipment performance points, represents the numbered index of the uniformly randomly generated tree, represents the number of uniformly randomly generated trees, Represents the parametric marginal probability distribution of the target node with index k in a uniformly randomly generated tree.

[0121] In steps S401 to S402 shown in the present embodiment, by merging the parameter marginal probability distribution of the device node in each uniform random spanning tree, the target marginal probability distribution of the device node can be obtained, thereby achieving the use of the parameter marginal probability distribution of each spanning tree node in the uniform random spanning tree to replace the marginal probability distribution of each device node in the Markov random field, solving the problem of non-convergence caused by the closed loop when the Markov random field performs belief propagation, thereby improving the accuracy of device parameter configuration.

[0122] In step S106 of some embodiments, after determining the target edge probability distribution of each device node in the target Markov random field, it is necessary to adjust the values ​​of the device parameters according to the network performance index of the target network, a pre-set network performance evaluation function and the edge probability distribution of each node to obtain a target parameter value combination that makes the network performance of the target network closest to the network performance index or equal to the network performance index.

[0123] It is important to know that network performance includes the state of the target network and the cost consumption of the target network. Therefore, by merging the first mapping relationship and the second mapping relationship, the mapping relationship between the value combination of the device parameters and the network performance can be obtained, and then the function is fitted according to the mapping relationship to obtain the network performance evaluation function.

[0124] See also Figure 5 In some embodiments, step S105 may also include but is not limited to steps S501 to S504:

[0125] Step S501, performing Gibbs sampling on device parameters based on target edge probability distribution to obtain a first parameter combination;

[0126] Step S502, greedily sampling the device parameters based on the target edge probability distribution to obtain a second parameter combination;

[0127] Step S503, based on the network performance evaluation function and the network performance index, screening out a candidate parameter combination from the first parameter combination and the second parameter combination;

[0128] Step S504, screening out a target parameter value combination from the candidate parameter combination and the preset candidate parameter combination, wherein the candidate parameter combination refers to the parameter combination selected by the target network in the previous round of parameter configuration.

[0129] In step S501 of some embodiments, Gibbs sampling is performed on the device parameters in the device represented by the device node according to the target edge probability distribution of each device node to obtain a combination configuration of the device parameters of the device, i.e., a first parameter combination. Specifically, a random number is generated according to the edge probability distribution of the states that can be selected by the device parameters in each device node, and a target state is selected from the states that can be selected by the device parameters according to the generated random number, thereby obtaining the first parameter combination. For example, device parameter 1 has a first state and a second state, the edge probability distribution of the first state is 40%, and the edge probability distribution of the second state is 60%. By generating any random number from 1 to 5, the target state is selected from the first state and the second state as the first parameter combination of device parameter 1 (when the random number is 1 or 2, the first state is selected as the first parameter combination of device parameter 1, and when the random number is any number from 3 to 5, the second state is selected as the first parameter combination of device parameter 1).

[0130] In step S502 of some embodiments, based on the target edge probability distribution of each device node, greedy sampling is performed on the device parameters in the device represented by the device node to obtain a combination configuration of the device parameters of the device, i.e., a second parameter combination. Specifically, based on the edge probability distribution of the states that can be selected by the device parameters in each device node, the state with the largest edge probability distribution is selected as the second parameter combination of the device parameters.

[0131] In step S503 of some embodiments, the first parameter combination is substituted into the network performance evaluation function for calculation, so as to obtain the network performance of the target network under the first parameter combination. Then, the second parameter combination is substituted into the network performance evaluation function for calculation, so as to obtain the network performance of the target network under the second parameter combination. The network performance of the target network under the first parameter combination is compared with the network performance of the target network under the second parameter combination, and a candidate parameter combination whose network performance is closest to the network performance index or equal to the network performance index is selected.

[0132] See also Figure 6 In some embodiments, step S503 includes but is not limited to steps S601 to S603:

[0133] Step S601, substituting the first parameter combination into the network performance evaluation function to obtain the first network performance;

[0134] Step S602, substituting the second parameter combination into the network performance evaluation function to obtain the second network performance;

[0135] Step S603, based on the network performance indicator, the first network performance and the second network performance, determine a candidate parameter combination from the first parameter combination and the second parameter combination, wherein the network performance of the candidate parameter combination is closest to the network performance indicator.

[0136] In steps S601 and S602 of some embodiments, a network performance evaluation function is used to calculate the network performance of a target network under a first parameter combination of device parameters to obtain a first network performance, and a network performance evaluation function is used to calculate the network performance of the target network under a second parameter combination of device parameters to obtain a second network performance.

[0137] In step S603 of some embodiments, by comparing the difference between the first network performance and the network performance indicator with the difference between the second network performance and the network performance indicator, a candidate network performance that is closest to or equal to the network performance indicator can be obtained, and then a candidate parameter combination can be obtained based on the parameter combination corresponding to the candidate network performance.

[0138] In steps S601 to S603 shown in this embodiment, by comparing the network performance of the first parameter combination with the network performance of the second parameter combination, a candidate parameter combination whose network performance is closest to the network performance indicator or equal to the network performance indicator is selected from the first parameter combination and the second parameter combination, thereby improving the accuracy of device parameter configuration.

[0139] In step S504 of some embodiments, after obtaining the candidate parameter combination, the network performance of the candidate parameter combination is compared with the network performance of the alternative parameter combination selected by the target network in the previous round of device parameter configuration, and the target parameter value combination whose network performance is closest to the network performance index or equal to the network performance index can be screened out from the candidate parameter combination and the preset alternative parameter combination.

[0140] It should be noted that if there is no candidate parameter combination, the candidate parameter combination obtained in the current round needs to be used as the candidate parameter combination for the next round.

[0141] In steps S501 to S504 shown in this embodiment, Gibbs sampling is performed on the device parameters according to the edge probability distribution to obtain a first parameter combination, greedy sampling is performed on the device parameters to obtain a second parameter combination, and then candidate parameter combinations are screened out from the first parameter combination and the second parameter combination according to the network performance evaluation function and the network performance index. Finally, a target parameter value combination is screened out from the candidate parameter combination and the alternative parameter combination, which can improve the diversity of device parameter selection, thereby improving the probability of obtaining a target parameter value combination whose network performance is closest to the network performance index or equal to the network performance index.

[0142] The embodiment of the present application constructs a target Markov random field by utilizing the acquired device quantity, device association relationship and device parameters of the devices in the target network, thereby improving the parameter configuration efficiency of a large-scale network. Furthermore, spanning tree sampling is performed on the target Markov random field to obtain multiple uniform random spanning trees, and belief propagation is performed on the multiple uniform random spanning trees, thereby breaking the closed-loop state of the target Markov random field and enabling belief propagation to converge, thereby obtaining accurate device parameter configuration and improving the efficiency of device parameter configuration in large-scale networks.

[0143] See also Figure 7 The embodiment of the present application also provides a device parameter configuration apparatus, which can implement the above-mentioned device parameter configuration method, and the apparatus includes:

[0144] The network device information acquisition module 701 is used to obtain the device quantity, device association relationship, device parameters and network performance indicators of the target network.

[0145] A random field construction module 702 is used to construct a target Markov random field based on the number of devices, device associations, and device parameters, wherein the target Markov random field includes a plurality of device nodes;

[0146] A spanning tree sampling module 703 is used to perform spanning tree sampling on the target Markov random field to obtain multiple uniform random spanning trees;

[0147] The spanning tree belief propagation module 704 is used to perform belief propagation on the uniform random spanning tree to obtain the marginal probability distribution of the parameters of the device nodes;

[0148] The edge probability calculation module 705 is used to perform probability merging on the parameter edge probability distribution to obtain the target edge probability distribution of each device node;

[0149] The device parameter configuration module 706 is used to configure the device parameters based on the edge probability distribution, the network performance index and the preset network performance evaluation function to obtain the target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

[0150] The specific implementation of the device parameter configuration apparatus is substantially the same as the specific implementation of the device parameter configuration method described above, and will not be described in detail herein.

[0151] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above device parameter configuration method when executing the computer program. The electronic device can be any smart terminal including a tablet computer, a car computer, etc.

[0152] See also Figure 8 , Figure 8 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0153] The processor 801 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0154] The memory 802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 802, and the processor 801 calls and executes the device parameter configuration method of the embodiment of this application;

[0155] Input / output interface 803, used to implement information input and output;

[0156] The communication interface 804 is used to realize the communication interaction between the device and other devices. The communication can be realized through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.);

[0157] A bus 805 that transmits information between various components of the device (e.g., the processor 801, the memory 802, the input / output interface 803, and the communication interface 804);

[0158] The processor 801 , the memory 802 , the input / output interface 803 and the communication interface 804 are connected to each other in communication within the device via a bus 805 .

[0159] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned device parameter configuration method is implemented.

[0160] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0161] The device parameter configuration method, device parameter configuration apparatus, electronic device and storage medium provided in the embodiments of the present application can improve the parameter configuration efficiency of large-scale networks by constructing a target Markov random field using the acquired device quantity, device association relationship and device parameters of the devices in the target network. Furthermore, spanning tree sampling is performed on the target Markov random field to obtain multiple uniform random spanning trees, and belief propagation is performed on the multiple uniform random spanning trees, thereby breaking the closed-loop state of the target Markov random field and enabling belief propagation to converge, thereby obtaining accurate device parameter configuration and improving the efficiency of device parameter configuration in large-scale networks.

[0162] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0163] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0164] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0165] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0166] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0167] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0168] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. The mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0169] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0170] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0171] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0172] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A device parameter configuration method, characterized in that: The method comprises: Obtaining the number of devices in the target network, device association relationships, device parameters, and network performance indicators of the target network; Based on the number of devices, the device association relationship, and the device parameters, a target Markov random field is constructed, wherein the target Markov random field includes a plurality of device nodes; Performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees, wherein the uniform random spanning trees include spanning tree nodes; Perform belief propagation on each of the uniform random spanning trees to obtain a parameter marginal probability distribution of the spanning tree nodes on each of the uniform random spanning trees; Probabilistically merging the parameter edge probability distributions to obtain a target edge probability distribution for each of the device nodes; Based on the target edge probability distribution, the network performance index and a preset network performance evaluation function, the device parameters are configured to obtain a target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

2. The method according to claim 1, characterized in that The target Markov random field includes device association edges, and the target Markov random field is subjected to spanning tree sampling to obtain a plurality of uniform random spanning trees, including: Calculating the equivalent resistance value of the device-associated edge to obtain an estimated probability of the device-associated edge, wherein the estimated probability represents the probability of the device-associated edge participating in constructing a uniform random spanning tree; Based on the estimated probability and the device node, uniformly and randomly sample the device-related edges to obtain target relationship edges; Based on the device nodes and the target relationship edges, a plurality of the uniformly randomly generated trees are constructed.

3. The method according to claim 2, characterized in that The uniform random spanning tree includes a spanning tree edge, and the belief propagation is performed on each of the uniform random spanning trees to obtain the parameter edge probability distribution of the spanning tree nodes on each of the uniform random spanning trees, including: Obtaining a first mapping relationship and a second mapping relationship, wherein the first mapping relationship represents a mapping relationship between a value combination of the device parameter and a state of the target network, and the second mapping relationship is used to represent a mapping relationship between a parameter value of the device parameter and a cost consumption of a device in the target network; Based on the first mapping relationship and the second mapping relationship, calculating the node joint probability distribution of the spanning tree nodes on each of the uniform random spanning trees; Calculating a connection edge belief of each spanning tree edge based on the node joint probability distribution of two spanning tree nodes on the spanning tree edge; Based on the estimated probability, modifying the connection edge belief to obtain a modified belief; The modified belief is propagated to obtain the marginal probability distribution of the parameters of the spanning tree nodes.

4. The method according to claim 3, characterized in that The step of modifying the connection edge belief based on the estimated probability to obtain the modified belief includes: The device nodes are numbered to obtain device node numbers, and the tree node numbers of the spanning tree nodes on each of the uniform random spanning trees are determined according to the device node numbers; Based on the tree node number, the revised belief is calculated using the following formula: ； in, represents the modified belief of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j, Indicates that the tree node is numbered i, Indicates that the tree node number is j, represents the estimated probability of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j, Represents the connecting edge belief of the spanning tree edge between the spanning tree node with the tree node number i and the spanning tree node with the tree node number j.

5. The method according to claim 1, characterized in that The probabilistic merging of the parameter edge probability distributions to obtain the target edge probability distribution of each device node includes: Searching for the spanning tree node of each of the device nodes in each of the uniform random spanning trees to obtain a tree node set, wherein the tree node set includes one or more of the spanning tree nodes; The parameter edge probability distributions of the spanning tree nodes in the tree node set are multiplied and merged to obtain the target edge probability distribution.

6. The method according to any one of claims 1 to 5, characterized in that: The configuring of the device parameters based on the target edge probability distribution, the network performance index and the preset network performance evaluation function to obtain a target parameter value combination includes: Based on the target edge probability distribution, performing Gibbs sampling on the device parameters to obtain a first parameter combination; Based on the target edge probability distribution, greedily sample the device parameters to obtain a second parameter combination; Based on the network performance evaluation function and the network performance indicator, screening out a candidate parameter combination from the first parameter combination and the second parameter combination; The target parameter value combination is screened out from the candidate parameter combination and the preset alternative parameter combination, wherein the alternative parameter combination refers to the parameter combination selected by the target network in the previous round of parameter configuration.

7. The method according to claim 6, characterized in that The selecting a candidate parameter combination from the first parameter combination and the second parameter combination based on the network performance evaluation function and the network performance indicator comprises: Substituting the first parameter combination into the network performance evaluation function to obtain a first network performance; Substituting the second parameter combination into the network performance evaluation function to obtain a second network performance; Based on the network performance indicator, the first network performance and the second network performance, the candidate parameter combination is determined from the first parameter combination and the second parameter combination, wherein the network performance of the candidate parameter combination is closest to the network performance indicator.

8. A device parameter configuration device, characterized in that: The device comprises: A network device information acquisition module is used to obtain the number of devices in the target network, device association relationships, device parameters and network performance indicators of the target network; A random field construction module, used to construct a target Markov random field based on the number of devices, the device association relationship, and the device parameters, wherein the target Markov random field includes a plurality of device nodes; A spanning tree sampling module, used for performing spanning tree sampling on the target Markov random field to obtain a plurality of uniform random spanning trees, wherein the uniform random spanning trees include spanning tree nodes; A spanning tree belief propagation module, used to perform belief propagation on each of the uniform random spanning trees to obtain the parameter marginal probability distribution of the spanning tree nodes on each of the uniform random spanning trees; An edge probability calculation module, used for probabilistically merging the parameter edge probability distributions to obtain a target edge probability distribution for each of the device nodes; The device parameter configuration module is used to configure the device parameters based on the target edge probability distribution, the network performance indicators and a preset network performance evaluation function to obtain a target parameter value combination, wherein the network performance evaluation function is used to characterize the mapping relationship between the value combination of the device parameters and the performance of the target network.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the device parameter configuration method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the device parameter configuration method according to any one of claims 1 to 7 is implemented.

Images