Dynamic arrangement method and device

Abstract

The invention relates to the technical field of mobile communication, in particular to a dynamic arrangement method and device. The method comprises: in response to a received network configuration request, obtaining a service attribute identifier of each type of to-be-configured network resource information, the network configuration request carrying service information, device information and to-be-configured network resource information; acquiring a configuration instruction set according to the equipment information, the to-be-configured network resource information and an instruction library; extracting real-time configuration information and the configuration instruction set by adopting a target neural network algorithm to obtain a configuration relation structure corresponding to the to-be-configured network resource information; and performing part-of-speech tagging and information retrieval on the real-time configuration information by adopting the configuration instruction set and the service attribute identifier, completing adjustment of the configuration relation structure, and outputting a dynamic instruction model. By adopting the method, the obtaining efficiency and the obtaining accuracy of the dynamic instruction can be improved, and the failure probability of issuing the dynamic instruction model can be reduced.

Description

Technical Field

[0001] The present disclosure relates to the field of mobile communication technology, and in particular to a dynamic scheduling method and device. Background Art

[0002] With the development of science and technology, network orchestration has emerged. Network orchestration refers to the orderly arrangement and organization of various network service units for the purpose of user needs to generate services that meet user requirements. For example, you can rely on manual combing of preset business models, analyze the rules of business models, and design a set of orchestration rules for each business based on the analyzed topological structure. However, since the preset business model cannot cover all scenarios and cannot be dynamically adapted according to the real-time situation of the existing network, the probability of failure to issue is high. When there are many types of business models, the workload of the analysis task is large and it is not easy to update, which makes the efficiency of network orchestration low. Summary of the invention

[0003] The present disclosure provides a dynamic arrangement method and device, which can improve the acquisition efficiency and accuracy of dynamic instructions and reduce the failure probability of dynamic instruction model issuance. The technical solution of the present disclosure is as follows:

[0004] According to a first aspect of an embodiment of the present disclosure, a dynamic scheduling method is provided, including:

[0005] In response to the received network configuration request, obtaining a service attribute identifier of each type of network resource information to be configured, wherein the network configuration request carries service information, device information, and network resource information to be configured;

[0006] Acquire a configuration instruction set according to the device information, the network resource information to be configured and an instruction library;

[0007] Using a target neural network algorithm to extract the real-time configuration information and the configuration instruction set, and obtaining a configuration relationship structure corresponding to the network resource information to be configured;

[0008] The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

[0009] According to some embodiments, the method further comprises:

[0010] According to the network resource information to be configured, the device information and the business scenario information, the business attribute identifier of each type of network resource information to be configured is mapped to obtain tokens data corresponding to the business attribute identifier.

[0011] According to some embodiments, acquiring a configuration instruction set according to the device information, the network resource information to be configured and an instruction library includes:

[0012] According to the device information and the network resource information to be configured, obtaining a mapping relationship set corresponding to the network resource information to be configured in an instruction library;

[0013] The device information, the to-be-configured network resource information and the mapping relationship set are used to acquire a configuration instruction set.

[0014] According to some embodiments, the method further comprises:

[0015] Acquire an initial neural network model and training data, wherein the training data includes configuration information of at least one managed device, a historical configuration instruction set, and at least one historical dynamic instruction model;

[0016] Obtaining a network weight and a threshold value of the initial neural network at a first moment, wherein the first moment is a moment of initializing the initial neural network;

[0017] Obtaining input information and output information of neurons in each layer of the initial neural network model;

[0018] According to the output information of the neurons in each layer, an energy function corresponding to the initial neural network model is obtained;

[0019] According to the energy function, the input information and the output information of the neurons in each layer, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold;

[0020] Determine a network weight and a threshold at a second moment according to the energy function, the first adjustment amount, and the second adjustment amount, wherein the second moment is adjacent to the first moment and close to the current moment;

[0021] According to the network weights and thresholds at the second moment, the initial neural network model is adjusted to obtain a target neural network model.

[0022] According to some embodiments, wherein each layer of neurons includes three layers of neurons, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold value according to the energy function and the input information and output information of each layer of neurons includes:

[0023] Obtaining a first adjustment amount of a network weight between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, and the adjustment coefficient;

[0024] Obtaining a second adjustment amount of a threshold between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the all-1 vector and the adjustment coefficient;

[0025] Obtaining a first adjustment amount of a network weight between the second-layer neurons and the first-layer neurons according to the output data of the initial neural network model, the output information of the third-layer neurons, the output information of the second-layer neurons, the input data of the initial neural network model and the adjustment coefficient;

[0026] According to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, the input data of the initial neural network model, the all-1 vector and the adjustment coefficient, a second adjustment amount of the threshold between the second layer of neurons and the first layer of neurons is obtained.

[0027] According to some embodiments, the step of using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, completing the adjustment of the configuration relationship structure, and outputting a dynamic instruction model includes:

[0028] Input the real-time configuration information and the configuration relationship structure into the Word2Vec model, obtain first word segmentation vector representation information and first part-of-speech vector representation information corresponding to the real-time configuration information, and obtain second word segmentation vector representation information and second part-of-speech vector representation information corresponding to the configuration relationship structure;

[0029] According to the first word segmentation vector representation information, the first part-of-speech vector representation information, the second word segmentation vector representation information, and the second part-of-speech vector representation information, obtaining the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure;

[0030] According to the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure, the target position corresponding to the implementation configuration information and the configuration relationship structure is obtained, wherein the target position is the position with the highest similarity between the position information corresponding to the real-time configuration information and the position information corresponding to the configuration relationship structure;

[0031] Obtaining a word segmentation vector representation corresponding to the target position;

[0032] Obtaining a resource conflict identification result according to the business attribute identifier corresponding to the word segmentation vector representation;

[0033] When the resource conflict identification result indicates that there is no resource conflict, the configuration relationship structure is adjusted and a dynamic instruction model is output.

[0034] According to some embodiments, the method further comprises:

[0035] Using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, to obtain processed real-time configuration information;

[0036] The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the configuration relationship structure to obtain a processed configuration relationship structure.

[0037] According to a second aspect of the present disclosure, there is provided

[0038] an identification acquisition unit, configured to acquire, in response to a received network configuration request, a service attribute identification of each type of network resource information to be configured, wherein the network configuration request carries service information, device information, and network resource information to be configured;

[0039] A set acquisition unit, used to acquire a configuration instruction set according to the device information, the network resource information to be configured and an instruction library;

[0040] A structure acquisition unit, used to extract the real-time configuration information and the configuration instruction set using a target neural network algorithm to acquire a configuration relationship structure corresponding to the network resource information to be configured;

[0041] The model output unit is used to use the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model. According to a third aspect of the embodiment of the present disclosure, a network device is provided, including:

[0042] processor;

[0043] a memory for storing instructions executable by the processor;

[0044] The processor is configured to execute the instructions to implement the dynamic orchestration method described in any one of the aforementioned aspects.

[0045] According to a fourth aspect of an embodiment of the present disclosure, a storage medium is provided. When instructions in the storage medium are executed by a processor of a network device, the network device is enabled to execute the dynamic orchestration method described in any one of the preceding aspects.

[0046] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program, wherein the computer program implements any one of the methods described in the preceding aspects when executed by a processor.

[0047] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0048] In some or related embodiments, by responding to the received network configuration request, obtaining the service attribute identifier of each type of network resource information to be configured, wherein the network configuration request carries service information, device information and network resource information to be configured; obtaining the configuration instruction set according to the device information, the network resource information to be configured and the instruction library; extracting the real-time configuration information and the configuration instruction set by using the target neural network algorithm, obtaining the configuration relationship structure corresponding to the network resource information to be configured; using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, completing the adjustment of the configuration relationship structure, and outputting a dynamic instruction model. Therefore, an adaptive, self-optimizing, and self-learning instruction arrangement mechanism can be provided, which can dynamically arrange the network resources to be configured through the network configuration request, and there is no need to manually comb the preset business model, which can reduce labor costs and, and can dynamically adjust the network resource information of the existing network according to the network configuration request, when the number of services is large or the business model changes, reduce the workload of manual analysis, reduce the complexity of instruction arrangement, update the dynamic instruction model, improve the acquisition efficiency and accuracy of dynamic instructions, and reduce the failure probability of dynamic instruction model issuance.

[0049] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0051] Figure 1 is a flow chart of a dynamic arrangement method according to an exemplary embodiment;

[0052] Figure 2 is a flow chart of a dynamic arrangement method according to an exemplary embodiment;

[0053] Figure 3 is a schematic diagram of a dynamic orchestration system according to an exemplary embodiment;

[0054] Figure 4is a schematic diagram of training a network model according to an exemplary embodiment;

[0055] Figure 5 is a block diagram of a dynamic arrangement device according to an exemplary embodiment;

[0056] Figure 6 The invention is a block diagram of a network device according to an exemplary embodiment. DETAILED DESCRIPTION

[0057] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0058] The embodiments of the present disclosure propose a dynamic orchestration method and apparatus. In some embodiments, the terms such as dynamic orchestration method, information processing method, communication method, etc. can be replaced with each other, the terms such as dynamic orchestration apparatus, information processing apparatus, communication apparatus, etc. can be replaced with each other, and the terms such as information processing system, communication system, etc. can be replaced with each other.

[0059] The embodiments of the present disclosure are not exhaustive, but are only illustrative of some embodiments, and are not intended to be a specific limitation on the scope of protection of the present disclosure. In the absence of contradiction, each step in a certain embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain embodiment can also be implemented as an independent embodiment, and the order of the steps in a certain embodiment can be arbitrarily exchanged. In addition, the optional implementation methods in a certain embodiment can be arbitrarily combined; in addition, the embodiments can be arbitrarily combined, for example, some or all of the steps of different embodiments can be arbitrarily combined, and a certain embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0060] In each embodiment of the present disclosure, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between the embodiments are consistent and can be referenced to each other, and the technical features in different embodiments can be combined to form a new embodiment based on their internal logical relationships.

[0061] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure.

[0062] In the embodiments of the present disclosure, unless otherwise specified, elements expressed in the singular form, such as "a", "an", "the", "above", "said", "aforementioned", "this", etc., may mean "one and only one", or "one or more", "at least one", etc. For example, when using articles such as "a", "an", "the" in English in translation, the noun after the article may be understood as a singular expression or a plural expression.

[0063] In the embodiments of the present disclosure, “plurality” refers to two or more.

[0064] In some embodiments, the terms “at least one,” “one or more,” “a plurality of,” “multiple,” etc. may be used interchangeably.

[0065] The prefixes such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different description objects, and do not constitute restrictions on the position, order, priority, quantity or content of the description objects. The statement of the description object refers to the description in the context of the claims or embodiments, and should not constitute unnecessary restrictions due to the use of prefixes. For example, if the description object is a "field", the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields", and the "first" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the description object is a "level", the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of description objects is not limited by the ordinal number, and can be one or more. Taking the "first device" as an example, the number of "devices" can be one or more. In addition, the objects modified by different prefixes may be the same or different. For example, if the description object is "device", then the "first device" and the "second device" may be the same device or different devices, and their types may be the same or different. For another example, if the description object is "information", then the "first information" and the "second information" may be the same information or different information, and their contents may be the same or different.

[0066] In some embodiments, "terminal" or "terminal device" can be referred to as "user equipment (UE)", "user terminal" "mobile station (MS)", "mobile terminal (MT)", subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, etc.

[0067] In some embodiments, data, information, etc. may be obtained with the user's consent.

[0068] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure 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 interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0069] Figure 1 is a flow chart of a dynamic arrangement method according to an exemplary embodiment. Figure 1 As shown, the dynamic orchestration method can be used to determine the root cause information scenario corresponding to the abnormal information when abnormal information occurs in the system, including the following steps:

[0070] In step S11, in response to the received network configuration request, a service attribute identifier of each type of network resource information to be configured is obtained, wherein the network configuration request carries service information, device information and network resource information to be configured;

[0071] According to some embodiments, the execution subject of the embodiments of the present disclosure is a network device. The name of the network device is not limited. For example, the network device can also be called a server, or a server cluster, etc.

[0072] In some embodiments, the network configuration request may be, for example, a request sent by a terminal for configuring network resource information. The network configuration request does not specifically refer to a fixed request. For example, when the network configuration request sending end changes, the network configuration request may also change accordingly. For example, the terminal sending the network configuration request may specifically be a business order application end. For example, when the request time point of the network configuration request changes, the network configuration request may also change accordingly. For example, when the information included in the network configuration request changes, the network configuration request may also change accordingly.

[0073] According to some embodiments, the network resource information may be used to indicate network resources, for example. The network resource information does not specifically refer to a fixed resource information. For example, when the resource type included in the network resource information changes, the network resource information may also change accordingly. For example, when the amount of information corresponding to the network resource information changes, the network resource information may also change accordingly.

[0074] In some embodiments, the service attribute identifier can be used to uniquely identify the service attribute. The service attribute does not specifically refer to a fixed service attribute or a type of service attribute. For example, when the method for determining the service attribute changes, the service attribute identifier can also change accordingly. For example, when the classification method for network resource information changes, the service attribute identifier can also change accordingly.

[0075] According to some embodiments, the service information may be used to indicate the service information executed by the current network configuration request sender, or information about the service corresponding to the network resource information to be configured. The service information may include, for example, a service interface, a service interface description, a physical gateway address, etc.

[0076] In some embodiments, the device information may be used to indicate information of the device that sends the network configuration request. The device information includes, but is not limited to, device identification, device type, device manufacturer, etc. The device information does not specifically refer to a fixed information. For example, when the amount of information included in the device information changes, the device information may also change accordingly. For example, when the specific information corresponding to the device information changes, the device information may also change accordingly.

[0077] According to some embodiments, the network resource information to be configured may be used to indicate network resource information that has not been configured and needs to be configured. The network resource information to be configured does not specifically refer to a fixed network resource information. For example, when the network configuration request changes, the network resource information to be configured may also change accordingly. Different categories may correspond to different network resource information to be configured.

[0078] According to some embodiments, in response to a received network configuration request, a service attribute identifier of each type of network resource information to be configured is obtained, wherein the network configuration request carries service information, device information, and network resource information to be configured.

[0079] In step S12, a configuration instruction set is obtained according to the device information, the network resource information to be configured and the instruction library;

[0080] In some embodiments, the instruction library may be, for example, a collection of at least one instruction. The instruction library does not specifically refer to a fixed library. For example, when the number of instructions corresponding to the instruction library changes, the instruction library may also change accordingly. For example, when a certain instruction in the instruction library changes, the instruction library may also change accordingly. For example, when a modification instruction for the instruction library is received, the instruction library may also change accordingly.

[0081] In some embodiments, the configuration instruction set may be, for example, a configuration instruction set obtained after translating the network resource information to be configured. The configuration instruction set may be, for example, a collection of at least one configuration instruction. The configuration instruction set does not specifically refer to a fixed set. For example, when the configuration instructions included in the configuration instruction set change, the configuration instruction set may also change accordingly. For example, when a configuration instruction in the configuration instruction set changes, the configuration instruction set may also change accordingly.

[0082] In some embodiments, a configuration instruction set is obtained based on the device information, the to-be-configured network resource information, and an instruction library.

[0083] In step S13, a target neural network algorithm is used to extract the real-time configuration information and the configuration instruction set to obtain a configuration relationship structure corresponding to the network resource information to be configured;

[0084] In some embodiments, the target neural network algorithm may be, for example, an algorithm that has been trained and can be used for information extraction. The target neural network algorithm does not specifically refer to a fixed algorithm. For example, when a parameter in the target neural network algorithm changes, the target neural network algorithm may also change accordingly. For example, when a training method corresponding to the target neural network algorithm changes, the target neural network algorithm may also change accordingly.

[0085] According to some embodiments, the implementation configuration information may be, for example, the actual configuration information of the device to be operated obtained through an interface. The implementation configuration information does not specifically refer to a fixed information. For example, when the time point of obtaining the implementation configuration information changes, the implementation configuration information also changes accordingly. For example, when the device to be operated changes, the implementation configuration information may also change accordingly.

[0086] In some embodiments, the coordination relationship structure may be composed of a group of ordered, view-level relationship configuration instructions. The coordination relationship structure does not specifically refer to a fixed structure. For example, when the relationship configuration instruction changes, the coordination relationship structure may also change accordingly.

[0087] According to some embodiments, a target neural network algorithm is used to extract the real-time configuration information and the configuration instruction set to obtain a configuration relationship structure corresponding to the network resource information to be configured.

[0088] In step S14, the configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the real-time configuration information, the configuration relationship structure is adjusted, and a dynamic instruction model is output.

[0089] According to some embodiments, part-of-speech tagging may be, for example, a process of tagging the part-of-speech after analyzing the implementation configuration information and the configuration relationship structure. The part-of-speech tagging process, for example, does not specifically refer to a fixed process. For example, different tagging methods may result in different part-of-speech tagging results.

[0090] According to some embodiments, information retrieval may be, for example, a process of retrieving real-time configuration information and configuration relationship structures.

[0091] In some embodiments, the configuration instruction set and the service attribute identifier may be used to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

[0092] In some or related embodiments, by responding to the received network configuration request, obtaining the service attribute identifier of each type of network resource information to be configured, wherein the network configuration request carries service information, device information and network resource information to be configured; obtaining the configuration instruction set according to the device information, the network resource information to be configured and the instruction library; extracting the real-time configuration information and the configuration instruction set by using the target neural network algorithm, obtaining the configuration relationship structure corresponding to the network resource information to be configured; using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, completing the adjustment of the configuration relationship structure, and outputting the dynamic instruction model. Therefore, the network resources to be configured can be dynamically arranged through the network configuration request, and there is no need to manually comb the preset business model, which can reduce the labor cost, and the network resource information of the existing network can be dynamically adjusted according to the network configuration request. When the number of services is large or the business model changes, the workload of manual analysis is reduced, the dynamic instruction model can be updated, the acquisition efficiency and accuracy of dynamic instructions can be improved, and the failure probability of dynamic instruction model issuance can be reduced.

[0093] Figure 2 is a flow chart of a dynamic arrangement method according to an exemplary embodiment. Figure 2 As shown, the dynamic arrangement method can be used in a dynamic arrangement scenario, and includes the following steps:

[0094] In step S21, in response to the received network configuration request, a service attribute identifier of each type of network resource information to be configured is obtained, wherein the network configuration request carries service information, device information and network resource information to be configured;

[0095] The specific process is as described above and will not be repeated here.

[0096] According to some embodiments, Figure 3 FIG. 1 is a schematic diagram of a dynamic scheduling system according to an exemplary embodiment. Figure 3 As shown, after the business order application receives the business information, device information and information of the network resources to be configured, it forwards the network configuration request to the resource identification module of the network device through the northbound interface to complete the business attribute identification of the network resources to be configured.

[0097] In step S22, according to the device information and the network resource information to be configured, a mapping relationship set corresponding to the network resource information to be configured is obtained in the instruction library;

[0098] The specific process is as described above and will not be repeated here.

[0099] In step S23, the device information, the to-be-configured network resource information and the mapping relationship set are used to obtain a configuration instruction set;

[0100] The specific process is as described above and will not be repeated here.

[0101] According to some embodiments, the method further comprises:

[0102] According to the network resource information to be configured, the device information and the business scenario information, the business attribute identifier of each type of network resource information to be configured is mapped to obtain tokens data corresponding to the business attribute identifier.

[0103] According to some embodiments, the resource identification layer of the network device can identify the business attributes of each type of network resource to be configured so that the instruction configuration model can process and understand them; the classification dimensions corresponding to the information of the network resource to be configured may include business interface, gateway interface, physical gateway address, customer IP address, VLAN, etc. Among them, for the same type of network resources, the corresponding business attributes are different for different business scenarios and different device types. For example, for the "business interface", device A can support the configuration data of multiple customers to be connected to the same "business interface", while device B only supports a single customer to have exclusive use of 1 "business interface"; in order for the instruction configuration model to calculate and process these business attributes, for example, these natural language-approximate business attributes can be mapped to tokens data, and the specific mapping method is as follows:

[0104] BIZ i = {R j , D k , S p},(i,j,k,p∈[1,2,3...N]}

[0105] Among them, BIZ i Represents a set of business attributes; R j Indicates a network resource set; D k Indicates a device type set; S p Represents a set of business scenarios.

[0106] According to some embodiments, a configuration instruction set is mapped based on an instruction library according to device information and network resource information to be configured; wherein the instruction library is used to store configuration instructions for various types of devices in the entire network; the configuration instruction set can be, for example, a batch of unordered, non-view-level, and non-ambiguous instruction sets, and the mapping method is:

[0107] INS q1 = {R j1 , D k}={<network resource configuration instruction 1>};

[0108] INS q2 = {R j2 , D k} = {<network resource configuration instruction 2.1> | <network resource configuration instruction 2.2>};

[0109] INS q3 = {R j3 , D k}={<network resource configuration instruction 3>, <network resource configuration instruction 4>};

[0110] Among them, INSq is the configuration instruction set of the current network resource to be configured; Rj represents the network resource set; Dk represents the device type set; “|” indicates that the network resource corresponds to multiple configuration instructions;

[0111] For example, for device A access service, the network resources to be configured on the BNG side are service interface, service interface description, physical gateway address, VLAN, customer IP address, direct routing strategy, etc. Through the device information and the network resource information to be configured, the configuration instruction set mapped is:

[0112] INS1={R1,D1}={<interface[Eth-Trunk108.1603]>};

[0113] INS2 = {R2, D1} = {<description[FoShanSanShui]>};

[0114] INS3 = {R3, D1} = {<ip address[183.238.198.33 255.255.255.248]|ip address[183.238.198.33 255.255.255.248]secondary>};

[0115] INS4 = {R4, D1} = {<qinq termination pe-vid

[1603] ce-vid

[3004] >};

[0116] INS5 = {R5, D1} = {<dhcp snooping bind-table static ip-address[183.238.198.34]vlan

[1603] ce-vlan

[3004] >};

[0117] INS6 = {R6, D1} = {<ip ip-prefix[Direct_Route]index[]permit[183.233.6.22832]>};

[0118] It should be noted that under the same device, the same type of network resources may correspond to multiple configuration instructions. For example, the instructions for configuring the physical gateway of device A are:<ip address[]> and<ip address[]secondary> The instruction policy library is only responsible for mapping out the configuration instructions of network resources, and is not responsible for clarifying unique instructions.

[0119] In step S24, a target neural network algorithm is used to extract the real-time configuration information and the configuration instruction set to obtain a configuration relationship structure corresponding to the network resource information to be configured;

[0120] The specific process is as described above and will not be repeated here.

[0121] According to some embodiments, the method further comprises:

[0122] Acquire an initial neural network model and training data, wherein the training data includes configuration information of at least one managed device, a historical configuration instruction set, and at least one historical dynamic instruction model;

[0123] Obtaining a network weight and a threshold value of the initial neural network at a first moment, wherein the first moment is a moment of initializing the initial neural network;

[0124] Obtaining input information and output information of neurons in each layer of the initial neural network model;

[0125] According to the output information of the neurons in each layer, an energy function corresponding to the initial neural network model is obtained;

[0126] According to the energy function, the input information and the output information of the neurons in each layer, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold;

[0127] Determine a network weight and a threshold at a second moment according to the energy function, the first adjustment amount, and the second adjustment amount, wherein the second moment is adjacent to the first moment and close to the current moment;

[0128] According to the network weight and threshold at the second moment, the initial neural network model is adjusted to obtain the target neural network model. Therefore, the target neural network model can be obtained according to the initial neural network model and training data, thereby improving the accuracy of model acquisition.

[0129] According to some embodiments, wherein each layer of neurons includes three layers of neurons, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold value according to the energy function, the input information and the output information of each layer of neurons includes:

[0130] Obtaining a first adjustment amount of a network weight between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, and the adjustment coefficient;

[0131] Obtaining a second adjustment amount of a threshold between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the all-1 vector and the adjustment coefficient;

[0132] Obtaining a first adjustment amount of a network weight between the second-layer neurons and the first-layer neurons according to the output data of the initial neural network model, the output information of the third-layer neurons, the output information of the second-layer neurons, the input data of the initial neural network model and the adjustment coefficient;

[0133] According to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, the input data of the initial neural network model, the all-1 vector and the adjustment coefficient, a second adjustment amount of the threshold between the second layer of neurons and the first layer of neurons is obtained.

[0134] According to some embodiments, based on the model adaptation layer of the network device, the configuration relationship structure of the network resources to be configured can be learned and extracted through a neural network algorithm; an ordered configuration instruction set is output through the neural network to automatically realize the routing selection between the devices to be configured.

[0135] According to some embodiments, the BP neural network learns the configuration schemes of a large number of existing network devices, and can extract the configuration relationship structure of the corresponding network resources for the configuration information of a certain device; the training process of the BP neural network is as follows:

[0136] 1. Determine the BP network structure and training data:

[0137] The configuration information of all managed devices in the network is obtained, and the device configuration information and configuration instruction set are set as the input samples of the BP neural network, and the dynamic instruction model is set as the output sample of the BP neural network to form a 5-input 1-output training network; the maximum number of training times is set to 5000, the number of hidden layer neurons is 5, the network learning rate is 0.035, the target error of training is 0.65x10-3, and the momentum factor is added by default. The obtained training data can be shown in Table 1, for example.

[0138] Table 1

[0139]

[0140]

[0141] 2. Initialize network weights and thresholds

[0142] The network device can set the preconditions: ① Set the training steps: train in batches; ② Set the threshold to be similar to the weight, that is, treat the threshold as a random number with a sample input of 1; ③ Transpose the original training data to input by row; ④ Initialize the network and assign a random number matrix to the network weights and thresholds;

[0143] Because it is a network result of 2 input factors, 3 hidden layer neurons and 1 output, the weight Wij(t) of the first and second layers is set to a 2*3 random number matrix; the threshold Bij(t) of the second layer neurons is set to a 3*1 matrix; the weight Wjk(t) between the second and third layers is set to a 3*1 random matrix; and the threshold Bjk(t) of the third layer is set to a 1*1 matrix.

[0144] 3. Calculate the input and output of neurons

[0145] The input and output of the first layer are equal to the value of the actual input sample, O1=X (X is the input sample, O1 is the first layer output sample), which is a 2*30 matrix.

[0146] The input of the second layer of neurons, I2, is equal to the sum of the values ​​of all neurons in the first layer and the threshold, that is:

[0147] I2=Wij*X+Bij*ones, which is a 3*30 matrix, where ones is a matrix with all 1s;

[0148] The output of the second layer of neurons, It is a 2*3 data matrix;

[0149] The input of the third layer of neurons, I3 = Wjk*O2+Bjk*ones, is a 1*30 matrix;

[0150] The output of the third layer of neurons is O3=I3.

[0151] 4. Calculate the energy function E

[0152] The purpose of calculating the energy function (the sum of squared errors between the network output and the actual output samples) is to stop training the network when the predetermined error is reached. Assuming that the actual output sample is Y, the definition of E shows that:

[0153] E=∑(Y-O3) 2

[0154] 5. Calculate the adjustment of weights and thresholds between the second and third layers

[0155]

[0156] 6. Calculate the weight and threshold adjustment between the first and second layers

[0157] According to the idea of ​​error back propagation, the weight and threshold adjustment between the second and third layers are calculated first, and then the weight and threshold adjustment between the first and second layers are calculated;

[0158]

[0159] 7. Calculate the adjusted weights and thresholds to obtain the optimal BP neural network model

[0160] The weights and thresholds of each layer at time t are added with their respective adjustments and assigned to the weights and thresholds at time t+1:

[0161]

[0162] In step S25, the configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

[0163] The specific process is as described above and will not be repeated here.

[0164] According to some embodiments, real-time configuration information of the device to be operated can be obtained through the southbound interface; the real-time configuration information and configuration instruction set of the device to be operated are input into the trained BP neural network to extract the configuration relationship structure of the network resources to be configured; the configuration relationship structure is composed of a batch of ordered configuration instructions with a view hierarchy relationship.

[0165] According to some embodiments, Figure 4 FIG. 1 is a schematic diagram showing a training of a network model according to an exemplary embodiment. Figure 4As shown, in order to improve the accuracy of configuration relationship structure extraction, the configuration of all network devices is collected at a daily granularity, and the training samples are updated to enable the BP network to relearn.

[0166] According to some embodiments, the step of using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, completing the adjustment of the configuration relationship structure, and outputting a dynamic instruction model includes:

[0167] Input the real-time configuration information and the configuration relationship structure into the Word2Vec model, obtain first word segmentation vector representation information and first part-of-speech vector representation information corresponding to the real-time configuration information, and obtain second word segmentation vector representation information and second part-of-speech vector representation information corresponding to the configuration relationship structure;

[0168] According to the first word segmentation vector representation information, the first part-of-speech vector representation information, the second word segmentation vector representation information, and the second part-of-speech vector representation information, obtaining the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure;

[0169] According to the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure, the target position corresponding to the implementation configuration information and the configuration relationship structure is obtained, wherein the target position is the position with the highest similarity between the position information corresponding to the real-time configuration information and the position information corresponding to the configuration relationship structure;

[0170] Obtaining a word segmentation vector representation corresponding to the target position;

[0171] Obtaining a resource conflict identification result according to the business attribute identifier corresponding to the word segmentation vector representation;

[0172] When the resource conflict identification result indicates that there is no resource conflict, the configuration relationship structure is adjusted and a dynamic instruction model is output.

[0173] According to some embodiments, the method further comprises:

[0174] Using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, to obtain processed real-time configuration information;

[0175] The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the configuration relationship structure to obtain a processed configuration relationship structure.

[0176] According to some embodiments, based on the model self-optimization layer of the network device, the real-time configuration information of the device to be operated can be tagged with parts of speech and information retrieved in combination with the service attributes and configuration instruction set of the network resources to be configured, the configuration relationship structure can be adjusted, and the dynamic instruction model can be output. Therefore, the instruction arrangement can be combined with AI algorithms, part of speech tagging and information retrieval technology, which can improve the matching of the dynamic instruction model with the existing network configuration, improve the existing network adaptation in service activation, and have a high configuration accuracy, reducing the new service launch cycle.

[0177] The real-time configuration information and configuration relationship structure of the operating device are segmented and tagged with parts of speech respectively, and then the real-time configuration information and the configuration relationship structure are input into the Word2Vec model respectively to obtain the segmentation vector representation and the part of speech vector representation of the real-time configuration information and the configuration relationship structure; according to the segmentation vector representation and the part of speech vector representation of the real-time configuration information and the configuration relationship structure, the position information and the similarity matrix corresponding to the real-time configuration information and the configuration relationship structure are obtained; finally, the similarity of the real-time configuration information and the configuration relationship structure is judged, and the position in the real-time configuration information that is highly similar to the configuration relationship structure is found, and the highly similar segmentations are extracted to judge the business attributes. If a resource conflict is judged according to the business attributes, a conflict prompt is output and the output of the dynamic instruction model is refused; if it is judged according to the business attributes that the resources do not conflict, the configuration relationship structure is further adjusted.

[0178] According to some embodiments, in an Internet dedicated line shared gateway service scenario, the service interface of device A supports accessing configuration information of multiple customers, and determines based on word segmentation annotation and service attributes that the service interface to be configured has been configured on the device to be operated.

[0179] According to some embodiments, the configuration model for the A device is as follows:

[0180] INS1={R1,D1}={<interface[Eth-Trunk108.1603]>}

[0181] INS2 = {R2, D1} = {<description[FoShanSanShui]>}

[0182] INS3 = {R3, D1} = {<ip address[183.238.198.33 255.255.255.248]|ipaddress[183.238.198.33 255.255.255.248]secondary>}

[0183] INS4 = {R4, D1} = {<qinq termination pe-vid

[1603] ce-vid

[3004] >}

[0184] INS5 = {R5, D1} = {<dhcp snooping bind-table static ip-address[183.238.198.34]vlan

[1603] ce-vlan

[3004] >}

[0185] INS6 = {R6, D1} = {<ip ip-prefix[Direct_Route]index[]permit[183.233.6.228 32]>};

[0186] According to some embodiments, the vector model fine-tunes the instruction and the corresponding operation data and then converts the above instruction into a vector, which may be specifically shown in Table 2, for example.

[0187] Table 2

[0188]

[0189] According to some embodiments, the A device has been configured with the following instructions:

[0190] INS1={R1,D1}={<interface[Eth-Trunk108.1603]>}

[0191] INS2 = {R2, D1} = {<description[FoShanSanShui]>}

[0192] Then the identification needs to be configured to enter the sub-interface operation, and the conflict configuration in the instruction set needs to be INS2 = {R2, D1} = {<description[]>} discarded, and finally output the optimal dynamic instruction model. As follows:

[0193] INS1={R1,D1}={<interface[Eth-Trunk108.1603]>}

[0194] INS3 = {R3, D1} = {<ip address[183.238.198.33 255.255.255.248]|ipaddress[183.238.198.33 255.255.255.248]secondary>}

[0195] INS4 = {R4, D1} = {<qinq termination pe-vid

[1603] ce-vid

[3004] >}

[0196] INS5 = {R5, D1} = {<dhcp snooping bind-table static ip-address[183.238.198.34]vlan

[1603] ce-vlan

[3004] >}

[0197] INS6 = {R6, D1} = {<ip ip-prefix[Direct_Route]index[]permit[183.233.6.228 32]>}

[0198] In one or related embodiments, a mapping relationship set corresponding to the network resource information to be configured can be obtained in the instruction library according to the device information and the network resource information to be configured; the configuration instruction set is obtained using the device information, the network resource information to be configured and the mapping relationship set; therefore, the mapping relationship set can be screened according to the instruction library to improve configuration efficiency and accuracy, reduce single-order jamming, and improve the accuracy of obtaining the configuration instruction set and the accuracy of obtaining the dynamic orchestration model.

[0199] A block diagram of a dynamic arrangement device is shown according to an exemplary embodiment. Figure 5 , the device 500 comprises:

[0200] The identification acquisition unit 501 is used to obtain the service attribute identification of each type of network resource information to be configured in response to the received network configuration request, wherein the network configuration request carries service information, device information and network resource information to be configured;

[0201] A set acquisition unit 502 is used to acquire a configuration instruction set according to the device information, the network resource information to be configured and the instruction library;

[0202] The structure acquisition unit 503 is used to extract the real-time configuration information and the configuration instruction set using a target neural network algorithm to obtain a configuration relationship structure corresponding to the network resource information to be configured;

[0203] The model output unit 504 is used to use the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

[0204] According to some embodiments, the identification acquisition unit 501 is further configured to:

[0205] According to the network resource information to be configured, the device information and the business scenario information, the business attribute identifier of each type of network resource information to be configured is mapped to obtain tokens data corresponding to the business attribute identifier.

[0206] According to some embodiments, the set acquisition unit 502 is used to acquire the configuration instruction set according to the device information, the network resource information to be configured and the instruction library, specifically to:

[0207] According to the device information and the network resource information to be configured, obtaining a mapping relationship set corresponding to the network resource information to be configured in an instruction library;

[0208] The device information, the to-be-configured network resource information and the mapping relationship set are used to acquire a configuration instruction set.

[0209] According to some embodiments, the model output unit 504 is further configured to:

[0210] Acquire an initial neural network model and training data, wherein the training data includes configuration information of at least one managed device, a historical configuration instruction set, and at least one historical dynamic instruction model;

[0211] Obtaining a network weight and a threshold value of the initial neural network at a first moment, wherein the first moment is a moment of initializing the initial neural network;

[0212] Obtaining input information and output information of neurons in each layer of the initial neural network model;

[0213] According to the output information of the neurons in each layer, an energy function corresponding to the initial neural network model is obtained;

[0214] According to the energy function, the input information and the output information of the neurons in each layer, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold;

[0215] Determine a network weight and a threshold at a second moment according to the energy function, the first adjustment amount, and the second adjustment amount, wherein the second moment is adjacent to the first moment and close to the current moment;

[0216] According to the network weights and thresholds at the second moment, the initial neural network model is adjusted to obtain a target neural network model.

[0217] According to some embodiments, wherein each layer of neurons includes three layers of neurons, the model output unit 504 is used to obtain a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold according to the energy function and the input information and output information of each layer of neurons, including:

[0218] Obtaining a first adjustment amount of a network weight between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, and the adjustment coefficient;

[0219] Obtaining a second adjustment amount of a threshold between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the all-1 vector and the adjustment coefficient;

[0220] Obtaining a first adjustment amount of a network weight between the second-layer neurons and the first-layer neurons according to the output data of the initial neural network model, the output information of the third-layer neurons, the output information of the second-layer neurons, the input data of the initial neural network model and the adjustment coefficient;

[0221] According to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, the input data of the initial neural network model, the all-1 vector and the adjustment coefficient, a second adjustment amount of the threshold between the second layer of neurons and the first layer of neurons is obtained.

[0222] According to some embodiments, the model output unit 504 is used to use the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information to complete the adjustment of the configuration relationship structure, and when outputting the dynamic instruction model, is specifically used to:

[0223] Input the real-time configuration information and the configuration relationship structure into the Word2Vec model, obtain first word segmentation vector representation information and first part-of-speech vector representation information corresponding to the real-time configuration information, and obtain second word segmentation vector representation information and second part-of-speech vector representation information corresponding to the configuration relationship structure;

[0224] According to the first word segmentation vector representation information, the first part-of-speech vector representation information, the second word segmentation vector representation information, and the second part-of-speech vector representation information, obtaining the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure;

[0225] According to the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure, the target position corresponding to the implementation configuration information and the configuration relationship structure is obtained, wherein the target position is the position with the highest similarity between the position information corresponding to the real-time configuration information and the position information corresponding to the configuration relationship structure;

[0226] Obtaining a word segmentation vector representation corresponding to the target position;

[0227] Obtaining a resource conflict identification result according to the business attribute identifier corresponding to the word segmentation vector representation;

[0228] When the resource conflict identification result indicates that there is no resource conflict, the configuration relationship structure is adjusted and a dynamic instruction model is output.

[0229] According to some embodiments, the model output unit 504 is further used to perform part-of-speech tagging and information retrieval on the real-time configuration information using the configuration instruction set and the service attribute identifier to obtain the processed real-time configuration information;

[0230] The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the configuration relationship structure to obtain a processed configuration relationship structure.

[0231] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0232] In some or related embodiments, an identification acquisition unit is used to respond to a received network configuration request to obtain a service attribute identification of each type of network resource information to be configured, wherein the network configuration request carries service information, device information and network resource information to be configured; a set acquisition unit is used to obtain a configuration instruction set according to the device information, the network resource information to be configured and the instruction library; a structure acquisition unit is used to extract the real-time configuration information and the configuration instruction set using a target neural network algorithm to obtain a configuration relationship structure corresponding to the network resource information to be configured; a model output unit is used to perform part-of-speech tagging and information retrieval on the real-time configuration information using the configuration instruction set and the service attribute identification, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model. Therefore, the network resources to be configured can be dynamically arranged through the network configuration request, and there is no need to manually comb the preset business model, which can reduce labor costs, and the network resource information of the existing network can be dynamically adjusted according to the network configuration request. When the number of services is large or the business model changes, the workload of manual analysis can be reduced, the dynamic instruction model can be updated, the acquisition efficiency and accuracy of dynamic instructions can be improved, and the failure probability of dynamic instruction model issuance can be reduced.

[0233] Figure 6 600 is a block diagram of a network device 600 provided in an embodiment of the present disclosure. For example, the network device 600 may be provided as a network device. Figure 6The network device 600 includes a processing component 622, which further includes at least one processor, and a memory resource represented by a memory 632 for storing instructions executable by the processing component 622, such as an application. The application stored in the memory 632 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 622 is configured to execute instructions to perform any of the methods described above in the network device.

[0234] The network device 600 may also include a power supply component 627 configured to perform power management of the network device 600, a wired or wireless network interface 650 configured to connect the network device 600 to a network, and an input / output (I / O) interface 658. The network device 600 may operate based on an operating system stored in the memory 632, such as Windows Server TM, Mac OS X TM, Unix TM, Linux TM, Free BSD TM or the like.

[0235] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0236] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0237] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0238] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0239] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0240] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0241] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0242] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A dynamic arrangement method, characterized in that: include: In response to the received network configuration request, obtaining a service attribute identifier of each type of network resource information to be configured, wherein the network configuration request carries service information, device information, and network resource information to be configured; Acquire a configuration instruction set according to the device information, the network resource information to be configured and an instruction library; Using a target neural network algorithm to extract the real-time configuration information and the configuration instruction set, and obtaining a configuration relationship structure corresponding to the network resource information to be configured; The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

2. The method according to claim 1, characterized in that The method further comprises: According to the network resource information to be configured, the device information and the business scenario information, the business attribute identifier of each type of network resource information to be configured is mapped to obtain tokens data corresponding to the business attribute identifier.

3. The method according to claim 1, characterized in that The acquiring a configuration instruction set according to the device information, the network resource information to be configured and the instruction library includes: According to the device information and the network resource information to be configured, obtaining a mapping relationship set corresponding to the network resource information to be configured in an instruction library; The device information, the to-be-configured network resource information and the mapping relationship set are used to acquire a configuration instruction set.

4. The method according to claim 1, characterized in that The method further comprises: Acquire an initial neural network model and training data, wherein the training data includes configuration information of at least one managed device, a historical configuration instruction set, and at least one historical dynamic instruction model; Obtaining a network weight and a threshold value of the initial neural network at a first moment, wherein the first moment is a moment of initializing the initial neural network; Obtaining input information and output information of neurons in each layer of the initial neural network model; According to the output information of the neurons in each layer, an energy function corresponding to the initial neural network model is obtained; According to the energy function, the input information and the output information of the neurons in each layer, obtaining a first adjustment amount of the network weight between two adjacent layers of neurons and a second adjustment amount of the threshold; Determine a network weight and a threshold at a second moment according to the energy function, the first adjustment amount, and the second adjustment amount, wherein the second moment is adjacent to the first moment and close to the current moment; According to the network weights and thresholds at the second moment, the initial neural network model is adjusted to obtain a target neural network model.

5. The method according to claim 4, characterized in that in, The layers of neurons include three layers of neurons, and obtaining the first adjustment amount of the network weight between two adjacent layers of neurons and the second adjustment amount of the threshold according to the energy function, the input information and the output information of the layers of neurons includes: Obtaining a first adjustment amount of a network weight between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, and the adjustment coefficient; Obtaining a second adjustment amount of a threshold between the second layer of neurons and the third layer of neurons according to the output data of the initial neural network model, the output information of the third layer of neurons, the all-1 vector and the adjustment coefficient; Obtaining a first adjustment amount of a network weight between the second-layer neurons and the first-layer neurons according to the output data of the initial neural network model, the output information of the third-layer neurons, the output information of the second-layer neurons, the input data of the initial neural network model and the adjustment coefficient; According to the output data of the initial neural network model, the output information of the third layer of neurons, the output information of the second layer of neurons, the input data of the initial neural network model, the all-1 vector and the adjustment coefficient, a second adjustment amount of the threshold between the second layer of neurons and the first layer of neurons is obtained.

6. The method according to claim 1, characterized in that The step of using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, completing the adjustment of the configuration relationship structure, and outputting a dynamic instruction model includes: Input the real-time configuration information and the configuration relationship structure into the Word2Vec model, obtain first word segmentation vector representation information and first part-of-speech vector representation information corresponding to the real-time configuration information, and obtain second word segmentation vector representation information and second part-of-speech vector representation information corresponding to the configuration relationship structure; According to the first word segmentation vector representation information, the first part-of-speech vector representation information, the second word segmentation vector representation information, and the second part-of-speech vector representation information, obtaining the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure; According to the position information and the similarity matrix corresponding to the real-time configuration information and the position information and the similarity matrix corresponding to the configuration relationship structure, the target position corresponding to the implementation configuration information and the configuration relationship structure is obtained, wherein the target position is the position with the highest similarity between the position information corresponding to the real-time configuration information and the position information corresponding to the configuration relationship structure; Obtaining a word segmentation vector representation corresponding to the target position; Obtaining a resource conflict identification result according to the business attribute identifier corresponding to the word segmentation vector representation; When the resource conflict identification result indicates that there is no resource conflict, the configuration relationship structure is adjusted and a dynamic instruction model is output.

7. The method according to claim 6, characterized in that The method further comprises: Using the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, to obtain processed real-time configuration information; The configuration instruction set and the service attribute identifier are used to perform part-of-speech tagging and information retrieval on the configuration relationship structure to obtain a processed configuration relationship structure.

8. A dynamic arrangement device, characterized in that: include: an identification acquisition unit, configured to acquire, in response to a received network configuration request, a service attribute identification of each type of network resource information to be configured, wherein the network configuration request carries service information, device information, and network resource information to be configured; A set acquisition unit, used to acquire a configuration instruction set according to the device information, the network resource information to be configured and an instruction library; A structure acquisition unit, used to extract the real-time configuration information and the configuration instruction set using a target neural network algorithm to acquire a configuration relationship structure corresponding to the network resource information to be configured; The model output unit is used to use the configuration instruction set and the service attribute identifier to perform part-of-speech tagging and information retrieval on the real-time configuration information, complete the adjustment of the configuration relationship structure, and output a dynamic instruction model.

9. A network device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the dynamic orchestration method according to any one of claims 1 to 7.

10. A storage medium storing instructions, characterized in that: When the instructions are executed on a communication device, the communication device is caused to execute the dynamic scheduling method according to any one of claims 1 to 7.

Images