An intent execution method, apparatus and communication device

Abstract

The application provides an intention execution method, device and communication equipment. The method comprises: obtaining intention input information; performing template matching in a plurality of saved intention templates according to the intention input information to determine a target intention template; and executing a target intention based on the intention input information and the target intention template. That is, a plurality of intention templates are constructed in advance. After obtaining the intention input information, the intention input information is matched in the plurality of intention templates to determine a target intention template matched with the intention input information. Moreover, strategy information or learning model information is added in the target intention template. The corresponding strategy or learning model can be used to select information of the target intention template. Thus, the intention input information and the target intention template matched with the intention input information are used to execute the target intention, so that the flexibility of intention execution can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to an intention execution method, device and communication equipment. BACKGROUND

[0002] An intention can be understood as a demand, target and constraint reflecting a demand and conveying it to an autonomous system, can be knowledge created by an intention management function and sent to another intention management function so as to consider the intention in autonomous operation. For an intention, it is necessary to define the expressiveness of the intention through intention modeling, and a vocabulary and structure required for coding intention knowledge can be introduced, and an intention model or intention template obtained through modeling can define common semantics so as to enable both parties to reach an agreement on the meaning of the intention.

[0003] In the prior art, for an intention model used in intention modeling of a communication network, the object, action, expectation and the like of the intention are mainly described. For a simple intention, the above information can basically complete the definition of the intention. However, for a complex intention, such as a multi-layer network designed in a communication network (for example, the intention of a communication service customer is composed of the intention of a communication service provider, and the intention of the communication service provider is composed of the intention of a network operator), the intention template obtained through the prior intention modeling does not fully consider the differences that may exist in different environments, and it is possible that for the same intention scenario, the intention model normally operated in some environment cannot normally operate in another environment due to differences in some factors (such as the security level of the communication network), and another set of intention model needs to be designed, resulting in poor universality and low reuse rate of the intention model. Thus, in the intention execution process, even for the same intention scenario, different intention templates may need to be used to execute the intention, which easily leads to poor flexibility of intention execution. SUMMARY

[0004] The technical problem to be solved by the present application is to provide an intention execution method, device and communication equipment to enhance the flexibility of intention execution.

[0005] To solve the above technical problem, an embodiment of the present application provides an intention execution method, which comprises:

[0006] obtaining intention input information;

[0007] performing template matching in a plurality of saved intention templates according to the intention input information to determine a target intention template, the target intention template comprising intention description information of a target intention and further comprising strategy information or learning model information associated with the target intention, the strategy information corresponding to a strategy or the learning model information corresponding to a learning model being used for selecting information of the target intention template;

[0008] perform the target intent based on the intent input information and the target intent template.

[0009] Embodiments of the present application also provide an intent execution device, the device comprising:

[0010] a first obtaining module, configured to obtain intent input information;

[0011] a matching module, configured to perform template matching in a plurality of saved intent templates according to the intent input information, to determine a target intent template, wherein the target intent template comprises intent description information of a target intent, and further comprises strategy information or learning model information associated with the target intent, and the strategy corresponding to the strategy information or the learning model corresponding to the learning model information is used to select information of the target intent template;

[0012] an execution module, configured to perform the target intent based on the intent input information and the target intent template.

[0013] Embodiments of the present application also provide a communication device, comprising a processor and a memory storing a computer program, wherein the computer program is run by the processor to execute the method described above.

[0014] Embodiments of the present application also provide a computer readable storage medium storing instructions, wherein the instructions are run on a computer to make the computer execute the method described above.

[0015] The above scheme of the present application has at least the following advantages:

[0016] In the present embodiment, a plurality of intent templates are constructed in advance, and after obtaining the intent input information, matching is performed in the plurality of intent templates to determine a target intent template matched with the intent input information, and the target intent template is added with strategy information or learning model information, and the corresponding strategy or learning model can be used to select information of the target intent template, so that the target intent is executed by using the intent input information and the target intent template matched with the intent input information, and the flexibility of intent execution is improved. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is one of the intent execution method flowcharts provided by the embodiments of the present application;

[0018] Figure 2 is an intent expectation information schematic diagram provided by the embodiments of the present application;

[0019] Figure 3 is an intent scope information schematic diagram provided by the embodiments of the present application;

[0020] Figure 4is a schematic diagram of output information provided by an embodiment of the present application;

[0021] Figure 5 is a second flowchart of an intent execution method provided by an embodiment of the present application;

[0022] Figure 6 is a principle diagram of dynamic construction of an intent based on a policy library framework provided by an embodiment of the present application;

[0023] Figure 7 is a principle diagram of dynamic construction of an intent based on an experience library framework provided by an embodiment of the present application;

[0024] Figure 8 is a hierarchical form structure diagram of an intent provided by an embodiment of the present application;

[0025] Figure 9 is a module schematic diagram of an intent execution apparatus provided by an embodiment of the present application. DETAILED DESCRIPTION

[0026] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood and the scope of the present application can be accurately conveyed to those skilled in the art.

[0027] As shown in Figure 1 , an embodiment of the present application provides an intent execution method, the method comprising:

[0028] Step 101: obtaining intent input information;

[0029] Step 102: performing template matching in a plurality of saved intent templates according to the intent input information to determine a target intent template;

[0030] The target intent template includes intent description information of a target intent, and further includes strategy information or learning model information associated with the target intent, the strategy information corresponding to a strategy or the learning model information corresponding to a learning model, which are used to select information of the target intent template;

[0031] Step 103: executing the target intent based on the intent input information and the target intent template.

[0032] The intent execution method of the embodiment of the present application can be applied in a communication network, for example, can be specifically applied in a network slice service scenario, etc. In the actual application of the communication network, the user has certain demands for the communication network, and after introducing the intent into the network, the intent can be converted into a corresponding network configuration, the corresponding network configuration is applied in the communication network, the network is changed, and the demands of the user are met.

[0033] It should be noted that the above intention input information can be understood as the intention information input by the user, can express the intention of the user, can be transmitted to the intention management system, and can be analyzed by the intention management system. Template matching can be performed in the saved multiple intention templates. It can be understood that multiple intention templates (which can also be understood as intention models) are pre-constructed and can be saved in the intention management system. After obtaining the intention input information, template matching can be performed in the multiple intention templates to obtain a target intention template matched with the intention input information. Subsequently, in the process of executing the intention, the intention input information and the target intention template are used to realize the execution of the target intention. In addition, strategy information or learning model information is added to the target intention template, which can be used to select the information of the target intention template to facilitate the execution of the intention.

[0034] In the embodiment, multiple intention templates are pre-constructed, and after obtaining the intention input information, matching can be performed in the multiple intention templates to determine a target intention template matched with the intention input information. Moreover, strategy information or learning model information is added to the target intention template, and the corresponding strategy or learning model can be used to select the information of the target intention template. In this way, the intention input information and the target intention template matched with the intention input information are used to execute the target intention, which can improve the flexibility of intention execution.

[0035] In one embodiment, the intention description information includes at least one of the following:

[0036] Intention expectation information;

[0037] Intention scope information;

[0038] Intention implementation mode;

[0039] Sub-intention information;

[0040] Successor intention information;

[0041] Intention output information.

[0042] In one embodiment, the strategy corresponding to the strategy information includes at least one of the following:

[0043] A first strategy for selecting an intention expectation;

[0044] A second strategy for selecting an intention scope;

[0045] A third strategy for selecting an intention implementation mode;

[0046] A fourth strategy for selecting a sub-intention;

[0047] A fifth strategy for selecting a first sub-intention;

[0048] A sixth strategy for selection of successor intents;

[0049] A seventh strategy for selection of transfer information between parent intents and child intents;

[0050] An eighth strategy for selection of transfer information between predecessor intents and successor intents, the predecessor intents and the successor intents being child intents of the same level having predecessor and successor relationship.

[0051] In one embodiment, the learning model information corresponding learning model comprises at least one of:

[0052] A first model for selection of intent expectation;

[0053] A second model for selection of intent scope;

[0054] A third model for selection of intent implementation;

[0055] A fourth model for selection of child intents;

[0056] A fifth model for selection of first child intent;

[0057] A sixth model for selection of successor intents;

[0058] A seventh model for selection of transfer information between parent intents and child intents;

[0059] An eighth model for selection of transfer information between predecessor intents and successor intents, the predecessor intents and the successor intents being child intents of the same level having predecessor and successor relationship.

[0060] It should be noted that the intent description information can also include basic information (basic_info) and intent execution information. For example, the basic information can include, but is not limited to, intent identification (id) information, intent designer information (designer), intent name (name), intent version (version), intent description (description), and the like. If an intent includes a sub-intent, that is, the intent is composed of a sub-intent, in the sub-intent information of the intent, the information related to the sub-intent can be obtained by using the id information of the sub-intent. For the intent execution information, it mainly describes the information that needs to be saved in the actual running process of the intent instance. For the parent intent, the execution process thereof involves the execution of each corresponding sub-intent. By defining the intent execution information, the execution of the parent intent and each corresponding sub-intent can be described, including but not limited to at least one of the following information: intent execution phase, intent execution state, and intent execution result. The intent execution phase is mainly used to express which step in the intent execution process is specifically executed, such as which sub-intent is specifically executed. The intent execution state is mainly used to express the state of the intent execution, such as activation, deactivation, execution, execution completion, and the like. The intent execution result is mainly used to express the result of the intent execution, such as success, failure, and the like.

[0061] Intent expectation information: For an intent, whether it is a parent intent or a sub-intent, all expectation (expectation_info) information of the intent needs to be described, such as intent target and intent implementation effect information. For each intent, there can be multiple dimensions of expectations, so a set of expectation lists (expectation_list) is formed by multiple dimensions of expectations. For example, the expectations can be expectations for achieving the effect of a certain function, expectations for deploying a certain service, and expectations for certain index thresholds, and the like. For each expectation, it can be composed of an expectation target (target) and a set of state lists (state_list) that the expectation achieves. A possible organization form is as follows Figure 2The expected intent can be different in different scenarios in actual business, and can be a flexible combination of multiple different expectations in the above list of intent expectations. Therefore, in the embodiment of the application, a first strategy (an intent expectation construction strategy) can be introduced, so that the intent can dynamically select the intent expectation according to different factors in the intent running stage, that is, the first strategy can be used for selection of the intent expectation, and the first strategy can be used to select the intent in the intent expectation information of the target intent. Or in the design here, machine learning can be introduced, for example, a first model (an intent expectation construction model) is introduced, the relationship that the user selects different expectation list combinations for different factors in the running process is trained, and subsequently the corresponding first model can be used to dynamically select the intent expectation in the intent running stage, that is, the first model can be used for selection of the intent expectation, and the first model can be used to select the intent in the intent expectation information of the target intent.

[0062] Intent scope information: For each intent, there can be multiple dimensions of scope information (scope_info), and the intent scope information can include scope information of each dimension, that is, a set of scope lists (scope_list) composed of multiple dimensions of scope information. For example, the scope information can include spatial information, that is, spatial scope information, such as information about the service deployment location; or can include time information, that is, time scope information, such as information about the service running time. Through the above information, the time, space, and other limitation information of the intent can be determined, facilitating subsequent intent conflict resolution in the running stage based on the information. For each scope information, a target and a set of conditions (condition_list) describing the target can be composed, and a possible organization form is as shown in Figure 3 The scope can include, but is not limited to, coverage area, use period, and the like in the communication network. In different scenarios, the scope of the intent can be different, and can be a flexible combination of multiple different scopes in the above list of intent scopes. Therefore, in the embodiment of the application, a second strategy (an intent scope information construction strategy) can be introduced, so that the intent can dynamically select the intent scope information according to different factors in the intent running stage, that is, the second strategy can be used for selection of the intent scope, and the second strategy can be used to select the scope in the intent scope information of the target intent. Or in the design here, for example, a second model (an intent scope information construction model) is introduced, the relationship that the user selects different scope list combinations for different factors in the running process is trained, and subsequently the corresponding second model can be used to dynamically select the scope in the intent running stage, that is, the second model can be used for selection of the intent scope, and the second model can be used to select the scope in the intent scope information of the target intent.

[0063] Implementation info: for an intent, to achieve its expectation, it also needs to finally map to corresponding implementation info, such as implementation workflows, implementation policy, etc. In the running stage, the corresponding intent is implemented by executing the corresponding implementation workflows or implementation policy. Therefore, it is necessary to specify the implementation workflows or implementation policy corresponding to the intent in the intent template. In addition, for an intent, it can also have multiple specific implementation ways. Here, the third strategy (intent implementation way selection strategy) can be introduced to design the implementation workflow or implementation policy selection logic under the intent, so that in the running stage, according to the difference of some factors, the pre-set implementation workflow or implementation policy selection strategy under the intent can be dynamically mapped to different implementation workflows or implementation policies, that is, the third strategy can be used for the selection of intent implementation way, and the third strategy can be used for the selection of intent implementation way in the intent implementation way of the target intent; or in the design here, machine learning can also be introduced, and the third model (intent implementation way selection model) is introduced, according to the relationship between the selection of different intent implementation ways by the user in the running process for different factors, the corresponding model training is carried out, and subsequently the actual intent implementation way can be obtained based on the corresponding intent implementation way selection model in the intent running stage, that is, the third model can be used for the selection of intent implementation way, and the third model can be used for the selection in the intent implementation way of the target intent.

[0064] Intent output information: for an intent, its execution result can be used as output information (output_info), which can include multiple dimensions of output information, that is, a set of output information lists (output_list) is composed of multiple dimensions of output information, and one possible organization form is as shown in Figure 4 , that is, one form that can represent the intent output information is as shown in Figure 4 . For the information transmission between parent and child intents and the information transmission between predecessor and successor intents, the information transmission can be carried out based on the intent output information.

[0065] Sub-intent information:

[0066] 1. Parent-child intent composition relationship description: In the case of an intent with sub-intents, the parent-child intent composition relationship (sub_intent_info) needs to be specified, i.e., the sub-intent information corresponding to the intent needs to be specified, and the intent is the parent intent of the corresponding sub-intent. If the parent intent is composed of multiple sub-intents, all sub-intents can be listed, and the execution order of the sub-intents in the intent running phase is the default order listed in the template. For the parent intent, all sub-intents may not need to be executed in the intent running phase. Here, the fourth strategy (sub-intent selection strategy or sub-intent combination strategy (select_sub_isolation_intent)) can be introduced to select sub-intents. In this way, in the intent running phase, different sub-intent combinations can be dynamically selected according to different factors based on the pre-set sub-intent combination strategy; or in the design here, machine learning can also be introduced, for example, the fourth model (sub-intent selection model or sub-intent combination model) is introduced, and different sub-intent combinations are selected for different factors in the actual running process by the user to form a data set, and the corresponding model is trained, and in the subsequent intent running phase, the sub-intent combination can be obtained based on the corresponding sub-intent combination model.

[0067] 2. Parent-child intent information transmission description: For the parent intent, in the intent running phase, the information carried by the parent intent needs to be transmitted to the sub-intent as the information source of the sub-intent expectation description. Here, the seventh strategy (parent-child intent information transmission strategy) can be introduced to design the information transmission logic of the parent intent and each sub-intent under it. In this way, in the running phase, different parent intent information can be transmitted to the sub-intent according to different factors based on the pre-set information transmission strategy of the parent intent and each sub-intent under it; or in the design here, machine learning can also be introduced, for example, the seventh model (parent-child intent information transmission model) is introduced, and different parent-child intent information transmission relationships are selected for different factors in the running process by the user to form a data set, and the corresponding model is trained, and in the subsequent intent running phase, the actual parent intent transmission information can be obtained based on the corresponding parent intent information transmission model.

[0068] For the first sub-intent dynamic selection:

[0069] The first sub-intent under a parent intent is the first sub-intent in the list of sub-intents under the parent intent. For a parent intent, its first sub-intent (which can be one or multiple) needs to be specified, i.e., the first or first batch of sub-intents that need to be executed among all sub-intents of the parent intent. Here, the fifth strategy (first sub-intent selection strategy) can be introduced to design the first sub-intent selection logic (select_first_sub_isolation_intent). In this way, in the running phase, different first sub-intents can be dynamically selected according to different factors based on the preset first sub-intent selection strategy corresponding to the parent intent; or in the design here, machine learning can be introduced, for example, the fifth model (first sub-intent selection model) is introduced, and according to the relationship that the user selects different first sub-intents for different factors in the running process, the corresponding model training is performed, and subsequently the actual first sub-intent of the parent intent can be obtained based on the corresponding first sub-intent selection model in the intent running phase.

[0070] Successor intent information:

[0071] 1. Precursor successor intent composition relationship description: if a parent intent has multiple sub-intents, the execution order among all sub-intents under the parent intent also needs to be specified. Therefore, for each sub-intent, the information (successor_intent_info) of its possible successor intent (successor sub-intent) needs to be specified. For each intent, its successor intent can be one or multiple (successor_intent_list). The successor intent of the last sub-intent in all sub-intents points to its parent intent. When specifying the successor intent of an intent, the sixth strategy (successor intent selection strategy) can be introduced to design the successor intent selection logic (select_successor_intent). In this way, in the running phase, different successor intents can be dynamically selected according to different factors based on the preset successor intent selection strategy; or in the design here, machine learning can be introduced, for example, the sixth model (successor intent selection model) is introduced, and according to the relationship that the user selects different successor intents for different factors in the running process, the corresponding model training is performed. Subsequently, the actual successor intent can be selected based on the corresponding successor intent selection model in the intent running phase.

[0072] 2. The predecessor successor intention information transmission description: for an intention, if there is a successor intention, the information carried needs to be transmitted to the successor intention in the intention running phase as the information source of the successor intention expectation description. Here, the eighth strategy (information transmission strategy between predecessor and successor intentions) can be introduced to design the information transmission logic of the predecessor and successor intentions. Thus, in the running phase, different information can be transmitted to the successor intention according to the preset information transmission strategy between the predecessor and successor intentions according to different factors; or in the design here, machine learning can be introduced, for example, the eighth model (information transmission model between predecessor and successor intentions) is introduced, according to the user in the running process, the relationship of the selection of different information transmission between the predecessor and successor intentions for different factors is trained, and the actual intention transmission information of the successor intention can be obtained based on the corresponding information transmission model between the predecessor and successor intentions in the intention running phase.

[0073] In the embodiment of the application, in the intention template, not only the sub-intention information and the successor intention information can be introduced, but also the corresponding strategy information or learning model information can be introduced, the selection of information can be realized, and the flexibility of the dynamic construction of the intention can be improved, thereby improving the flexibility of the intention execution.

[0074] In one embodiment, based on the intention input information and the target intention template, the step 103 of executing the target intention comprises:

[0075] Step 1031: performing intention expectation selection in the intention expectation information of the target intention based on the strategy corresponding to the strategy information of the target intention or the learning model corresponding to the learning model information of the target intention, and the network environment of the execution object of the intention input information and the target intention, to determine the target expectation information of the target intention;

[0076] Step 1032: performing intention scope selection in the intention scope information of the target intention based on the strategy corresponding to the strategy information of the target intention or the learning model corresponding to the learning model information of the target intention, and the network environment of the execution object of the intention input information and the target intention, to determine the target scope information of the target intention;

[0077] Step 1033: performing sub-intention selection in the sub-intention of the target intention based on the strategy corresponding to the strategy information of the target intention or the learning model corresponding to the learning model information of the target intention, and the network environment of the execution object of the intention input information and the target intention, to determine the sub-intention list of the target intention;

[0078] Step 1034: based on the strategy corresponding to the strategy information of the target intent or the learning model corresponding to the learning model information of the target intent, and the network environment of the execution object of the intent input information and the target intent, the first sub-intent selection is performed in the sub-intent list of the target intent to determine the first sub-intent of the target intent;

[0079] Step 1035: based on the strategy corresponding to the strategy information of the target intent or the learning model corresponding to the learning model information of the target intent, and the network environment of the execution object of the intent input information and the target intent, the selection of the transfer information between the parent and child intents is performed in the target expectation information of the target intent and the target scope information of the target intent to obtain the transfer information between the target intent and each sub-intent of the target intent.

[0080] Step 1036: in the case that the first sub-intent does not include a sub-intent, based on the strategy corresponding to the strategy information of the first sub-intent or the learning model corresponding to the learning model information of the first sub-intent, and the network environment of the execution object of the first sub-intent, the transfer information between the target intent and the first sub-intent, the target expectation information of the first sub-intent and the target scope information of the first sub-intent, the intent implementation mode selection is performed to determine the intent implementation mode of the first sub-intent.

[0081] Step 1037: the intent implementation mode of the first sub-intent is executed in the execution object of the first sub-intent to obtain the execution result of the first sub-intent.

[0082] It should be noted that in the process of determining the target expectation information of the target intent, the first strategy or the first model can be used, that is, the intent input information and the network environment of the execution object of the target intent are input into the first strategy or the first model for intent expectation selection to determine the target expectation information of the target intent. In the process of determining the target scope information of the target intent, the second strategy or the second model can be used, that is, the intent input information and the network environment of the execution object of the target intent are input into the second strategy or the second model for scope selection to determine the target scope information of the target intent. In the process of determining the sub-intent list of the target intent, the fourth strategy or the fourth model can be used, that is, the intent input information and the network environment of the execution object of the target intent are input into the fourth strategy or the fourth model for sub-intent selection of the target intent to determine the sub-intent list of the target intent. In the process of determining the first sub-intent of the target intent, the fifth strategy or the fifth model can be used, that is, the intent input information and the network environment of the execution object of the target intent are input into the fifth strategy or the fifth model for first sub-intent selection to determine the first sub-intent of the target intent. In the process of selecting the information transmission between the parent and child intents, the seventh strategy or the seventh model can be used, that is, the intent input information and the network environment of the execution object of the target intent are input into the seventh strategy or the seventh model, and the seventh strategy or the seventh model is used to select the information transmission between the parent and child intents in the target expectation information of the target intent and the target scope information of the target intent to obtain the transmission information between the target intent and each sub-intent of the target intent. In the process of selecting the intent implementation mode of the first sub-intent, the third strategy or the third model can be used, the network environment of the intent execution object of the first sub-intent, the transmission information between the target intent and the first sub-intent, the target expectation information of the first sub-intent, and the target scope information of the first sub-intent are input into the third strategy or the third model for intent implementation mode selection to determine the intent implementation mode of the first sub-intent. As an example, the target expectation information of the first sub-intent is the expectation information determined in the intent expectation information of the first sub-intent in the first sub-intent template, which is the first strategy corresponding to the first strategy information in the first sub-intent template or the first template corresponding to the first template information, and the target scope information of the first sub-intent is the scope information determined in the intent scope information of the first sub-intent in the first sub-intent template, which is the second strategy corresponding to the second strategy information in the first sub-intent template or the second template corresponding to the second template information.

[0083] The intent execution object can be understood as a managed entity, and the corresponding intent needs to be executed on the intent execution object, so that the intent execution object can meet the user's demand. In addition, it should be noted that all sub-intents in the sub-intent list of an intent are executed, which indicates that the intent is executed. For example, the target intent matched is intent A, the corresponding sub-intent list includes sub-intent B and sub-intent C, the first sub-intent of intent A is sub-intent B, and sub-intent B has no corresponding sub-intent. According to the selected intent implementation mode of sub-intent B, the execution in the intent execution object of sub-intent B is performed, and the execution result of sub-intent B is obtained.

[0084] In the embodiment, the selection of the intent expectation, the intent scope, the first sub-intent, the intent implementation mode, and the transfer information of the target intent can be implemented by the strategy corresponding to the strategy information of the target intent in the target intent template or the model corresponding to the learning model information of the strategy, the dynamic construction of the intent is implemented, the flexibility of the dynamic construction of the intent is improved, the execution result of the first sub-intent is obtained by using the selected intent implementation mode of the first sub-intent to execute in the intent execution object of the first sub-intent, and the flexibility of the execution of the first sub-intent is improved.

[0085] In one embodiment, it further includes:

[0086] In the case where the first sub-intent includes a sub-intent, the first sub-intent of the target intent is taken as the target intent, the target intent expectation information of the target intent, the target scope information of the target intent, the sub-intent list of the target intent, the first sub-intent of the target intent, and the transfer information between the target intent and each sub-intent of the target intent are reselected until the first sub-intent of the target intent does not include a sub-intent.

[0087] If the first sub-intention further includes a sub-intention, the first sub-intention is indicated as being executed only when the sub-intention list of the first sub-intention is executed. In this embodiment, if the first sub-intention includes a sub-intention, the first sub-intention of the target intention and the target intention expected information, the target scope information, the sub-intention list of the target intention, and the transfer information between the target intention and its sub-intention of the first sub-intention need to be selected, and the process is similar to the process of selecting the first sub-intention of the target intention, the target intention expected information of the target intention, the target scope information, the sub-intention list of the target intention, and the transfer information between the target intention and its sub-intention, with the difference being that the intention is different, the former is the first sub-intention of the target intention, and the latter is the target intention. Therefore, the execution of the first sub-intention of the target intention can be completed in a loop. That is, in the case where the first sub-intention includes a sub-intention, the first sub-intention of the target intention is taken as the target intention, the target intention is updated, and steps 1031-1035 are repeatedly executed. If the target intention is updated, the first sub-intention of the target intention is also updated, until the first sub-intention of the target intention does not include a sub-intention. Based on the strategy information corresponding to the strategy of the first sub-intention of the target intention or the learning model information corresponding to the learning model of the first sub-intention, the network environment of the intention execution object of the first sub-intention, the transfer information between the target intention and the first sub-intention, the target expected information of the first sub-intention, and the target scope information of the first sub-intention, the intention implementation mode is selected, the intention implementation mode of the first sub-intention is determined, and the intention implementation mode of the first sub-intention is executed to obtain the execution result of the first sub-intention. For example, the matched target intention is intention A, the corresponding sub-intention list includes sub-intention B and sub-intention C, sub-intention B has corresponding sub-intentions, such as sub-intention D and sub-intention E, the first sub-intention of intention A is sub-intention B, and since sub-intention B has a sub-intention, the target intention can be updated subsequently, the first sub-intention B of intention A is taken as the target intention, its corresponding sub-intentions include sub-intentions D and E, the first sub-intention of the target intention can be selected again, for example, the updated target intention is sub-intention B, and the selected first sub-intention is sub-intention D. Since the first sub-intention D does not include a sub-intention, the intention execution mode of sub-intention D can be determined, the intention implementation mode of the first sub-intention is executed in the intention execution object of the first sub-intention to obtain the execution result of the first sub-intention. Since the updated first sub-intention is sub-intention D at this time, the intention execution mode of sub-intention D is executed in the intention execution object of sub-intention D to obtain the execution result of sub-intention D.

[0088] In one embodiment, after the intention implementation mode of the first sub-intention is executed in the intention execution object of the first sub-intention to obtain the execution result of the first sub-intention, it further includes:

[0089] Step 1038: based on the strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or the learning model corresponding to the learning model information of the first sub-intention, and the network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention, the successor intention of the first sub-intention is determined by selecting the successor intention in the sub-intention list, and the first sub-intention template is the intention template corresponding to the first sub-intention;

[0090] Step 1039: based on the strategy corresponding to the strategy information of the first sub-intention or the learning model corresponding to the learning model information of the first sub-intention, and the network environment of the intention execution object of the first sub-intention, the transfer information between the first sub-intention and the successor intention of the first sub-intention is obtained by selecting the transfer information between the predecessor and the successor intention in the execution result of the first sub-intention;

[0091] Step 10310: in the case that the successor intention of the first sub-intention does not include a sub-intention, based on the strategy corresponding to the strategy information of the successor intention in the successor intention template or the learning model corresponding to the learning model information of the successor intention, and the network environment of the intention execution object of the successor intention, the transfer information between the first sub-intention and the successor intention of the first sub-intention, the target intention expectation information of the successor intention and the target scope information of the successor intention, the intention implementation mode of the successor intention of the first sub-intention is determined by selecting the intention implementation mode, and the successor intention template is the intention template corresponding to the successor intention of the first sub-intention;

[0092] Step 10311: the intention implementation mode of the successor intention of the first sub-intention is executed in the intention execution object of the successor intention, and the execution result of the successor intention is obtained.

[0093] After the first sub-intention implementation manner is executed in the intention execution object of the first sub-intention, and the execution result of the first sub-intention is obtained, the successor intention of the first sub-intention can be selected. It should be noted that in the process of determining the successor intention of the first sub-intention, the sixth strategy in the strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or the sixth model in the learning model corresponding to the learning model information of the first sub-intention can be used, that is, the network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention are input into the sixth strategy or the sixth model to select the successor intention, and the successor intention of the first sub-intention is determined. In the process of determining the transfer information between the first sub-intention and the successor intention of the first sub-intention, the eighth strategy in the strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or the eighth model in the learning model corresponding to the learning model information of the first sub-intention can be used, that is, the network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention are input into the eighth strategy or the eighth model to select the transfer information between the predecessor and the successor intention, and the transfer information between the first sub-intention and the successor intention of the first sub-intention is determined. In the process of selecting the intention implementation manner of the successor intention of the first sub-intention, based on the third strategy in the strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or the third model in the model corresponding to the model information of the first sub-intention in the first sub-intention template, the network environment of the intention execution object of the successor intention, the transfer information between the first sub-intention and the successor intention of the first sub-intention, the target intention expectation information of the successor intention, and the target scope information of the successor intention are input into the third strategy or the third model to select the intention implementation manner, and the intention implementation manner of the first sub-intention is determined. In one example, the target expectation information of the successor intention of the first sub-intention is the expectation information determined in the intention expectation information of the successor intention in the successor intention template through the first strategy corresponding to the first strategy information or the first model corresponding to the first model information in the successor intention template, and the target scope information of the successor intention is the scope information determined in the intention scope information of the first sub-intention in the successor intention template through the second strategy corresponding to the second strategy information or the second model corresponding to the second model information in the successor intention template.

[0094] As an example, the matched target intent is intent A, the corresponding sub-intent list includes sub-intent B and sub-intent C, the first sub-intent of intent A is sub-intent B, and sub-intent B has no corresponding sub-intent. Therefore, the execution of the selected sub-intent B is performed in the intent execution object according to the selected intent implementation manner of sub-intent B, and the execution result of sub-intent B is obtained. After the execution of sub-intent B is completed, the successor intent of sub-intent B, for example, sub-intent C, can be selected. Sub-intent C does not include sub-intent, and the intent implementation manner of sub-intent C can be determined. Then, the execution result of the successor intent C is obtained by executing the intent implementation manner of sub-intent C in the intent execution object of sub-intent C. At this time, the two sub-intents of intent A are executed, indicating that the execution of intent A is completed, and the execution result of intent A is obtained. For another example, the matched target intent is intent A, the corresponding sub-intent list includes sub-intent B and sub-intent C, and sub-intent B has corresponding sub-intents, for example, sub-intents D and E. The first sub-intent of intent A is sub-intent B. Since sub-intent B has sub-intents, i.e., sub-intents D and E, the first sub-intent of sub-intent is sub-intent D, and sub-intent D has no sub-intent, the execution of sub-intent D is performed first, and then the successor intent, for example, sub-intent E, is executed in the intent execution process. After the execution of sub-intent E is completed, the two sub-intents of sub-intent B are executed, indicating that the execution of sub-intent B is completed. Then, the successor intent of sub-intent B, for example, sub-intent C, is executed. Since C has no sub-intent, the execution result of sub-intent C is obtained by executing the intent implementation manner of sub-intent C in the intent execution object of sub-intent C. After the execution of sub-intent C is completed, the two sub-intents of intent A are executed.

[0095] In one embodiment, after obtaining the execution result of the successor intent, the method further includes:

[0096] In step 10312, when the sub-intent list of the target intent further includes an unexecuted sub-intent, the successor intent of the first sub-intent is taken as the first sub-intent, the strategy corresponding to the strategy information of the first sub-intent in the first sub-intent template or the learning model corresponding to the learning model information of the first sub-intent, and the network environment of the intent execution object and the execution result of the first sub-intent are returned. The sub-intent list is selected, and the step of determining the successor intent of the first sub-intent is re-determined. The successor intent is selected until the sub-intents in the sub-intent list of the target intent are executed, the execution of the target intent is determined to be completed, and the execution result of the target intent is obtained.

[0097] It can be understood that the target intent is the latest target intent, and the sub-intent list of any intent can be selected by the sub-intent selection strategy or the sub-intent selection model in the intent template corresponding to the intent. In the case where the sub-intent list of the target intent does not include an unexecuted sub-intent, it indicates that the target intent is executed, and the execution result of the target intent is obtained. In the case where the sub-intent list of the target intent still includes an unexecuted sub-intent, the successor intent in the unexecuted sub-intent needs to be selected, the successor intent of the first sub-intent is taken as the first sub-intent, the first sub-intent is updated, and then steps 1038-10311 can be repeatedly executed until the sub-intents in the sub-intent list of the target intent are executed, it is determined that the target intent is executed, and the execution result of the target intent is obtained.

[0098] In one embodiment, after obtaining the execution result of the target intent, the method further includes:

[0099] In the case where the intent corresponding to the target intent template determined by the template matching of the target intent does not match, the target intent is taken as the first sub-intent, the parent intent of the first sub-intent is taken as the target intent, and in the case where the sub-intent list of the target intent still includes an unexecuted sub-intent, the strategy corresponding to the strategy information of the first sub-intent in the first sub-intent template or the learning model corresponding to the learning model information of the first sub-intent, and the intent input information of the first sub-intent, the network environment of the intent execution object of the first sub-intent and the execution result of the first sub-intent are returned, the successor intent is selected in the sub-intent list, and the step of determining the successor intent of the first sub-intent is re-determined until the sub-intent list of the target intent determined by the template matching is executed.

[0100] Since the target intent may have been updated in the middle, the current target intent may not be consistent with the target intent determined by template matching based on the intent input information. This indicates that the matched target intent has not yet been completed. In this case, the target intent can be taken as the first child intent, and the parent intent of the first child intent can be taken as the target intent. For example, as in the example above, the matched target intent is intent A, and the corresponding child intent list includes child intents B and C. Child intent B has corresponding child intents, such as child intents D and E. The first child intent of intent A is child intent B. Since child intent B also has child intents, namely child intents D and E, if the currently updated target intent is child intent B, and the corresponding child intents D and E have been completed, then child intent B is completed. However, the matched intent A still has child intent C that has not been completed. In this case, the parent intent of child intent B can be taken as the first child intent, the first child intent can be updated, and then the parent intent of the first child intent can be taken as the first child intent. The diagram (Intent A) serves as the target intent, meaning the target intent is updated to Intent A at this point. The first sub-intent is then updated to Sub-intent B. Since Sub-intent B has been executed, we can further select the successor intent of Sub-intent B from the unexecuted sub-intents in the target intent's intent list. For example, we can select Sub-intent C as the successor intent of Sub-intent B. Through a similar process, Sub-intent C is executed. After Sub-intent C is executed, it indicates that the target intent has been executed. At this point, the target intent is Intent A, which matches the matched Intent A, indicating that the matched Intent A has been executed. At this point, the entire intent execution process ends.

[0101] like Figure 5 The process of the above method is specifically illustrated using a particular embodiment. The specific flow is as follows:

[0102] 1: Users express their intent by inputting intent input information, which is then transmitted to the intent management system.

[0103] 2. The intent management system parses the intent input information and matches the corresponding target intent template from multiple saved intent templates. It can then transmit the intent input information and the strategy or model information from the target intent template to the strategy or model system. The strategy or model system obtains the corresponding information through the pre-set corresponding strategy or machine learning model, including the following information:

[0104] (1) Obtain the actual target expectation information of the target intent based on the first strategy or first model of the target intent;

[0105] (2) Obtain the actual target scope information of the target intent based on the second strategy or second model of the target intent;

[0106] (3) Obtain the actual list of sub-intents of the target intent based on the fourth strategy or fourth model, for example, in Figure 5 In this context, the target intention originally had four sub-intentions (sub-intentions)Figure 1 , child intention Figure 2 , child intention Figure 3 and child intention Figure 4 ). Based on the fourth strategy or the fourth model in the target intention template, the child intention selection is performed, for example, in the current intention execution process, the list of child intentions that need to be executed is only child intention Figure 1 , child intention Figure 2 and child intention Figure 4 , without executing child intention Figure 3 .

[0107] (4) Based on the fifth strategy or the fifth model in the target intention, the first child intention that needs to be executed is obtained, in which the first child intention is child intention Figure 5 , which does not include child intention Figure 1 .

[0108] (5) Based on the seventh strategy or the eighth model in the target intention, the information actually required by the target intention to be passed to each child intention is obtained.

[0109] 3: The information in step 2 is passed to the intention management system.

[0110] 4 The intention management system passes the corresponding parent intention information to the intention system corresponding to the first child intention.

[0111] 5-6: The first child intention actual intention implementation mode (execution strategy or workflow) is obtained through the third strategy or the third model of the target intention in the strategy or template system, and the strategy or template system returns the obtained first child intention actual intention implementation mode to the intention system of the first child intention.

[0112] 7-8: The intention system of the first child intention performs corresponding operations on the managed entity corresponding to the first child intention according to the obtained first child intention actual intention implementation mode. The managed entity realizes the expectation of the first child intention by completing the corresponding operation process, and returns the corresponding execution result to the intention system of the first child intention.

[0113] 9-10: Through the sixth strategy or the sixth model of the first child intention in the strategy or model system, and the eighth strategy or the eighth model of the first child intention, the corresponding information is obtained, and the obtained corresponding information is returned to the intention system of the first child intention, including the following information:

[0114] (1) Based on the sixth strategy or the sixth model of the first child intention, the successor intention of the first child intention is obtained, as shown in Figure 5 , the successor intention of the first child intention is child intention Figure 4 , which does not include child intention

[0115] (2) Based on the eighth strategy or the eighth model of the first child intention, the information passed by the first child intention to its successor intention, i.e. child intention Figure 1 , is obtained.

[0116] 11: The intent system of the sub-intent passes the information to the intent system of the sub-intent. Figure 4

[0117] 12-13: The sub-intent passes the information to the third strategy or the third model in the strategy or model system. Figure 4 Figure 4 Figure 4

[0118] 14-15: The intent system of the sub-intent performs corresponding operations on the managed entity of the sub-intent based on the acquired intent implementation of the sub-intent. Figure 4 Figure 4 Figure 4 Figure 4 Figure 4 Figure 4

[0119] 16-17: The sub-intent passes the information to the sixth strategy or the sixth model in the strategy or model system. Figure 1 Figure 4 Figure 1 Figure 4 Figure 2

[0120] 18: The intent system of the sub-intent passes the corresponding information to the intent system of the sub-intent. Figure 4 Figure 2

[0121] 19-20: The sub-intent passes the information to the third strategy or the third model in the strategy or model system. Figure 2 Figure 2 Figure 2

[0122] 21-22: The intent system of the sub-intent performs corresponding operations on the managed entity of the sub-intent based on the acquired intent implementation of the sub-intent. Figure 2 Figure 2 Figure 2 Figure 2 Figure 2 Figure 2

[0123] 23-24: The sub-intent passes the information to the sixth strategy or the sixth model in the strategy or model system.​​​​​​​​​​​​​​​​​​​​​​​​​​Figure 2 The sixth strategy or sixth model in the process of obtaining sub-meaning Figure 2 The corresponding subsequent intent and the information to be conveyed, here is the sub-intent. Figure 2 It is the last sub-meaning, and its successor meaning is the sub-meaning. Figure 1 The parent intent will return the acquired information to the intent management system.

[0124] 25: The Son's Intention of the Father Figure 4 Ziyi Figure 2 He Ziyi Figure 6 Once all operations are completed, the parent intent is executed, meaning the user's target intent has been achieved, and the corresponding result is returned to the user.

[0125] The following description uses an embodiment of the present invention applied to a communication network as an example. Based on the intent template of the method in this embodiment, the problem of poor intent decomposition and combination in the intent application process of a communication network can be solved. Furthermore, by introducing strategies or models into the intent template, the flexible reuse of intents is enhanced. Specifically, to address the issue that existing intent templates in communication networks cannot adequately support intent decomposition and combination, this problem is improved by introducing intent decomposition and combination descriptions into the intent template. For intent decomposition and combination, two factors need to be considered: first, if the intent has sub-intents, the information of all sub-intents contained in the intent needs to be considered; second, if the intent is also a sub-intent, the relationship between the intent and other sub-intents belonging to the same parent intent needs to be considered. Therefore, the corresponding template introduces information such as parent-child intent relationship descriptions, first-child intent relationship descriptions, and predecessor-successor intent relationship descriptions. Through the above-mentioned newly added information, the model expression problem of intent decomposition and combination is solved. The enhancement of the flexible reuse of intent templates is achieved by introducing strategies or machine learning modules into multiple processes and information transmission of intents, thereby improving the flexibility of dynamic intent construction.

[0126] Introduction strategy:

[0127] like Figure 8 As shown, a policy framework is introduced into the intent flow of the communication network. During multiple flows or information transmission processes of the intent, corresponding policies can be preset or dynamically updated. Through policy control, the intent can be mapped to different execution flows in different environments based on different factors, enhancing the reusability of the intent. The dynamically constructed intent flow and information transmission flow involved here are as follows:

[0128] Dynamic construction of intended and desired information;

[0129] Dynamic construction of intent scope information;

[0130] Dynamic construction of the means to achieve the intent;

[0131] Dynamic construction of parent-child intention relationship;

[0132] Dynamic construction of predecessor-successor intention relationship.

[0133] In each of the above processes, different strategies can be introduced to achieve dynamic construction of the strategies. In the intention template, when describing each of the above processes, the process in the intention is associated with the strategy system by adding corresponding strategy information description. In the actual execution process of the intention, the intention execution process is associated with the strategy system based on the strategy information in each of the processes described in the template, thereby achieving dynamic construction of the intention process.

[0134] Introducing a machine learning model:

[0135] That is, model training technology can be introduced, and the intention process and information transmission process involved in the dynamic construction are similar to the above-mentioned introduction of strategies, except that this is a machine learning model, and the above is a strategy, which will not be described here. In the intention model, when describing each of the above processes, the process in the intention is associated with the model training system in the same way as the introduction of the strategy, by adding corresponding model training information description. In the actual execution process of the intention, the intention execution process is associated with the model training based on the model training information in each of the processes described in the template, thereby achieving dynamic construction of the intention process.

[0136] In a communication network, the hierarchical form of intention is as shown in Figure 9 The communication service provider intention is a sub-intention of the communication service customer intention, the network operation intention: wireless side intention, the network operation intention: bearer network side intention and the network operation intention: core network side intention are all sub-intentions of the communication service provider intention, and the sub-intentions of the wireless side intention can include multiple network service related intentions on the wireless network side, the sub-intentions of the bearer network side intention can include multiple network service related intentions on the bearer network side, and the sub-intentions of the core network side intention can include multiple network service related intentions on the core network side.

[0137] For the communication service customer intention template:

[0138] Desired information description: This describes the customer's expectation of the service provided by the communication service provider, such as providing reliable service guarantee for key customers during certain traffic peak periods. This can include, but is not limited to, expectations for business types, expectations for service quality, etc.

[0139] Scope information description: This describes the restrictions and constraints on the provision of services, such as including but not limited to restrictions on service time information, service area information, etc.

[0140] Parent-child intent relationship description: the sub-intent of describing the communication service client's intent is the intent of the communication service provider, and the corresponding intent information is transmitted to the intent of the communication service provider, which can include but is not limited to service type, service quality requirement, time and space scope information, etc.

[0141] For communication service provider intent template:

[0142] Expected information description: it describes the information transmitted by the communication service provider according to the communication service client's intent, which is converted into the expectation of what to do to the network, such as the expectation of bandwidth, reliability, service level agreement (SLA) (such as delay, packet loss, jitter), inter-site link requirements (such as topology relationship, access protocol), quality of service (Qos) policy information (such as platinum policy) of private line service, etc.

[0143] Scope information description: it describes the limitation and constraint to the network, such as network area information, site access reliability constraint (such as different devices in the same network service provider point (POP point)), etc.

[0144] Parent-child intent relationship description: in the communication network, the communication service provider intent will be decomposed into the intent of the network operator, which will be decomposed into wireless side intent, bearer network side intent and core network side intent according to the division of network networking different domains, and the expectation and scope information of different domains need to be transmitted to each sub-intent according to the different domains.

[0145] First sub-intent information description: in the above sub-intent, the first executed sub-intent needs to be described, such as the wireless side sub-intent can be executed first.

[0146] For network operator intent template:

[0147] The network operator intent is the management and control that the network operator can provide to a group of network elements (i.e. sub-network), and this paper takes the bearer network side network operator intent as an example to explain the intent template of network operation intent. The wireless side intent and core network side intent model are similar to the bearer network, which will not be described here.

[0148] Expected information description: describes the expected information provided by the bearer network domain service, which can include but is not limited to service type (L3VPN), access point device and interface identifier, UNI side configuration information, bandwidth (such as 100M), SLA (delay, packet loss, jitter), service topology, Qos policy information (such as platinum policy) and the like. The foregoing policy or model training information description can be introduced here, and the expected information is corresponded to the policy or model. When the intent is running, the corresponding expected information can be matched according to the training result of the policy or model for different levels of customers or different Qos policy information;

[0149] Scope information description: describes the limitation and constraint on the bearer network side, such as deployment network area information, service providing time information, deployment policy (such as error rollback), protection policy (such as 1:1), tunnel policy and the like. The foregoing policy or model training information description can be introduced here, and the scope information is corresponded to the policy or model. When the intent is running, the corresponding scope information can be matched according to the training result of the policy or model for different levels of customers;

[0150] Parent-child intent relationship description: the intent of the network operator on the bearer network side can be further divided into each sub-intent related to the network service, such as the network element deployment or expansion sub-intent based on the performance index expectation, the sub-intent for providing service on the bearer network side, the sub-intent for providing security reinforcement or security isolation of the network element. The foregoing policy or model training information description can be introduced here, and the sub-intent information is corresponded to the policy or model. When the intent is running, the corresponding different sub-intent combination information can be matched according to the training result of the policy or model for different levels of customers;

[0151] First sub-intent information description: in the foregoing sub-intent, the first executed sub-intent needs to be described, such as the network element deployment or expansion sub-intent, and the specific requirements for deployment, customer level, Qos policy information and the like need to be sent to each sub-intent. The foregoing policy or model training information description can be introduced here, and the first sub-intent selection is corresponded to the policy or model. When the intent is running, the corresponding first sub-intent information can be matched according to the training result of the policy or model for different levels of customers;

[0152] Predecessor-successor intent relationship description: mainly describes the predecessor-successor relationship of the bearer network side network operator intent and the wireless side intent, the core network side intent and the corresponding information transmission process.

[0153] For network service intent template:

[0154] As above, the network operator's intention on the bearer network side can be further decomposed into various sub-intentions related to network services, such as network element deployment or expansion sub-intentions based on performance index requirements, which can be sub-intentions for providing services and ensuring the detection and security of the network on the bearer network side, or sub-intentions for providing security reinforcement or security isolation of the network element. For the network service intention template, the intention implementation description and the predecessor-successor intention relationship description information are mainly described, and the remaining information is similar to the foregoing intention template, which will not be described again.

[0155] Intention implementation description: This describes the specific implementation of various network service-related intentions, which ultimately correspond to different workflows, strategies, or other implementation methods. For example, in the network service detection sub-intention, we can choose to use digital twin technology (high accuracy but complex implementation) or choose conventional analysis and evaluation technology (simple implementation but less accurate than digital twin technology) for detection. Here, the foregoing strategy or model training information can also be introduced to correspond the intention implementation to the strategy or model. During the intention running, the corresponding implementation can be matched according to the results of the strategy or model training for different levels of customers, such as selecting digital twin technology for high-level customers and conventional analysis and evaluation technology for low-level customers.

[0156] Predecessor-successor intention relationship description: This mainly describes the predecessor-successor relationship between various network service sub-intentions. For example, for the network service deployment sub-intention, the successor sub-intentions can include the security isolation sub-intention. Here, the foregoing strategy or model training information can also be introduced to correspond the intention implementation to the strategy or model. During the intention running, the corresponding successor intention and information transmission can be matched according to the results of the strategy or model training for different levels of customers, such as including the security isolation sub-intention for high-level customers and not including the security isolation sub-intention for low-level customers.

[0157] Take the network service-related intention on the core network side as an example. To prevent mutual influence and attack behavior between network services running on the unified infrastructure resources on the core network side, isolation control is needed between network services on the network element, data, resource, data, network access, and other levels. In this embodiment, based on the network service security isolation technology on the core network side, the application of the template in the network service security isolation on the core network side is introduced.

[0158] Network service intention model design on the core network side based on security isolation technology:

[0159] The network service security isolation intention on the core network side is composed of network element isolation sub-intention, data isolation sub-intention, and network isolation sub-intention, and the design of each part is as follows.

[0160] Intention expectation design:

[0161] The core network side network service security isolation intention is composed of a network element isolation sub-intention, a data isolation sub-intention, and a network isolation sub-intention. Based on the design scheme of the embodiment of the application, the intention expectation of the core network side network service security isolation intention can be designed. For example, three example expectations of the core network side network service security isolation intention can be designed, which are about customer level, network element isolation requirement, and data isolation requirement. The expectation of the customer level is preferential, the expectation of the network element isolation level is high-level isolation, and the expectation of the data isolation requirement has two states, one is whether to require data encryption, and the other is whether to require identity authentication. Default values are given here, that is, data encryption and identity authentication are required by default, and the default values of the corresponding expectations can be changed according to the specific requirements of the user in the running state. An intention expectation selection strategy (for example, a unique-policy-id-1-select-expection-list strategy) is also configured here, which is used to dynamically select (construct) the intention expectation information in the intention running phase.

[0162] Intention scope design:

[0163] For example, two example scopes of the core network side network service security isolation intention can be designed, which are about isolation area and isolation period scope information. The scope information of the isolation area gives the default value of the specific area latitude and longitude information, and the isolation period scope information also gives the default isolation period of T (for example, 20) days. A scope selection model based on a machine learning model (unique-ml-model-id-1-select-scope-list) is also configured here, which is used to dynamically select the intention scope information in the intention running phase.

[0164] Intention implementation method design:

[0165] For example, in the intention implementation method design, by designing an intention implementation method selection strategy (for example, a unique-policy-id-1-select-implementation-method), the specific implementation method of the core network side network service security isolation intention can be dynamically obtained in the intention running phase, which can be a workflow or an execution strategy.

[0166] Sub-intention information design:

[0167] For example, the intent consists of a network element isolation sub-intent (ne-isolation-intent-id-1), a data isolation sub-intent (data-isolation-intent-id-2), and a network isolation sub-intent (network-isolation-intent-id-3), and the sub-intent list (sub_intent_list) lists the three sub-intents. The sub-intent selection model (select_sub_isolation_intent) gives a scheme for selecting the sub-intent list based on a machine learning model in the intent running phase, the first sub-intent selection model (select_first_sub_isolation_intent) gives a scheme for selecting the first sub-intent based on a machine learning model in the intent running phase, and the parent-to-sub-intent information transmission model (transmit_info_to_sub_intent) gives a scheme for transmitting information between the parent and the sub-intent based on a machine learning model in the intent running phase.

[0168] Intent output information design:

[0169] For example, for the network service security isolation intent on the core network side, after the intent execution is completed, the corresponding information can be output, such as the level of network element isolation, the algorithm selected for data encryption, etc. In this example, the network service security isolation intent on the bearer network side can include two output information, one is the network element isolation level (ne_isolation_level), which represents which of the physical machine dedicated, virtual machine dedicated network element isolation levels is used, and the other is the data encryption algorithm (data_encryption_algorithm), which represents which data encryption algorithm is used, such as DEA, RSA, DSA algorithm, etc.

[0170] Successor intent information design:

[0171] After the network service security isolation intent at the core network side is executed, the network service network element security configuration intent at the core network side or the network service management security intent at the core network side can be executed. The network service network element security configuration intent at the core network side is used for virtual network element security configuration and network service network element access control configuration. The network service management security intent at the core network side is used for ensuring network service operation security, including tenant separation, security protocol configuration, cross-domain security configuration, etc. In this example, the successor intent list (successor_intent_list) lists two successor sub-intents of the network service security isolation intent at the core network side. The select successor intent (select_successor_intent) gives a scheme for selecting a successor intent based on a strategy in an intent running phase. The transmit info to successor intent (transmit_info_to_successor_intent) gives a scheme for transmitting information between predecessors and successors based on a strategy in an intent running phase.

[0172] Actual running intent template:

[0173] In the intent running phase, the actual intent information is obtained based on matching of user intent input information to a corresponding target intent template, in combination with the intent input information and environment, etc. In this example, the intent template is described as follows:

[0174] 1) basic_info: basic information description of the intent instance;

[0175] 2) intent_model_info: represents intent template information of the intent instance in the intent knowledge base, for example, can include an identifier (ID) representing the intent template in the intent knowledge base, a name representing the intent template in the intent knowledge base, and a version representing the intent template in the intent knowledge base, etc.

[0176] 3) expectation_info: describes actual expectation information of the intent in a running state obtained based on an intent expectation selection strategy, in this example, the intent has three expectations, one is about customer level expectation (customer-level), one is about network element isolation requirement (ne-isolation-requirement), and one is about data isolation requirement (date-isolation-requirement).

[0177] 4) scope_info: the actual scope information description obtained according to the intent scope selection strategy of the intent in the running state, in this example, the intent has two scopes, one is the scope information about the isolated area (isolated-area), and the other is the scope information about the network element isolation period (service_availability_period).

[0178] 5) implementation_info: the actual intent implementation selected according to the intent implementation selection strategy in the running process, in this example, the intent implementation selected according to the strategy can be unique-workflow-id-1.

[0179] 6) output_info: the actual output information of the intent, in this example, there are two intent output information is_satisfied (whether the isolation requirement is satisfied) and estimated_development_period (estimated development period).

[0180] 7) sub_intent_info: the sub-intent information of the intent, for example, a sub-intent list (sub_intent_list) representing all actual sub-intents of the intent, a first intent (first_intent) representing the first executed sub-intent, and transmit_intent_information representing the information transmitted by the parent intent to the sub-intent.

[0181] 8) successor_intent_info: the successor intent information of the intent, for example, a successor intent list (successor_intent_list) representing all actual successor intents of the intent, and transmit_intent_information representing the information transmitted by the intent to the successor intent.

[0182] 9) intent_execution_info: represents the intent instance execution information description, and each factor is described as follows:

[0183] execution_stage: represents the intent instance execution stage, in this example, the intent has executed to the network service data isolation sub-intent (data-isolation-intent-id-2) stage.

[0184] execution_status: represents the intent execution status, in this example, the intent has been executed (finished).

[0185] execution_result: represents an intention execution result, in this example, the intention execution is failed.

[0186] The scheme of the embodiment of the present application proposes an enhanced intention template and use flow in a communication network, and detailed template information description is given based on the structure of the communication network. In the intention template, the idea of designing the intention template based on intention decomposition and combination is more completely realized by introducing the schemes of parent intention and child intention relationship description, intention predecessor and successor relationship description, etc.

[0187] Moreover, a model description is proposed in which a strategy or machine learning model is introduced in the flow of child intention combination, first child intention selection under parent intention, successor intention selection, etc. in the communication network. In the actual running environment, different selections can be made based on the strategy or machine learning model according to different environmental factors, thereby enhancing the flexibility of intention design and execution.

[0188] In addition, a model description is proposed in which a strategy or machine learning model is introduced in the flow of intention expectation information selection, intention scope information selection, information transmission between parent and child intentions, information transmission between predecessors and successors, intention implementation mode selection, etc. in the communication network. In the actual running environment, different information transmissions and selections can be made based on the strategy or machine learning model according to different environmental factors, thereby enhancing the flexibility of intention design and execution.

[0189] In the existing intention scheme of the communication network, the idea of decomposing and combining intentions is not considered, only simple intention expectation description is concerned, and complex composite intentions cannot be designed and executed, and the intention cannot be used in complex scenarios similar to multi-layer network structures. The intention template for decomposing and combining intentions proposed in the scheme of the embodiment of the present application can well complete the hierarchical design and execution of complex intentions. The intention model of the communication network proposed in the prior art does not consider enhancing the flexibility of intention design and execution by introducing a strategy and a machine learning model. The scheme of the embodiment can introduce a strategy and machine learning related technology in the template description of multiple flows of the intention, so that the entire flow of supporting the intention is more automated, and the support for the automatic design and running flow of the communication network can be better realized.

[0190] As shown in ​ The embodiment of the present application further provides an intention execution device 900, and the device 900 comprises:

[0191] A first acquisition module 901 is configured to acquire intention input information.

[0192] The matching module 902 is configured to perform template matching in the saved multiple intent templates according to the intent input information, determine a target intent template, and the target intent template includes intent description information of a target intent, and further includes policy information or learning model information associated with the target intent, the policy information corresponds to a policy or the learning model information corresponds to a learning model, and the policy or the learning model is used for selecting information of the target intent template.

[0193] The execution module 903 is configured to execute the target intent based on the intent input information and the target intent template.

[0194] In an embodiment, the intent description information includes at least one of the following:

[0195] intent expectation information;

[0196] intent scope information;

[0197] intent implementation mode;

[0198] sub-intent information;

[0199] successor intent information;

[0200] intent output information.

[0201] In an embodiment, the policy corresponding to the policy information includes at least one of the following:

[0202] a first policy for selection of intent expectation;

[0203] a second policy for selection of intent scope;

[0204] a third policy for selection of intent implementation mode;

[0205] a fourth policy for selection of sub-intent;

[0206] a fifth policy for selection of first sub-intent;

[0207] a sixth policy for selection of successor intent;

[0208] a seventh policy for selection of transfer information between parent intent and sub-intent;

[0209] an eighth policy for selection of transfer information between predecessor intent and successor intent, the predecessor intent and the successor intent being sub-intents of the same level and having a predecessor and successor relationship.

[0210] In an embodiment, the learning model corresponding to the learning model information includes at least one of the following:

[0211] a first model for selection of intent expectation;

[0212] The second model is used for selecting the scope of intent;

[0213] The third model is used to select the method of achieving the intent;

[0214] The fourth model is used for selecting sub-intents;

[0215] The fifth model is used for the selection of the primary intention;

[0216] The sixth model is used for selecting subsequent intentions;

[0217] The seventh model is used for selecting the information to be passed between parent and child intentions;

[0218] The eighth model is used to select the information to be passed between the preceding and succeeding intentions. The preceding and succeeding intentions are sub-intentions of the same level that have a preceding and succeeding relationship.

[0219] In one embodiment, the execution module 903 includes:

[0220] The expectation selection module is used to select the expectation information of the target intention based on the strategy information corresponding to the strategy information of the target intention or the learning model information corresponding to the learning model information of the target intention, as well as the network environment of the intention input information and the execution object of the target intention.

[0221] The scope selection module is used to select the scope of intent based on the policy information corresponding to the policy information of the target intent or the learning model information corresponding to the learning model information of the target intent, as well as the network environment of the intent input information and the execution object of the target intent, and to determine the target scope information of the target intent.

[0222] The sub-intent selection module is used to select sub-intents from the sub-intents of the target intent based on the policy information corresponding to the policy of the target intent or the learning model information corresponding to the learning model of the target intent, as well as the network environment of the intent input information and the execution object of the target intent, and to determine the sub-intent list of the target intent.

[0223] The first sub-intention selection module is used to select the first sub-intention from the list of sub-intentions of the target intention based on the policy information corresponding to the policy of the target intention or the learning model information corresponding to the learning model of the target intention, as well as the network environment of the intention input information and the execution object of the target intention.

[0224] The first transmission information determination module is configured to determine transmission information between the target intent and each of the child intents of the target intent based on a policy corresponding to the policy information of the target intent or a learning model corresponding to the learning model information of the target intent, the intent input information, and a network environment of the execution object of the target intent, the target expectation information of the target intent, and the target scope information of the target intent.

[0225] The first mode selection module is configured to, in a case where the first child intent does not include a child intent, select an intent implementation mode based on a policy corresponding to the policy information of the first child intent or a learning model corresponding to the learning model information of the first child intent, a network environment of an execution object of the first child intent, the transmission information between the target intent and the first child intent, the target expectation information of the first child intent, and the target scope information of the first child intent, and determine the intent implementation mode of the first child intent.

[0226] The first execution module is configured to execute the intent implementation mode of the first child intent in the execution object of the first child intent, and obtain an execution result of the first child intent.

[0227] In one embodiment, the target expectation information of the first child intent is determined from the intent expectation information of the first child intent in a first template of the first child intent, the first template being determined by a first policy corresponding to first policy information or a first template corresponding to first template information, the first child intent template being an intent template corresponding to the first child intent, and the target scope information of the first child intent being determined from the intent scope information of the first child intent in a second template of the first child intent template, the second template being determined by a second policy corresponding to second policy information or a second template corresponding to second template information.

[0228] In one embodiment, the apparatus further comprises:

[0229] The first processing module is configured to, in a case where the first child intent includes a child intent, reselect the target expectation information of the target intent, the target scope information of the target intent, the list of child intents of the target intent, the first child intent of the target intent, and the transmission information between the target intent and each of the child intents of the target intent, until the first child intent of the target intent does not include a child intent.

[0230] In one embodiment, the apparatus further comprises:

[0231] a successor intention selection module configured to, after the first sub-intention is executed in the intention execution object of the first sub-intention to obtain an execution result of the first sub-intention, select a successor intention of the first sub-intention from the sub-intention list based on a strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or a learning model corresponding to the learning model information of the first sub-intention, and a network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention, the first sub-intention template being an intention template corresponding to the first sub-intention;

[0232] a second transfer information determination module configured to, based on the strategy corresponding to the strategy information of the first sub-intention or the learning model corresponding to the learning model information of the first sub-intention, and the network environment of the intention execution object of the first sub-intention, select transfer information between the first sub-intention and the successor intention of the first sub-intention from the execution result of the first sub-intention;

[0233] a second mode selection module configured to, in a case where the successor intention of the first sub-intention does not include a sub-intention, select an intention implementation manner of the successor intention of the first sub-intention based on a strategy corresponding to the strategy information of the successor intention in the successor intention template or a learning model corresponding to the learning model information of the successor intention, and a network environment of the intention execution object of the successor intention, the transfer information between the first sub-intention and the successor intention of the first sub-intention, target intention expectation information of the successor intention, and target scope information of the successor intention, the successor intention template being an intention template corresponding to the successor intention of the first sub-intention;

[0234] a second execution module configured to execute the intention implementation manner of the successor intention of the first sub-intention in the intention execution object of the successor intention to obtain an execution result of the successor intention.

[0235] In an embodiment, the apparatus further comprises:

[0236] a second processing module configured to, after the execution result of the successor intention is obtained, in a case where the sub-intention list of the target intention further includes an unexecuted sub-intention, determine the successor intention of the first sub-intention as the first sub-intention, and return to the step of selecting the successor intention from the sub-intention list based on the strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or the learning model corresponding to the learning model information of the first sub-intention, and the network environment of the intention execution object and the execution result of the first sub-intention, and determine the successor intention of the first sub-intention until the sub-intentions in the sub-intention list of the target intention are executed, determine that the target intention is executed, and obtain an execution result of the target intention.

[0237] In one embodiment, the target expected information of the successor intention of the first sub-intention is expected information determined in the intention expected information of the successor intention of the successor intention template through a first strategy corresponding to first strategy information in the successor intention template or a first template corresponding to first template information, and the target scope information of the successor intention is scope information determined in the intention scope information of the first sub-intention of the successor intention template through a second strategy corresponding to second strategy information in the successor intention template or a second template corresponding to second template information.

[0238] In one embodiment, the device further comprises:

[0239] The third processing module is configured to, after obtaining the execution result of the target intention, in a case where the target intention does not match the intention corresponding to the target intention template determined by the template matching, taking the target intention as the first sub-intention, taking the parent intention of the first sub-intention as the target intention, in a case where the sub-intention list of the target intention further includes an unexecuted sub-intention, returning a strategy corresponding to the strategy information of the first sub-intention in the first sub-intention template or a learning model corresponding to the learning model information of the first sub-intention, and the intention input information of the first sub-intention, the network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention, performing the successor intention selection in the sub-intention list, and re-determining the successor intention in the step of determining the successor intention of the first sub-intention until the sub-intention list of the template matching target intention is executed completely.

[0240] It should be noted that the device corresponds to the above method, and all implementation manners in the above method embodiments are applicable to the embodiments of the device and can achieve the same technical effects.

[0241] Embodiments of the present application also provide a communication device, comprising a processor and a memory storing a computer program, wherein the computer program is executed by the processor to perform the above method. All implementation manners in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0242] Embodiments of the present application also provide a computer readable storage medium comprising instructions, which, when executed on a computer, cause the computer to perform the above method. All implementation manners in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0243] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0244] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0245] In the embodiments provided by the present application, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. There can be another division during actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0246] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0247] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0248] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the technical solutions that essentially contribute to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk, and various storage media that can store program codes.

[0249] Furthermore, it is to be understood that the components and / or steps of the device and method of the present application can be decomposed and / or recombined. Such decompositions and / or recombinations are to be considered as equivalent to the present application. Also, the steps of performing the series of processes described above can be executed in time series naturally following the order of description, but do not necessarily have to be executed in time series. Some steps can be executed in parallel or independently of each other. It is to be understood by those skilled in the art that all or any of the steps or components of the method and device of the present application can be implemented in hardware, firmware, software, or a combination thereof, in any computing device (including a processor, a storage medium, etc.) or network of computing devices, using the basic programming skills of those skilled in the art upon reading the description of the present application.

[0250] Therefore, the object of the present application can also be achieved by running a program or a set of programs on any computing device. The computing device can be a commonly known general-purpose device. Therefore, the object of the present application can also be achieved only by providing a program product containing program code for implementing the method or device. That is, such a program product also constitutes the present application, and a storage medium storing such a program product also constitutes the present application. Obviously, the storage medium can be any commonly known storage medium or any storage medium developed in the future. It is to be understood that the components and / or steps of the device and method of the present application can be decomposed and / or recombined. Such decompositions and / or recombinations are to be considered as equivalent to the present application. Also, the steps of performing the series of processes described above can be executed in time series naturally following the order of description, but do not necessarily have to be executed in time series. Some steps can be executed in parallel or independently of each other.

[0251] The above is the preferred embodiment of the present application. It should be noted that for those skilled in the art, several improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements are also to be considered as within the scope of the present application.

Claims

1. An intent execution method, characterized in that, The method includes: Obtain intent input information; Based on the intent input information, template matching is performed on multiple saved intent templates to determine the target intent template. The target intent template includes intent description information of the target intent, as well as strategy information or learning model information associated with the target intent. The strategy corresponding to the strategy information or the learning model corresponding to the learning model information is used to select information from the target intent template. Based on the intent input information and the target intent template, execute the target intent; Executing the target intent based on the intent input information and the target intent template includes: Based on the strategy corresponding to the strategy information of the target intent or the learning model corresponding to the learning model information of the target intent, as well as the network environment of the intent input information and the execution object of the target intent, the following processing is performed: Intent expectation selection is performed in the intent expectation information of the target intent to determine the target expectation information of the target intent; intent scope selection is performed in the intent scope information of the target intent to determine the target scope information of the target intent; sub-intent selection is performed in the sub-intents of the target intent to determine the sub-intent list of the target intent; first sub-intent selection is performed in the sub-intent list of the target intent to determine the first sub-intent of the target intent; and information transmission between parent and child intents is selected in the target expectation information and the target scope information of the target intent to obtain the transmission information between the target intent and each of its sub-intents. When the first sub-intention does not include a sub-intention, the intention implementation method of the first sub-intention is determined by selecting an intention implementation method based on the policy corresponding to the policy information of the first sub-intention or the learning model corresponding to the learning model information of the first sub-intention, as well as the network environment of the intention execution object of the first sub-intention, the transmission information between the target intention and the first sub-intention, the target expectation information of the first sub-intention, and the target scope information of the first sub-intention. The intent implementation method of the first sub-intent is executed in the intent execution object of the first sub-intent to obtain the execution result of the first sub-intent.

2. The intent execution method according to claim 1, characterized in that, The intent description information includes at least one of the following: Intended expected information; Intent scope information; How the intent is realized; Sub-intent information; Subsequent intent information; Intended to output information.

3. The intention execution method according to claim 1, characterized in that, The strategy corresponding to the strategy information includes at least one of the following: The first strategy is used for selecting intended expectations; The second strategy is used for selecting the scope of intent; The third strategy is used to select the method of achieving the intention; The fourth strategy is used for selecting sub-intents; The fifth strategy is used for selecting the primary intention; The sixth strategy is used for selecting subsequent intentions; The seventh strategy is used to select the information to be passed between parent and child intentions; The eighth strategy is used for selecting the information to be transmitted between a preceding intention and a succeeding intention, wherein the preceding intention and the succeeding intention are sub-intentions of the same level that have a preceding and succeeding relationship.

4. The intent execution method according to claim 1, characterized in that, The learning model information corresponds to at least one of the following: The first model is used for selecting intention expectations; The second model is used for selecting the scope of intent; The third model is used to select the method of achieving the intent; The fourth model is used for selecting sub-intents; The fifth model is used for the selection of the primary intention; The sixth model is used for selecting subsequent intentions; The seventh model is used for selecting the information to be passed between parent and child intentions; The eighth model is used for selecting the information to be transmitted between a preceding intention and a succeeding intention, wherein the preceding intention and the succeeding intention are sub-intentions of the same level that have a preceding and succeeding relationship.

5. The intention execution method according to claim 1, characterized in that, The target expectation information of the first sub-intention is the expectation information determined in the intention expectation information of the first sub-intention in the first sub-intention template through the first strategy corresponding to the first strategy information or the first template corresponding to the first template information in the first sub-intention template. The first sub-intention template is the intention template corresponding to the first sub-intention. The target scope information of the first sub-intention is the scope information determined in the intention scope information of the first sub-intention in the first sub-intention template through the second strategy corresponding to the second strategy information or the second template corresponding to the second template information in the second strategy information or the second template corresponding to the second template information in the first sub-intention template.

6. The intent execution method according to claim 1, characterized in that, Also includes: If the first sub-intent includes sub-intents, the first sub-intent of the target intent is taken as the target intent, and the target intent expectation information, the target scope information, the sub-intent list, the first sub-intent, and the transmission information between the target intent and each sub-intent of the target intent are reselected until the first sub-intent of the target intent does not include sub-intents.

7. The intention execution method according to claim 1, characterized in that, After executing the intent implementation method of the first sub-intent in the intent execution object of the first sub-intent and obtaining the execution result of the first sub-intent, the process further includes: Based on the policy corresponding to the policy information of the first sub-intention in the first sub-intention template or the learning model corresponding to the learning model information of the first sub-intention, as well as the network environment of the intention execution object of the first sub-intention and the execution result of the first sub-intention, the subsequent intention is selected in the sub-intention list to determine the subsequent intention of the first sub-intention, and the first sub-intention template is the intention template corresponding to the first sub-intention. Based on the strategy corresponding to the strategy information of the first sub-intention or the learning model corresponding to the learning model information of the first sub-intention, and the network environment of the intention execution object of the first sub-intention, the information transmission between the predecessor and successor intentions is selected in the execution result of the first sub-intention to obtain the transmission information between the first sub-intention and the successor intention of the first sub-intention. When the successor intent of the first sub-intent does not include a sub-intent, the intent implementation method is selected based on the policy corresponding to the policy information of the successor intent in the successor intent template or the learning model corresponding to the learning model information of the successor intent, as well as the network environment of the intent execution object of the successor intent, the transmission information between the first sub-intent and the successor intent of the first sub-intent, the target intent expectation information of the successor intent, and the target scope information of the successor intent, and the intent implementation method of the successor intent is determined. The successor intent template is the intent template corresponding to the successor intent of the first sub-intent. The intent implementation method of the successor intent of the first sub-intent is executed in the intent execution object of the successor intent to obtain the execution result of the successor intent.

8. The intention execution method according to claim 7, characterized in that, After obtaining the execution result of the subsequent intent, the method further includes: If the sub-intent list of the target intent also includes unexecuted sub-intents, the successor intent of the first sub-intent is taken as the first sub-intent. The policy corresponding to the policy information of the first sub-intent in the first sub-intent template or the learning model corresponding to the learning model information of the first sub-intent, as well as the network environment of the intent execution object and the execution result of the first sub-intent, are returned. The successor intent is selected in the sub-intent list. The step of determining the successor intent of the first sub-intent is repeated until all sub-intents in the sub-intent list of the target intent are executed. The execution result of the target intent is then obtained.

9. The intention execution method according to claim 7, characterized in that, The target expectation information of the successor intention of the first sub-intention is the expectation information determined in the intention expectation information of the successor intention in the successor intention template through the first strategy corresponding to the first strategy information or the first template corresponding to the first template information in the successor intention template. The target scope information of the successor intention is the scope information determined in the intention scope information of the first sub-intention in the successor intention template through the second strategy corresponding to the second strategy information or the second template corresponding to the second template information in the successor intention template.

10. The intention execution method according to claim 7, characterized in that, After obtaining the execution result of the target intent, the process further includes: If the target intent does not match the intent corresponding to the target intent template determined by template matching, the target intent is taken as the first sub-intent, and the parent intent of the first sub-intent is taken as the target intent. If the sub-intent list of the target intent still includes unexecuted sub-intents, the strategy corresponding to the strategy information of the first sub-intent in the first sub-intent template or the learning model corresponding to the learning model information of the first sub-intent, as well as the intent input information of the first sub-intent, the network environment of the intent execution object of the first sub-intent, and the execution result of the first sub-intent are returned. The subsequent intent is selected in the sub-intent list, and the step of determining the subsequent intent of the first sub-intent is re-determined until the sub-intent list of the target intent matched by the template is executed.

11. An intent execution device, characterized in that, The device includes: The first acquisition module is used to acquire intent input information; The matching module is used to perform template matching among multiple saved intent templates based on the intent input information to determine the target intent template. The target intent template includes intent description information of the target intent, as well as strategy information or learning model information associated with the target intent. The strategy corresponding to the strategy information or the learning model corresponding to the learning model information is used to select information from the target intent template. The execution module is used to execute the target intent based on the intent input information and the target intent template; The execution module includes: The expectation selection module is used to select the target expectation information of the target intention based on the strategy information corresponding to the strategy information of the target intention or the learning model information corresponding to the learning model information of the target intention, as well as the intention input information and the network environment of the execution object of the target intention. The scope selection module is used to select the scope of intent in the scope information of the target intent and determine the scope information of the target intent. The sub-intent selection module is used to select sub-intents from the sub-intents of the target intent and determine the sub-intent list of the target intent; The first sub-intent selection module is used to select the first sub-intent from the list of sub-intents of the target intent and determine the first sub-intent of the target intent; The first transmission information determination module is used to select the transmission information between parent and child intentions from the target expectation information of the target intention and the target scope information of the target intention, so as to obtain the transmission information between the target intention and each of the child intentions of the target intention. The first method selection module is used to select the intention implementation method of the first sub-intention based on the strategy corresponding to the strategy information of the first sub-intention or the learning model corresponding to the learning model information of the first sub-intention, as well as the network environment of the intention execution object of the first sub-intention, the transmission information between the target intention and the first sub-intention, the target expectation information of the first sub-intention and the target scope information of the first sub-intention, when the first sub-intention does not include a sub-intention. The first execution module is used to execute the intent implementation method of the first sub-intent in the intent execution object of the first sub-intent, and obtain the execution result of the first sub-intent.

12. A communication device, characterized in that, include: A processor, a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any one of claims 1 to 10.

13. A computer-readable storage medium, characterized in that, Store instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 10.

Images