A communication method and apparatus

Abstract

The application discloses a communication method and device, and belongs to the technical field of communication. In the method, a first network element receives first information, and the first information is used for requesting to perform a first task. The first network element acquires experience information associated with the first task, and the experience information is used for scheduling the first task. The experience information comprises value information and / or risk information. The value information comprises a scheduling method corresponding to a successful task and a service function list determined based on a service function participating in the successful task. The risk information comprises risk characteristic information determined based on a failure factor of a failed task. The successful task is a historical successful task associated with the first task, and the failed task is a historical failed task associated with the first task. The method is beneficial to improving the success rate of the first task performed by the network element.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a communication method and apparatus. Background Technology

[0002] With the rapid development and widespread application of communication technologies, network intelligence has become crucial for improving network performance, optimizing user experience, and enhancing network management efficiency. In network intelligence, network agents can collaborate with various network functions to form a closed-loop generation system, flexibly orchestrating, deploying, executing, and reclaiming end-to-end network slices to perform specific network tasks.

[0003] Improving the success rate of network intelligent agents in performing tasks is a problem that needs to be solved. Summary of the Invention

[0004] This application provides a communication method and apparatus that can help improve the success rate of network intelligent agents in performing tasks.

[0005] In a first aspect, embodiments of this application provide a communication method. This method can be applied to a first network element, a chip within the first network element, or a logic module or software capable of implementing all or part of the functions of the first network element. The following description uses a first network element as an example. The method includes: the first network element receiving first information, the first information being used to request the execution of a first task; the first network element acquiring experience information associated with the first task, the experience information being used to orchestrate the first task. The experience information includes value information and / or risk information. The value information includes an orchestration method corresponding to successful tasks and a list of service functions determined based on the service functions participating in successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks. Successful tasks are historical successful tasks associated with the first task, and failed tasks are historical failed tasks associated with the first task.

[0006] In this method, after receiving first information for requesting the execution of a first task, the first network element acquires experience information for orchestrating the first task, enabling the first network element or other network elements to orchestrate the first task based on the experience information associated with it. This experience information includes value information determined based on orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on failure factors of historically failed tasks associated with the first task. Therefore, orchestrating the first task based on this experience information allows the network element to learn from historical tasks, thereby improving the success rate of executing the first task.

[0007] In one optional implementation, the first network element obtains experience information associated with the first task by: sending second information to a second network element, the second information being used to request the acquisition of experience information associated with the first task, the second information including information for instructing the first task; and receiving experience information from the second network element.

[0008] Based on the above scheme, the first network element obtains the experience information associated with the first task by requesting it from the second network element. This is beneficial for the network element to arrange the first task based on the experience information associated with the first task, thereby improving the success rate of the network element in executing the first task.

[0009] In another optional implementation, the first network element may also perform the following operations: send third information to the second network element, the third information being used to request subscription to experience information associated with the first task, the third information including subscription type and information for instructing the first task, the subscription type including value type and / or risk type.

[0010] Based on the above scheme, the first network element requests the second network element to subscribe to the experience information associated with the first task, which is beneficial for the second network element to provide the first network element with the experience information associated with the first task in a timely manner. This, in turn, is beneficial for the first network element or other network elements to arrange the first task based on the latest experience information associated with the first task.

[0011] In one optional implementation, the first network element receives fourth information from the second network element, which indicates whether the subscription was successful or failed. This method enables the first network element to know whether the experience information associated with the first task has been successfully subscribed to.

[0012] In one optional implementation, when the first network element requests the second network element to subscribe to the experience information associated with the first task, obtaining the experience information associated with the first task includes: receiving experience information from the second network element.

[0013] Based on the above scheme, the first network element obtains the experience information associated with the first task by requesting a subscription from the second network element. Therefore, after receiving the first information used to request the execution of the first task, it can directly use the experience information associated with the first task obtained through subscription to schedule the first task. Compared to the first network element receiving the first information used to request the execution of the first task and then requesting the experience information associated with the first task from the second network element, this method reduces the number of requests made by the first network element.

[0014] In one alternative implementation, the experience information further includes experience type and / or information for indicating a second task, the experience type including value type and / or risk type, the second task being a historical task associated with the first task.

[0015] When the experience type includes a value type, it helps to indicate that the corresponding experience information is directly applicable / referenced / used; when the experience type includes a risk type, it helps to indicate that the corresponding experience information needs to be avoided. Furthermore, when the experience information also includes information used to indicate a second task, it can present the task referenced in obtaining the experience information, which helps network elements scheduling the first task determine whether to use the experience information.

[0016] In one optional implementation, the first network element sends inference information of the first task to the second network element. The inference information of the first task is obtained based on the orchestration result of the first task and the execution of the first task by the generative network instance.

[0017] Based on the above scheme, the first network element sends the inference information after the first task is executed to the second network element, which is beneficial for the second network element to use the inference information as an experience reference for the next task, such as as an experience reference for the next task related to the first task.

[0018] In one optional implementation, the reasoning information of the first task includes at least one of the following: an identifier of the first task, information for indicating the first task, an identifier of the first network element, a list of subtasks in the first task, a list of service functions corresponding to the first task, an identifier of the generative network, and feedback on the operation status of the generative network.

[0019] Secondly, embodiments of this application also provide a communication method. This method can be applied to a second network element, a chip within the second network element, or a logic module or software capable of implementing all or part of the functions of the second network element. The following description uses a second network element as an example. The method includes: the second network element receiving second information from a first network element, the second information being used to request the acquisition of experience information associated with a first task, the second information including information for instructing the first task. The second network element sends the experience information associated with the first task to the first network element, the experience information being obtained based on historical reasoning information associated with the first task. The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to successful tasks in the historical reasoning information and a list of service functions determined based on the service functions participating in successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks in the historical reasoning information. Successful tasks are historical successful tasks associated with the first task, and failed tasks are historical failed tasks associated with the first task.

[0020] As can be seen, in this method, when the first network element requests the experience information associated with the first task, the second network element sends the experience information associated with the first task to the first network element. This is beneficial for the first network element or other network elements to orchestrate the first task based on the experience information. The experience information includes value information determined based on the orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on the failure factors of historically failed tasks associated with the first task. Therefore, when a network element orchestrates the first task based on this experience information, it is beneficial for the network element to learn from historical tasks, thereby improving the success rate of the network element in executing the first task.

[0021] In one alternative implementation, the experience information further includes experience type and / or information for indicating a second task, the experience type including value type and / or risk type, the second task being a historical task associated with the first task.

[0022] When the experience type includes a value type, it helps to indicate that the corresponding experience information is directly applicable / referenced / used; when the experience type includes a risk type, it helps to indicate that the corresponding experience information needs to be avoided. Furthermore, when the experience information also includes information used to indicate a second task, it can present the task referenced in obtaining the experience information, which helps network elements scheduling the first task determine whether to use the experience information.

[0023] In one optional implementation, the second network element obtains the experience information associated with the first task based on the historical reasoning information associated with the first task, thereby enabling it to provide the experience information associated with the first task to the first network element.

[0024] In one optional implementation, the second network element obtains historical reasoning information associated with the first task from an experience set. This historical reasoning information includes k valuable experiences and / or l risk experiences, where the k valuable experiences are associated with successful tasks and the l risk experiences are associated with failed tasks. The experience set includes valuable and / or risk experiences from multiple tasks, including tasks associated with the first task, and k and l are positive integers. In this approach, the second network element obtains experience information associated with the first task based on the historical reasoning information, including: obtaining value information based on the k valuable experiences, and / or, obtaining risk information based on the l risk experiences.

[0025] Based on the above scheme, the value information in the experience information is obtained by the second network element based on k value experiences associated with the first task, and the risk information in the experience information is obtained by the second network element based on l risk experiences associated with the first task. Furthermore, the k value experiences are associated with successful tasks, and the l risk experiences are associated with failed tasks. Therefore, the value information can serve as a positive reference for scheduling the first task, and the risk information can serve as a negative reference for scheduling the first task. This is beneficial to improving the success rate of scheduling the first task by the first network element or other network elements based on the experience information.

[0026] In one alternative implementation, the value information is determined based on the success rate and task satisfaction of k value experiences, which is beneficial to improving the success rate of the first network element or other network elements in arranging the first task based on the value information.

[0027] In one optional implementation, the second network element may further perform the following operations: receiving inference information for multiple tasks, the inference information including at least one of the following: task identifier, information for indicating the task, network element identifier, list of subtasks in the task, list of service functions corresponding to the task, identifier of the generative network corresponding to the task, and feedback on the operation status of the generative network, wherein the multiple tasks include tasks associated with the first task; and determining an experience set based on the inference information of the multiple tasks.

[0028] Based on the above scheme, the second network element can divide the reasoning information of multiple tasks into a valuable set of value experience and a risky set of risk experience. This is beneficial for retrieving the value experience associated with the first task from the value experience set and / or retrieving the risk experience associated with the first task from the risk experience set, thereby facilitating the acquisition of experience information associated with the first task.

[0029] In one optional implementation, when the experience information includes value information, the second network element sends fifth information to the third network element. This fifth information instructs the verification of the service function list in the value information. The fifth information includes information indicating the first task and the service function list in the value information. In this method, the second network element sending experience information to the first network element includes: sending the experience information to the first network element if the service function list in the value information has passed verification.

[0030] Based on the above scheme, the second network element requests the third network element to verify the service function list in the acquired value information. Only if the service function list in the value information is verified successfully will the second network element send the experience information to the first network element. This method ensures that the service function list sent by the second network element to the first network element guarantees the successful execution of the first task, thereby increasing the probability of the first task being successfully executed.

[0031] Thirdly, embodiments of this application also provide a communication method. This method can be applied to a second network element, a chip within the second network element, or a logic module or software capable of implementing all or part of the functions of the second network element. The following description uses a second network element as an example. The method includes: the second network element receiving third information from a first network element. The third information is used to request subscription to experience information associated with a first task. The third information includes a subscription type and information indicating the first task. The subscription type includes a value type and / or a risk type. The second network element sends the experience information to the first network element. The experience information includes value information and / or risk information. The value information includes an orchestration method corresponding to a successful task and a list of service functions determined based on the service functions participating in the successful task. The risk information includes risk characteristic information determined based on the failure factors of a failed task. A successful task is a historical successful task associated with the first task, and a failed task is a historical failed task associated with the first task.

[0032] As can be seen, in this method, when the first network element requests to subscribe to the experience information associated with the first task, the second network element sends the experience information associated with the first task to the first network element. This is beneficial for the first network element or other network elements to orchestrate the first task based on the experience information associated with the first task. The experience information includes value information determined based on the orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on the failure factors of historically failed tasks associated with the first task. Therefore, when a network element orchestrates the first task based on this experience information, it is beneficial for the network element to learn from historical tasks, thereby improving the success rate of the network element in executing the first task.

[0033] Optionally, other embodiments of the third aspect can be found in the embodiments of the second aspect described above, and will not be repeated here.

[0034] Fourthly, embodiments of this application also provide a communication device. This communication device has some or all of the functions of the first network element described in the first aspect, or it has implemented some or all of the functions of the second network element described in the second or third aspect. For example, the communication device may possess some or all of the functions of the first network element described in the first aspect of this application, or it may possess the functions of any one of the embodiments of this application implemented individually. The functions can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions.

[0035] In one possible design, the communication device may include a processing unit and a communication unit. The processing unit is configured to support the communication device in performing the corresponding functions described in the above method. The communication unit is used to support communication between the communication device and other communication devices. The communication device may also include a storage unit coupled to the processing unit and the communication unit, which stores necessary program instructions and data for the communication device.

[0036] In one embodiment, the communication device includes a processing unit and a communication unit, and the device is applied to a first network element;

[0037] The communication unit is used to receive first information, which is used to request the execution of a first task;

[0038] The processing unit is used to obtain experience information associated with the first task, and the experience information is used to arrange the first task.

[0039] The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to successful tasks and a list of service functions determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0040] In addition, other alternative implementations of the communication device in this regard can be found in the relevant content of the first aspect above, and will not be described in detail here.

[0041] In another embodiment, the communication device includes a processing unit and a communication unit, and the device is applied to a second network element;

[0042] The communication unit is configured to receive second information from the first network element, the second information being used to request the acquisition of experience information associated with the first task, the second information including information for instructing the first task;

[0043] The communication unit is further configured to send the experience information to the first network element, wherein the experience information is obtained based on the historical reasoning information associated with the first task;

[0044] The experience information includes value information and / or risk information. The value information includes the arrangement method corresponding to successful tasks in the historical reasoning information and the service function list determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks in the historical reasoning information. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0045] In addition, other alternative implementations of the communication device in this regard can be found in the relevant content of the second aspect above, and will not be described in detail here.

[0046] In another embodiment, the communication device includes a processing unit and a communication unit, the device being applied to a second network element, and the processing unit being used to process signals / signaling;

[0047] The communication unit is configured to receive third information from the first network element, the third information being used to request subscription to experience information associated with the first task, the third information including a subscription type and information for instructing the first task, the subscription type including a value type and / or a risk type;

[0048] The communication unit is also used to send the experience information to the first network element;

[0049] The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to successful tasks and a list of service functions determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0050] In addition, other alternative implementations of the communication device in this regard can be found in the relevant content of the third aspect above, and will not be described in detail here.

[0051] As an example, the processing unit can be a processor, and the communication unit can be a transceiver unit, transceiver, or communication interface. It is understood that when the communication device is a device (e.g., a first network element or a second network element), the communication unit can be a transceiver within the communication device (e.g., a transceiver includes a transmitter and a receiver), implemented, for example, through an antenna, feeder, and codec within the communication device. Alternatively, if the communication device is a chip located within a device, the processing unit can be the chip's processing circuitry, logic circuitry, etc., and the communication unit can be the chip's input / output interface, such as input / output circuitry, pins, etc.

[0052] In another embodiment, the communication device is a chip or chip system. The processing unit may also be a processing circuit or logic circuit; the communication unit may be an input / output interface, interface circuit, output circuit, input circuit, pin, or related circuit on the chip or chip system.

[0053] In implementation, the processor can be used for, but is not limited to, baseband-related processing, and the transceiver can be used for, but is not limited to, radio frequency transceiver. These devices can be disposed on separate chips, or at least partially or entirely on the same chip. For example, the processor can be further divided into analog baseband processors and digital baseband processors. The analog baseband processor can be integrated with the transceiver on the same chip, while the digital baseband processor can be disposed on a separate chip. With the continuous development of integrated circuit technology, more and more devices can be integrated on the same chip. For example, a digital baseband processor can be integrated with multiple application processors (e.g., but not limited to graphics processors, multimedia processors, etc.) on the same chip. Such a chip can be called a system-on-a-chip (SoC). Whether the various devices are disposed independently on different chips or integrated on one or more chips often depends on the needs of the product design. This application does not limit the implementation form of the above-mentioned devices.

[0054] Fifthly, embodiments of this application also provide a processor for executing the various methods described above. During the execution of these methods, the processes of sending and receiving the aforementioned information can be understood as the processor outputting the aforementioned information and the processor receiving the input information. When outputting the aforementioned information, the processor outputs the information to a transceiver for transmission. After being output by the processor, the information may require further processing before reaching the transceiver. Similarly, when the processor receives the input information, the transceiver receives the information and inputs it to the processor. Furthermore, after the transceiver receives the information, the information may require further processing before being input to the processor.

[0055] Unless otherwise specified, or unless it contradicts its actual function or internal logic in the relevant description, the transmission and reception operations involved by the processor can be more generally understood as processor output and reception, input and other operations, rather than transmission and reception operations directly performed by radio frequency circuits and antennas.

[0056] In implementation, the processor can be a dedicated processor for executing these methods, or it can be a processor that executes computer instructions stored in memory to execute these methods, such as a general-purpose processor. The memory can be a non-transitory memory, such as read-only memory (ROM), which can be integrated with the processor on the same chip or disposed on different chips. This application does not limit the type of memory or the arrangement of the memory and processor.

[0057] Sixthly, embodiments of this application also provide a communication system, which includes a first network element for performing the method described in the first aspect and a second network element for performing the method described in the second or third aspect. In another possible design, the system may further include other means / functional network elements that interact with at least one of the first and second network elements.

[0058] In a seventh aspect, embodiments of this application provide a computer-readable storage medium for storing instructions that, when executed on a communication device, implement the method described in any one of the first to third aspects.

[0059] Eighthly, embodiments of this application also provide a computer program product including instructions that, when executed on a communication device, implement the method described in any one of the first to third aspects.

[0060] Ninthly, this application provides a chip including a processor (or logic circuit) for implementing the methods in any possible implementation of any of the first to third aspects. Optionally, the chip may also include a communication interface (or interface). In one possible implementation, if the chip is the smallest processing unit in the entire machine, the chip may be a processor, or it may include a processor and a memory, or it may include a processor, a memory, and a transceiver for implementing the methods in any possible implementation of any of the first to third aspects.

[0061] Tenthly, this application provides a chip system. The chip system includes a processor and an interface, the interface being used to acquire programs or instructions, and the processor being used to invoke the programs or instructions to implement the functions involved in any of the first to third aspects. In one possible design, the chip system further includes a memory for storing necessary program instructions and data. The chip system may be composed of chips or may include chips and other discrete devices.

[0062] The beneficial effects of the fourth to tenth aspects mentioned above can be specifically referred to in the description of the beneficial effects of any of the first to third aspects, and will not be repeated here. Attached Figure Description

[0063] Figure 1 It is a schematic diagram of a system architecture;

[0064] Figure 2 This is a flowchart illustrating a communication method provided in an embodiment of this application;

[0065] Figure 3 This is a schematic diagram illustrating the division of a value experience set and a risk experience set according to an embodiment of this application;

[0066] Figure 4 This is an interactive schematic diagram of another communication method provided in an embodiment of this application;

[0067] Figure 5 This is a schematic diagram of a communication method provided in an embodiment of this application;

[0068] Figure 6 This is an interactive schematic diagram of another communication method provided in the embodiments of this application;

[0069] Figure 7 This is an interactive schematic diagram of another communication method provided in the embodiments of this application;

[0070] Figure 8 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0071] Figure 9 This is a schematic diagram of another communication device provided in an embodiment of this application. Detailed Implementation

[0072] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0073] Hereinafter, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this embodiment, unless otherwise stated, "a plurality of" means two or more.

[0074] Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0075] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner to facilitate understanding.

[0076] It should be understood that in the embodiments of this application, "at least one (item)" refers to one or more. "More than one" refers to two or more. "At least two (items)" refers to two or three or more. "And / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" can represent three cases: only A exists, only B exists, and A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple. Both "...when" and "if" indicate that a corresponding action will be taken under certain objective circumstances. They are not time limits, nor do they require a judgment action to be taken when the action is taken, nor do they imply any other limitations.

[0077] In the embodiments of this application, "instruction" may include: direct instruction, or indirect instruction, or explicit instruction, or implicit instruction.

[0078] In the embodiments of this application, "including" may include: direct inclusion, or indirect inclusion, or explicit inclusion, or implicit inclusion.

[0079] It should be understood that the prior art may change as the technical solutions evolve, and the technical solutions provided in the embodiments of this application are not limited to the prior art provided.

[0080] It should be noted that different embodiments or some steps (e.g., any one or more steps) in different embodiments of this application can be combined with each other to form new embodiments. It should also be noted that the steps or any one or more steps in different embodiments are not limited to include optional steps in a certain embodiment, mandatory steps in a certain embodiment, or both optional and mandatory steps in a certain embodiment; the embodiments of this application are not limited in this way.

[0081] It should be noted that, unless otherwise specified or in case of logical conflict, the terminology and / or descriptions between different embodiments are consistent and can be referenced in each other.

[0082] It should be noted that the order of the steps in the embodiments of this application is not limited by this application.

[0083] It should be noted that the order of judgment of different conditions in the embodiments of this application is not limited in the embodiments of this application.

[0084] It should be noted that the terms "after" and "time" in the embodiments of this application do not strictly limit the time point.

[0085] It should be noted that the nouns and terms used in the embodiments of this application are merely examples, and may be other names. The embodiments of this application are not limited to these.

[0086] To better understand the communication methods disclosed in the embodiments of this application, the systems applicable to the embodiments of this application are described below.

[0087] Please see Figure 1 , Figure 1 This is a schematic diagram of a system architecture. For example... Figure 1 As shown, the system architecture includes network agents and common components. The common components include network memory. The network agents and common components are connected through an agent-based interface (ABI). The network agents include, but are not limited to, planning agents, assembling agents, connection agents, and execution agents.

[0088] Optionally, common components also include a network actor and / or a network toolbox.

[0089] Optional, Figure 1 The system architecture shown also includes a network generative pre-trained transformer (NetGPT), which can also be referred to as a large network model. Optional, Figure 1 The system architecture shown also includes a network sandbox. Optionally, NetGPT and the network sandbox can be included in common components or exist independently outside of them.

[0090] Optional, Figure 1 The system architecture shown also includes terminals. Optional, Figure 1The system architecture shown also includes a radio access network (RAN). Optional, Figure 1 The system architecture shown also includes a gateway. Optional, Figure 1 The system architecture shown also includes generative networks (GN).

[0091] In this embodiment, the RAN can be an evolved universal terrestrial radio access (E-UTRA) system, a new radio (NR) system, or a future radio access system as defined in the 3rd generation partnership project (3GPP). The RAN can also include two or more of the aforementioned different radio access systems. The RAN can also be an open RAN (O-RAN). The RAN can be a terrestrial network communication system or a non-terrestrial network (NTN) communication system. The NTN system can be an NTN system integrated with 4th Generation (4G), 5th Generation (5G), and any future communication system, such as NR NTN, Internet of Things (IoT) NTN, etc. The NTN communication system can be, for example, a satellite communication system, and can also include unmanned aerial vehicles (UAVs), high altitude platform stations (HAPS), and other aerial access network equipment; this application does not limit this.

[0092] In this application embodiment, the RAN node, also known as network device, radio access network device, RAN entity, or access node, is used to help terminal devices access the communication system wirelessly. In one application scenario, the RAN node can be a base station, an evolved NodeB (eNodeB), a transmission reception point (TRP), a next-generation NodeB (gNB) in a 5G mobile communication system, or a base station in a future mobile communication system. The RAN node can be a macro base station, a micro base station, an indoor station, a relay node, or a donor node. In a satellite communication system, the RAN node can be a satellite or a base station device mounted on a satellite. The RAN node can also be a gateway station (or ground station, earth station, signal gateway, gateway, or gateway station), a high-altitude platform (HAP), a drone, a hot air balloon, a low-Earth orbit satellite, a medium-Earth orbit satellite, a high-Earth orbit satellite, etc., without limitation.

[0093] In another application scenario, multiple RAN nodes can collaborate to help terminals achieve wireless access, with different RAN nodes implementing different functions of the base station. For example, a RAN node can be a central unit (CU), a distributed unit (DU), or a radio unit (RU). Here, the CU performs the functions of the base station's Radio Resource Control (RRC) and Packet Data Convergence Protocol (PDCP), and can also perform the functions of the Service Data Adaptation Protocol (SDAP). The DU performs the functions of the base station's Radio Link Control (RANC) and Medium Access Control (MAC) layers, and can also perform some or all of the physical layer functions. For specific descriptions of these protocol layers, refer to the relevant 3GPP technical specifications. The RU can be used to implement radio frequency signal transmission and reception. The CU and DU can be two independent RAN nodes, or they can be integrated into the same RAN node, such as within a baseband unit (BBU). RUs can be included in radio frequency equipment, such as remote radio units (RRUs) or active antenna units (AAUs). CUs can be further divided into two types of RAN nodes: CU-control plane and CU-user plane.

[0094] In different systems, RAN nodes may have different names. For example, in an O-RAN system, a CU can be called an open CU (O-CU), a DU can be called an open DU (O-DU), and an RU can be called an open RU (O-RU). The RAN nodes in the embodiments of this application can be implemented through software modules, hardware modules, or a combination of software and hardware modules. For example, a RAN node can be a server loaded with the corresponding software modules. The embodiments of this application do not limit the specific technology or device form used in the RAN nodes. For ease of description, a base station is used as an example of a RAN node in the following description.

[0095] In this embodiment, the terminal is a device with wireless transceiver capabilities, capable of sending signals to or receiving signals from a base station. The terminal can also be referred to as a terminal device, user equipment (UE), mobile station, mobile terminal, etc. Terminals can be widely used in various scenarios, such as NTN, device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), IoT, virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, etc. Terminals can be mobile phones, tablets, computers with wireless transceiver capabilities, wearable devices, vehicles, airplanes, ships, robots, robotic arms, smart home devices, etc. Terminals can also be satellite communication terminals, such as very small aperture terminals (VSAT) (commonly referred to as VSAT terminals), portable stations, fixed stations, vehicle-mounted or airborne satellite communication terminals, etc. It should be understood that the satellite communication terminal communicates with satellites and can also act as a micro base station to provide a data interface to accessed user equipment. The embodiments of this application do not limit the specific technology or device form used in the terminal.

[0096] Optional, Figure 1 The system architecture shown is the A-Core architecture, which is a next-generation core network based on network agents. A-Core can leverage the reasoning capabilities and collaborative division of labor of multiple network agents to flexibly handle various user intentions and needs, generating generative network instances on demand without requiring users to provide detailed network configuration information.

[0097] Network intelligent agents are artificial intelligence (AI) agents designed for next-generation core networks. By working collaboratively with various network functional modules, network intelligent agents can construct a closed-loop generative system, enabling flexible orchestration, rapid deployment, efficient execution, and resource reclamation of generative network instances. Simultaneously, the system can continuously optimize performance based on real-time network feedback, achieving self-reflection and self-evolution.

[0098] The planning agent is responsible for receiving user-defined slice requests. Its main task is to understand the user's intent and break it down into multiple executable tasks. The assembly agent receives the output of the planning agent and selects appropriate network or application functions for each task.

[0099] The assembly agent is an entity that receives the output from the planning agent and selects appropriate network or application functions for each task. The assembly agent's output consists of two parts: a first part for configuring terminals and / or radio access networks, and a second part for deploying functions in the core network. The connection agent receives the first part of the assembly agent's output, determines which terminals and / or radio access networks will constitute GN instances, and determines their configuration parameters and resources. The execution agent receives the second part of the assembly agent's output and manages and controls the lifecycle of network slices.

[0100] Network memory, also known as public memory, is a shared component designed to enhance collaboration among multiple network agents. It is responsible for collecting, vectorizing, and storing knowledge and experience information not available to NetGPT, and for creating vectorized indexes of this data. Furthermore, network memory can respond to data retrieval requests from agents and provide them with relevant data. The data stored in network memory is divided into short-term memory and long-term memory. Long-term memory encompasses user registration information, network topology details (including the capabilities, location, and status data of user devices and wireless access networks), general network knowledge, and expert insights. Short-term memory includes the results of temporary tasks or functional coordination, the operational status of general network instances, the output of various functions, and the utilization rate of network resources.

[0101] NetGPT provides advanced network inference capabilities by fine-tuning large language models and optimizing them for mobile network-specific data. Based on input prompts, NetGPT assists each agent in generating inference results relevant to network slices, such as task decomposition and function selection. NetGPT typically employs appropriate prompting techniques to guide the model in solving complex inference problems, thereby generating higher-quality inference results.

[0102] ABI is the interface between network agents and public components. It can understand the purpose of a message based on NetGPT and decide how to forward the message to the appropriate network agent or public component.

[0103] Common components such as network actuators, network toolkits, and network sandboxes are used to assist agents in task orchestration, function assembly, connectivity configuration, and execution scheduling.

[0104] In this embodiment of the application, the first network element can be Figure 1 In the network agent shown in the system, the second network element can be... Figure 1 The network memory in the system shown can be a third network element. Figure 1The system shown also includes a network sandbox as a common component. For example, the first network element could be a planning agent. Another example is that the first network element could be an assembly agent. Yet another example is that the first network element could be a connection agent. And yet another example is that the first network element could be an execution agent.

[0105] Optionally, the first network element, the second network element, and the third network element can also be other network elements, and this application embodiment does not limit this.

[0106] Furthermore, when the solutions of this application embodiment are applied to future communication systems, the names of the corresponding network function entities may change, and this application does not limit this.

[0107] The embodiments disclosed in this application will be presented to illustrate various aspects, embodiments, or features of this application in relation to systems including multiple devices, components, modules, etc. It should be understood and appreciated that individual systems may include additional devices, components, modules, etc., and / or may not include all the devices, components, modules, etc. discussed in conjunction with the accompanying drawings. Furthermore, combinations of these approaches may also be used.

[0108] 5G networks can leverage network slicing technology, allowing operators to build dedicated, virtualized, and logical networks on common physical infrastructure to meet diverse communication needs. However, in practical applications, they still face numerous challenges and pain points. First, 5G network management primarily relies on manual configuration and preset rules, which is not only time-consuming and labor-intensive but also difficult to adapt to rapidly changing network environments and user demands. With the convergence of technologies such as the Internet of Things, big data, and artificial intelligence, user demands for network services are becoming increasingly personalized and diversified, requiring networks to provide more flexible, intelligent, and efficient services. However, existing network management methods struggle to meet these demands, leading to issues such as inconsistent service quality, uneven distribution of network resources, and complex management. Second, dynamic optimization of 5G network resources and personalized service customization are significant challenges facing current network intelligence. Traditional fixed network functions and service methods are difficult to dynamically adjust based on real-time user needs and feedback, which not only affects user experience but also wastes substantial network resources.

[0109] Future networks will face higher demands for intelligence, requiring not only the inheritance and optimization of 5G network functions but also the introduction of more advanced technologies and architectures to cope with increasingly complex and diverse communication needs. Network-native intelligence will become an inevitable trend in future network development, meaning that networks will no longer rely on manual configuration and preset rules but will be able to dynamically adapt to the network environment and user needs through self-learning, self-reflection, and self-optimization.

[0110] Network intelligence requires the network to possess the following key capabilities: (1) Adaptability: The network needs to be able to dynamically adjust network configuration and service strategies based on real-time data and user feedback to provide optimal network performance and user experience. (2) Personalized services: The network needs to be able to provide customized service solutions based on users' individual needs. This requires the network to have strong data analysis and processing capabilities, be able to extract valuable information from massive amounts of data, and apply it to service optimization. (3) On-demand generation: The network needs to be able to generate the required services and content in real time based on user needs. This requires the network to have strong intelligent orchestration and decision-making capabilities, be able to dynamically generate service solutions that meet user expectations.

[0111] Furthermore, large language models (LLMs) will play a crucial role in future networks. Through natural language processing and generation technologies, LLMs can understand and generate complex textual information, thereby providing intelligent decision support and personalized services to the network. For example, LLMs can analyze users' communication behavior and preferences to generate personalized network service plans; they can also dynamically adjust network configurations and optimize resource allocation by analyzing network data in real time, improving network performance and user experience. LLMs not only effectively address the current pain points in 5G network intelligence but also provide new ideas and solutions for future network development. Through continuous optimization and evolution of LLMs, future networks will be able to more intelligently, efficiently, and flexibly meet diverse user needs, providing users with a superior network experience.

[0112] This application provides a communication method 100. In this communication method 100, after a first network element receives first information for requesting the execution of a first task, it acquires experience information for orchestrating the first task, thereby enabling the first network element or other network elements to orchestrate the first task based on the experience information. The experience information includes value information determined based on orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on failure factors of historically failed tasks associated with the first task. Therefore, when a network element orchestrates the first task based on this experience information, it is beneficial to improve the success rate of the network element in executing the first task.

[0113] The embodiments of this application are described in detail below with reference to the accompanying drawings. The embodiments of this application use a first network element, a second network element, and a third network element as the implementing entities to illustrate the corresponding methods. For example, the first network element is... Figure 1 In the system shown, the network agent is, for example, a planning agent, and the second network element is... Figure 1 The network memory in the system shown, the third network element is Figure 1The network sandbox in the system shown. For example, the first network element in the method can also be a processor, module, chip, chip system, or software module that supports the implementation of the corresponding method; the second network element in the method can also be a processor, module, chip, chip system, or software module that supports the implementation of the corresponding method; and the third network element in the method can also be a processor, module, chip, chip system, or software module that supports the implementation of the corresponding method.

[0114] This application provides a communication method 100. Figure 2 This is a flowchart illustrating the communication method 100. The communication method 100 is described from the perspective of the first network element and includes, but is not limited to, the following steps:

[0115] S201. The first network element receives first information, which is used to request the execution of a first task.

[0116] Optionally, the first network element receives first information from the terminal device. For example, the first information may be intent information from the terminal device, which is used to request the execution of a first task, and the intent information is equivalent to a request to request the execution of the first task.

[0117] Accordingly, the terminal device sends first information to the first network element. Optionally, if the terminal device has a first task to be executed, it sends first information to the first network element to request the first network element to execute the first task. Optionally, the first task to be executed in the terminal device can be user-inputted, so the terminal device determines the first information based on the user-inputted first task to request the execution of the first task.

[0118] Optionally, the first network element receives first information from the cloud server. For example, the first information may be intent information from the cloud server, which is used to request the execution of a first task. In this mode, the first task may be initiated by a user of the cloud server to the cloud server, and the cloud server determines the first information based on the first task initiated by the user to request the execution of the first task.

[0119] Optionally, the first information includes information for indicating the first task, such as semantics for indicating the first task. For example, the first task is to collect statistics on the daily traffic flow of park A over the past week. The information for the first task includes a query for indicating the first task. For example, if the first information sent by the terminal device to the first network element includes a query for collecting statistics on the daily traffic flow of park A over the past week, then the first information is used to request statistics on the daily traffic flow of park A over the past week.

[0120] Optionally, the intent information may also be referred to as task request information or query information, etc., and this application embodiment does not limit it.

[0121] S202. The first network element obtains the experience information associated with the first task, and the experience information associated with the first task is used to orchestrate the first task.

[0122] The experiential information associated with the first task can be historical experiential information related to the first task, or experiential information determined based on inference information from historical tasks associated with the first task. This experiential information can serve as a reference / guide for orchestrating the first task; that is, the experiential information associated with the first task is used to orchestrate the first task. The inference information for the task includes information generated from task design and execution. Task design includes the planning agent and the assembly agent's processing of the task, such as the planning agent's decomposition of the task, and the assembly agent's deployment of functions in the terminal device and / or radio access network, as well as the core network, based on the planning agent's output, and outputting a service function list. Task execution includes the connecting agent and the executing agent deploying the task to the network based on the service function list output by the assembly agent, and constructing the topological connection between the radio access network and the generative network to execute the task.

[0123] The historical tasks associated with the first task can be historical tasks that match or are similar to the first task. For example, a historical task associated with the first task can be a task in the historical tasks whose matching degree / similarity with the first task is greater than a preset value. For example, a historical task associated with the first task can be a historical task whose semantic description matches or is similar to the semantic description of the first task to a preset value. The preset value can be pre-set or configured, and this application embodiment does not limit it.

[0124] For example, the first task is to count the daily traffic flow of park A over the past week. Historical tasks include Task A and Task B. Task A counts the daily traffic flow in the city center of city X last week, and Task B counts the traffic flow on bridge M of city X during the morning rush hour this week. If the semantic descriptions of Task A and Task B match the semantic description of the first task more than a preset value, then Task A and Task B can be considered historical tasks associated with the first task.

[0125] Additionally, the experiential information associated with the first task includes information indicating its relevance to historical tasks, such as value information and / or risk information. Value information is determined based on successful tasks associated with the first task, and risk information is determined based on failed tasks associated with the first task. Successful tasks are historically successful tasks associated with the first task, and failed tasks are historically failed tasks associated with the first task. Historically successful tasks associated with the first task refer to successful tasks in the historical tasks whose matching degree / similarity with the first task is greater than a preset value; for example, successful tasks in the historical tasks whose semantic description matches / similares with the semantic description of the first task to a preset value. Similarly, historically failed tasks associated with the first task refer to failed tasks in the historical tasks whose matching degree / similarity with the first task is greater than a preset value.

[0126] Alternatively, the experiential information associated with the first task includes value information determined based on historical successful tasks associated with it, and / or risk information determined based on historical failed tasks associated with it. Therefore, value information can serve as a positive reference for orchestrating the first task, while risk information can serve as a negative reference. Or, it can be understood that value information and risk information provide guidance and reference for the task orchestration, function assembly, connection configuration, and function deployment processes of the first task.

[0127] Optionally, the experience information may also include experience types and / or information for indicating a second task. The experience type may include value types and / or risk types, and the second task is a historical task associated with the first task. Wherein, if the experience information includes value information, the experience type is a value type, indicating that the value information is a successful experience determined based on a historical successful task associated with the first task; if the experience information includes risk information, the experience type is a risk type, indicating that the risk information is a failure experience determined based on a historical failed task associated with the first task. Furthermore, if the experience information also includes information for indicating a second task, it may present the task referenced in obtaining the experience information, facilitating the network elements orchestrating the first task to determine whether to use the experience information.

[0128] Optionally, the experience information may also include the identifier of the first network element. The identifier of the first network element is used by the first network element to determine that the received experience information was sent to itself.

[0129] It is evident that after receiving the intent information, the first network element acquires the experience information associated with and used to orchestrate the first task. This facilitates the first network element's orchestration of the first task based on the experience information associated with it, or enables other network elements to orchestrate the first task based on the experience information associated with it. Since the experience information associated with the first task is determined based on historical successful and / or historical failed tasks associated with it, orchestrating the first task based on this experience information helps improve the success rate of the network element in executing the first task.

[0130] In this embodiment, the first network element can obtain the experience information associated with the first task through various implementation methods, including but not limited to the following two implementation methods:

[0131] Implementation method 1: The first network element obtains the experience information associated with the first task by requesting the second network element.

[0132] In one optional implementation, the first network element acquires experience information associated with the first task by: sending second information to a second network element, the second information being used to request the acquisition of experience information associated with the first task, the second information including information for instructing the first task; and receiving experience information from the second network element. Correspondingly, the second network element performs the following operations: receiving the second information from the first network element; and sending the experience information associated with the first task to the first network element.

[0133] Optionally, the second information further includes at least one of the following: the identifier of the first network element and the identifier of the first task. Wherein, the second information including the identifier of the first network element enables the second network element to determine, based on the identifier of the first network element, the experience information associated with the first task to be fed back to the first network element. The second information including the identifier of the first task enables the second network element to retrieve the executed first task based on the identifier of the first task, and then obtain the experience information associated with the first task based on this identifier.

[0134] As can be seen, the first network element requests the experience information associated with the first task from the second network element through the second information. Consequently, the second network element responds to the second information by providing the experience information associated with the first task to the first network element, enabling the first network element to obtain the experience information associated with the first task.

[0135] Implementation method 2: The first network element obtains the experience information associated with the first task by subscribing to the second network element.

[0136] In one optional implementation, the first network element performs the following operations: sending third information to the second network element, the third information being used to request subscription to experience information associated with the first task, the third information including a subscription type and information indicating the first task, the subscription type including a value type and / or a risk type. Optionally, the first network element receives fourth information from the second network element, the fourth information being used to provide feedback on the subscription status, i.e., indicating whether the subscription was successful or failed. Correspondingly, the second network element receives the third information from the first network element. Optionally, the second network element sends the fourth information to the first network element.

[0137] When the subscription type includes a value type, it means that the experience information requested by the first network element from the second network element is value information, i.e., the request is for the value information associated with the first task. When the subscription type includes a risk type, it means that the experience information requested by the first network element from the second network element is risk information, i.e., the request is for the risk information associated with the first task. When the subscription type includes both value and risk types, it means that the first network element requests both value and risk information associated with the first task from the second network element.

[0138] As can be seen, the first network element requests subscription to the experience information associated with the first task from the second network element via the third information. The second network element can respond with the third information and send back the fourth information to the first network element, so that the first network element knows whether the subscription was successful. This method is beneficial for the second network element to provide the experience information associated with the first task to the first network element in a timely manner.

[0139] Optionally, the third information may also include at least one of the following: subscription identity (subscription ID), the identifier of the first network element, and the subscription period. The subscription identity represents the identifier of this subscription task. The identifier of the first network element is used by the second network element to confirm that this subscription was initiated by the first network element, so that it can subsequently provide feedback on the subscription status and content to the first network element. The subscription period represents the period during which the second network element needs to provide feedback on the experience information associated with the first task to the first network element, such as 24 hours or 7 days.

[0140] Optionally, the fourth information may also include the identifier of the first network element. The identifier of the first network element is used by the first network element to determine whether the fourth information is being fed back to itself.

[0141] Optionally, if the fourth information indicates that the subscription has failed, the fourth information may also be used to indicate the reason for the subscription failure, so that the first network element may be aware of the reason for the subscription failure.

[0142] Optionally, the third information may also be referred to as a subscription request or subscription information, and the fourth information may also be referred to as a subscription response or subscription reply. The naming of the third and fourth information is not limited in this application embodiment.

[0143] In one optional implementation, if the first network element has already subscribed to the experience information associated with the first task from the second network element, the first network element obtains the experience information associated with the first task by receiving the experience information associated with the first task from the second network element. Correspondingly, the second network element sends the experience information associated with the first task to the first network element.

[0144] Optionally, if the fourth information indicates successful subscription, the first network element receives the experience information associated with the first task from the second network element. Correspondingly, if the fourth information indicates successful subscription, the second network element sends the experience information associated with the first task to the first network element.

[0145] It is evident that if the first network element successfully subscribes to the experience information associated with the first task from the second network element, it can obtain the experience information associated with the first task provided by the second network element.

[0146] Optionally, when the second network element obtains / detects the experience information associated with the first task for the first time after the fourth information indicates that the subscription was successful, the second network element sends the experience information associated with the first task to the first network element.

[0147] Optionally, after the second network element sends the experience information associated with the first task to the first network element for the first time, if the experience information associated with the first task is updated, the second network element sends the updated experience information associated with the first task to the first network element. This is beneficial for the first network element to obtain the latest experience information, which in turn is beneficial for the first network element or other network elements to arrange the first task based on the updated experience information, thereby improving the success rate of the network element in executing the first task.

[0148] Optionally, if the subscription type in the third information is value-based, the experience information sent by the second network element to the first network element is value-based; if the subscription type in the third information is risk-based, the experience information sent by the second network element to the first network element is risk-based; if the subscription type in the third information includes both value-based and risk-based information, the experience information sent by the second network element to the first network element includes both value-based and risk-based information. Therefore, the second network element can provide the first network element with the corresponding type of experience information based on the first network element's subscription requirements.

[0149] Optionally, the second network element can send the experience information associated with the first task to the first network element via a subscription notification. For example, the second network element sends a subscription notification to the first network element, and the subscription notification includes the experience information associated with the first task.

[0150] Optionally, the subscription notification may also include at least one of the following: a subscription identifier, an experience identifier, a user intent list, a subtask decomposition list, a function matching set, and context information. The subscription identifier identifies the subscription task being fed back. The experience identifier identifies the feedback experience information, enabling the agent to distinguish between different experience information when multiple pieces of experience information are fed back. The user intent list is a list of user intents corresponding to historical tasks associated with the first task. The subtask decomposition list is a list of subtasks that can be broken down into when the experience information includes value information. The function matching set is a list of service functions that can participate in the first task when the experience information includes value information. The context information includes at least one of the following: time, location, network topology information, and resource utilization. Time and location are the time and location information included in the experience information recorded by the second network element, and may be included in the user's query information. For example, if the user's query information instructs the statistical analysis of traffic flow in Park A between 8:00 AM and 6:00 PM on Saturday, then the time in the context information is 8:00 AM to 6:00 PM on Saturday, and the location is Park A. The time and location context helps the network agent determine the similarity between the time and location information of the current task and the time and location information in its experience, thereby generating orchestrated reasoning results that are more in line with the user scenario. Furthermore, for the same query intent, different time and location information may correspond to different experience information. Network topology information refers to the network devices selected by the connecting agent in the first task, including base stations and terminal devices. Resource utilization refers to the computing resource information used in the deployment of the executing agent in the first task, including server equipment and its corresponding identification information, server memory utilization, and energy consumption.

[0151] As can be seen, the first network element can pre-subscribe to the experience information associated with the first task from the second network element to obtain the experience information associated with the first task. This method enables the first network element to obtain the latest experience information associated with the first task in a timely manner. In addition, compared with the first network element requesting the experience information associated with the first task from the second network element after receiving the first information for requesting the execution of the first task, this method can reduce the number of times the first network element requests the experience information associated with the first task from the second network element.

[0152] Optionally, the first network element can also request the second network element to subscribe to experience information related to other tasks besides the first task through third information. For example, it can request the second network element to subscribe to experience information related to the third task. In this case, the third information also includes information indicating the third task and the subscription type corresponding to the experience information related to the third task. The subscription type includes value type and / or risk type. In this method, the first network element can also obtain experience information related to other tasks besides the first task by subscribing to the second network element.

[0153] In one optional implementation, for implementation 1, before the second network element sends the experience information associated with the first task to the first network element, it also obtains the experience information associated with the first task.

[0154] In another optional implementation, for implementation 2, before the second network element sends the experience information associated with the first task to the first network element, for example, before sending the experience information associated with the first task to the first network element in the case of the fourth information indicating successful subscription, the experience information associated with the first task is also obtained.

[0155] It is evident that, regardless of the implementation method by which the first network element obtains the experience information associated with the first task, the second network element must also obtain the experience information associated with the first task before providing it to the first network element.

[0156] In one optional implementation, the second network element obtains the experience information associated with the first task, including: obtaining the experience information associated with the first task based on the historical reasoning information associated with the first task.

[0157] Optionally, the second network element obtains historical reasoning information associated with the first task from the experience set. This historical reasoning information includes k valuable experiences and / or l risk experiences. The k valuable experiences are associated with successful tasks, and the l risk experiences are associated with failed tasks. The experience set includes valuable experiences and / or risk experiences from multiple tasks, including tasks associated with the first task, where k and l are positive integers. In this approach, the second network element obtains experience information associated with the first task based on the historical reasoning information, including: obtaining value information from the experience information based on the k valuable experiences, and / or obtaining risk information from the experience information based on the l risk experiences.

[0158] In one optional implementation, the second network element further performs the following operations: receiving inference information from multiple tasks; and determining an experience set based on the inference information from the multiple tasks. The inference information from the multiple tasks is sent to the second network element by each of the multiple network elements after executing a task, thus allowing the second network element to store the inference information from the multiple tasks and determine the experience set based on the inference information from the multiple tasks.

[0159] In addition, the reasoning information for a task includes at least one of the following: the task identifier, information used to indicate the task, the identifier of the network element, a list of subtasks in the task, a list of service functions corresponding to the task, the identifier of the generative network corresponding to the task, and feedback on the operation status of the generative network.

[0160] Optionally, the second network element stores the task's inference information in the following order: task ID, query ID, agent ID, subtask list, service function list (SFlist), generative network ID, generative network feedback, and user mark. For example, Table 1 shows one storage format for inference information.

[0161] Table 1. Reasoning Information for Historical Tasks

[0162]

[0163] Among them, task ID represents the task identifier, used to identify the task; Query ID represents the identifier of the query information corresponding to the task; Agents ID represents the identifier of the Agent participating in the task, that is, the identifier of the network element participating in the task; subtask list is the task chain output after the network element orchestrates the task (that is, the list of subtasks in the task); SF list is the list of service functions determined when the network element orchestrates the task to execute the task, that is, the list of service functions corresponding to the task, including one or more service functions; GN ID is used to identify the generative network executing the task; GN feedback is used to provide feedback on the running status of the generative network instance, such as whether the generative network instance has been deployed successfully, success indicates that the GN instance has been successfully deployed and the output result has been obtained, failure indicates that an error occurred during the deployment of the GN instance, that is, the function assembly was unsuccessful, the network deployment failed, etc.

[0164] Optionally, GN feedback can also be used to provide feedback on the performance parameters of the GN instance, such as, but not limited to, network performance metrics and service performance metrics. Network performance metrics include, but are not limited to, throughput, jitter, latency, packet loss rate, and energy consumption. Service performance metrics include, but are not limited to, accuracy, F1 score, and recall. The F1 score is used to evaluate the performance of the AI ​​model, and recall represents the proportion of correctly predicted positive examples out of all actual positive examples.

[0165] Additionally, user ratings are feedback scores from users after the GN instance is operational, reflecting user satisfaction with the GN instance. For example, user feedback could be given to the GN instance, which then relays the feedback to the second network element. User ratings can range from 0 to 100 points, representing a subjective evaluation factor. Optionally, if the GN instance fails to deploy successfully, users will not provide a rating; NA in Table 1 indicates that no user rating was provided.

[0166] Optionally, the second network element determines the experience set based on the reasoning information from multiple tasks, including: determining the reasoning information from multiple tasks as the experience set. Alternatively, the second network element determines that the experience set includes the reasoning information from multiple tasks.

[0167] In one optional implementation, the second network element further determines a value experience set and a risk experience set based on the experience set. Specifically, the second network element divides the reasoning information of multiple tasks in the experience set into a value experience set and a risk experience set based on user ratings and success or failure rates of multiple tasks. The value experience set includes one or more value experiences associated with successful tasks, and the risk experience set includes one or more risk experiences associated with failed tasks.

[0168] In one alternative implementation, please refer to Figure 3 , Figure 3 This is a diagram illustrating the division between a set of value experiences and a set of risk experiences. For example... Figure 3 As shown, the second network element can classify the inference information of tasks with user ratings greater than P1, or the inference information of tasks with historical success rates greater than F1, into a value experience set. Therefore, one or more value experiences in the value experience set represent inference information for one or more successful tasks; in other words, one or more value experiences in the value experience set are associated with successful tasks.

[0169] In addition, such as Figure 3 As shown, the second network element can classify the inference information of tasks with user ratings less than P2, or the inference information of tasks with historical failure rates greater than F2, into the risk experience set. Therefore, one or more risk experiences in the risk experience set are inference information for one or more failed tasks; in other words, one or more risk experiences in the risk experience set are associated with failed tasks.

[0170] Wherein, P1, P2, F1, and F2 can be values ​​preset by the second network element, values ​​preset by the first network element for the second network element, or values ​​negotiated in advance by the first and second network elements; this application embodiment does not limit these values. The historical success rate and historical failure rate of the task are determined based on the success or failure feedback from multiple historical tasks corresponding to that task.

[0171] For example, P1 is 80, P2 is 60, F1 is 70, F2 is 30, and the historical tasks include Task 1, Task 2, Task 3, and Task 4. The user ratings for Task 1, Task 2, Task 3, and Task 4 are 88, 70, 58, and none, respectively. The historical success rates for Task 1, Task 2, Task 3, and Task 4 are 83%, 80%, 75%, and 60%, respectively. Therefore, since the user rating for Task 1 is higher than 80 points and its historical success rate is higher than 70%, the second network element allocates the reasoning information for Task 1 from the experience set to the value experience set. Similarly, since the user rating for Task 2 is lower than 80 points, but its historical success rate is higher than 70%, the second network element allocates the reasoning information for Task 2 from the experience set to the value experience set. Since the user rating for Task 3 is lower than 60 points, the second network element allocates the reasoning information for Task 3 from the experience set to the risk experience set. Furthermore, since Task 4 has no user rating, indicating a failure, and its historical failure rate is 40%, higher than 30%, the second network element allocates the reasoning information for Task 4 from the experience set to the risk experience set. Thus, the value experience set includes the reasoning information for Task 1 and Task 2, and the risk experience set includes the reasoning information for Task 3 and Task 4.

[0172] It is evident that the second network element receives reasoning information from multiple executed tasks, identifies this information into an experience set, and then divides the reasoning information from multiple tasks in the experience set into two categories: one is a set of valuable experiences associated with successful tasks, and the other is a set of risky experiences associated with failed tasks. This is beneficial for providing valuable and risky experience information for subsequent tasks.

[0173] As can be seen, the experience set includes valuable experiences associated with multiple successful tasks and / or risky experiences associated with multiple failed tasks, and these multiple tasks include tasks associated with the first task. Therefore, to provide the first network element with experience information associated with the first task, the second network element obtains / retrieves k valuable experiences and / or l risky experiences from the historical inference information associated with the first task from the experience set. This allows it to obtain value information based on the k valuable experiences when the historical inference information includes them, and to obtain risk information based on the l risky experiences when the historical inference information includes them. Furthermore, the value information can serve as a positive reference for orchestrating the first task, and the risk information can serve as a negative reference for orchestrating the first task, which helps increase the probability of the first task being successfully orchestrated.

[0174] In one optional implementation, the value information is determined based on the success rate and / or task satisfaction of k valuable experiences, for example, task satisfaction is represented by user ratings. Optionally, the value information is obtained by ranking the k valuable experiences based on their success rate and / or task satisfaction, and by performing a weighted attention mechanism on the ranked scores of the k valuable experiences.

[0175] Alternatively, the second network element obtains value information based on k valuable experiences, including: ranking the k valuable experiences according to their success rate and / or task satisfaction, and performing a weighted attention mechanism on the ranked scores of the k valuable experiences to obtain value information.

[0176] Optionally, the second network element sorts the k valuable experiences based on the success rate of the k tasks and / or user ratings, with each of the k valuable experiences corresponding to a sorting score. Optionally, the sorting score S of each of the k valuable experiences satisfies the following formula (1):

[0177] S=w1*P+w2*F (1)

[0178] Where P represents the user rating of the task, i.e., the User mark in Table 1; F represents the historical success rate of the task, which is determined based on the GN feedback in Table 1. w1 is the weight value of the user rating, and w2 is the weight value of the success rate.

[0179] Optionally, if the second network element sorts the k valuable experiences based on the success rate of the k tasks and user ratings, then w1 + w2 = 1. Alternatively, if the second network element sorts the k valuable experiences based on the success rate of the k tasks, then w1 = 0 and w2 = 1. Alternatively, if the second network element sorts the k valuable experiences based on user ratings of the k tasks, then w1 = 1 and w2 = 0.

[0180] Optionally, the specific values ​​of w1 and w2 can be preset or dynamically adjusted, such as dynamically increasing the weight of user rating w1 to reflect user preferences.

[0181] In another optional implementation, where GN feedback also provides network performance metrics and service performance metrics, the value information is determined based on at least one of the success rate, task satisfaction, network performance metrics, and service performance metrics of the k value experiences. Optionally, the value information is obtained by ranking the k value experiences based on at least one of the success rate, task satisfaction, network performance metrics, and service performance metrics, and by applying a weighted attention mechanism to the ranked scores of the k value experiences. Alternatively, the second network element obtains value information based on the k value experiences by: ranking the k value experiences based on at least one of the success rate, task satisfaction, network performance metrics, and service performance metrics, and by applying a weighted attention mechanism to the ranked scores of the k value experiences.

[0182] Optionally, the ranking score S of each of the k value experiences satisfies the following formula (2):

[0183] S=w1*P+w2*F+w3*NI+w4*SI (2)

[0184] Wherein, P, F, w1 and w2 can be referred to in the above formula (1), and will not be repeated. In addition, NI represents network performance indicators, which can be the average of multiple indicators such as throughput, jitter, latency, packet loss rate and energy consumption; SI represents service performance indicators, which can be the average of multiple indicators such as accuracy, F1 score and recall rate; w3 is the weight value of network performance indicators, and w4 is the weight value of service performance indicators.

[0185] Optionally, if the second network element sorts the k valuable experiences according to the success rate, task satisfaction, network performance indicators, and service performance indicators, then w1+w2+w3+w4=1.

[0186] Optionally, when the second network element ranks the k valuable experiences based on a portion of the success rate, task satisfaction, network performance indicators, and service performance indicators, the weights corresponding to the unused portions of the success rate, task satisfaction, network performance indicators, and service performance indicators are 0, and the sum of the weights corresponding to the used portions equals 1. For example, if the second network element ranks the k valuable experiences based on the success rate, task satisfaction, and network performance indicators, the value of w4 is 0, and w1 + w2 + w3 = 1. As another example, if the second network element ranks the k valuable experiences based on the network performance indicators and service performance indicators, the values ​​of w1 and w2 are 0, and w3 + w4 = 1.

[0187] Optionally, the specific values ​​of w1, w2, w3, and w4 can be preset or dynamically adjusted. For example, the weight of user rating w1 can be dynamically increased to reflect user preferences, or the weight of business performance indicators w4 can be dynamically increased to improve business performance.

[0188] After ranking the k valuable experiences, the second network element applies a weighted attention mechanism to the scores of these ranked experiences to obtain valuable information. In other words, the second network element increases the information weight of high-ranking valuable experiences in its summary of the k valuable experiences, so that they can serve as direct references in reasoning for similar tasks in the future. This increase in the weight of high-ranking valuable experiences in the summary of the k valuable experiences requires a weighted attention mechanism; that is, in the generative model, the ranking score is used as a weight to adjust the importance of the experiences.

[0189] In one approach, in the NetGPT Transformer model, the attention weight in the weighted attention mechanism can be expressed as: attention weight = softmax(ranking score × similarity between query and key), which is expressed as the following formula (3):

[0190]

[0191] Where Q, K, and V represent the query, key, and value, respectively. The "query" is the input information, the "key" is used to calculate the attention distribution, and the "value" is used to generate the selection information. S is the ranking score of k valuable experiences. d k This is the normalization scaling factor, a constant value. Furthermore, after the user's intent information is input into the Transformer model, it is embedded into a vector through word embeddings. Therefore, Q specifically represents the vector corresponding to the query after word embedding.

[0192] As can be seen, the second network element can rank the k valuable experiences based on their success rates and user ratings, obtaining a ranking score for each of the k valuable experiences. Then, a weighted attention mechanism is applied based on this ranking score, meaning that historical experiences with high success rates and high ratings are given increased information weight, thereby summarizing the value information associated with the first task. Therefore, the value information is determined based on the k valuable experiences, and since these k valuable experiences are associated with historical successful tasks related to the first task, the value information can be considered as being determined based on these historical successful tasks associated with the first task.

[0193] Additionally, the value information includes the orchestration method corresponding to the successful task and a list of service functions determined based on the service functions participating in the successful task. The successful task is a historical successful task associated with the first task. The orchestration method corresponding to the successful task can be an orchestration method determined by the orchestration method used by the successful task. An orchestration method refers to the method used to implement a task, such as splitting the task into multiple sub-tasks. The list of service functions includes the service functions used to execute multiple sub-tasks and the connection relationships between the multiple service functions, which represent the order in which the multiple sub-tasks are executed. A service function can be any one of the multiple network elements / agents in the core network; this application embodiment does not limit this.

[0194] In another optional implementation, the risk information is obtained by extracting risk features from l risk experiences. Alternatively, the second network element analyzes the failure experiences in the l risk experiences using a feature extraction algorithm, extracts the features leading to the failure, obtains the failure factors, and thus obtains the risk information. The risk information includes risk feature information determined based on the failure factors of the failed task, which is a historical failed task associated with the first task. Feature extraction algorithms include, but are not limited to, term frequency–inverse document frequency (TF-IDF), decision trees, and random forests.

[0195] Optionally, the risk feature information in the risk information can implicitly indicate that when arranging / reasoning the first task in the future, care should be taken not to include features similar to those in the risk feature information. For example, when arranging the first task, the semantics / keywords in the risk feature information should not appear.

[0196] Optionally, the risk information may also include risk warning information, which directly instructs that when arranging / reasoning for the first task, care should be taken to avoid features similar to those in the risk characteristic information. The risk warning information is determined based on the risk characteristic information.

[0197] It is evident that after the second network element obtains the historical reasoning information associated with the first task, it extracts value information from the experience information based on k value experiences in the historical reasoning information, and / or extracts risk information from the experience information based on l risk experiences in the historical reasoning information. Alternatively, it can be understood that after the second network element obtains the historical reasoning information associated with the first task, it extracts the reasoning features for task decomposition, function matching, and assembly by the agent from the k value experiences in the historical reasoning information, i.e., extracts historical successful experiences to obtain value information; and / or, it extracts the features leading to failure from the l risk experiences in the historical reasoning information, i.e., extracts historical failure experiences to obtain risk information. The value information includes the orchestration method and service function list, and the risk information includes risk feature information.

[0198] Optionally, the second network element can invoke NetGPT, using k value experiences as input to NetGPT to obtain the value information output by NetGPT, and / or using l risk experiences as input to NetGPT to obtain the risk information output by NetGPT.

[0199] The implementation method of NetGPT determining value information based on k value experiences can be found in the implementation method of the second network element determining value information based on k value experiences, and will not be repeated here. Similarly, the implementation method of NetGPT determining risk information based on l risk experiences can be found in the implementation method of the second network element determining risk information based on l risk experiences, and will not be repeated here.

[0200] Therefore, the second network element obtains value information based on k value experiences either by determining the value information itself based on the k value experiences, or by calling NetGPT to obtain the value information based on the k value experiences. Similarly, the second network element obtains risk information based on l risk experiences either by determining the risk information itself based on the l risk experiences, or by calling NetGPT to obtain the risk information based on the l risk experiences.

[0201] In summary, the second network element can continuously collect inference information from multiple tasks, and divide this information according to task success rate and / or user ratings, into a set of value experiences associated with successful tasks and a set of risk experiences associated with failed tasks. Therefore, when the first network element requests / subscribes to experience information associated with the first task, the second network element retrieves the top k value experiences from the value experience set whose similarity to the first task is greater than a preset value, and / or retrieves the top l risk experiences from the risk experience set whose similarity to the first task is greater than a preset value. The second network element then retrieves historical successful experiences based on the k value experiences to obtain value information; and / or retrieves historical failed experiences based on the l risk experiences to obtain risk information. Finally, the second network element provides the first network element with the experience information associated with the first task, which includes value information and / or risk information.

[0202] Optionally, after obtaining the experience information associated with the first task, the first network element orchestrates the first task based on this experience information. The orchestration of the first task by the first network element includes, but is not limited to: breaking down the first task into multiple sub-tasks, determining the network elements / agents that will execute the multiple sub-tasks, and determining the logical order in which the multiple network elements / agents will execute the multiple network elements. Since the experience information associated with the first task includes value information determined based on historical successful tasks associated with the first task, and / or risk information determined based on historical failed tasks associated with the first task, the first network element can directly reference the value information during the orchestration process, for example, by directly adopting the orchestration methods and service function lists from the value information; and / or by proactively avoiding the risk characteristics in the risk information. This approach can improve the success rate of the first network element orchestrating the first task, thereby improving the success rate of the first network element executing the first task.

[0203] Optionally, after the first network element obtains the experience information associated with the first task, it sends the experience information associated with the first task to other network elements used for orchestrating the first task, so that other network elements can orchestrate the first task based on the experience information associated with the first task, thereby increasing the probability that the first task is successfully orchestrated and thus increasing the probability that the first task is successfully executed.

[0204] Optionally, the first network element sends inference information for the first task to the second network element. This inference information includes at least one of the following: an identifier for the first task, information indicating the first task, an identifier for the first network element, a list of subtasks within the first task, a list of service functions corresponding to the first task, an identifier for the generative network, and feedback on the operational status of the generative network. The inference information for the first task is obtained based on the orchestration results of the first task and the execution of the first task by a generative network instance. For example, the inference information for the first task may be obtained by the first network element executing the first task based on the orchestration results of the first task and a generative network instance; or, alternatively, the inference information for the first task may be obtained by other network elements besides the first and second network elements executing the first task based on the orchestration results of the first task and a generative network instance.

[0205] It is evident that the first network element can send the reasoning information of the first task to the second network element, so that the second network element can use the reasoning information of the first task as a reference for subsequent experience. For example, the second network element stores the reasoning information of the first task and divides the reasoning information of the first task into a value experience set or a risk experience set.

[0206] In this embodiment, after receiving first information for requesting the execution of a first task, the first network element obtains experience information for orchestrating the first task. This facilitates the first network element or other network elements in orchestrating the first task based on the experience information associated with it. The experience information includes value information determined based on the orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on the failure factors of historically failed tasks associated with the first task. Therefore, when a network element orchestrates the first task based on this experience information, it is doing so by leveraging successful and / or failed experiences from historical tasks. This allows the network element to learn from historical tasks, thereby increasing the success rate of executing the first task and improving the quality of the generated results.

[0207] In another approach, the network agent can collect reasoning information from previously executed historical tasks into an experience pool. During the training of the large language model, partial experience can be extracted from the experience pool as training samples. During task reasoning, the network agent extracts experience from the experience pool through experience recall and obtains the experience results obtained from the training of the large language model through insight guidance, using these as references for reasoning. Compared with this approach, the embodiments of this application refine the experience extraction process for historical tasks and eliminate the need to train a large language model that continuously retries tasks, thus reducing cost overhead.

[0208] This application embodiment uses the example of a first network element as a network agent and a second network element as a network memory, and the network agent, after receiving information for requesting the execution of a first task, requests experience information associated with the first task from the network memory to exemplify the above communication method 100. Figure 4 This is a schematic diagram of the interaction of the communication method, and its interaction process includes, but is not limited to:

[0209] S401. The network agent receives intent information, which is used to request the execution of the first task.

[0210] The intent information can be found in S201 above and will not be repeated here.

[0211] Optionally, prior to S401, any one of the planning agent, assembly agent, connection agent, execution agent, and GN agent in the network agent can send the reasoning information of the task to the network memory. The network memory determines the value experience set and the risk experience set, that is, it divides the reasoning information of multiple tasks into the value experience set and the risk experience set. The implementation method can be found in S202 above, and will not be repeated here.

[0212] S402. The network agent sends experience retrieval information to the network memory, the experience retrieval information being used to request the retrieval of experience information related to the first task. Correspondingly, the network memory receives the experience retrieval information from the network agent.

[0213] Among them, the experience retrieval information can be referred to as the second information in S202 above, and the experience information associated with the first task can be referred to as described in S202 above, which will not be repeated here.

[0214] S403. Retrieve experience information associated with the first task from network memory.

[0215] Optionally, the network memory acquires the experience information associated with the first task, including: acquiring the historical reasoning information associated with the first task; and acquiring the experience information associated with the first task based on the historical reasoning information associated with the first task. The implementation method can be found in S202 above, and will not be repeated here.

[0216] Optionally, the network memory obtains the experience information associated with the first task based on the historical reasoning information associated with the first task, including: calling NeTGPT based on the historical reasoning information associated with the first task to obtain the experience information associated with the first task. The implementation method can be found in S202 above, and will not be repeated here.

[0217] S404. The network memory sends the experience information associated with the first task to the network agent. Correspondingly, the network agent receives the experience information associated with the first task from the network memory.

[0218] As can be seen, after receiving the experience retrieval information from the network agent, the network memory obtains the experience information associated with the first task and provides the experience information associated with the first task to the network agent, thereby enabling the network agent to obtain the experience information associated with the first task.

[0219] For example, the first task is to calculate the daily traffic flow of park A over the past week. The experience information associated with the first task acquired by the network memory includes value information, which includes agent ID, experience type, relative query information, and content. The agent ID is "Agent-001," indicating a network agent identified as "Agent-001." The experience type is "HVE," indicating that the experience type is valuable experience. The relative queries include: "Please calculate the daily traffic flow in area B of city A last week" and "Please calculate the traffic flow on bridge B of city A during the morning rush hour this week," representing queries corresponding to tasks whose semantic similarity to the first task is greater than a preset value. The content includes an orchestration method and a list of service functions. The orchestration method includes: first, obtaining the time and location from the user's query; then, attempting to instruct base stations within the current location to activate sensing capabilities and track vehicle trajectories; and finally, collecting statistical data on the number of vehicles. The list of service functions includes: named entity recognition, environmental perception, high-precision positioning, dynamic tracking, motion detection, and behavior perception.

[0220] It is evident that when the experiential information associated with the first task acquired by the network memory includes value information, this value information includes orchestration methods and a list of service functions. The orchestration method is a summary of the methods used in historically successful tasks associated with the first task. For example, it might be an orchestration method derived from summarizing the orchestration experience of the K most similar historical tasks to the first task, reflecting orchestration methods with high ranking scores. The list of service functions is a summary of the service functions used in historically successful tasks associated with the first task. Therefore, the orchestration method and the list of service functions in the value information can serve as a direct reference / guide for orchestrating the first task.

[0221] For example, the first task is to calculate the daily traffic flow in park A over the past week. The experience information associated with the first task acquired by the network memory includes risk information, which includes the agent's identifier, experience type, relative query information, and content. The agent's identifier is "Agent-001," indicating a network agent with identifier 001. The experience type is "HRE," indicating a risky experience. The relative queries include: "Please calculate the daily traffic flow in area B of city A last week" and "Please construct a digital twin model of area C of city A and detect pedestrian / vehicle traffic behavior." The content includes risk characteristic information and risk warning information (or alert information). The risk characteristic information includes: 3D environment modeling, pedestrian behavior perception, and abnormal behavior statistics. The risk warning information includes: "Please do not divide the task into modeling and behavioral awareness."

[0222] As can be seen, when the experience information associated with the first task acquired by the network memory includes risk information, the risk information includes risk characteristic information and risk warning information. Among them, risk characteristic information indicates that when the orchestration result contains these characteristic keywords, it is likely to lead to failure; risk warning information is used to remind users not to classify the current task into the sub-task category mentioned in the prompt, so as to reduce the failure rate of task orchestration.

[0223] S405. The network agent orchestrates the first task based on the experience information associated with the first task.

[0224] The implementation method of the first task of network intelligent agent orchestration can be found in S202 above, and will not be repeated here.

[0225] Optionally, the planning agent in the network agent combines the assembly agent, the connection agent, and the execution agent to perform the functional assembly, connection configuration, and functional deployment of the first task, so as to perform the first task through the GN agent.

[0226] Optionally, the GN agent in the network agent performs the first task and obtains the reasoning information of the first task. The reasoning information of the first task can be referred to in S202 above, and will not be repeated here.

[0227] S406. The network agent sends the reasoning information for the first task to the network memory.

[0228] Optionally, the GN agent in the network agent sends the reasoning information for the first task to the network memory.

[0229] It is evident that after the GN agent executes the first task, it also sends the reasoning information of the first task to the network memory, which allows the network memory to use the reasoning information of the first task as a reference for the next experience information, thereby improving the accuracy of the experience information.

[0230] For example, Figure 5 This is a schematic diagram of a communication method. Figure 5 In this context, the network agent includes an agent GPT, a toolbox, a memory, and an actuator. The network agent connects to the GN instance via the ABI, and the GN instance, in turn, connects to the network memory via the ABI. This GN instance is created by the network agent to perform the primary task. Figure 5As shown, upon receiving intent information, the network agent requests experience information associated with the first task from the network memory through experience retrieval. After obtaining the experience information associated with the first task, the network memory feeds back this experience information to the network agent. Thus, the network agent orchestrates the first task based on the experience information associated with it and creates a GN (Service Controller) for executing the first task. The GN includes SF1, SF2, and SF3, etc., for executing the first task. These GN agents are determined by the network agent based on the service function list in the experience information fed back by the network memory; that is, the network agent refers to the service function list in the experience information fed back by the network memory to determine the service functions for executing the first task. The network agent executes the first task through the GN instance and feeds back the reasoning information of the first task to the network memory through GN feedback, so that the network memory can use the reasoning information of the first task as new experience for the next experience summary.

[0231] As can be seen, in this embodiment, after receiving the intent information, the network agent requests the network memory to obtain the experience information associated with the first task from the intent information. Therefore, the network memory obtains the experience information associated with the first task according to the user intent transmitted by the network agent and feeds back the experience information associated with the first task to the network agent. Then, the network agent orchestrates the first task based on the experience information associated with the first task. The experience information associated with the first task includes value information and / or risk information. The value information includes an orchestration method and service function list determined based on historical successful tasks associated with the first task, which can serve as a positive reference for orchestrating the first task; the risk information includes risk characteristic information determined based on historical failed tasks associated with the first task, which can serve as a negative reference for orchestrating the first task. Therefore, by orchestrating the first task based on the experience information associated with the first task, the network agent can directly reference the value information in the experience information and / or actively avoid the risk information in the experience information, thereby improving the success rate of the first task orchestration and thus improving the success rate of the network agent executing the first task.

[0232] In addition, the network agent arranges the first task based on the experience information associated with the first task, which does not require repeated scheduling of large language models for reasoning on the same or highly similar tasks, thus saving computing resources and improving response speed.

[0233] This application embodiment further illustrates the communication method 100 by taking the first network element as a network agent and the second network element as a network memory, and by illustrating that after the network agent receives information for requesting the execution of the first task, it requests the network memory to obtain experience information associated with the first task. If the service function list in the value information is verified as valid, the network memory provides the experience information associated with the first task to the network agent. Figure 6 This is a schematic diagram of the interaction of the communication method, and its interaction process includes, but is not limited to:

[0234] S601. The network agent receives intent information, which is used to request the execution of the first task.

[0235] S602. The network agent sends experience retrieval information to the network memory, the experience retrieval information being used to request the retrieval of experience information related to the first task. Correspondingly, the network memory receives the experience retrieval information from the network agent.

[0236] S603. Retrieve experience information associated with the first task from network memory.

[0237] In this application embodiment, the implementation methods of S601 to S603 can be referred to the implementation methods of S401 to S403 described above, and will not be repeated here.

[0238] S604. The network memory sends a fifth message to the network sandbox, which instructs the verification of the service function list in the value information. Correspondingly, the network sandbox receives the fifth message from the network memory.

[0239] The fifth piece of information includes information indicating the first task and a list of service functions within the value information. The network sandbox can be viewed as a black box, used to verify the list of service functions within the value information.

[0240] Optionally, the fifth information may also include at least one of the identifier of the first task and the identifier of the network agent.

[0241] As can be seen, the network memory responds to the experience retrieval information from the network agent and obtains the experience information associated with the first task. If the obtained experience information includes value information, the network memory requests the network sandbox to verify the service function list in the value information, ensuring that the service function list in the value information can fulfill the first task.

[0242] S605. The network sandbox sends a sixth message to the network memory, which indicates whether the verification passed or failed.

[0243] Optionally, the network sandbox also responds to a fifth message from memory to verify the list of service functions in the value information. Optionally, the network sandbox verifies the list of service functions in the value information by: performing a first task based on the list of service functions in the value information.

[0244] Optionally, the network sandbox determines the verification result of the service function list based on the execution result of the first task. For example, the network sandbox executes the first task in the order of the service functions in the service function list. If the first task is executed successfully, the service function list is determined to have passed verification; if the first task is not executed successfully, i.e., the first task fails, the service function list is determined to have failed verification, i.e., it has not passed verification.

[0245] Optionally, if the network sandbox determines that the service function list verification is successful, the sixth message is used to indicate that the verification is successful; if the network sandbox determines that the service function list verification is unsuccessful, the sixth message is used to indicate that the verification is unsuccessful.

[0246] Optionally, the sixth piece of information can indicate whether the verification passed or failed using bit information. For example, a bit value of 1 in the sixth piece of information indicates that the service function list has passed verification; a bit value of 0 in the sixth piece of information indicates that the service function list has failed verification.

[0247] Optionally, the fifth information may also be called verification information, and the sixth information may also be called verification response information. This application embodiment does not limit this.

[0248] As can be seen, after receiving the fifth message from the network memory, the network sandbox responds to the fifth message by executing the first task based on the service function list in the value information to verify the service function list. The network sandbox also sends the verification result of the service function list back to the network memory via the sixth message.

[0249] S606. If the service function list verification is successful, the network memory sends the experience information associated with the first task to the network agent. Correspondingly, the network agent receives the experience information associated with the first task from the network memory.

[0250] It is evident that the network memory only feeds back the experience information associated with the first task to the first network element after the service function list has been verified. This ensures the accuracy of the experience information associated with the first task and helps improve the success rate of the network agent in arranging the first task based on this experience information.

[0251] S607. The network agent orchestrates the first task based on the experience information associated with the first task.

[0252] S608. The network agent sends the reasoning information for the first task to the network memory.

[0253] In this application embodiment, the implementation methods of S607 and S608 can be referred to the implementation methods of S605 and S606 described above, and will not be repeated here.

[0254] visible, Figure 6 The communication method shown is Figure 4 Compared to the communication methods shown, the difference lies in: Figure 6 In the communication method shown, after the network memory obtains the experience information associated with the first task, it requests the network sandbox to verify the service function list in the experience information to verify the feasibility of assembling the service function list. This reduces repeated retrieval and reasoning caused by errors in the service function list. Furthermore, the network sandbox only feeds back the experience information associated with the first task to the network agent after the service function list has been verified. This ensures the suitability of the value information in the experience information with the current user needs, making the value information fed back by the network memory directly usable. This guarantees the success rate of the network agent's arrangement of the first task based on the experience information, thereby improving the accuracy of the reasoning results for the first task.

[0255] This application embodiment further illustrates the above communication method 100 by taking the first network element as a network agent and the second network element as a network memory, and the network agent obtaining experience information related to the first task from the network memory through a subscription method. Figure 7 This is a schematic diagram of the interaction of the communication method, and its interaction process includes, but is not limited to:

[0256] S701. The network agent sends subscription information to the network memory, the subscription information being used to request subscription to the experience information associated with the first task. Correspondingly, the network memory receives the subscription information from the network agent.

[0257] The experience information associated with the first task can be found in the communication method 100 described above, and will not be repeated here.

[0258] S702. The network memory sends a subscription response to the network agent, indicating whether the subscription was successful or failed. Correspondingly, the network agent receives the subscription response from the network memory.

[0259] The subscription information can be found in the third information in S202 above, and the subscription response can be found in the fourth information in S202 above, which will not be repeated here.

[0260] It is evident that the network agent requests subscription information related to the first task from the network memory by subscribing to the information, and learns whether the subscription was successful by receiving the subscription response from the memory.

[0261] S703. The network memory sends the experience information associated with the first task to the network agent. Correspondingly, the network agent receives the experience information associated with the first task from the network memory.

[0262] Optionally, if the network memory detects experience information associated with the first task that matches the subscription information for the first time upon receiving a subscription response indicating successful subscription, it sends the experience information associated with the first task to the network agent. Correspondingly, if the network agent receives the experience information associated with the first task from the network memory upon receiving a subscription response indicating successful subscription.

[0263] Optionally, if the subscription response indicates successful subscription and the experience information associated with the first task has been updated, the network memory sends the updated experience information associated with the first task to the network agent, so that the network agent can obtain the latest experience information.

[0264] Optionally, the network memory may also acquire experience information associated with the first task, the implementation of which can be found in S202 above and will not be repeated here.

[0265] It is evident that the network memory obtains the latest experience information related to the first task in a timely manner by requesting subscription from the network agent.

[0266] For example, the subscription information sent by the network agent to the network memory includes agent identifier, subscription ID, subscription type, query information, subscription scope, and subscription duration. The agent identifier is "Agent-001," representing the network agent with identifier "Agent-001." The subscription ID is "Sub-001," representing the subscription event with identifier "Sub-001." The subscription type is "HVE," indicating that the subscribed experience type is a value experience, i.e., it requests the value information within the subscribed experience information. The query information is "Please detect the traffic flow in the city center of City A during the morning rush hour (8:00 to 10:00) to optimize the traffic light timing," representing the first task. The subscription scope is "Global," indicating that the subscription for the experience information associated with the first task is performed globally. The subscription duration is "7 days," indicating that the subscription for the experience information associated with the first task is performed within a 7-day period.

[0267] The experience information sent from the network memory to the network agent includes a subscription identifier, experience ID, experience type, query list, subtask list, service function list, and context information. The subscription identifier is "Sub-001," indicating that the feedback is for a subscription event with identifier "Sub-001." The experience ID is "Exp-001," indicating that the feedback is for an experience with identifier "Exp-001." The experience type is "HVE," indicating that the feedback is value information of the value type. The query list includes "Please optimize traffic flow and safety at tourist attraction A by monitoring visitor flow and density," representing a historical task associated with the first task and used to determine the experience information. The subtask list is used for orchestration methods and includes: Subtask 1: Use location functionality to perceive the real-time location of visitors within the park; Subtask 2: Call base stations to track the movement of each visitor; Subtask 3: Identify the behavior of each visitor in real time, especially abnormal behavior; Subtask 4: Analyze abnormal visitor behavior to prevent potential safety risks. The service function list includes high-precision positioning, dynamic tracking, target detection, action recognition, and security analysis. Contextual information includes time, location, and resources. The time is 2023-10-01 8:00-12:00, the location is the area with the identifier "Area-001", and resources include network resources and topology resources. Network resources are server entities responsible for computing and function deployment, such as graphics processing units (GPUs). Network topology resources are base stations and terminal resources related to network connections, such as base station-001 (BS-001) and base station-002 (BS-002).

[0268] S704. The network agent receives intent information, which is used to request the execution of a first task.

[0269] S705. The network agent orchestrates the first task based on the experience information associated with the first task.

[0270] S704 can be referred to in S201 above, and the implementation of S705 can be referred to in S405 above, and will not be repeated here.

[0271] Optionally, after orchestrating the first task, the network agent may also perform the following operations: create a GN instance for executing the first task; execute the first task through the created GN instance to obtain the inference information of the first task; and send the inference information of the first task to the network memory. The implementation method can be found in S202 above and will not be repeated here.

[0272] As can be seen, in this embodiment, the network agent requests subscription information from the network memory to subscribe to the experience information associated with the first task. Upon successful subscription, the network agent obtains the experience information associated with the first task from the network memory. Therefore, after receiving the intent information, the network agent can directly use the obtained experience information associated with the first task to orchestrate the first task. Since the experience information associated with the first task includes value information determined based on the orchestration methods and service functions used in historically successful tasks associated with the first task, and / or risk information determined based on the failure factors of historically failed tasks associated with the first task, the network agent can learn from historical tasks and improve the success rate of orchestrating the first task, thereby increasing the success rate of executing the first task.

[0273] In addition, the subscription mechanism of the network agent to the network memory for experience information related to the first task enables the network memory to provide available functional information in a timely manner according to the functional capability expectations put forward by the network agent.

[0274] The following section further describes the corresponding device implementation scheme in relation to the technical solution described above.

[0275] To achieve the functions of the methods provided in the embodiments of this application, the first network element, the second network element, and the third network element may include hardware structures and / or software modules, and implement the above functions in the form of hardware structures, software modules, or hardware structures plus software modules. Whether a particular function is executed in the form of hardware structures, software modules, or hardware structures plus software modules depends on the specific application and design constraints of the technical solution.

[0276] Figure 8 This is a schematic diagram of the structure of a communication device 800 provided in this application. The communication device 800 may include modules corresponding to the methods / operations / steps / actions described in any embodiment of the communication method 100. These modules may be hardware circuits, software, or a combination of hardware circuits and software.

[0277] The communication device 800 includes a communication unit 801 and a processing unit 802, used to implement the methods executed by the various devices in the foregoing embodiments. The communication unit 801 is also called a transceiver unit, which includes a sending unit and a receiving unit. The sending unit is used to send signals, and the receiving unit is used to receive signals.

[0278] In one possible implementation, the communication device 800 is, for example, a first network element. Specifically, the communication unit 801 is used to receive first information, which is used to request the execution of a first task; the processing unit 802 is used to obtain experience information associated with the first task, which is used to orchestrate the first task.

[0279] The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to successful tasks and a list of service functions determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0280] In another possible implementation, the communication device 800 is, for example, a second network element. Specifically, the communication unit 801 is configured to receive second information from a first network element, the second information being used to request the acquisition of experience information associated with a first task, the second information including information for instructing the first task; the communication unit 801 is also configured to send the experience information to the first network element, the experience information being obtained based on historical reasoning information associated with the first task;

[0281] The experience information includes value information and / or risk information. The value information includes the arrangement method corresponding to successful tasks in the historical reasoning information and the service function list determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks in the historical reasoning information. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0282] In another possible implementation, the communication device 800 is, for example, a second network element. Specifically, the communication unit 801 is configured to receive third information from a first network element, the third information being used to request subscription to experience information associated with a first task, the third information including a subscription type and information indicating the first task, the subscription type including a value type and / or a risk type; the communication unit 801 is also configured to send the experience information to the first network element;

[0283] The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to successful tasks and a list of service functions determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks. The successful tasks are historical successful tasks associated with the first task, and the failed tasks are historical failed tasks associated with the first task.

[0284] Optionally, other embodiments of the communication device 800 can be found in the above method embodiments, and will not be repeated here.

[0285] In one possible implementation, when the communication device 800 is a chip, the chip includes a transceiver unit and a processing unit. The transceiver unit can be an input / output circuit or a communication interface; the processing unit is a processor, microprocessor, integrated circuit, or logic circuit integrated on the chip.

[0286] This application also provides a communication device 900, please refer to [link to relevant documentation]. Figure 9 , Figure 9 This is a schematic diagram of another communication device 900. The communication device 900 can be used to perform the steps executed by the first network element, the second network element, and the third network element in the method embodiments described above, as can be found in the relevant descriptions in the above method embodiments.

[0287] The communication device 900 includes a processor 901. Optionally, the communication device 900 may also include a memory 902 and a transceiver 903.

[0288] In one possible implementation, the processor 901, memory 902, and transceiver 903 are connected via a bus, and the memory stores computer instructions. Optionally, the processor 901 and memory 902 can also be integrated together.

[0289] Optionally, the processing unit 802 in the foregoing embodiments may specifically be the processor 901 in this embodiment, therefore the specific implementation of the processor 901 will not be described in detail. The communication unit 801 in the foregoing embodiments may specifically be the transceiver 903 in this embodiment, therefore the specific implementation of the transceiver 903 will not be described in detail.

[0290] In this application, the processor can be a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in this application can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0291] In this application, the memory can be non-volatile memory, such as a hard disk drive (HDD) or a solid-state drive (SSD), or it can be volatile memory, such as random-access memory (RAM). Memory is any other medium capable of carrying or storing desired program code in the form of instructions or data structures, and accessible by a computer, but is not limited to this. The memory in this application can also be a circuit or any other device capable of implementing storage functions for storing program instructions and / or data.

[0292] This application provides another communication device, which includes a processor and an interface. Optionally, it also includes a memory, with the processor coupled to the memory, the processor being used to read and execute computer instructions stored in the memory to implement a network service instantiation method as shown in the embodiment of communication method 100.

[0293] This application also provides a communication system, which includes a first network element and a second network element. The first network element is used to perform all or part of the steps performed by the first network element in the preceding embodiments. The second network element is used to perform all or part of the steps performed by the second network element in the preceding embodiments. In another possible design, the system may further include other device / functional network elements that interact with at least one of the first network element and the second network element, such as a third network element.

[0294] This application provides a computer-readable storage medium. The computer-readable storage medium stores a program or instructions. When the instructions are executed on a communication device, they implement a network service instantiation method as shown in the embodiment of communication method 100.

[0295] This application provides a computer program product. The computer program product includes instructions. When the instructions are executed on a communication device, they implement a network service instantiation method as shown in the embodiment of communication method 100.

[0296] This application provides a chip or chip system including at least one processor and an interface, the interface and at least one processor being interconnected via a line, the at least one processor being used to run computer programs or instructions to perform a network service instantiation method as shown in the embodiment of communication method 100.

[0297] The interfaces in the chip can be input / output interfaces, pins, or circuits, etc.

[0298] The aforementioned chip system can be a System-on-a-Chip (SoC) or a baseband chip, etc. The baseband chip can include a processor, channel encoder, digital signal processor, modem, and interface module, etc.

[0299] In one implementation, the chip or chip system described above in this application further includes at least one memory, which stores instructions. The memory can be an internal storage unit of the chip, such as a register or cache, or it can be a storage unit of the chip itself (e.g., read-only memory, random access memory, etc.).

[0300] The technical solutions provided in this application can be implemented in whole or in part through software, hardware, firmware, or any combination thereof. When implemented using software, they can be implemented in whole or in part as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a network device, a terminal, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media, etc.

[0301] In this application, provided there is no logical contradiction, the various embodiments may reference each other. For example, the methods and / or terms between method embodiments may reference each other, the functions and / or terms between device embodiments may reference each other, and the functions and / or terms between device embodiments and method embodiments may reference each other.

[0302] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A communication method, characterized in that, The method includes: Receive first information, which is used to request the execution of a first task; Obtain the experience information associated with the first task, and the experience information is used to orchestrate the first task; The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to the successful task and the service function list determined based on the service functions participating in the successful task. The risk information includes risk characteristic information determined based on the failure factors of the failed task. The successful task is a historical successful task associated with the first task, and the failed task is a historical failed task associated with the first task.

2. The method according to claim 1, characterized in that, The step of obtaining the experience information associated with the first task includes: Send a second message to the second network element. The second message is used to request the acquisition of experience information associated with the first task. The second message includes information for instructing the first task. Receive the experience information from the second network element.

3. The method according to claim 1, characterized in that, The method further includes: Send a third message to the second network element. The third message is used to request subscription to the experience information associated with the first task. The third message includes a subscription type and information for instructing the first task. The subscription type includes a value type and / or a risk type.

4. The method according to claim 3, characterized in that, The system receives a fourth message from the second network element, which indicates whether the subscription was successful or failed.

5. The method according to claim 3 or 4, characterized in that, The step of obtaining the experience information associated with the first task includes: Receive the experience information from the second network element.

6. The method according to any one of claims 1 to 4, characterized in that, The experience information also includes experience type and / or information for indicating the second task, the experience type including value type and / or risk type, the second task being a historical task associated with the first task.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The inference information of the first task is sent to the second network element. The inference information of the first task is obtained based on the orchestration result of the first task and the execution of the first task by the generative network instance.

8. The method according to claim 7, characterized in that, The reasoning information for the first task includes at least one of the following: the identifier of the first task, information for indicating the first task, the identifier of the first network element, a list of subtasks in the first task, a list of service functions corresponding to the first task, the identifier of the generative network, and feedback on the operation status of the generative network.

9. A communication method, characterized in that, The method includes: Receive second information from a first network element, the second information being used to request the acquisition of experience information associated with a first task, the second information including information for instructing the first task; send the experience information to the first network element, the experience information being acquired based on historical reasoning information associated with the first task; The experience information includes value information and / or risk information. The value information includes the arrangement method corresponding to successful tasks in the historical reasoning information and the service function list determined based on the service functions participating in the successful tasks. The risk information includes risk characteristic information determined based on the failure factors of failed tasks in the historical reasoning information. The successful task is a historical successful task associated with the first task, and the failed task is a historical failed task associated with the first task.

10. The method according to claim 9, characterized in that, The experience information also includes experience type and / or information for indicating the second task, the experience type including value type and / or risk type, the second task being a historical task associated with the first task.

11. The method according to claim 9 or 10, characterized in that, The method further includes: Based on the historical reasoning information associated with the first task, obtain the experience information associated with the first task.

12. The method according to claim 11, characterized in that, The method further includes: Obtain historical reasoning information associated with the first task from the experience set. The historical reasoning information includes k valuable experiences and / or l risk experiences. The k valuable experiences are associated with the successful task, and the l risk experiences are associated with the failed task. The experience set includes valuable and / or risky experiences from multiple tasks, the multiple tasks include tasks associated with the first task, and k and l are positive integers; The step of obtaining the experience information associated with the first task based on the historical reasoning information associated with the first task includes: The value information is obtained based on the k value experiences, and / or the risk information is obtained based on the l risk experiences.

13. The method according to claim 12, characterized in that, The value information is determined based on the success rate and task satisfaction of the k value experiences.

14. The method according to claim 12 or 13, characterized in that, The method further includes: Receive inference information for multiple tasks, the inference information including at least one of the following: task identifier, information for indicating the task, network element identifier, list of subtasks in the task, list of service functions corresponding to the task, identifier of the generative network corresponding to the task, and operation status feedback of the generative network, wherein the multiple tasks include tasks associated with the first task. The experience set is determined based on the reasoning information from the multiple tasks.

15. The method according to any one of claims 9 to 14, characterized in that, If the experiential information includes the value information, the method further includes: Send a fifth message to a third network element. The fifth message is used to instruct the verification of the service function list in the value information. The fifth message includes information for instructing the first task and the service function list in the value information. Sending the experience information to the first network element includes: If the service function list in the value information is verified, the experience information is sent to the first network element.

16. A communication method, characterized in that, The method includes: Receive third information from a first network element, the third information being used to request subscription to experience information associated with a first task, the third information including a subscription type and information for instructing the first task, the subscription type including a value type and / or a risk type; The experience information is sent to the first network element; The experience information includes value information and / or risk information. The value information includes the orchestration method corresponding to the successful task and the service function list determined based on the service functions participating in the successful task. The risk information includes risk characteristic information determined based on the failure factors of the failed task. The successful task is a historical successful task associated with the first task, and the failed task is a historical failed task associated with the first task.

17. A communication device, characterized in that, The communication device includes a module for performing the method according to any one of claims 1 to 8, or includes a module for performing the method according to any one of claims 9 to 15, or includes a module for performing the method according to claim 16.

18. A communication device, characterized in that, The communication device includes a processor configured to perform the method according to any one of claims 1 to 8, or to perform the method according to any one of claims 9 to 15, or to perform the method according to claim 16.

19. A chip, characterized in that, It includes at least one processor, the processor being configured to execute instructions to cause a communication device including the chip to perform the method as claimed in any one of claims 1 to 8, or the method as claimed in any one of claims 9 to 15, or the method as claimed in claim 16.

20. The chip according to claim 19, characterized in that, The chip also includes an interface circuit for receiving the executed instructions and transmitting them to the processor.

21. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store instructions that, when executed on a communication device, implement the method according to any one of claims 1 to 8, or the method according to any one of claims 9 to 15, or the method according to claim 16.

22. A computer program product containing instructions, characterized in that, When the instructions are executed on the communication device, they implement the method according to any one of claims 1 to 8, or the method according to any one of claims 9 to 15, or the method according to claim 16.

Images