Communication method, device and system

Abstract

The invention provides a communication method, device and system, and relates to the technical field of mobile communication. The communication method is executed by a DCCF (Data Collection Coordination Function) network element, and comprises the following steps: receiving a data acquisition request sent by a NWDAF (Network Data Analysis Function) network element; judging whether training data corresponding to the data acquisition request is collected or not; under the condition that the training data corresponding to the data acquisition request is not collected, sending a data subscription request to a perception function SF network element; and receiving training data sent by the SF network element in response to the data subscription request, and sending the training data sent by the SF network element to the NWDAF network element. By introducing the DCCF network element, the DCCF network element, the NWDAF network element and the SF network element cooperatively execute the training data collection process, the efficiency of training data collection is improved, and the performance and the response speed of the communication and inductance integrated network are further improved.

Description

Technical Field

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

[0002] With the development of wireless communication technology and the increasing demand for high-quality data transmission, communication systems are evolving towards a higher level. In this context, the integration of sensing and communication technology has become a highly anticipated cutting-edge technology. The core idea of ​​synaesthesia integration is to use wireless radio frequency (RF) signals to simultaneously realize sensing and communication functions. This allows us to achieve real-time monitoring of the environment and rapid data transmission without adding additional hardware burdens. It can be used to optimize network resource allocation and improve the overall efficiency of the network.

[0003] In addition, synaesthesia integration technology can also provide users with richer and more accurate services, such as location-based personalized recommendations, intelligent traffic management, etc., thereby significantly improving user experience and providing strong technical support for various intelligent applications.

[0004] In the study of synaesthesia network architecture, the sensing function (SF) can be introduced into the core network to collect and process environmental perception data, and the network data analysis function (NWDAF) can be introduced to perform model training, reasoning judgment and intelligent prediction.

[0005] In related technologies, intelligent perception based on artificial intelligence / machine learning (AI / ML) models is achieved based on the collaboration of core network elements such as SF and NWDAF. Summary of the invention

[0006] The present disclosure provides a communication method, device and system.

[0007] According to the first aspect of the present disclosure, a communication method is provided, which is executed by a data collection coordination function DCCF network element, including: receiving a data acquisition request sent by a network data analysis function NWDAF network element, wherein the data acquisition request is used to request to obtain training data required for training a machine learning model; judging whether the training data corresponding to the data acquisition request has been collected; if the training data corresponding to the data acquisition request has not been collected, sending a data subscription request to a perception function SF network element; receiving the training data sent by the SF network element in response to the data subscription request, and sending the training data sent by the SF network element to the NWDAF network element, wherein the NWDAF network element is used to train the machine learning model.

[0008] In some embodiments, the communication method further includes: in a case where the training data has been collected, sending the collected training data to the NWDAF network element.

[0009] In some embodiments, the determining whether the training data corresponding to the data acquisition request has been collected includes: determining whether the collected user device corresponding to the training data agrees to have the data collected; and if the collected user device agrees to have the data collected, determining whether the training data corresponding to the data acquisition request has been collected.

[0010] In some embodiments, the determining whether the collected user device corresponding to the training data agrees to have the data collected includes: determining the identifier of the collected user device corresponding to the training data according to the data acquisition request; querying the unified data management (UDM) network element according to the identifier of the collected user device to obtain the contract information of the collected user device; and determining whether the collected user device agrees to have the data collected according to the contract information of the collected user device.

[0011] In some embodiments, the contract information includes the data type and the purpose of data collection. Judging whether the collected user device agrees to the data being collected based on the contract information of the collected user device includes: when the data type indicates that the user device agrees to the data collection behavior and the data collection purpose indicates that the user device agrees to use the collected data for model training, determining that the collected user device agrees to the data being collected; otherwise, determining that the collected user device does not agree to the data being collected.

[0012] In some embodiments, the communication method further includes: sending a subscription message to the UDM network element when the collected user equipment agrees to have data collected, and the subscription message is used to instruct the UDM network element to notify the DCCF network element when the contract information of the collected user equipment changes.

[0013] In some embodiments, the communication method further includes: if the collected user equipment does not agree to have the data collected, sending a response of refusing to obtain the data to the NWDAF network element, wherein the response of refusing to obtain the data includes a reason for refusal.

[0014] In some embodiments, the communication method also includes: when the data acquisition request includes first storage indication information or the DCCF network element is pre-configured with second storage indication information, after receiving the training data from the SF network element, storing the training data in an analysis data storage function ADRF network element, wherein the first storage indication information and the second storage indication information are used to indicate that the training data is stored in the ADRF network element.

[0015] According to a second aspect of the present disclosure, a communication device is provided, which is arranged in a data collection and coordination function DCCF network element, and includes: a module for executing the communication method as described above.

[0016] According to a third aspect of the present disclosure, a communication device is provided, which is arranged in a data collection and coordination function DCCF network element, comprising: a memory; and a processor coupled to the memory, wherein the processor is configured to execute the communication method as described above based on instructions stored in the memory.

[0017] According to a fourth aspect of the present disclosure, a communication system is provided, comprising: a DCCF network element, configured to include the communication device as described above; a SF network element, configured to collect training data in response to a data subscription request sent by the DCCF network element, and send the training data to the DCCF network element; a NWDAF network element, configured to send a data acquisition request to the DCCF network element, and train a machine learning model based on the training data acquired from the DCCF network element.

[0018] In some embodiments, the SF network element is configured to: when the data subscription request carries the sensing mode indication information or the SF network element is pre-configured with the sensing mode indication information, collect the training data according to the sensing mode indication information.

[0019] In some embodiments, the SF network element is also configured to send a model subscription request to the NWDAF network element after receiving a service perception request sent by an access and mobility management function AMF network element, so as to trigger the NWDAF network element to perform the operation of sending the data acquisition request to the DCCF network element; the NWDAF network element is also configured to send the trained machine learning model to the SF network element; the SF network element is also configured to use the trained machine learning model and the collected perception measurement data to obtain a perception result corresponding to the service perception request, and send the perception result to the AMF network element.

[0020] According to a fifth aspect of the present disclosure, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the communication method as described above is implemented.

[0021] According to a sixth aspect of the present disclosure, a computer program product is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the communication method as described above is implemented.

[0022] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0024] Figure 1 is a flow chart of a communication method according to some embodiments of the present disclosure;

[0025] Figure 2 is a flowchart of a communication method according to some other embodiments of the present disclosure;

[0026] Figure 3 is a flow chart of a communication method performed by multiple parties according to some embodiments of the present disclosure;

[0027] Figure 4 is a flowchart of a communication method performed by multiple parties according to other embodiments of the present disclosure;

[0028] Figure 5 is a schematic diagram of the structure of a communication device according to some embodiments of the present disclosure;

[0029] Figure 6 is a schematic diagram of the structure of a communication device according to other embodiments of the present disclosure;

[0030] Figure 7 is a structural diagram of a communication system according to some embodiments of the present disclosure.

[0031] The present disclosure can be more clearly understood from the following detailed description with reference to the accompanying drawings. DETAILED DESCRIPTION

[0032] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless otherwise specifically stated.

[0033] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0034] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0035] Technologies, methods, and apparatus known to ordinary technicians in the relevant field may not be discussed in detail, but where appropriate, such technologies, methods, and apparatus should be considered part of the authorization specification.

[0036] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0037] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0038] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0039] The inventors of the present disclosure discovered that in the intelligent perception process based on artificial intelligence / machine learning models, there is a waste of resources and network overload caused by repeated data collection, which affects the performance and response speed of the synaesthesia integrated network.

[0040] In view of this, the present disclosure proposes a communication method, device and system, which can solve the problems existing in the related technologies and improve the performance and response speed of the synaesthesia integrated network.

[0041] Figure 1 FIG. 1 is a flow chart of a communication method according to some embodiments of the present disclosure. Figure 1 As shown, the communication method is executed by a data collection coordination function (DCCF) network element, including steps S11 to S14.

[0042] Step S11: receiving a data acquisition request sent by the NWDAF network element.

[0043] In some examples, before training the machine learning model, the network data analysis function (NWDAF) network element sends a data acquisition request to the DCCF network element. The data acquisition request is used to request the acquisition of training data required for training the machine learning model.

[0044] The data acquisition request may include at least one of an event identifier, a target user equipment identifier, and a target range. The event identifier is used to indicate the type of perception service, the target user equipment identifier is used to indicate which user equipment provides perception services, and the target range is used to indicate the cell where the target user equipment is located. By making the data acquisition request carry the above information, it is convenient to assist the DCCF network element in collecting the required training data. In addition, the data acquisition request may also include first storage indication information. The first storage indication information is used to indicate that in addition to returning the training data to the NWDAF network element, the training data is also stored in the Analytics Data Repository Function (ADRF) network element.

[0045] Step S12: Determine whether the training data corresponding to the data acquisition request has been collected.

[0046] In some embodiments, step S12 includes: determining whether there is training data corresponding to the data acquisition request locally; if there is training data corresponding to the data acquisition request locally, it indicates that the training data corresponding to the data acquisition request has been collected; otherwise, it indicates that the training data corresponding to the data acquisition request has not been collected.

[0047] For example, the DCCF network element retrieves local data according to the event identifier carried by the data acquisition request; if training data corresponding to the above information carried by the data acquisition request exists locally, it indicates that the training data corresponding to the data acquisition request has been collected; otherwise, it indicates that the training data corresponding to the data acquisition request has not been collected.

[0048] For another example, the DCCF network element retrieves local data based on the event identifier, target user equipment identifier, and target range carried in the data acquisition request; if training data corresponding to the above information carried in the data acquisition request exists locally, it indicates that the training data corresponding to the data acquisition request has been collected; otherwise, it indicates that the training data corresponding to the data acquisition request has not been collected.

[0049] In other embodiments, step S12 includes: determining whether there is training data corresponding to the data acquisition request locally; if so, determining whether the training data corresponding to the data acquisition request is available; if the training data existing locally is available, it indicates that the training data corresponding to the data acquisition request has been collected; otherwise, it indicates that the training data corresponding to the data acquisition request has not been collected.

[0050] For example, the DCCF network element retrieves local data according to the event identifier, target user equipment identifier, and target range carried by the data acquisition request. If there is training data corresponding to the above information carried by the data acquisition request locally, it can then be determined whether the training data is available based on at least one of the information such as the collection time of the training data and whether the status field is available. For example, when the interval between the collection time of the training data and the current time is greater than a set threshold, it is determined that the training data is unavailable; otherwise, it is determined that the training data is available. For another example, when the status field of the training data is a first value, it is determined that the training data is available; when the status field of the training data is a second value, it is determined that the training data is unavailable.

[0051] Step S13: If the training data corresponding to the data acquisition request is not collected, a data subscription request is sent to the SF network element.

[0052] In some examples, the DCCF network element may send a data subscription request to the SF network element based on the Nnf_EventExposure_Subscribe service.

[0053] The data subscription request may carry the address indication information of the DCCF network element. By making the data subscription request carry the address indication information of the DCCF network element, it is convenient for the SF network element to return the collected training data to the DCCF network element accurately and timely.

[0054] In some examples, the data subscription request may also carry perception mode indication information. The perception mode indication information is used to indicate which perception mode the SF network element adopts to collect training data. For example, the perception mode indication information can be specifically represented by the value of the perception mode identification field. When the perception mode identification field is the first value, it indicates that the SF network element adopts the base station self-transmitting and self-receiving perception mode to collect data; when the perception mode identification field is the second value, it indicates that the SF network element adopts the base station-to-base station transceiver perception mode to collect data; when the perception mode identification field is the third value, it indicates that the SF network element adopts the user equipment-to-base station-to-base station perception mode to collect data; when the perception mode identification field is the fourth value, it indicates that the SF network element adopts the user equipment-to-self-transmitting and self-receiving perception mode to collect data; when the perception mode identification field is the fifth value, it indicates that the SF network element adopts the user equipment-to-user equipment transceiver perception mode to collect data; when the perception mode identification field is the sixth value, it indicates that the SF network element adopts the base station-to-user equipment-to-receive perception mode to collect data. By making the data subscription request carry the perception mode indication information, it is convenient for the SF network element to adopt the perception mode indicated by the DCCF network element to collect training data, which helps to improve the efficiency of training collection.

[0055] In some examples, the data subscription request may also carry the indication information of the device to be collected. The indication information of the device to be collected is used to indicate which user devices the SF network element collects data from. By making the data subscription request carry the indication information of the device to be collected, it is convenient for the SF network element to collect training data for the device specified by the DCCF network element without executing the device selection process to be collected, which helps to improve the efficiency of training collection.

[0056] Step S14: receiving the training data sent by the SF network element, and sending the training data sent by the SF network element to the NWDAF network element.

[0057] After receiving the data subscription request, the SF network element collects the required training data and sends the collected training data to the DCCF network element. The DCCF network element receives the training data sent by the SF network element and sends it to the NWDAF network element so that the NWDAF network element can train the machine learning model based on the training data.

[0058] In some embodiments, the communication method further includes: in a case where it is determined that the training data corresponding to the data acquisition request has been collected, the DCCF network element sending the collected training data to the NWDAF network element.

[0059] In the disclosed embodiment, by introducing the DCCF network element, cooperating with the NWDAF network element and the SF network element to execute the above training data collection process, it helps to improve the efficiency of training data collection, thereby improving the performance and response speed of the synaesthesia integrated network.

[0060] Figure 2 FIG. 1 is a flow chart of a communication method according to other embodiments of the present disclosure. Figure 2 As shown, the communication method is executed by a DCCF network element, including steps 201 to 210.

[0061] Step 201: Receive a data acquisition request sent by an NWDAF network element.

[0062] In some examples, the data acquisition request includes an event identifier, a target user device identifier, and a target range. In addition, the data acquisition request may also include first storage indication information.

[0063] Step 202: Determine whether the user device agrees to have its data collected.

[0064] Step 202 can be implemented in a variety of ways, and two implementations are described below for exemplary purposes.

[0065] In the first implementation, step 202 includes: determining the identifier of the collected user device corresponding to the training data according to the data acquisition request; querying the Unified Data Management (UDM) network element according to the identifier of the collected user device to obtain the contract information of the collected user device; and judging whether the collected user device agrees to the data collection according to the contract information of the collected user device.

[0066] In some examples, the contract information includes a data type. In these examples, judging whether the collected user device agrees to have the data collected based on the contract information includes: if the data type indicates that the user device agrees to the data collection behavior, determining that the collected user device agrees to have the data collected; otherwise, determining that the collected user device does not agree to have the data collected.

[0067] In other examples, the contract information includes the data type and the purpose of data collection. In these examples, judging whether the collected user device agrees to have the data collected based on the contract information includes: if the data type indicates that the user device agrees to the data collection behavior, and the data collection purpose indicates that the user device agrees to use the collected data for model training, determining that the collected user device agrees to have the data collected; otherwise, determining that the collected user device does not agree to have the data collected.

[0068] In the second implementation, step 202 includes: determining the identifier of the collected user device corresponding to the training data according to the data acquisition request; instructing the SF network element to send a collection authorization request to the collected user device according to the identifier of the collected user device; and judging whether the collected user device agrees to have the data collected according to the collection authorization response returned by the SF network element.

[0069] For example, the DCCF network element first determines the identifier of the collected user device corresponding to the training data based on the event identifier, target UE identifier and other information carried in the data acquisition request; then instructs the SF network element to send a collection authorization request to these collected user devices; if the SF network element receives a response from these collected user devices agreeing to data collection, it sends a first collection authorization response to the DCCF network element to let the DCCF network element know that these collected user devices agree to have their data collected; if the SF network element receives a response from these collected user devices rejecting data collection, it sends a second collection authorization response to the DCCF network element to let the DCCF network element know that these collected user devices do not agree to have their data collected.

[0070] In the specific implementation, considering that there may be multiple user devices to be collected, the judgment result of step 202 can be confirmed to be yes when at least some of the collected user devices agree to have their data collected. The DCCF network element can then instruct the SF network element to collect data from these collected user devices that agree to have their data collected. In addition, in other embodiments, the judgment result of step 202 can also be confirmed to be yes when all the collected user devices agree to have their data collected.

[0071] If the judgment result of step 202 is yes, execute step 203; otherwise, execute step 210.

[0072] In some embodiments, the communication method further includes: if the judgment result of step 202 is yes, sending a subscription message to the UDM network element. The subscription message is used to instruct the UDM network element to notify the DCCF network element when the contract information of the collected user equipment changes. For example, if the user equipment agrees to data collection, the DCCF network element uses the Nudm_SDM_Subscribe service to subscribe to the UDM network element for change notifications of the data type in the contract information of the user equipment and the purpose of data collection. Through the above operations, it is possible to respond in a timely manner after the user equipment changes the contract information, further improving the security of training data collection.

[0073] In the disclosed embodiment, by having the DCCF network element determine whether the user device agrees to data collection, and performing subsequent steps such as data collection after the user device agrees to data collection, and returning the collected data to the NWDAF network element, the security of collecting training data in the intelligent perception process based on the machine learning model can be improved, the privacy of user data can be better protected, and it helps to improve the user experience of the integrated network service.

[0074] Step 203: Determine whether training data has been collected.

[0075] In step 203, it is determined whether the training data corresponding to the data acquisition request has been collected. For details on how to perform step 203, reference may be made to the relevant descriptions in the above embodiments.

[0076] If the judgment result of step 203 is yes, execute steps 204 and 205; otherwise, execute steps 206 to 209.

[0077] In some embodiments, the communication method further includes: if the judgment result of step 203 is no, adding the NWDAF network element to a list of data consumers that have subscribed to the training data.

[0078] Step 204: Send the collected training data to the NWDAF network element.

[0079] In some embodiments, the communication method further includes: when it is determined that the training data corresponding to the data acquisition request has been collected, the DCCF network element adds the NWDAF network element to a data consumer list corresponding to the training data. By maintaining the data consumer list, it is easy to understand the consumption of various types of training data.

[0080] Step 205: Store the collected training data in the ADRF network element.

[0081] In some examples, when the data acquisition request includes the first storage indication information, the DCCF network element, in addition to returning the collected training data corresponding to the data acquisition request to the NWDAF network element, also stores the collected training data to the ADRF network element according to the first storage indication information. The first storage indication information is used to indicate that the training data is to be stored in the ADRF network element.

[0082] In other examples, when the DCCF network element is pre-configured with the second storage indication information, in addition to returning the collected training data corresponding to the data acquisition request to the NWDAF network element, the DCCF network element also stores the collected training data to the ADRF network element according to the second storage indication information. The second storage indication information is used to indicate that the collected training data is stored in the ADRF network element. By storing the training data in the ADRF network element, it is convenient to carry out long-term storage and analysis of the training data, thereby improving the flexibility of the synaesthesia integrated network service.

[0083] It should be noted that the embodiment of the present disclosure is described by taking the example of executing step 204 first and then executing step 205. In specific implementation, step 205 may be executed first and then step 204. Alternatively, step 204 and step 205 may be executed simultaneously.

[0084] Step 206: Send a data subscription request to the SF network element.

[0085] In some examples, the DCCF network element uses the Nnf_EventExposure_Subscribe service to send a data subscription request to the SF network element, wherein the data subscription request may include address indication information of the DCCF network element.

[0086] Step 207: Receive training data sent by the SF network element.

[0087] After receiving the data subscription request, the SF network element collects data for model training from the user equipment and / or the radio access network (RAN) according to the sensing mode (e.g., base station self-transmitting and self-receiving sensing mode, inter-base station transceiving sensing mode, user equipment transmitting and base station receiving sensing mode, user equipment self-transmitting and self-receiving sensing mode, inter-user equipment transceiving sensing mode, and base station transmitting and user equipment receiving sensing mode). Afterwards, the SF network element sends the collected training data to the DCCF network element.

[0088] Step 208: Send the training data sent by the SF network element to the NWDAF network element.

[0089] In some examples, the DCCF network element uses the Ndccf_DataManagement_Notify service to send the training data collected by the SF network element to the NWDAF network element.

[0090] Step 209: Store the training data sent by the SF network element in the ADRF network element.

[0091] In some examples, when the data acquisition request includes the first storage indication information, the DCCF network element, in addition to returning the training data corresponding to the data acquisition request collected by the SF network element to the NWDAF network element, also stores the training data in the ADRF network element according to the first storage indication information. The first storage indication information is used to indicate that the training data is to be stored in the ADRF network element.

[0092] In other examples, when the DCCF network element is pre-configured with the second storage indication information, the DCCF network element, in addition to returning the training data corresponding to the data acquisition request collected by the SF network element to the NWDAF network element, also stores the above training data to the ADRF network element according to the second storage indication information. The second storage indication information is used to indicate that the training data is stored in the ADRF network element. By storing the training data in the ADRF network element, it is convenient to carry out long-term storage and analysis of the training data, and improve the flexibility of the synaesthesia integrated network service.

[0093] It should be noted that the embodiment of the present disclosure is described by taking the example of first executing step 208 and then executing step 209. In specific implementation, step 209 may be executed first and then step 208. Alternatively, step 208 and step 209 may be executed simultaneously.

[0094] Step 210: Send a response of refusing to obtain data to the NWDAF network element.

[0095] The response of refusing to obtain data carries the reason for refusal. By sending the response of refusing to obtain data to the NWDAF network element when the user equipment does not agree to data collection, the NWDAF network element can promptly learn of the data acquisition failure and take other related operations accordingly.

[0096] In the disclosed embodiment, by introducing the DCCF network element and coordinating the above training data collection process with the NWDAF network element and the SF network element, the following technical effects can be achieved: 1. It can effectively alleviate the resource waste and network overload problems caused by a large number of repeated training data acquisition requests, and optimize the performance and response speed of the synaesthesia integrated network; 2. By checking whether the user agrees to collect data before collecting training data, the privacy protection of user data is strengthened; 3. It supports different data collection and processing requirements and optional storage for ADRF network elements, provides a flexible data management strategy, and has good scalability to adapt to the development of future technologies.

[0097] Figure 3 FIG. 1 is a flow chart of a communication method performed by multiple parties according to some embodiments of the present disclosure. Figure 3 As shown, the communication method is jointly executed by multiple network elements such as NWDAF network element, SF network element, DCCF network element, etc., including steps 301 to 318.

[0098] Step 301: The AF network element sends a service perception request to the NEF network element.

[0099] In the disclosed embodiment, the intelligent perception process is triggered by an application function (AF) network element. In step 301, the AF network element sends a service perception request to a network exposure function (NEF) network element.

[0100] In some examples, the service perception request includes service type, service requirements, and designated perception node information. The service type is used to indicate the type of perception service, such as dynamic map, vehicle tracking, positioning, etc. The service requirements are used to indicate the requirements for the perception service. For example, the service requirements include one or more of the service requirements such as perception resolution, perception accuracy, perception duration, target area information, and latency. The designated perception node information is used to indicate which nodes are to be perceived. For example, the designated perception node information is the user equipment information to be perceived.

[0101] In some embodiments, before step 301, the communication method further includes: the NWDAF network element and the SF network element are respectively registered with a network data analytics function (NRF) network element.

[0102] In specific implementation, when the NWDAF network element registers with the NRF network element, the registered information may include the network element's support identification for artificial intelligence / machine learning-based perception capabilities, machine learning model filtering information, and first machine learning model interoperability information. Among them, the first machine learning model interoperability information may include: a list of manufacturers that share machine learning models with the NWDAF network element, and the format and operating environment of the machine learning model.

[0103] When the SF network element registers with the NRF network element, the registered information may include the network element's support identification for the perception capability based on artificial intelligence / machine learning, and the second machine learning model interoperability information. Among them, the second machine learning model interoperability information includes the manufacturer identification of the SF network element itself and the format and operating environment of the machine learning model supported by the SF network element itself.

[0104] Step 302: The NEF network element sends a service perception request to the AMF network element.

[0105] In some embodiments, step 302 includes: the NEF network element performs an authorization check on the service perception request sent by the AF network element; if the authorization check passes, the NEF network element selects an access and mobility management function (Access and MobilityManagement Function, referred to as AMF) network element, and sends the service perception request to the selected AMF network element; if the authorization check fails, the NEF network element sends a response that the authorization check failed to the AF network element.

[0106] Among them, the authorization check can be implemented in a variety of ways. For example, the authorization check includes: determining the identifier of the AF network element that sends the service perception request according to the service perception request; querying the UDM network element according to the identifier of the AF network element to obtain the query result; determining whether the AF network element has the right to use a specific perception service according to the query result, wherein the specific perception service can be determined according to the service perception request; if the query result shows that the AF network element has the right to use the specific perception service, the authorization check passes; otherwise, the authorization check fails. After the authorization check passes, the NEF network element can select a suitable AMF network element according to the perception node information, target area information, etc. carried by the service perception request.

[0107] Step 303: The AMF network element selects the SF network element.

[0108] In this step, after receiving the service perception request, the AMF network element can select a suitable SF network element according to the target area information or target location information carried in the service perception request.

[0109] Step 304: The AMF network element sends a service perception request to the SF network element.

[0110] Step 305: The SF network element sends a model subscription request to the NWDAF network element.

[0111] In some examples, after receiving the service perception request, the SF network element first checks whether there is a trained model that matches the service perception request. If there is no trained model that matches the service perception request, the SF network element first discovers the NWDAF network element that supports model training (i.e., the NWDAF network element containing MTLF) through the NRF network element, and then executes step 305; if there is a trained model that matches the service perception request, the SF network element uses the model to perform reasoning perception to obtain the perception result, and feeds the perception result back to the AF network element through steps 316 to 318.

[0112] In other examples, after receiving the service perception request, the SF network element first determines whether to use a method based on a machine learning model to calculate the perception result. In the case where it is determined that a method based on a machine learning model is used to calculate the perception result, the SF network element then checks whether there is a trained model that matches the service perception request. If there is no trained model that matches the service perception request, the SF network element first discovers an NWDAF network element that supports model training (i.e., an NWDAF network element containing MTLF) through the NRF network element, and then executes step 305; if there is a trained model that matches the service perception request, the SF network element uses the model to perform inference perception to obtain the perception result, and feeds the perception result back to the AF network element through steps 316 to 318. In addition, in the case where it is determined that the perception result is not calculated based on a machine learning model, the perception result is calculated using the traditional algorithm configured inside the SF network element.

[0113] In 305, the SF network element sends a model subscription request to the NWDAF network element through the Nnwdaf_MLModelProvision_Subscribe service, wherein the model subscription request may include machine learning model filtering information and second machine learning model interoperability information.

[0114] Step 306: The NWDAF network element sends a data acquisition request to the DCCF network element.

[0115] After receiving the model subscription request, the NWDAF network element can send a data acquisition request to the DCCF network element through the Ndccf_DataManagement_Subscribe service to request the data required for training the machine learning model. The data acquisition request may include an event identifier, a target user device identifier, and a target range.

[0116] Step 307: The DCCF network element performs a data collection authorization check.

[0117] In this step, the DCCF network element may perform a data collection authorization check in the following manner: the DCCF network element determines the identification of the user device whose data is being collected based on the data acquisition request; based on the identification of the user device whose data is being collected, retrieves the contract information of these user devices whose data is being collected from the UDM network element; if the contract information indicates that these user devices agree to have their data collected and agree to use these data for model training, the data collection authorization check passes; otherwise, the data collection authorization check fails.

[0118] If the data collection authorization check passes, step 308 is executed; otherwise, the DCCF network element sends a response of refusing to obtain data to the NWDAF network element.

[0119] In addition, if the data collection authorization check is passed, the DCCF network element can also use the Nudm_SDM_Subscribe service to subscribe to the UDM network element for notifications of changes in the contract information of the user equipment whose data is being collected.

[0120] Step 308: The DCCF network element checks whether there is training data locally.

[0121] In this step, the DCCF network element checks whether there is training data corresponding to the data acquisition request locally; if not, it indicates that the training data corresponding to the data acquisition request has not been collected, and step 309 is executed; if so, it indicates that the training data corresponding to the data acquisition request has been collected, and step 312 is executed.

[0122] In addition, if the DCCF network element has collected the requested training data, the DCCF network element will also add the NWDAF network element that sent the data acquisition request to the data consumer list corresponding to the training data.

[0123] Step 309: The DCCF network element sends a data collection request to the SF network element.

[0124] In this step, the DCCF network element may send a data collection request to the SF network element through the Nnf_EventExposure_Subscribe service. The data collection request may include address indication information of the DCCF network element. In addition, after executing step 309, the DCCF network element may also add the NWDAF network element that sends the data acquisition request to the list of data consumers that have subscribed to the training data.

[0125] Step 310: The SF network element collects training data.

[0126] In this step, the SF network element may collect the required training data according to the specified sensing mode, wherein the specified sensing mode may be a base station self-transmitting and self-receiving sensing mode, a base station-to-base station transceiving sensing mode, a user equipment transceiving and base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-base station-to-user equipment ...

[0127] For example, if the sensing mode is a base station spontaneous transmission and self-reception sensing mode, the SF network element may collect training data on the base station side; if the sensing mode is an inter-base station transmission and reception sensing mode, the SF network element may collect training data on multiple base station sides; if the sensing mode is a user equipment transmission and base station reception sensing mode, the SF network element may collect training data on the user equipment and base station sides; if the sensing mode is a user equipment spontaneous transmission and self-reception sensing mode, the SF network element may collect sensing measurement data on the user equipment side; if the sensing mode is inter-user equipment transmission and reception sensing, the SF network element may collect sensing measurement data on multiple user equipment sides; if the sensing mode is a base station transmission and user equipment reception sensing mode, the SF network element may collect sensing measurement data on both the user side and the base station side.

[0128] Step 311: The SF network element sends training data to the DCCF network element.

[0129] When data collection is completed, the SF network element can send the collected training data to the DCCF network element through the Nnf_EventExposeure_Notify service.

[0130] Step 312: The DCCF network element sends the training data to the NWDAF network element.

[0131] In this step, the DCCF network element can use the Ndccf_DataManagement_Notify service to send the training data to the NWDAF network element. In addition, if the data acquisition request sent by the NWDAF network element carries the first storage indication information or the DCCF network element is pre-set with the second storage indication information, the DCCF network element will also store the collected training data in the ADRF network element. The first storage indication information and the second storage indication information are used to indicate that the collected training data is stored in the ADRF network element.

[0132] Step 313: The NWDAF network element trains the machine learning model.

[0133] In this step, the NWDAF network element trains and updates the machine learning model based on the acquired training data.

[0134] Step 314: The NWDAF network element provides the trained model to the SF network element.

[0135] The NWDAF NE can provide the trained machine learning model to the SF NE through the Nnwdaf_MLModelProvision_Notify service.

[0136] Step 315: SF network element collects data and performs model inference.

[0137] After receiving the trained machine learning model, the SF network element collects the data required for perception reasoning, and uses the trained machine learning model to process the data required for perception reasoning to obtain the perception results corresponding to the business perception request.

[0138] Step 316: The SF network element sends the sensing result to the AMF network element.

[0139] Step 317: The AMF network element sends the sensing result to the NEF network element.

[0140] Step 318: The NEF network element sends the sensing result to the AF network element.

[0141] In the disclosed embodiment, by collaboratively executing the above process through SF network elements, NWDAF network elements, and DCCF network elements, intelligent perception based on machine learning can be realized, the data fusion and intelligent analysis capabilities of the synaesthesia integrated network can be improved, and the possibility of obtaining more accurate perception results is provided. At the same time, collaborative data collection through SF network elements and DCCF network elements helps to optimize network resource allocation, thereby improving the overall efficiency of the network, effectively alleviating the network overload problem caused by a large number of repeated data acquisition requests, and helping to optimize network performance and response speed; moreover, by introducing a data collection authorization check process before collecting data, the privacy protection of user data can be strengthened. In addition, the communication method of the disclosed embodiment supports different data collection and processing requirements, as well as optional storage of ADRF network elements, provides a flexible data management strategy, and has good scalability to adapt to the development of future technologies.

[0142] Figure 4 FIG. 1 is a flow chart of a communication method performed by multiple parties according to other embodiments of the present disclosure. Figure 4 As shown, the communication method is jointly executed by multiple network elements such as NWDAF network element, SF network element, DCCF network element, etc., and includes steps 401 to 416. The communication method of the embodiment of the present disclosure is Figure 3 The main difference between the communication methods shown in the figure is that the communication method of the embodiment of the present disclosure is triggered by the user equipment, while Figure 3 The communication method shown is triggered and executed by the AF network element.

[0143] Step 401: The UE sends a service awareness request to the AMF network element.

[0144] Step 402: The AMF network element selects the SF network element.

[0145] Step 403: The AMF network element sends a service perception request to the SF network element.

[0146] Steps 404 to 414 and Figure 3 Steps 305 to 315 in the illustrated process are consistent and will not be described in detail below.

[0147] Step 415: The SF network element sends the sensing result to the AMF network element.

[0148] Step 416: The AMF network element sends the sensing result to the UE.

[0149] In the disclosed embodiment, by combining the perception data collection and reasoning capabilities of the perception network element SF with the model training capabilities of the NWDAF network element and the data collection and coordination capabilities of the DCCF network element, it is possible to more efficiently realize intelligent perception based on machine learning, improve the performance and response speed of the integrated synaesthesia network, support the development of various intelligent perception applications, and promote the intelligent evolution of the network.

[0150] Figure 5 is a schematic diagram of the structure of a communication device according to some embodiments of the present disclosure. Figure 5 As shown, the communication device 50 is arranged in the DCCF network element, and is used to execute the communication method as described above. The device specifically includes a receiving module 51, a judging module 52, a subscription module 53, and a data transceiving module 54.

[0151] The receiving module 51 is configured to receive a data acquisition request sent by the NWDAF network element, wherein the data acquisition request is used to request to acquire the training data required for training the machine learning model.

[0152] The determination module 52 is configured to determine whether the training data corresponding to the data acquisition request has been collected.

[0153] The subscription module 53 is configured to send a data subscription request to the SF network element when the training data corresponding to the data acquisition request is not collected.

[0154] The data transceiver module 54 is configured to receive the training data sent by the SF network element in response to the data subscription request, and send the training data sent by the SF network element to the NWDAF network element. The NWDAF network element is used to train the machine learning model.

[0155] In some embodiments, the data transceiver module 54 is further configured to: in the case where the training data corresponding to the data acquisition request has been collected, send the collected training data to the NWDAF network element.

[0156] In the embodiments of the present disclosure, the above device helps to improve the efficiency of training data collection, thereby improving the performance and response speed of the synaesthesia integration network.

[0157] Figure 6 is a schematic diagram of the structure of a communication device according to some other embodiments of the present disclosure. Figure 6 As shown, the communication device 60 includes a memory 61 and a processor 62 coupled to the memory 61. The memory 61 is used to store instructions for executing the corresponding embodiments of the communication method. The processor 62 is configured to execute the communication method in any of the embodiments of the present disclosure based on the instructions stored in the memory 61.

[0158] Figure 7 is a schematic diagram of the structure of a communication system according to some embodiments of the present disclosure. Figure 7 As shown, the communication system 70 includes a NWDAF network element 71, a DCCF network element 72, and a SF network element 73.

[0159] The NWDAF network element 71 is configured to send a data acquisition request to the DCCF network element 72. The data acquisition request is used to request the acquisition of training data required for training the machine learning model.

[0160] The DCCF network element 72 is configured to include the communication device as described above, and is used to perform the communication method as described above. For example, the DCCF network element 72 receives the data acquisition request sent by the NWDAF network element 71; determines whether the training data corresponding to the data acquisition request has been collected; if the training data corresponding to the data acquisition request has not been collected, sends a data subscription request to the SF network element 73; receives the training data sent by the SF network element 73 in response to the data subscription request, and sends the training data sent by the SF network element 73 to the NWDAF network element 71.

[0161] The SF network element 73 is configured to collect training data in response to the data subscription request sent by the DCCF network element 72 , and send the training data to the DCCF network element 72 .

[0162] In some examples, the SF network element 73 is configured to: collect training data according to the perception mode indication information when the data subscription request carries the perception mode indication information or the SF network element is pre-configured with the perception mode indication information.

[0163] The NWDAF network element 71 is also configured to train the machine learning model based on the training data obtained from the DCCF network element 72.

[0164] In some embodiments, the SF network element 73 is also configured to send a model subscription request to the NWDAF network element 71 after receiving the service perception request sent by the AMF network element, so as to trigger the NWDAF network element 71 to execute the operation of sending a data acquisition request to the DCCF network element; the NWDAF network element 71 is also configured to send the trained machine learning model to the SF network element 73; the SF network element 73 is also configured to use the trained machine learning model and the collected perception measurement data to obtain the perception result corresponding to the service perception request, and send the perception result to the AMF network element.

[0165] In some embodiments, the communication system may further include a device for performing the following steps: Figure 3 or Figure 4 Other network elements or devices in the process shown.

[0166] In the disclosed embodiment, the above system can improve the efficiency of training data collection, thereby improving the performance and response speed of the synaesthesia integration network.

[0167] Here, various aspects of the present disclosure are described with reference to flowcharts and / or block diagrams of methods, devices, and computer program products according to embodiments of the present disclosure. It should be understood that each frame of the flowchart and / or block diagram and the combination of frames can be implemented by computer-readable program instructions.

[0168] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable device to produce a machine, so that the processor executes the instructions to produce means for implementing the functions specified in one or more blocks in the flowchart and / or block diagram.

[0169] These computer-readable program instructions may also be stored in a computer-readable memory, which cause the computer to work in a specific manner to produce an article of manufacture, including instructions for implementing the functions specified in one or more blocks in the flowchart and / or block diagram.

[0170] The present disclosure can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0171] So far, the communication method, device and system according to the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed here.

Claims

1. A communication method, performed by a data collection coordination function DCCF network element, comprising: Receiving a data acquisition request sent by a network data analysis function NWDAF network element, wherein the data acquisition request is used to request to obtain training data required for training a machine learning model; Determining whether training data corresponding to the data acquisition request has been collected; In case that the training data corresponding to the data acquisition request is not collected, sending a data subscription request to the perception function SF network element; Receive the training data sent by the SF network element in response to the data subscription request, and send the training data sent by the SF network element to the NWDAF network element, wherein the NWDAF network element is used to train the machine learning model.

2. The communication method according to claim 1, further comprising: In the case where the training data has been collected, the collected training data is sent to the NWDAF network element.

3. The communication method according to claim 1, wherein: The determining whether the training data corresponding to the data acquisition request has been collected includes: Determine whether the user device corresponding to the training data agrees to have the data collected; In a case where the collected user equipment agrees to have data collected, it is determined whether training data corresponding to the data acquisition request has been collected.

4. The communication method according to claim 3, wherein: The step of determining whether the collected user equipment corresponding to the training data agrees to have the data collected comprises: Determining, according to the data acquisition request, an identifier of a user device corresponding to the training data; Querying a unified data management (UDM) network element according to the identifier of the collected user equipment to obtain the contract information of the collected user equipment; According to the contract information of the collected user equipment, it is determined whether the collected user equipment agrees to have its data collected.

5. The communication method according to claim 4, wherein: The contract information includes the data type and the purpose of data collection. The determining, based on the contract information of the collected user device, whether the collected user device agrees to have the data collected includes: When the data type indicates that the user device agrees to the data collection and the data collection purpose indicates that the user device agrees to use the collected data for model training, it is determined that the collected user device agrees to the data collection; otherwise, it is determined that the collected user device does not agree to the data collection.

6. The communication method according to claim 4, further comprising: In the case where the collected user equipment agrees to have its data collected, a subscription message is sent to the UDM network element, wherein the subscription message is used to instruct the UDM network element to notify the DCCF network element when the contract information of the collected user equipment changes.

7. The communication method according to claim 3, further comprising: In the case that the collected user equipment does not agree to have the data collected, a response of refusing to obtain the data is sent to the NWDAF network element, wherein the response of refusing to obtain the data includes a reason for refusal.

8. The communication method according to any one of claims 1 to 7, further comprising: In a case where the data acquisition request includes first storage indication information or the DCCF network element is pre-configured with second storage indication information, after receiving the training data from the SF network element, the training data is stored in an analysis data storage function ADRF network element, wherein the first storage indication information and the second storage indication information are used to indicate that the training data is stored in the ADRF network element.

9. A communication device, arranged in a data collection and coordination function DCCF network element, comprising: A module for executing the communication method according to any one of claims 1 to 8.

10. A communication device, arranged in a data collection and coordination function DCCF network element, comprising: Memory; as well as A processor coupled to the memory, wherein the processor is configured to execute the communication method according to any one of claims 1 to 8 based on instructions stored in the memory.

11. A communication system comprising: A DCCF network element, configured to include the communication device according to claim 9 or 10; The SF network element is configured to collect training data in response to the data subscription request sent by the DCCF network element, and send the training data to the DCCF network element; The NWDAF network element is configured to send a data acquisition request to the DCCF network element, and train the machine learning model based on the training data obtained from the DCCF network element.

12. The communication system according to claim 11, wherein: The SF network element is configured as follows: In a case where the data subscription request carries the perception mode indication information or the SF network element is pre-configured with the perception mode indication information, the training data is collected according to the perception mode indication information.

13. The communication system according to claim 11, wherein: The SF network element is further configured to, after receiving the service perception request sent by the access and mobility management function AMF network element, send a model subscription request to the NWDAF network element to trigger the NWDAF network element to perform the operation of sending the data acquisition request to the DCCF network element; The NWDAF network element is further configured to send the trained machine learning model to the SF network element; The SF network element is also configured to use the trained machine learning model and the collected perception measurement data to obtain a perception result corresponding to the service perception request, and send the perception result to the AMF network element.

14. A computer-readable storage medium having computer instructions stored thereon, wherein the instructions, when executed by a processor, implement the communication method according to any one of claims 1 to 8.

15. A computer program product having computer instructions stored thereon, wherein the instructions are executed by a processor to implement the communication method according to any one of claims 1 to 8.

Images