Communication control method and communication device based on user intention prediction

Abstract

The invention provides a communication control method and a communication device based on user intention prediction, and the method comprises the steps: carrying out the semantic analysis of the semantic data of the content and / or service of a historical request of a user, and obtaining the intention data corresponding to the content / service of the historical request of the user; after the intention data and the content and / or service (namely the historical data of the content and / or service requested by the user) of the historical request of the user are fused, a deep learning algorithm is used for analysis, intention prediction information of the user is obtained, advanced deployment of a network architecture and advanced configuration of network resources are completed through the intention prediction information of the user, and the user experience is improved. According to the invention, correct, reasonable and timely distribution of network resources is realized, and rationality and effectiveness of network architecture deployment are ensured, so that communication efficiency is improved.

Description

Communication control method and communication device based on user intention prediction Technical Field

[0001] The present application relates to the field of communications, and more specifically, to a communication control method and a communication device based on user intention prediction. Background Art

[0002] Over the past few decades, wireless communication systems have evolved and been researched, from first-generation analog communication systems to 5G NR communication systems and the future 6G communication systems. 5G communication systems, by integrating the development of technologies such as artificial intelligence and big data, are gradually becoming a reality with emerging services such as autonomous driving, drones, and immersive extended reality. To better serve users and conserve computing and bandwidth resources, these services often require flexible, on-demand, and high-quality resource allocation. Overall, future wireless communication networks need to strengthen their integration with artificial intelligence (AI), leveraging AI's powerful predictive and decision-making capabilities to create user-centric, intelligent, and endogenous 6G networks.

[0003] To achieve more intelligent network deployment and resource allocation, the industry has proposed a number of new network architectures. Among them, intent-based networking (IBN) has attracted much attention. With the introduction of user intent, network autonomy is becoming possible through analyzing user intent, managing policies, and optimizing resources.

[0004] In recent years, the use of natural language processing (NLP) to analyze user intent has become a major research topic. While NLP models can identify user intent to a certain extent, they are currently limited by the highly dynamic and time-varying network environment, making it difficult to pre-deploy network architectures and configure network resources, thus hindering communication efficiency.

[0005] Summary of the Invention

[0006] The present application provides a communication control method and a communication device based on user intention prediction. According to the user's intention prediction information, the advance deployment of the network architecture and the advance configuration of network resources are completed, thereby realizing the correct, reasonable and timely allocation of network resources, ensuring the rationality and effectiveness of the network architecture deployment, and thus improving communication efficiency.

[0007] On the first aspect, a communication control method based on user intention prediction is provided, and the execution subject of the method can be a core network device or a chip applied to the core network device. The method includes: determining user intention data based on semantic data corresponding to the content and / or services historically requested by the user, and the user intention data reflects the historical intention of the user; fusing the historical data of the content and / or services requested by the user with the user intention data; using a deep learning network model to process the fused data to obtain user intention prediction information, and the intention prediction information includes: the probability of the user making a request for a certain content and / or service at a certain time in the future; using the intention prediction information to complete the advance deployment of the network architecture and the advance configuration of network resources.

[0008] The communication control method based on user intention prediction provided in the first aspect performs semantic analysis on the semantic data of the content and / or services historically requested by the user to obtain the intention data corresponding to the content / service historically requested by the user, fuses the intention data with the content and / or services historically requested by the user (i.e., the historical data of the content and / or services requested by the user), and then uses a deep learning algorithm to perform analysis to obtain the user's intention prediction information. The user's intention prediction information is used to complete the advance deployment of the network architecture and the advance configuration of network resources, thereby realizing the correct, reasonable and timely allocation of network resources, ensuring the rationality and effectiveness of the network architecture deployment, and thus improving communication efficiency.

[0009] Exemplarily, the historical data of the content and / or services requested by the user includes:

[0010] The user ID of the user who initiated the request within the preset time length, the identification ID of the requested content and / or service within the preset time length, and the time information of the request initiated by the user within the preset time length.

[0011] For example, the semantic data corresponding to the content and / or services requested by the user in the past may include keywords of the requested content and / or services, such as the genre of the requested content (e.g., a movie), the purpose of using the smart service, etc. This keyword information can indirectly reflect the user's preferences and purposes.

[0012] In any possible implementation of the first aspect, determining user intent data based on semantic data corresponding to content and / or services historically requested by a user includes: semantically encoding keywords included in the semantic data using a natural language processing model to obtain multiple semantic feature vectors, where the semantic feature vectors are used to represent semantic similarity and analogy between the keywords; and semantically aggregating the multiple semantic feature vectors to obtain the user intent data. This implementation improves the accuracy of the obtained user intent data.

[0013] In any possible implementation of the first aspect, the multiple semantic feature vectors are semantically aggregated to obtain the user intention data using at least one of the following methods: addition, averaging, a deep learning model based on a recurrent neural network, or a deep learning model based on an attention mechanism.

[0014] Exemplarily, for example, for any semantic data, an NLP training model can be used to encode each keyword in the semantic data to obtain a vector representation of semantic features such as keyword semantic similarity and analogy (i.e., a semantic feature vector), thereby achieving semantic encoding. For multiple keyword information (i.e., multiple semantic feature vectors) contained in a content and / or service, aggregation operations (aggregation of multiple semantic feature vectors) can be performed using addition, averaging, and deep learning models based on recurrent neural networks, attention mechanisms, etc. to obtain user intent data (expressed in the form of a vector) for the requested content and / or service. The user intent data can reflect the user's intention and can be used as input for subsequent prediction tasks.

[0015] In any possible implementation of the first aspect, the historical data of the content and / or service requested by the user and the user intention data are integrated, including: integrating the historical data of the content and / or service requested by the user and the user intention data by adding or splicing.

[0016] In any possible implementation of the first aspect, fusing historical data of user-requested content and / or services with the user intent data includes: preprocessing the historical data of the user-requested content and / or services; fusing the preprocessed historical data of the user-requested content and / or services with the user intent data by adding or concatenating them; the preprocessing including at least one of averaging the historical data of the user-requested content and / or services, using a recurrent neural network or an attention mechanism deep learning model, or filtering the historical data of the user-requested content and / or services using a time parameter. In this implementation, preprocessing the historical data of the user-requested content and / or services can enhance the deep learning network model's perception of the historical data, thereby improving the accuracy and precision of the deep learning network model's data processing.

[0017] In any possible implementation of the first aspect, processing the fused data using a deep learning network model to obtain user intent prediction information includes: using the deep learning network model to obtain node information in historical data of content and / or services requested by the user, the node including: the user and the content and / or service requested by the user; using the deep learning network model to generate a vector representation for each node containing structural features, temporal features, and semantic features, the temporal feature being the time at which the user requested the content and / or service; using the deep learning network model to calculate a matching value between the user and the content and / or service; and obtaining the intent prediction information based on the matching value. In this implementation, the accuracy of the obtained intent prediction information can be improved.

[0018] In any possible implementation of the first aspect, the intention prediction information is used to complete the advance deployment of the network architecture and the advance configuration of network resources, including: based on the intention prediction information, using the monitoring network element to determine whether the intention prediction information meets the preset threshold, determine whether the current policy meets the requirements, record the historical data of the content and / or services requested by the user, and the semantic data corresponding to the content and / or services historically requested by the user; using the selection network element to update and optimize the current policy; using the deployment network element to convert the current policy or the updated policy into a control instruction for the wireless network infrastructure according to the configuration instruction specification of the network parameters; using the execution network element to execute the control instruction to complete the advance deployment of the network architecture and the advance configuration of the computing resources for the content and / or services requested by the user.

[0019] For example, corresponding operations such as reservation, addition, and deletion can be performed on the computing resources (network resources) of the content and / or services that the user will request (in the future), thereby completing the advance configuration of the network resources.

[0020] Exemplarily, the advance deployment of the network architecture may include the advance deployment of the PCRF.

[0021] Exemplarily, the current corresponding strategies may include: cache update strategy, slice resource management strategy, air interface time and frequency resource allocation strategy, core network computing resource allocation management strategy, service function chain resource management strategy, etc.

[0022] In any possible implementation of the first aspect, the monitoring network element, the selection network element, the deployment network element, and the execution network element are core network elements.

[0023] In a second aspect, a communication device is provided, which includes a unit for executing each step in the above first aspect or any possible implementation manner of the first aspect.

[0024] In a third aspect, a communication device is provided, comprising at least one processor and a memory, wherein the at least one processor is configured to execute the method in the above first aspect or any possible implementation of the first aspect.

[0025] In a fourth aspect, a communication device is provided, which includes at least one processor and an interface circuit, and the at least one processor is used to execute the method in the above first aspect or any possible implementation of the first aspect.

[0026] In a fifth aspect, a core network device is provided, which includes any communication device provided in the second aspect, third aspect or fourth aspect.

[0027] In a sixth aspect, a computer program product is provided, which includes a computer program. When the computer program is executed by a processor, it is used to execute the method in the first aspect or any possible implementation of the first aspect.

[0028] In the seventh aspect, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed, it is used to execute the method in the first aspect or any possible implementation of the first aspect.

[0029] In an eighth aspect, a chip is provided, comprising: a processor for calling and running a computer program from a memory, so that a communication device equipped with the chip executes the method in the first aspect or any possible implementation of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] FIG1 is a schematic diagram of an example of a communication system architecture provided in this application.

[0031] FIG2 is a schematic diagram of a network architecture provided in the present application for implementing a communication control method based on user intention prediction provided in the present application.

[0032] FIG3 is a schematic flowchart of an example of a communication control method based on user intention prediction provided in this application.

[0033] FIG4 is a schematic flowchart of another example of a communication control method based on user intention prediction provided in the present application.

[0034] FIG5 is a schematic flowchart of an example of an algorithm provided by the present application for fusing user intention data with historical data of content and / or services requested by the user and processing the fused data.

[0035] FIG6 is a schematic flowchart of another example of a communication control method based on user intention prediction provided in the present application.

[0036] FIG7 is a schematic flowchart of another example provided by the present application of an algorithm for fusing user intention data with historical data of content and / or services requested by the user and processing the fused data.

[0037] FIG8 is a schematic block diagram of a communication device according to an embodiment of the present application.

[0038] FIG9 is a schematic block diagram of a communication device according to another embodiment of the present application.

[0039] FIG10 is a schematic block diagram of a communication device according to an embodiment of the present application. DETAILED DESCRIPTION

[0040] The technical solution in this application will be described below with reference to the accompanying drawings.

[0041] In the description of the embodiments of this application, unless otherwise specified, " / " represents or. For example, A / B can represent A or B. "And / or" in this article is simply a description of the association relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of this application, "plurality" means two or more than two.

[0042] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this embodiment, unless otherwise specified, "plurality" means two or more.

[0043] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: Long Term Evolution (LTE) system, LTE Frequency Division Duplex (FDD) system, LTE Time Division Duplex (TDD) system, fifth generation (5G) system, new radio (NR) or sixth generation (6G) communication systems evolved after 5G, etc.

[0044] The terminal device mentioned in the embodiments of the present application may be a device with wireless transceiver functions, specifically user equipment (UE), access terminal, subscriber unit, user station, mobile station, customer-premises equipment (CPE), remote station, remote terminal, mobile device, user terminal, wireless communication equipment, user agent or user device. The terminal device may also be a satellite phone, a cellular phone, a smart phone, a wireless data card, a wireless modem, a machine type communication device, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a communication device carried on a high-altitude aircraft, a wearable device, a drone, a robot, a smart point of sale (POS) machine, a terminal in device-to-device communication (D2D), a terminal in vehicle to everything (V2X), a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical care, a wireless terminal in a smart grid, a wireless terminal in transportation security (TSV), ... self-driving, a wireless terminal in a smart grid, a wireless terminal in transportation security (TSV), a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in self-driving, a wireless terminal in a smart grid, a wireless terminal in transportation security (TSV), a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in self-driving, a wireless terminal in a smart grid, a wireless terminal This application does not limit the wireless terminals in smart cities, smart homes, or terminal devices in future communication networks.

[0045] In the embodiments of the present application, the device for implementing the function of the terminal device can be the terminal device; it can also be a device that can support the terminal device to implement the function, such as a chip system. The device can be installed in the terminal device or used in conjunction with the terminal device. In the embodiments of the present application, the chip system can be composed of a chip or include a chip and other discrete devices.

[0046] The network device mentioned in the embodiments of the present application is a device with wireless transceiver functions, which is used to communicate with a terminal device, and can also be a device that connects a terminal device to a wireless network. The network device can be a node in a radio access network, which can also be called a base station, or a radio access network (RAN) node (or device). The network device can be a network composed of multiple 5G-AN / 5G-RAN nodes. For example, the 5G-AN / 5G-RAN node can be: an access point (AP), a base station (BS), a next-generation base station (NR nodeB, gNB), a gNB with a central unit (CU) and a distributed unit (DU) separated, a transmission receive point (TRP), a transmission point (TP) or some other access node. It can also be a base station (NodeB, NB) in a Wideband Code Division Multiple Access (WCDMA) system, an evolved base station (Evolutional NodeB, eNB or eNodeB) in an LTE system, or a wireless controller in a Cloud Radio Access Network (CRAN) scenario, or the access network device can be a relay station, an access point (wireless fidelity access point, WiFi AP) in wireless fidelity technology, a global microwave access point (worldwide interoperability for microwave access, WiMAX), and a network device in a 5G network or an access network device in a future evolved public land mobile network (public land mobile network, PLMN), etc., and the embodiments of the present application are not limited.

[0047] Optionally, the network equipment in the embodiments of the present application may also include various forms of base stations, such as: macro base stations, micro base stations (also called small stations), relay stations, equipment that implements base station functions in communication systems evolved after 5G, transmission points (transmitting and receiving points, TRP), transmitting points (transmitting points, TP), mobile switching centers, and equipment that assumes base station functions in device-to-device (D2D), vehicle-to-everything (V2X), and machine-to-machine (M2M) communications, etc., and may also include centralized units (CU) and distributed units (DU) in cloud radio access network (C-RAN) systems, and network equipment in non-terrestrial network (NTN) communication systems, that is, they can be deployed on high-altitude platforms or satellites. The embodiments of the present application do not specifically limit this.

[0048] In an embodiment of the present application, the network device can communicate and interact with the core network device to provide communication services to the terminal device. For example, the network device can implement the wireless link maintenance function, maintain the wireless link with the terminal device, and be responsible for the protocol conversion of wireless link data and IP data quality monitoring; in addition, the network device can also implement wireless resource management functions, including the establishment and release of wireless links, the scheduling and allocation of wireless resources, etc. Optionally, the network device can also implement some mobility management functions, including configuring the terminal device to perform cell switching measurements, evaluating the quality of the terminal device's wireless link, and making decisions on the handover of the terminal device between cells.

[0049] In the embodiments of the present application, the core network device is, for example, a device in the 5G core network (CN). As a bearer network, the core network provides an interface to the data network, providing communication connections, authentication, management, policy control, and data service carrying for terminal devices.

[0050] The embodiments of the present application do not specifically limit the specific structure of the execution subject of the method provided by the embodiments of the present application. As long as it is possible to communicate according to the method provided by the embodiments of the present application by running a program that records the code of the method provided by the embodiments of the present application, for example, the execution subject of the method provided by the embodiments of the present application may be a core network device, or a functional module in the core network device that can call and execute the program.

[0051] In addition, various aspects or features of the present application can be implemented as methods, devices or products using standard programming and / or engineering techniques. The term "product" as used in this application covers computer programs that can be accessed from any computer-readable device, carrier or medium. For example, computer-readable media may include, but are not limited to: magnetic storage devices (e.g., hard disks, floppy disks or magnetic tapes, etc.), optical disks (e.g., compact discs (CDs), digital versatile discs (DVDs), etc.), smart cards and flash memory devices (e.g., erasable programmable read-only memories (EPROMs), cards, sticks or key controllers, etc.). In addition, the various storage media described herein may represent one or more devices and / or other machine-readable media for storing information. The term "machine-readable medium" may include, but is not limited to, wireless channels and various other media capable of storing, containing and / or carrying instructions and / or data.

[0052] Over the past few decades, wireless communication systems have evolved and been researched, from first-generation analog systems to 5G NR systems and the future 6G systems. Throughout this complex evolution, high throughput and massive connections have always been core challenges for wireless communication networks. To address these challenges, 5G communication systems propose applications such as ultra-reliable, low latency communications (URLLC), enhanced mobile broadband (eMBB), and massive machine type communications (mMTC) as technical goals.

[0053] Furthermore, 5G communication systems, incorporating advancements in technologies like artificial intelligence and big data, are gradually enabling emerging services like autonomous driving, drones, and immersive extended reality. To better serve users and conserve computing and bandwidth resources, these services often require flexible, on-demand, and high-quality resource allocation. This has resulted in a surge in the amount of data that needs to be processed within the network. However, traditional statistical-based data processing and analysis methods are inefficient, necessitating an urgent need for efficient data mining technologies. At the same time, user requirements for network reliability, latency, speed, and other performance indicators are continuously increasing, requiring operators to adopt more intelligent and automated O&M approaches to optimize network management. Overall, future wireless communication networks will need to strengthen their integration with AI, leveraging AI's powerful predictive and decision-making capabilities to create user-centric, intelligent, and endogenous 6G networks.

[0054] Current research on endogenous intelligence focuses on the following areas: service-oriented perception and analysis, network-oriented analysis and optimization, and traffic-oriented predictive tuning. Taking network slicing as an example, traditional network slicing divides a fixed physical network into several logical networks, each corresponding to a different service or a number of similar services, enabling flexible handling of diverse application scenarios. Research on endogenous intelligence goes a step further, performing perception and analysis of service traffic, predicting services likely to be requested by users, and optimizing network resource allocation using methods such as deep reinforcement learning. While promising results have been achieved, the degree of network automation and intelligence remains to be strengthened, particularly in terms of optimizing the matching of service perception with corresponding policies.

[0055] To achieve more intelligent network deployment and resource allocation, the industry has proposed a number of new network architectures. Among them, intent-based networking (IBN) has garnered significant attention. In this architecture, network operations and maintenance no longer rely on specific execution instructions. Instead, network deployment and resource allocation are accomplished by analyzing user intent—that is, "what the user wants the network to achieve." This intent is then translated and executed. Subsequently, the network must monitor the overall network status to ensure the correct implementation of "user intent" and complete closed-loop network management. With the introduction of user intent, network autonomy is becoming increasingly possible through intent analysis, policy management, and resource optimization.

[0056] Early research on user intent analysis and translation relied on specific intent representation languages. While these languages ​​can accurately describe user intent, their intelligence is generally low. Therefore, in recent years, the use of natural language processing (NLP) for user intent analysis has gradually become a major research focus. For example, NLP models are used to extract keywords from input user intent text, then network knowledge bases are used to match keywords, and reinforcement learning is then used to optimize network resource allocation strategies and network deployment.

[0057] Using NLP models, user intent can be identified to a certain extent. However, in the current highly dynamic and time-varying network environment, this approach lacks in-depth analysis of existing data (i.e., the content and / or services historically requested by users), ignores the prediction of user intent, and cannot complete the advance deployment of the network and the advance configuration of network resources, thus affecting communication efficiency.

[0058] In view of this, the present application provides a communication control method based on user intention prediction, which obtains intention data corresponding to the content / service requested by the user through semantic analysis of the semantic data of the content and / or services historically requested by the user, and fuses the intention data with the content and / or services historically requested by the user (i.e., the historical data of the content and / or services requested by the user), and then uses a deep learning algorithm to analyze the data to obtain the user's intention prediction information. The user's intention prediction information is used to complete the advance deployment of the network architecture and the advance configuration of network resources, thereby realizing the correct, reasonable and timely allocation of network resources, ensuring the rationality and effectiveness of the network architecture deployment, and thus improving communication efficiency.

[0059] First, a communication system to which the embodiments of the present application are applicable is introduced.

[0060] Figure 1 is a schematic diagram of a communication system applicable to an embodiment of the present application. As shown in Figure 1, the communication system includes: a terminal device, a network device, and a core network device, wherein the terminal device and the network device can be any of the above-mentioned terminal devices and network devices.

[0061] For example, in some possible implementations, in the example shown in FIG1 , the core network device may include a management device and a gateway device.

[0062] In some possible implementations, the management device may include: unified data management (UDM), access and mobility management function (AMF), session management function (SMF), policy control function (PCF), application function (AF), etc.

[0063] In some possible implementations, the gateway device may include a user plane function (UPF). In some embodiments, the gateway device may also include functional units such as a branching point (BP) and an uplink classifier (UL CL). These functional units can operate independently or be combined to implement certain control functions. For example, the AMF is primarily responsible for mobility management in mobile networks, such as user location updates, user network registration, and user handover. The SMF is primarily responsible for session management in mobile networks, such as session establishment, modification, and release. Specific functions include assigning IP addresses to users and selecting the UPF that provides message forwarding. The PCF is responsible for providing policies to the AMF and SMF, such as Quality of Service (QoS) policies and slice selection policies. The UDM is used to store user data, such as subscription information and authentication / authorization information. The AF is responsible for providing services to the 3GPP network, such as influencing service routing and interacting with the PCF for policy control. The UPF is primarily responsible for processing user messages, such as forwarding and billing.

[0064] For example, AMF, SMF, and PCF can be combined as a management device to complete access control and mobility management functions such as access authentication, security encryption, and location registration for terminal devices, as well as session management functions such as establishing, releasing, and changing user plane transmission paths, and analyzing some slice-related data (such as congestion) and terminal device-related data. UPF, as a gateway device, mainly completes functions such as routing and forwarding of user plane data, such as filtering data packets for terminal devices, data transmission / forwarding, rate control, and generating billing information.

[0065] In the communication system shown in Figure 1, the core network device can utilize the communication control method based on user intent prediction provided by the embodiments of the present application to complete the pre-deployment of the network architecture (i.e., network devices and various network elements included in the core network device) and the pre-configuration of network resources (e.g., communication resources and computing resources of the network device). For example, operations such as retaining, adding, and deleting content, services, and computing resources in the network can be implemented.

[0066] It should be understood that the communication system shown in Figure 1 may also include more network nodes, such as terminal devices or network devices. The network devices or terminal devices included in the communication system shown in Figure 1 may be the various forms of network devices or terminal devices described above. The embodiments of the present application are not shown one by one in the figures.

[0067] The communication control method based on user intent prediction provided by this application is described in detail below in conjunction with Figures 2 and 3. Figure 2 shows a schematic diagram of a network architecture for implementing the communication control method based on user intent prediction provided by this application. Figure 3 shows a schematic flow chart of a communication control method 300 based on user intent prediction provided by an embodiment of this application in the network architecture shown in Figure 2.

[0068] It should be understood that in the embodiments of the present application, the execution subject of the communication control method based on user intent prediction may be a core network device. As an example and not a limitation, the execution subject of the method may also be a chip, chip system, or processor applied to the core network device. For example, the core network device may be a PCF, AMF, or SMF, etc., and the embodiments of the present application are not limited thereto.

[0069] As shown in Fig. 3 , the method 300 shown in Fig. 3 may include S310 to S350. Each step in the method 300 will be described in detail below with reference to Fig. 3 .

[0070] S310: Acquire historical data of content and / or services requested by the user, and semantic data corresponding to the content and / or services requested by the user in the past.

[0071] As a possible implementation method, as shown in Figure 2, in an embodiment of the present application, the historical data of the content and / or services requested by the user (or it can also be called: the record of the content and / or services historically requested by the user) and the semantic data corresponding to the content and / or services historically requested by the user can be used as input data. The core network device can analyze and process the input data to obtain the user's intention prediction information.

[0072] It should be understood that in the embodiments of the present application, the historical data of content and / or services requested by a user may include historical data of content and / or services requested by one or more users, and the semantic data corresponding to the content and / or services historically requested by a user may also include semantic data corresponding to the content and / or services historically requested by one or more users. The embodiments of the present application are not limited thereto.

[0073] As a possible implementation, as shown in FIG2 , the core network device may obtain historical data of content and / or services requested by the user, as well as semantic data corresponding to the content and / or services historically requested by the user, from the server.

[0074] In an embodiment of the present application, the data (records) of the content and / or services historically requested by a user (also referred to as a terminal device) can be understood as the historical data of the content and / or services requested by the user. For example, the data of the content and / or services requested by the user within a preset time length (e.g., the previous week, the previous month, etc.) can be recorded as the historical data of the content and / or services requested by the user. For example, the historical data of the content and / or services requested by the user may include: the user ID of the user who initiated the request within the preset time length, the identification ID of the requested object within the preset time length, the timestamp information of the request occurring within the preset time length, etc. The embodiment of the present application is not limited here.

[0075] In the embodiments of the present application, the "content and / or service" requested by a user in the past may also be referred to as the "object" requested by the user in the past, and the semantic data corresponding to the content and / or service requested by the user in the past may also be referred to as the semantic data corresponding to the object requested by the user in the past. For example, one piece of semantic data may be associated with one piece of content or service requested by the user, and the content and / or service requested by the user in the past may include multiple pieces of content and / or services.

[0076] For example, the semantic data corresponding to the content and / or services requested by a user may include keywords for the requested content and / or services, such as the genre of the requested content (e.g., a movie), and the purpose of using the smart service. This keyword information can indirectly reflect the user's preferences and purpose. One piece of semantic data may include multiple keywords.

[0077] Optionally, as a possible implementation method, the core network device can actively query or return the user request to the content and / or service provider, and the content and / or service provider will provide the semantic data corresponding to the content and / or service historically requested by the user.

[0078] Optionally, as another possible implementation method, the core network device can analyze and process the content and / or services in the historical data of the content and / or services requested by the user, for example, using an NLP model, etc., to obtain semantic data corresponding to the content and / or services historically requested by the user.

[0079] S320: Perform semantic analysis on semantic data corresponding to the content and / or services requested by the user in the past to obtain user intention data.

[0080] In some possible implementations, during the semantic analysis of semantic data, as shown in Figure 2, semantic encoding can be performed on multiple semantic data of the user to obtain semantic feature vectors. After semantic encoding, semantic aggregation is performed on the obtained multiple semantic feature vectors to obtain user intent data.

[0081] For example, for any semantic data, an NLP training model can be used to encode each keyword in the semantic data to obtain a vector representation of semantic features such as keyword semantic similarity and analogy (i.e., a semantic feature vector), thereby achieving semantic encoding. For multiple keyword information (i.e., multiple semantic feature vectors) contained in a content and / or service, aggregation operations (aggregation of multiple semantic feature vectors) can be performed using addition, averaging, and deep learning models based on recurrent neural networks, attention mechanisms, etc. to obtain user intent data (expressed in the form of a vector) for the requested content and / or service. User intent data can reflect the user's intention and can serve as input for subsequent prediction tasks.

[0082] In the embodiments of this application, user intent data can be understood as data that reflects the user's wishes or preferences and can explain why the user initiated a request for the content and / or service. For example, this includes semantic feature data such as the style and introduction information of a piece of content, or the effects and expectations of a service. These features can reflect the reasons for the user's request and can be considered a type of intent data.

[0083] S330: Fusing the user intention data with the historical data of the content and / or services requested by the user, and using the fused data as input to the deep learning network model.

[0084] In an embodiment of the present application, as a possible implementation method, after obtaining the user intention data, the user intention data and the historical data of the content and / or services requested by the user can be fused by addition or splicing, and used as the input of a subsequent prediction model (deep learning network model). The output of the deep learning network model is the user's intention prediction information.

[0085] As another possible implementation method, after obtaining the user intent data, the historical data of the content and / or service requested by the user can also be preprocessed, the preprocessed historical data and the user intent data can be fused, and the fused data can be used as the input of the deep learning network model.

[0086] In embodiments of the present application, preprocessing historical data (also referred to as historical data) of user-requested content and / or services can enhance the deep learning network model's perception of this historical data. Exemplarily, preprocessing historical data of user-requested content and / or services can include aggregating historical records of content and / or services requested by users over a short period of time, thereby obtaining short-term user preferences. For content and / or services requested by users over a short period of time, the preprocessing process can obtain user demographic information that favors the service / content. Preprocessing of historical data can be achieved by averaging the records of all content and / or services requested by users or by using deep learning models such as recurrent neural networks and attention mechanisms. Alternatively, the "age of historical data" can be used as an indicator to filter out time-sensitive historical data, and the filtered historical data can be fused with user intent data. The age of historical data can be understood as the time when the user requested content and / or services, or alternatively, the time difference between a user's previous request for content and / or services and the user's most recent request. The embodiments of the present application do not limit the specific preprocessing process.

[0087] S340 , using a deep learning network model, processing historical data of user-requested content and / or services integrated with user intent data to obtain user intent prediction information.

[0088] For example, the user's intention prediction information may include: the probability of a certain user making a request for certain content and / or service at a certain time in the future; or the probability of each user among multiple users making a request for certain content and / or service at a certain time in the future, etc.

[0089] In some possible implementations, since the interaction process of historical data (i.e., historical data of content and / or services requested by users) can be abstracted as a dynamically evolving bipartite graph, the bipartite graph contains two types of nodes, namely users and objects requested by users (content and / or services). The edges of the bipartite graph are requests initiated by users for certain content and / or services, and the interaction process has a certain degree of dynamics, and the timestamp information of the interaction will be recorded. For this dynamic bipartite graph, a dynamic graph neural network model can be used to complete the task of analyzing and predicting user intentions. Taking the historical data of content and / or services requested by users that are integrated with user intention data as input, a specific vector representation containing structural features, temporal features, and semantic features is generated for each node, and then the matching value between the user and a certain requested content and / or service is calculated based on multi-layer perception, and the training of the dynamic graph neural network model is completed using the current network status as label information. After training is completed, the dynamic graph neural network model is used to predict the user's future intentions using the historical data of the user's recent requests for content and / or services and the corresponding user intention data as input. The prediction results of the user's future intentions can be used as a basis for determining whether the user will initiate a request for the content / service.

[0090] S350 utilizes user intention prediction information to achieve early deployment of network architecture and early configuration of network resources.

[0091] In an embodiment of the present application, the user's intention prediction information can be used as a basis for judging whether the user will initiate a request for a certain content and / or service in the future. After obtaining the user's intention prediction information (that is, the user will make a request for a certain content and / or service at a certain time in the future), the core network device issues instructions in combination with the current network status to complete the advance deployment of the network architecture (for example, the deployment of the policy and charging rules function unit (PCRF)), the advance deployment, optimization, issuance and execution of corresponding policies (for example: cache update policy, slice resource management policy, air interface time and frequency resource allocation policy, core network computing resource allocation management, service function chain resource management policy, etc.), and real-time monitoring and feedback of the network operation status to the upper structure to complete the closed-loop management of the network, ensuring that the computing resources of the content and / or service are correctly, reasonably and timely allocated through operations such as retention, addition and deletion to meet the user's future intentions, thereby better realizing the user's intentions and improving the user experience.

[0092] In some possible implementations, as shown in FIG2 , at the general computing control layer, based on user intent prediction information, monitoring network elements, selection network elements, deployment network elements, and execution network elements can be used to implement advance deployment of the network architecture (i.e., advance deployment of the infrastructure layer) and advance configuration of network resources. For example, based on the user intent prediction information, communication resources for the predicted user-requested content and / or services can be selected, deployed, executed, and monitored. In some possible implementations, the monitoring network element, selection network element, deployment network element, and execution network element can all be network elements included in the core network.

[0093] The monitoring network element is primarily responsible for monitoring and providing feedback on network status. For example, it can periodically analyze the accuracy of user intent predictions and the performance of the current deep learning network model to determine whether the deep learning network model needs to be retrained. If the user intent prediction result falls below a preset threshold, a command is sent to the model training unit to retrain the deep learning network model.

[0094] In some implementations, the monitoring network element can also combine the output of the deep learning network model (i.e., the user's intent prediction information) to determine whether existing policies (e.g., including: cache update policy, slice resource management, air interface time and frequency resource allocation policy, core network computing resource allocation management, service function chain resource management policy, etc.) need to be updated and optimized. If the current policy is difficult to meet the requirements (for example, by comparing the current network status with the specific content and / or service threshold standards included in the intent prediction information, such as whether the cache hit rate of the content and / or service cached by the current cache policy is higher than a certain threshold, etc., to determine whether the current policy meets the requirements), it sends an indication to the selected network element, and the selected network element optimizes the policy update based on the indication. After the selected network element completes the update, deployment, and execution of the policy and other content, the monitoring network element can also continuously perceive the network status to ensure the correct execution of the updated policy. If it is found that the network status is difficult to meet the predicted user intent, it sends an indication to the selected network element. After receiving the indication, the selected network element updates the policy. If the policy still fails to meet the requirements after multiple updates, it determines that the deep learning network model needs to be retrained. At the same time, the monitoring network element can also continuously record the requested content and / or service data, as well as the semantic data corresponding to the requested content and / or service, network status information, etc., and periodically send them to the server for storage after completing certain privacy protection operations.

[0095] The selection network element is mainly used to update and optimize the strategy based on the judgment results of the monitoring network element on the strategy and the user's intention prediction information, and select appropriate optimization algorithms, such as reinforcement learning and convex optimization algorithms.

[0096] The deployment network element is mainly used to combine the configuration instruction specifications of network parameters, convert the corresponding policies into control instructions for the wireless network infrastructure, and send them to the corresponding physical devices (such as base stations, etc.) to complete the early deployment of the network architecture.

[0097] The execution network element is mainly used to perform corresponding operations such as reservation, addition and deletion on the computing resources (network resources) of the content and / or services that the user will request (in the future), thereby completing the advance configuration of network resources.

[0098] The communication control method based on user intention prediction provided by the present application records the content and / or service data of the user's historical requests, as well as the semantic data of the requested object that can reflect the user's intention. Natural language processing is used to complete the encoding and aggregation of semantic data to obtain user intention data. The content and / or service data of the user's historical requests are integrated with the user intention data. Based on the historical data integrated with the user intention data, a deep learning model is used to complete the task of predicting the user's next intention. Based on the prediction results, the advance deployment of the network architecture and the advance configuration of network resources are completed to ensure the correct, reasonable and timely allocation of network resources, thereby ensuring the rationality and effectiveness of the network architecture deployment, thereby improving communication efficiency.

[0099] The communication control method based on user intention prediction provided by this application is explained below with reference to specific examples.

[0100] FIG4 is a schematic flow chart of another example of a communication control method 400 based on user intention prediction provided by an embodiment of the present application. As shown in FIG4 , the method 400 includes:

[0101] S401: Record historical data of content and / or services requested by a user, and semantic data corresponding to the content and / or services requested by the user in the past, in a server.

[0102] The monitoring network elements in the core network equipment can be used to record historical data on the content and / or services requested by users, as well as semantic data that can reflect user intentions, in the server for subsequent tasks such as model training.

[0103] For example, the historical data of the content and / or service requested by the user mainly includes: user ID, identification ID of the requested content and / or service, timestamp information of the request, etc.

[0104] The semantic data corresponding to the content and / or services requested by the user in the past may include: keywords for the requested content and / or services, such as the style of the content (music, video, news, etc.), the purpose of using smart services (catering, entertainment, medical care, etc.), etc. These keyword information can reflect the user's preferences, purposes, etc. from the side, so it can also be used as an input for analyzing user intentions. Different from the general intent network that passively receives and processes user input statements to obtain intentions, this application perceives the user's intentions by analyzing the semantic data of the user's requested object (content and / or service), which is an active intent analysis method. In addition, in an embodiment of the present application, the current network status can also be recorded as a label for deep learning network model training. The monitoring network element takes the above information as input and stores it in a specific storage server.

[0105] S402: Perform semantic analysis on semantic data corresponding to the content and / or services requested by the user in the past to obtain user intention data.

[0106] In some possible implementations, the semantic analysis module in the core network device can encode the recorded semantic data of the requested content and / or service during the process of performing semantic analysis on the semantic data, generate a vector containing semantic features, and aggregate the multi-layer semantic information contained therein to obtain user intent data. The user intent data can reflect the intention of the user's historical requests, that is, reflect the user's historical intention.

[0107] In some possible implementations, in order to implement the encoding operation of semantic data, some pre-trained natural language models such as ELMo, BERT, Glove, etc. can be used. The word embedding expressions obtained using these pre-trained models generally have good semantic features, such as semantic similarity and analogy between words. They can add a layer of implicit semantic connection between the user's request data, making it easier to analyze, perceive and predict the user's intentions. In addition, since the requested objects (content and / or services), such as movies and other content or various communication services, will contain multiple keyword information, in order to facilitate subsequent processing, these keyword information need to be aggregated. The aggregation operation of semantic data includes but is not limited to summing and averaging each semantic data, using deep learning models such as recurrent neural networks, attention mechanisms and their various variant models. After completing the aggregation operation, the semantic representation of the requested object (i.e., user intention data) will be obtained.

[0108] S403: The data fusion module in the core network device fuses the user intention data with the historical data of the content and / or service requested by the user, and uses the fused data as the input of the deep learning network model.

[0109] As a possible implementation method, in the process of fusing user intention data with historical data of content and / or services requested by the user, a direct fusion method can be used to fuse the user intention data reflecting the user intention with the feature vectors corresponding to the content and / or services requested by the user by splicing or adding them, and generate a specific vector representation containing semantic features for each content and / or service as input to the subsequent prediction model.

[0110] S404: The perception prediction model in the core network device performs perception prediction of the user intention based on the fused data and a deep learning network model.

[0111] In S404, the perception prediction model in the core network device uses a deep learning network model to process the historical data of the content and / or services requested by the user integrated with the user intention data to obtain the user intention prediction information (i.e., the prediction result).

[0112] In some possible implementations, the specific algorithm process of S403 and S404 is shown in FIG5 .

[0113] As shown in Figure 5, after semantic encoding of multiple keywords to obtain multiple pieces of semantic data, each piece of semantic data can be semantically aggregated using methods such as summation, averaging, or deep learning models. Direct fusion is then used to concatenate or add the aggregated data to the vector representation of the requested object (content and / or service) (i.e., historical data of the content and / or services requested by the user) for fusion.

[0114] Historical data integrated with semantic data is filtered based on its "information age" and aggregated using neural network techniques such as averaging or attention mechanisms to obtain a user's short-term interest representation. Using recurrent neural networks or their variants, such as gated recurrent units (GRUs) and long short-term memory (LSTMs), the user's short-term interest representation is used as input to update the user's long-term general interest representation. Repeating the above steps yields a vector representation of the user group representation of the requested object.

[0115] Taking the user's long-term general interest representation and the user group representation of the requested object as input, the interaction between the user's long-term general interest representation and the user group representation of the requested object is abstracted into a dynamically evolving bipartite graph, and the timestamp information of the corresponding content and / or service is spliced. The dynamic graph attention mechanism model (temporal graph attention, TGAT) is used to analyze the temporal and structural features of the interaction, and vector representations of the user and the requested object containing temporal features, structural features and semantic features are obtained; based on multi-layer perception, the matching value between the user and a certain requested content and / or service is calculated, and the current network state is used as label information to complete the training of the dynamic graph neural network model.

[0116] After training is completed, the dynamic graph neural network model is used to predict the user's future intentions using the historical data of the user's recent requests for content and / or services and the corresponding user intention data as input. The prediction results of the user's future intentions can be used as a basis for determining whether the user will initiate a request for the content / service.

[0117] After S404, the core network equipment performs general computing control based on the prediction results. This step is mainly performed in the general computing control layer. The four types of network elements in the general computing control layer can complete the deployment, optimization, issuance and execution of corresponding strategies in advance based on network status, prediction results and other information, and monitor and feedback the network operation status in real time to complete the closed-loop management of the network, ensure the correct, reasonable and timely allocation of general computing resources, and ensure that the corresponding content and / or services can meet the user's future intentions.

[0118] As shown in the process from S405 to S413 in Figure 4, the monitoring network element perceives that the accuracy of the deep learning network model is low, and after retraining, it is found that the strategy does not meet the requirements. After completing the selection deployment and execution, it is again perceived that the strategy does not meet the requirements, and the deep learning network model needs to be retrained and the strategy needs to be optimized multiple times.

[0119] S405, the monitoring network element periodically checks the accuracy of the user's intention prediction information. If the accuracy of the user's intention prediction (i.e., the prediction result) is lower than the preset threshold, it is determined that the deep learning network model needs to be retrained. In this case, the monitoring network element can send instructions to the perception prediction module to retrain the deep learning network model. If the retrained deep learning network model still does not meet the requirements, repeat this step.

[0120] S406: The monitoring network element analyzes the prediction results based on the updated deep learning network model and the network status to determine whether the current policy (e.g., cache update policy, slice resource management policy, air interface time and frequency resource allocation policy, core network computing resource allocation management, service function chain resource management policy, etc.) meets the requirements and whether policy optimization is needed. If it is found that the current policy does not meet the requirements, it is determined that the current policy still needs further optimization.

[0121] S407: The monitoring network element sends an instruction requiring policy optimization to the selected network element.

[0122] S408, selecting a network element based on the user's intention prediction information and selecting a suitable algorithm for strategy optimization, such as deep reinforcement learning, convex optimization algorithm, etc.

[0123] S409: After selecting a network element to complete the optimization of the policy, the updated policy is sent to the deployment network element.

[0124] S410, deploy network elements based on the updated policy, combined with the network parameter configuration instruction specifications, to generate control instructions for configuring the network infrastructure (such as base stations, etc.). For example, the control instructions may include: converting the updated policy into specific instructions, conflict resolution management instructions between the formulated policy and the current network management configuration, etc.

[0125] S411, the deployment network element sends the control instruction to the execution network element.

[0126] S412: The execution network element receives the control instruction and implements the specific execution of the control instruction, such as completing operations such as reservation, addition, and deletion of computing resources for content and / or services.

[0127] S413, the monitoring network element continues to perceive the current network status, and if it finds that the prediction accuracy of the deep learning network model is still insufficient and the strategy is still insufficient, then steps S406 to S412 are repeated.

[0128] S414 , the monitoring network element continuously records historical data of the content and / or service requested by the user, as well as semantic data corresponding to the content and / or service requested by the user in history, and periodically transmits the data back to the server for storage.

[0129] The communication control method based on user intention prediction provided by the present application records the content and / or service data of the user's historical requests, as well as the semantic data of the requested object that can reflect the user's intention. Natural language processing is used to complete the encoding and aggregation of semantic data to obtain user intention data. The content and / or service data of the user's historical requests are integrated with the user intention data. Based on the historical data integrated with the user intention data, a deep learning model is used to complete the task of predicting the user's next intention. Based on the prediction results, the monitoring network elements, the selection network elements, the deployment network elements and the execution network elements are combined with the current network status to issue instructions to complete the advance deployment of the network architecture, the deployment, optimization, issuance and execution of corresponding strategies (such as cache update strategies, slice resource management strategies, etc.), and real-time monitoring and feedback of the network operation status to the upper structure to complete the closed-loop management of the network, ensuring that the computing resources of the content and / or services are correctly, reasonably and timely allocated through operations such as retention, addition and deletion to meet the user's future intentions, thereby improving communication efficiency.

[0130] FIG6 is a schematic flow chart of another example of a communication control method 600 based on user intention prediction provided by an embodiment of the present application. As shown in FIG6 , the method 600 includes:

[0131] S601: Record historical data of content and / or services requested by a user, and semantic data corresponding to the content and / or services requested by the user in the past, in a server.

[0132] S602: Perform semantic analysis on semantic data corresponding to the content and / or services requested by the user in the past to obtain user intention data.

[0133] The specific description of S601 and S602 can refer to the corresponding specific description of S401 and S402 in method 400, and for the sake of brevity, they are not repeated here.

[0134] S603: The data fusion module in the core network device pre-processes the historical data of the content and / or service requested by the user.

[0135] In an embodiment of the present application, the preprocessing of historical data of content and / or services requested by users (also referred to as historical data) can enhance the perception of the deep learning network model on the historical data. Exemplarily, the preprocessing of historical data of content and / or services requested by users can include: aggregating historical records of content and / or services requested by users in a short period of time, so as to obtain the user's short-term preference expression. As for the content and / or services requested by users in a short period of time, the preprocessing process can obtain characteristic information of the user group that prefers the service / content. For the preprocessing of historical data, it can be achieved by averaging the records of all content and / or services requested by users, using recurrent neural networks, or deep learning models such as attention mechanisms. Alternatively, it is also possible to use the "age of historical data" as an indicator to filter out time-sensitive historical data, and fuse the filtered historical data with user intention data. The embodiment of the present application does not limit the specific process of preprocessing.

[0136] S604: The data fusion module in the core network device fuses the pre-processed historical data with the user intention data, and uses the fused data as input to the deep learning network model.

[0137] For example, user intention data reflecting user intention can be fused with user group feature information in preprocessed historical data by splicing or adding them together to generate a specific vector representation containing semantic features for each content and / or service as input to the subsequent prediction model.

[0138] S605: The perception prediction model in the core network device performs perception prediction of the user intention based on the fused data and a deep learning network model.

[0139] In some possible implementations, the specific algorithm process from S603 to S605 is shown in FIG7 .

[0140] As shown in FIG7 , after completing the semantic encoding of multiple keywords to obtain multiple semantic data, each piece of semantic data is semantically aggregated by summing, averaging, and using a deep learning model.

[0141] Historical data is filtered based on "information age" and pre-processed by aggregation operations using neural network methods such as averaging or attention mechanisms to obtain the user's short-term interest representation; a recurrent neural network or its various variants, such as GRU and LSTM, is used to update the user's long-term general interest representation with the user's short-term interest representation as input; the above steps are repeated to obtain the general user group representation of the requested object.

[0142] Afterwards, a direct fusion method is used to add the semantic data to the user group representation used to represent the requested object by splicing or adding them to perform a fusion operation.

[0143] Taking the user's long-term general interest representation and the user group representation of the requested object as input, the interaction between the two is abstracted into a dynamically evolving bipartite graph. TGAT is used to analyze the temporal and structural characteristics of the interaction, resulting in vector representations of the user and the requested object that include temporal, structural, and semantic features. Multi-layer perception is used to calculate the matching value between the user and the requested content and / or service, and the current network state is used as label information to complete the training of the dynamic graph neural network model.

[0144] After training is completed, the dynamic graph neural network model is used to predict the user's future intentions using the historical data of the user's recent requests for content and / or services and the corresponding user intention data as input. The prediction results of the user's future intentions can be used as a basis for determining whether the user will initiate a request for the content / service.

[0145] S606: The monitoring network element periodically checks the accuracy of the user's intention prediction information. If the accuracy of the user's intention prediction (i.e., the prediction result) is greater than or equal to a preset threshold, the monitoring network element determines that the accuracy of the deep learning network model meets the requirements. In this case, the monitoring network element can determine that retraining of the deep learning network model is not necessary.

[0146] S607 , the monitoring network element analyzes the prediction results based on the deep learning network model and the network status to determine whether the current strategy can meet the requirements and whether strategy optimization is needed. If the current strategy is found to meet the requirements, it is determined that the current strategy does not need to be updated.

[0147] S608: The monitoring network element sends the current policy to the deployment network element.

[0148] S609 , the deployment network element generates a control instruction for configuring the network infrastructure (eg, base station, etc.) based on the received current policy and in combination with the network parameter configuration instruction specification.

[0149] S610: The deployment network element sends a control instruction to the execution network element.

[0150] S611: The execution network element receives the control instruction and implements the specific execution of the control instruction, such as completing operations such as reservation, addition, and deletion of computing resources for content and / or services.

[0151] S612, the monitoring network element continuously senses the current network status and determines that the prediction accuracy of the deep learning network model can meet the requirements and the strategy also meets the requirements.

[0152] S613 , the monitoring network element continuously records historical data of the content and / or service requested by the user, as well as semantic data corresponding to the content and / or service requested by the user in history, and periodically transmits the data back to the server for storage.

[0153] The communication control method based on user intention prediction provided by the present application records the content and / or service data of the user's historical requests, as well as the semantic data of the requested object that can reflect the user's intention. Natural language processing is used to complete the encoding and aggregation of semantic data to obtain user intention data. The content and / or service data of the user's historical requests are preprocessed, and the preprocessed data is integrated with the user intention data. Based on the historical data integrated with the user intention data, a deep learning model is used to complete the task of predicting the user's next intention. Based on the prediction results, the network elements are monitored, the network elements are selected, the network elements are deployed, and the network elements are executed to complete the advance deployment of the network architecture and the advance configuration of network resources. The network operation status is monitored and fed back to the upper structure in real time to complete the closed-loop management of the network and ensure the correct, reasonable and timely allocation of network resources, thereby improving communication efficiency.

[0154] It should be understood that the above is only intended to help those skilled in the art better understand the embodiments of the present application, and is not intended to limit the scope of the embodiments of the present application. Based on the above examples given, those skilled in the art can obviously make various equivalent modifications or changes. For example, certain steps in each of the above methods may be unnecessary, or certain new steps may be added. Or a combination of any two or any multiple of the above embodiments. Such modifications, changes, or combined solutions also fall within the scope of the embodiments of the present application.

[0155] It should also be understood that the division of the modes, situations, categories and embodiments in the embodiments of the present application is only for the convenience of description and should not constitute a special limitation. The features of various modes, categories, situations and embodiments can be combined without contradiction.

[0156] It should also be understood that the various numerical numbers involved in the embodiments of this application are only for the convenience of description and are not intended to limit the scope of the embodiments of this application. The order of the sequence numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0157] It should also be understood that the above description of the embodiments of the present application focuses on emphasizing the differences between the various embodiments. The same or similar points that are not mentioned can be referenced with each other. For the sake of brevity, they will not be repeated here.

[0158] In this embodiment, the core network equipment (including the aforementioned monitoring network element, selection network element, deployment network element, and execution network element) can be divided into functional modules according to the above method. For example, the modules can be divided into functional modules corresponding to each function, or two or more functions can be integrated into a single processing module. The integrated modules can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and is merely a logical functional division. In actual implementation, other division methods may be used.

[0159] It should be noted that the relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module and will not be repeated here.

[0160] The core network device provided in the embodiment of the present application is used to execute any one of the communication control methods based on user intention prediction provided in the above method embodiment, so that the same effect as the above implementation method can be achieved. In the case of an integrated unit, the core network device (or it can also be called a core network network element) may include a processing module, a storage module and a communication module. Among them, the processing module can be used to control and manage the actions of the network element. For example, it can be used to support the network element to execute the steps performed by the processing unit. The storage module can be used to support the storage of program code and data, etc. The communication module can be used to support communication between the network element and other devices.

[0161] The processing module may be a processor or a controller. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, and so on. The storage module may be a memory. The communication module may specifically be a device that interacts with the device, such as a radio frequency circuit, a Bluetooth chip, or a Wi-Fi chip.

[0162] Figure 8 shows a schematic block diagram of an example communication device provided in an embodiment of the present application. This communication device 800 may correspond to the core network device described in the above method embodiment. It may also be a chip or component applied to the core network device. Furthermore, each module or unit in this communication device 800 is respectively configured to execute each action or process performed by the core network device described in the above method embodiment. As shown in Figure 8 , this communication device 800 may include a processing unit 810 and a communication unit 820. Optionally, this communication device 800 may also include a storage unit 830.

[0163] It should be understood that the specific process of each unit in the communication device 800 executing the above-mentioned corresponding steps can be referred to the relevant description of the core network equipment, monitoring network element, selecting network element, or deploying network element execution steps in the various embodiments shown in Figures 3 to 7 in the previous text. For the sake of brevity, they are not repeated here.

[0164] Optionally, the communication unit 820 may include a receiving unit (module) and a sending unit (module) for performing the steps of receiving information and sending information by the PSA, I-UPF, access network device, new I-SMF, or SMF in the aforementioned method embodiments. The storage unit 830 is used to store instructions executed by the processing unit 810 and the communication unit 820. The processing unit 810, the communication unit 820, and the storage unit 830 are communicatively connected, the storage unit 830 stores instructions, the processing unit 810 is used to execute the instructions stored in the storage unit, and the communication unit 820 is used to perform specific signal transmission and reception under the drive of the processing unit 810.

[0165] It should be understood that the communication unit 820 may be a transceiver, an input / output interface or an interface circuit, etc., the storage unit may be a memory, and the processing unit 810 may be implemented by a processor.

[0166] Figure 9 is a schematic block diagram of another example communication device provided herein. As shown in Figure 9 , communication device 900 may include a processor 910, a memory 920, a transceiver 930, and a bus system 940. The various components of communication device 900 are coupled together via bus system 940. In addition to a data bus, bus system 940 may also include a power bus, a control bus, and a status signal bus. However, for clarity, various buses are labeled as bus system 940 in Figure 9 . For ease of illustration, Figure 9 is only schematically drawn.

[0167] It should also be understood that the communication device 900 of Figure 9 can be a core network device, a monitoring network element, a selection network element, or a deployment network element, or the core network device, the monitoring network element, the selection network element, or the deployment network element can include the communication device 900 of Figure 9.

[0168] The communication device 800 of FIG. 8 or the communication device 900 of FIG. 9 can implement the steps performed by the core network device, monitoring network element, selection network element, or deployment network element in the embodiments shown in FIG. 3 to FIG. 7 . For similar descriptions, reference can be made to the descriptions of the corresponding methods described above. To avoid repetition, these steps will not be repeated here.

[0169] An embodiment of the present application also provides a chip system, as shown in Figure 10, which includes at least one processor 1010 and at least one interface circuit 1020. The processor 1010 and the interface circuit 1020 can be interconnected via lines. For example, the interface circuit 1020 can be used to receive signals from other devices. For another example, the interface circuit 1020 can be used to send signals to other devices. Exemplarily, the interface circuit 1020 can read instructions stored in the memory and send the instructions to the processor 1010. When the instructions are executed by the processor 1010, the chip system can execute the various steps executed by the core network equipment, monitoring network element, selection network element, or deployment network element in the above embodiments. Of course, the chip system can also include other discrete components, which are not specifically limited in the embodiments of the present application.

[0170] It should also be understood that the division of units in the above device is merely a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. Moreover, the units in the device can all be implemented in the form of software called through processing elements; or all be implemented in the form of hardware; or some units can be implemented in the form of software called through processing elements, and some units can be implemented in the form of hardware. For example, each unit can be a separately established processing element, or it can be integrated into a certain chip of the device. In addition, it can also be stored in a memory in the form of a program, and called by a certain processing element of the device to execute the function of the unit. Here, the processing element can also be called a processor, which can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each unit above can be implemented by the integrated logic circuit of the hardware in the processor element or in the form of software called through the processing element. In one example, the unit in any of the above devices may be one or more integrated circuits configured to implement the above method, such as one or more application specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms. For another example, when the unit in the device can be implemented in the form of a processing element scheduler, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call a program. For another example, these units can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0171] The present application also provides an apparatus, included in a core device, that implements the core network device, monitoring network element, selecting network element, or deploying network element functions described in any of the methods described in the above embodiments. This functionality can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes at least one module or unit corresponding to the above functionality. For example, a detection module or unit, a display module or unit, a determination module or unit, and a calculation module or unit.

[0172] An embodiment of the present application also provides a communication control system based on user intention prediction, which includes: the core network equipment, monitoring network elements, selection network elements, deployment network elements, etc. provided in the above method embodiments.

[0173] The present application also provides a computer-readable storage medium for storing computer program code, wherein the computer program includes instructions for executing any of the communication control methods based on user intent prediction provided in the embodiments of the present application. The computer-readable storage medium may be a read-only memory (ROM) or a random access memory (RAM), which is not limited in the present application.

[0174] The present application also provides a computer program product, which includes instructions. When the instructions are executed, the core network equipment, monitoring network elements, selection network elements, and deployment network elements perform corresponding operations corresponding to the above method.

[0175] An embodiment of the present application further provides a chip located in a communication device, comprising: a processing unit and a communication unit. The processing unit may be, for example, a processor, and the communication unit may be, for example, an input / output interface, a pin, or a circuit. The processing unit may execute computer instructions to cause the communication device to perform any of the communication control methods based on user intent prediction provided in the embodiments of the present application.

[0176] Optionally, the computer instructions are stored in a storage unit.

[0177] Optionally, the storage unit is a storage unit within the chip, such as a register, a cache, etc. The storage unit may also be a storage unit within the terminal that is located outside the chip, such as a ROM or other type of static storage device that can store static information and instructions, random access RAM, etc. The processor mentioned in any of the above may be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for executing a program for controlling the method of data packet transmission in the above-mentioned PDU session. The processing unit and the storage unit may be decoupled and respectively provided on different physical devices, and connected by wire or wireless means to implement the respective functions of the processing unit and the storage unit, so as to support the system chip to implement the various functions in the above-mentioned embodiments. Alternatively, the processing unit and the memory may also be coupled on the same device.

[0178] Among them, the communication control method based on user intention prediction, core network equipment, monitoring network element, selection network element, deployment network element, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0179] It will be appreciated that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a RAM, which is used as an external cache. There are many different types of RAM, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0180] Various objects such as various messages / information / equipment / network elements / systems / devices / actions / operations / processes / concepts that may appear in this application are named. It can be understood that these specific names do not constitute a limitation on the relevant objects. The names assigned may change with factors such as scenarios, contexts or usage habits. The understanding of the technical meaning of the technical terms in this application should be mainly determined from the functions and technical effects embodied / executed in the technical solutions.

[0181] In the various embodiments of the present application, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0182] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0183] The methods in the embodiments of the present application can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, they can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instruction is loaded and executed on a computer, the process or function described in the embodiments of the present application is executed in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer program or instruction can be stored in a computer-readable storage medium or transmitted via the computer-readable storage medium. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server that integrates one or more available media.

[0184] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0185] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0186] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0187] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0188] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a readable storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned readable storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

[0189] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A communication control method based on user intention prediction, characterized in that: The method comprises: Determining user intent data based on semantic data corresponding to content and / or services historically requested by the user, wherein the user intent data reflects the historical intent of the user; fusing historical data of content and / or services requested by the user with the user intent data; Processing the fused data using a deep learning network model to obtain user intention prediction information, the intention prediction information including: a probability of the user making a request for a certain content and / or service at a certain time in the future; The intention prediction information is used to complete the advance deployment of the network architecture and the advance configuration of network resources.

2. The method according to claim 1, characterized in that Determining user intent data based on semantic data corresponding to content and / or services historically requested by the user includes: Performing semantic encoding on the keywords included in the semantic data using a natural language processing model to obtain a plurality of semantic feature vectors, wherein the semantic feature vectors are used to represent the semantic similarity and analogy of the keywords; Semantic aggregation is performed on the multiple semantic feature vectors to obtain the user intention data.

3. The method according to claim 2, characterized in that Perform semantic aggregation on the multiple semantic feature vectors to obtain the user intention data using at least one of the following methods: Addition, averaging, deep learning models based on recurrent neural networks, or deep learning models based on attention mechanisms.

4. The method according to any one of claims 1 to 3, characterized in that The fusing of historical data of content and / or services requested by the user and the user intention data includes: The historical data of the content and / or service requested by the user and the user intention data are merged by adding or splicing.

5. The method according to any one of claims 1 to 3, characterized in that The fusing of historical data of content and / or services requested by the user and the user intention data includes: Preprocessing historical data of the content and / or service requested by the user; fusing the pre-processed historical data of the content and / or service requested by the user and the user intention data by adding or concatenating them; The preprocessing includes at least one of averaging the historical data of the content and / or service requested by the user, using a recurrent neural network or an attention mechanism deep learning model, or filtering the historical data of the content and / or service requested by the user using a time parameter.

6. The method according to any one of claims 1 to 5, characterized in that The method of processing the fused data using a deep learning network model to obtain user intention prediction information includes: Using the deep learning network model to obtain node information in historical data of the content and / or service requested by the user, the nodes include: the user, and the content and / or service requested by the user; Using the deep learning network model to generate a vector representation for each node that includes structural features, temporal features, and semantic features, wherein the temporal feature is the time when the user requested the content and / or service; Calculating a matching value between the user and content and / or service using the deep learning network model; The intention prediction information is obtained according to the matching value.

7. The method according to any one of claims 1 to 6, characterized in that The historical data of the content and / or services requested by the user includes: The user ID of the user who initiated the request within the preset time length, the identification ID of the requested content and / or service within the preset time length, and the time information of the request initiated by the user within the preset time length.

8. The method according to any one of claims 1 to 7, characterized in that The use of the intention prediction information to complete the advance deployment of the network architecture and the advance configuration of network resources includes: Based on the intention prediction information, using a monitoring network element to determine whether the intention prediction information meets a preset threshold, determine whether the current policy meets the requirements, record historical data of the content and / or services requested by the user, and semantic data corresponding to the content and / or services requested by the user in the past; Use selected network elements to update and optimize current strategies; Utilizing deployment network elements to convert current or updated policies into control instructions for wireless network infrastructure based on configuration instruction specifications for network parameters; The control instructions are executed by using the execution network element to complete the advance deployment of the network architecture and the advance configuration of the computing resources for the content and / or services requested by the user.

9. The method according to claim 8, characterized in that The monitoring network element, selection network element, deployment network element, and execution network element are core network elements.

10. A communication device, characterized in that: The device comprises means for performing the steps of the method according to any one of claims 1 to 9.

11. A core network device, characterized in that: The core network device includes a processor and a memory, the memory is used to store instructions, and the processor is used to read the instructions to execute the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 9.

13. A chip, characterized in that: include: A processor, configured to call and run a computer program from a memory, so that a communication device equipped with the chip executes the method according to any one of claims 1 to 9.

Images