Communication method and apparatus
By updating prediction models based on channel environments, the method addresses the complexity in wireless communication networks, improving communication performance and efficiency through beam prediction and model calibration.
Patent Information
- Application Number
- JP2024531192
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-11-24
- Filing Date
- 2022-11-23
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The increasing diversity of services in wireless communication networks, such as high rates, low latency, and large-scale connections, poses challenges in network planning, configuration, and resource scheduling, with energy conservation becoming a critical issue, necessitating the integration of artificial intelligence to enhance network intelligence.
A communication method and apparatus that updates prediction models based on channel environments to improve communication performance by predicting terminal device beams using downlink measurement information, allowing for calibration and retraining of prediction models when discrepancies are detected.
This approach reduces feedback overhead and ensures continuous improvement of predictive models, enhancing communication efficiency and reducing resource consumption in wireless networks.
Smart Images

Figure 2024541498000001_ABST
Abstract
Description
[Technical field]
[0001] The present application relates to the field of communication technologies, and in particular to communication methods and devices. [Background technology]
[0002] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Chinese Patent Application No. 202111404979.5, entitled "COMMUNICATION METHOD AND APPARATUS," filed with the China State Intellectual Property Office on November 24, 2021, which is incorporated herein by reference in its entirety.
[0003] In wireless communication networks, such as mobile communication networks, the services supported by the network are becoming more and more diverse, and therefore the requirements that need to be met are becoming more and more diverse. For example, the network needs to be able to support very high rates, very low latency, and / or very large connections. This characteristic makes network planning, network configuration, and / or resource scheduling more and more complex. Furthermore, as network capabilities become more and more powerful, such as supporting higher spectrum and new technologies such as high-order multiple-input multiple-output (MIMO) technology, beamforming, and / or beam management, energy conservation in the network has become a hot research topic. These new requirements, scenarios, and characteristics bring unprecedented challenges to network planning, operation and maintenance, and efficient operation. To address this challenge, artificial intelligence techniques may be introduced into wireless communication networks to implement network intelligence. Based on this, how to effectively implement artificial intelligence in networks has become a problem worth studying. Summary of the Invention
[0004] The present application provides a communication method and apparatus for updating a prediction model based on a channel environment to improve communication performance.
[0005] According to a first aspect, the present application provides a communication method. The method is applicable to a scenario in which a beam of a terminal device is predicted based on a prediction model. The method is executed by a terminal device or a module in the terminal device. In this specification, an example in which the method is executed by a terminal device is used for explanation. The method includes: the terminal device determining X pieces of downlink measurement information based on X downlink signals from the access network device, X being an integer greater than 1; determining a first downlink signal and a second downlink signal, the first downlink signal and the second downlink signal being downlink signals among the X downlink signals, the first downlink signal being a downlink signal determined based on an output of a first prediction model, an input of the first prediction model being determined based on Y pieces of downlink measurement information among the X pieces of downlink measurement information, the second downlink signal being determined based on the X pieces of downlink measurement information, Y being an integer less than or equal to X and greater than 0; and the terminal device sending first information to the access network device if the first downlink signal and the second downlink signal satisfy a first condition, the first information indicating the second downlink signal and the Y pieces of downlink measurement information, or the first information indicating that the first prediction model needs to be updated.
[0006] According to the above method, when the terminal device determines that the predicted first downlink signal and the second downlink signal determined based on the actual downlink measurement information satisfy a first condition due to a change in the channel environment or a mismatch between the prediction model and the channel environment, the terminal device may feed back the first information to the access network device, so as to enable the access network device to determine that the first prediction model needs to be calibrated, in other words, retrained, and thus the performance of the prediction model can be continuously improved.
[0007] In a possible design, if the first downlink signal and the second downlink signal satisfy a second condition, second information is sent to the access network device, where the second information indicates the second downlink signal or the second information indicates that the first prediction model does not need to be updated.
[0008] When the first downlink signal and the second downlink signal satisfy the second condition, the first prediction model may not be calibrated, thereby enabling the feedback overhead of the terminal device to be reduced and ensuring that the first prediction model can meet the communication requirements.
[0009] According to a second aspect, the present application provides a communication method. The method is applicable to a scenario in which a beam of a terminal device is predicted based on a prediction model. The method is executed by a terminal device or a module in the terminal device. In this specification, an example in which the method is executed by a terminal device is used for explanation. The method includes: a terminal device determining X downlink measurement information based on X downlink signals from an access network device, where X is an integer greater than 1; and a terminal device determining a first downlink signal and a second downlink signal, where the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, and an input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information, and the second downlink signal is a downlink signal determined based on an output of a first prediction model. determining a first information or a second information based on the first downlink signal and the second downlink signal, where Y is an integer less than or equal to X and greater than 0; and sending first information or second information to the access network device based on the first downlink signal and the second downlink signal, where the first information indicates that the first prediction model needs to be updated, or the first information indicates the second downlink signal and the Y downlink measurement information, and the second information indicates that the first prediction model is reliable, or the second information indicates that the first prediction model does not need to be updated, or the second information indicates that the second information does not include the Y downlink measurement information.
[0010] According to the above method, when the terminal device determines that the predicted first downlink signal does not match the second downlink signal determined based on the actual downlink measurement information due to a change in the channel environment or a mismatch between the prediction model and the channel environment, the terminal device may feed back the second downlink signal and the Y pieces of downlink measurement information to the access network device, so that the access network device can determine that the first prediction model needs to be calibrated, in other words, retrained, and thus the performance of the prediction model can be continuously improved.
[0011] In a possible design, the second information further indicates the second downlink signal.
[0012] In a possible design, first information is transmitted if a first condition is determined to be met, and second information is transmitted if a second condition is determined to be met.
[0013] For the first or second aspect, in a possible design, the X downlink measurement information are in one-to-one correspondence with the X downlink signals.
[0014] Regarding the first or second aspect, in a possible design, the first condition includes that differential information between the first downlink signal and the second downlink signal is greater than a first threshold.
[0015] According to the above method, when the differential information between the first downlink signal and the second downlink signal is greater than a first threshold, it can be determined that the prediction result of the first prediction model is inaccurate and cannot meet the communication requirements, and therefore the first prediction model needs to be updated.
[0016] Regarding the first or second aspect, in a possible design, the second condition includes that the first downlink signal and the second downlink signal are the same downlink signal, or that the first downlink signal and the second downlink signal are not the same downlink signal but differential information between the first downlink signal and the second downlink signal is less than or equal to a first threshold.
[0017] According to the above method, when the difference information between the first downlink signal and the second downlink signal is less than or equal to a first threshold, it can be determined that the prediction result of the first prediction model is accurate and can meet the communication requirements, and the first prediction model does not need to be updated.
[0018] Regarding the first or second aspect, in a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal.
[0019] Regarding the first or second aspect, in a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first position information corresponding to the first downlink signal and second position information corresponding to the second downlink signal.
[0020] Regarding the first or second aspect, in a possible design, the method further includes training a first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a second prediction model, and sending third information to the access network device, the third information indicating the second prediction model or indicating information about the second prediction model.
[0021] According to the above method, since the terminal device updates the first prediction model, the load on the access network device can be reduced.
[0022] Regarding the first or second aspect, in a possible design, the sending of the third information to the access network device may include: and sending third information to the access network device if the variation information between the second prediction model and the first prediction model is greater than or equal to a second threshold.
[0023] Regarding the first or second aspect, in a possible design, the information about the second predictive model includes variation information between the second predictive model and the first predictive model.
[0024] For the first or second aspect, in a possible design, the method further includes receiving fourth information from the access network device, the fourth information indicating information about a third prediction model, and updating the first prediction model to the third prediction model.
[0025] According to the above method, the access network device updates the first prediction model to ensure that communication efficiency is improved, and avoid communication interruptions caused by the access network device being unable to obtain an updated prediction model for a long period of time.
[0026] Regarding the first or second aspect, in a possible design, the third prediction model is obtained by training the first prediction model based on the Y downlink measurement information and the second downlink signal.
[0027] Regarding the first or second aspect, in a possible design, the first prediction model is the same as the prediction model configured in the access network device.
[0028] Since the same prediction model is set in the access network device and the terminal device, the terminal device can determine whether the prediction model set in the access network device is accurate based on downlink measurement information obtained by actual measurements, and calibrate the prediction model in the access network device in real time if the prediction model is inaccurate.
[0029] Regarding the first or second aspect, in a possible design, the first prediction model is agreed upon in a protocol or configured by the access network device.
[0030] According to a third aspect, the present application provides a communication method. The method is applicable to a scenario in which a beam of a terminal device is predicted based on a prediction model. The method is executed by an access network device or a module in the access network device. In this specification, an example in which the method is executed by an access network device is used for description. The method includes: an access network device transmits X downlink signals to a terminal device, where X is an integer greater than 1; and receives first information from the terminal device, where the first information indicates a second downlink signal and Y downlink measurement information of the X downlink measurement information, or the first information indicates that a first prediction model needs to be updated, where the second downlink signal is determined based on the X downlink measurement information (or the second downlink signal is an optimal signal of the X downlink signals), the X downlink measurement information is determined based on the X downlink signals (or the X downlink measurement information corresponds to the X downlink signals), and Y is an integer less than or equal to X and greater than 0. The method can be further described as follows: That is, the access network device transmits X downlink signals to the terminal device, where X is an integer greater than 1; the access network device receives first information from the terminal device, where the first information indicates a second downlink signal and Y downlink measurement information, or the first information indicates that the first prediction model needs to be updated, the second downlink signal is an optimal signal among the X downlink signals, and the Y downlink measurement information corresponds to Y downlink signals among the X downlink signals, where Y is an integer less than or equal to X and greater than 0.
[0031] In a possible design, the X downlink measurement information is in one-to-one correspondence with the X downlink signals.
[0032] According to a fourth aspect, the present application provides a communication method. The method is applicable to a scenario in which a beam of a terminal device is predicted based on a prediction model. The method is executed by an access network device or a module in the access network device. In this specification, an example in which the method is executed by an access network device is used for explanation. The method includes: an access network device transmitting X downlink signals to a terminal device, where X is an integer greater than 1; and receiving first information or second information from the terminal device, where the first information indicates a second downlink signal and Y downlink measurement information of the X downlink measurement information, or the first information indicates that a first prediction model needs to be updated, the second information indicates a second downlink signal, or the second information indicates that a first prediction model does not need to be updated, the second downlink signal is determined based on the X downlink measurement information (or the second downlink signal is an optimal signal of the X downlink signals), the X downlink measurement information is determined based on the X downlink signals (or the X downlink measurement information corresponds to the X downlink signals), and Y is an integer less than or equal to X and greater than 0. The method may be further described as follows: That is, the access network device transmits X downlink signals to the terminal device, where X is an integer greater than 1; the access network device receives first information or second information from the terminal device, where the first information indicates a second downlink signal and Y downlink measurement information, or the first information indicates that the first prediction model needs to be updated, the second information indicates the second downlink signal, or the second information indicates that the first prediction model does not need to be updated, the second downlink signal is an optimal signal among the X downlink signals, and the Y downlink measurement information corresponds to Y downlink signals among the X downlink signals, and Y is an integer less than or equal to X and greater than 0.
[0033] In a possible design, the X downlink measurement information is in one-to-one correspondence with the X downlink signals.
[0034] In a possible design, the first downlink signal and the second downlink signal satisfy a first condition, the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, and an input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information.
[0035] In a possible design, the method further includes receiving second information from the terminal device, where the second information indicates a second downlink signal or the second information indicates that the first prediction model does not need to be updated.
[0036] In a possible design, the first downlink signal and the second downlink signal satisfy a second condition.
[0037] Regarding the third or fourth aspect, in a possible design, the first condition includes that differential information between the first downlink signal and the second downlink signal is greater than a first threshold.
[0038] Regarding the third or fourth aspect, in a possible design, the second condition includes that the first downlink signal and the second downlink signal are the same downlink signal, or that the first downlink signal and the second downlink signal are not the same downlink signal but differential information between the first downlink signal and the second downlink signal is less than or equal to a first threshold.
[0039] Regarding the third or fourth aspect, in a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal.
[0040] Regarding the third or fourth aspect, in a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first position information corresponding to the first downlink signal and second position information corresponding to the second downlink signal.
[0041] Regarding the third or fourth aspect, in a possible design, the method further includes training a first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model, and transmitting fourth information to the terminal device, wherein the fourth information indicates the third prediction model or indicates information about the third prediction model.
[0042] Regarding the third or fourth aspect, in a possible design, the information about the third predictive model includes variation information between the third predictive model and the first predictive model.
[0043] Regarding the third or fourth aspect, in a possible design, the method further includes receiving third information from the terminal device, the third information indicating information about the second predictive model, and updating the first predictive model to the second predictive model.
[0044] Regarding the third or fourth aspect, in a possible design, the first prediction model is the same as the prediction model configured in the terminal device.
[0045] According to a fifth aspect, the present application provides a communication method. The method is applicable to a scenario in which a beam of a terminal device is predicted based on a prediction model. The method is executed by an access network device or a module in the access network device. In this specification, an example in which the method is executed by an access network device is used for explanation. The method includes: an access network device transmits X downlink signals to a terminal device, and receives X downlink measurement information from the terminal device, where the X downlink measurement information is determined based on the X downlink signals, where X is an integer greater than 1; and determining a first downlink signal and a second downlink signal, where the first downlink signal and the second downlink signal are downlink signals of the X downlink signals, and the first downlink signal is determined based on an output of a first prediction model. the input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information, and the second downlink signal is determined based on the X downlink measurement information, where Y is an integer less than or equal to X and greater than 0; training the first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model when the first downlink signal and the second downlink signal satisfy a first condition; and updating the first prediction model to the third prediction model.
[0046] According to the above method, only the first prediction model may be deployed in the access network device. When the access network device determines that the predicted first downlink signal does not match the second downlink signal determined based on the actual downlink measurement information due to the change of the channel environment or the mismatch between the prediction model and the channel environment, the access network device may calibrate the first prediction model based on the second downlink signal and the Y downlink measurement information, i.e., retrain the first prediction model to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0047] In a possible design, the X downlink measurement information is in one-to-one correspondence with the X downlink signals.
[0048] In a possible design, if the first downlink signal and the second downlink signal satisfy a second condition, the first prediction model is kept unchanged.
[0049] In one possible design, the first condition includes differential information between the first downlink signal and the second downlink signal being greater than a first threshold.
[0050] In a possible design, the second condition includes that the first downlink signal and the second downlink signal are the same downlink signal, or that the first downlink signal and the second downlink signal are not the same downlink signal but differential information between the first downlink signal and the second downlink signal is less than or equal to a first threshold.
[0051] In a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal.
[0052] In a possible design, the difference information between the first downlink signal and the second downlink signal includes an absolute value of a difference between first position information corresponding to the first downlink signal and second position information corresponding to the second downlink signal.
[0053] According to a sixth aspect, the present application provides a communication method. The method is applicable to a scenario in which uplink measurement information measured by an access network device is predicted based on a prediction model. The method is performed by an access network device or a module in the access network device. In this specification, an example in which the method is performed by an access network device is used for explanation. The method includes: an access network device receiving X uplink signals from a terminal device; determining X pieces of uplink measurement information based on the X uplink signals, where X is an integer greater than 1; determining a first uplink signal and a second uplink signal, where the first uplink signal and the second uplink signal are uplink signals among the X uplink signals, where the first uplink signal is an uplink signal determined based on an output of a fourth prediction model, where an input of the fourth prediction model is determined based on Y pieces of uplink measurement information among the X pieces of uplink measurement information, where the second uplink signal is determined based on the X pieces of uplink measurement information, where Y is an integer less than X and greater than 0; training a fourth prediction model based on the Y pieces of uplink measurement information and the second uplink signal to obtain a fifth prediction model if the first uplink signal and the second uplink signal satisfy a third condition; and updating the fourth prediction model to the fifth prediction model.
[0054] According to the above method, the fourth prediction model is disposed in the access network device. When the access network device determines that the predicted first uplink signal does not match the second uplink signal determined based on the actual uplink measurement information due to the change of the channel environment or the mismatch between the prediction model and the channel environment, the access network device can retrain the fourth prediction model based on the second uplink signal and the Y uplink measurement information to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0055] In a possible design, if the first uplink signal and the second uplink signal satisfy a fourth condition, the fourth prediction model is kept unchanged.
[0056] In a possible design, the third condition includes: differential information between the first uplink signal and the second uplink signal being greater than a third threshold.
[0057] In a possible design, the fourth condition includes that the first uplink signal and the second uplink signal are the same uplink signal, or that the first uplink signal and the second uplink signal are not the same uplink signal but differential information between the first uplink signal and the second uplink signal is less than or equal to a third threshold.
[0058] In a possible design, the difference information between the first uplink signal and the second uplink signal includes an absolute value of a difference between first uplink measurement information corresponding to the first uplink signal and second uplink measurement information corresponding to the second uplink signal.
[0059] In a possible design, the difference information between the first uplink signal and the second uplink signal includes an absolute value of a difference between position information corresponding to the first uplink signal and position information corresponding to the second uplink signal.
[0060] According to a seventh aspect, the present application further provides a communication device. The communication device can implement any method or any implementation provided in the first or second aspect. The communication device may be implemented by hardware, by software, or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions.
[0061] In a possible design, the communication device includes a processor. The processor is configured to support the communication device to execute the method according to the first aspect or the second aspect. The communication device may further include a memory. The memory may be coupled to the processor and may store program instructions and data required for the communication device. Optionally, the communication device further includes an interface circuit, the interface circuit configured to support the communication device to communicate with another communication device.
[0062] In a possible implementation, the structure of this communication device includes a processing unit and a communication unit. These units can perform corresponding functions in the above method examples. For details, please refer to the description of the first aspect or the second aspect. In this specification, details will not be described again.
[0063] According to an eighth aspect, the present application further provides a communication device. The communication device may implement any method or any implementation provided in any one of the third to sixth aspects. The communication device may be implemented by hardware, by software, or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the above functions.
[0064] In a possible implementation, the communication device includes a processor. The processor is configured to support the communication device to execute a method according to any one of the third to sixth aspects. The communication device may further include a memory. The memory may be coupled to the processor and may store program instructions and data required for the communication device. Optionally, the communication device further includes an interface circuit, the interface circuit configured to support the communication device to communicate with another communication device.
[0065] In a possible implementation, the structure of this communication device includes a processing unit and a communication unit. These units can perform corresponding functions in the above method examples. For details, please refer to the method description provided in any one of the third to sixth aspects. Details will not be described in this specification.
[0066] According to a ninth aspect, there is provided a communication device including a processor and an interface circuit. The interface circuit is configured to receive a signal from another communication device other than the communication device and transmit the signal to the processor, or transmit a signal from the processor to another communication device other than the communication device. The processor is configured to execute a computer program or instructions stored in a memory to perform the method of any one of the first aspect or the second aspect and possible implementations thereof, or to perform the method of any one of the third aspect to the sixth aspect and possible implementations thereof. Optionally, the device further includes a memory, the memory storing the computer program or instructions.
[0067] According to a tenth aspect, a computer readable storage medium is provided. The computer readable storage medium stores a computer program or instruction. When the computer program or instruction is executed on a computer, the computer can perform the method of any one of the first aspect or the second aspect and possible implementations thereof, or the computer can perform the method of any one of the third aspect to the sixth aspect and possible implementations thereof.
[0068] According to an eleventh aspect, there is provided a computer program product comprising computer readable instructions which, when executed on a computer, enable the computer to perform the method of any one of the first aspect or the second aspect and possible implementations thereof, or enable the computer to perform the method of any one of the third aspect to the sixth aspect and possible implementations thereof.
[0069] According to a twelfth aspect, a chip is provided. The chip includes a processor and may further include a memory. The processor is coupled to the memory and configured to execute a computer program or instructions stored in the memory to enable the chip to perform the method of any one of the first aspect or the second aspect and possible implementations thereof, or to enable the chip to perform the method of any one of the third aspect to the sixth aspect and possible implementations thereof.
[0070] According to a thirteenth aspect, there is provided a communication system, the system including an apparatus (such as a terminal device) implementing the first aspect or the second aspect, and an apparatus (such as an access network device) implementing the third aspect.
[0071] According to a fourteenth aspect, there is provided a communication system, wherein the system includes a terminal device and an apparatus (such as an access network device) implementing the fourth or fifth aspect. [Brief description of the drawings]
[0072] [Figure 1] FIG. 1 is a diagram of a communication system architecture applicable to the present disclosure.
[0073] [Diagram 2] FIG. 2 is a diagram of layer relationships of a neural network according to the present disclosure.
[0074] [Diagram 3] FIG. 1 is an exemplary diagram of an application framework of AI in a communication system according to the present disclosure.
[0075] [Figure 4(a)] FIG. 1 is an example diagram of a network architecture to which the methods provided in the present disclosure may be applied. [Figure 4(b)] FIG. 1 is an example diagram of a network architecture to which the methods provided in the present disclosure may be applied. [Figure 4(c)] FIG. 1 is an example diagram of a network architecture to which the methods provided in the present disclosure may be applied. [Figure 4(d)] FIG. 1 is an example diagram of a network architecture to which the methods provided in the present disclosure may be applied.
[0076] [Diagram 5] FIG. 1 is a diagram of a capability dialogue according to the present disclosure.
[0077] [Figure 6] FIG. 2 is a diagram of the structure of a model according to the present disclosure.
[0078] [Figure 7] 1 is a schematic flow diagram of a communication method according to the present disclosure.
[0079] [Figure 8] FIG. 1 is a diagram of beam positions according to the present disclosure.
[0080] [Figure 9]FIG. 1 is a diagram of beam positions according to the present disclosure.
[0081] [Figure 10] 1 is a schematic flow diagram of a model updating method according to the present disclosure.
[0082] [Figure 11] 1 is a schematic flow diagram of a communication method according to the present disclosure.
[0083] [Figure 12] 1 is a schematic flow diagram of a communication method according to the present disclosure.
[0084] [Figure 13] FIG. 2 is a diagram of the structure of a communication device according to the present disclosure.
[0085] [Figure 14] FIG. 2 is a diagram of the structure of a communication device according to the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0086] The present disclosure will now be described in detail with reference to the accompanying drawings herein.
[0087] The technical solutions of the present disclosure may be applied to various communication systems, such as long term evolution (LTE) systems, fifth generation (5G) mobile communication systems, or next generation mobile communication systems, which are not limited herein. The 5G system may also be referred to as a new radio (NR) system.
[0088] In this disclosure, the example of the interaction between a terminal device and an access network device is described.It should be noted that the method provided in this disclosure can be applied not only to the interaction between a terminal device and an access network device, but also to the interaction between two other devices, which is not limited in this disclosure.
[0089] In order to facilitate the understanding of the present disclosure, a communication system applicable to the present disclosure will be described in detail first using the communication system shown in Fig. 1 as an example. Fig. 1 is an architecture diagram of a communication system to which the present disclosure can be applied. The communication system includes an access network device and a terminal device. The terminal device may establish a connection with the access network device and communicate with the access network device. Fig. 1 is merely a diagram. The quantity of the access network devices and the terminal devices included in the communication system is not limited in the present disclosure.
[0090] To support artificial intelligence (AI) in wireless networks, a dedicated AI entity (or called an AI module) may be further introduced into the network. The AI entity may correspond to one independent network element or may be located within a network element, which may be a core network device, an access network device, or an operations, administration and maintenance (OAM), etc. For example, as shown in FIG. 1, the AI entity may be located outside the access network device and communicate with the access network device. For example, the access network device may forward data related to the AI model reported by the terminal device to the AI entity, which performs operations such as training dataset construction and model training, sends the trained artificial intelligence model to the access network device, and / or forwards the trained artificial intelligence model to each terminal device via the access network device. In the present disclosure, the OAM is configured to operate, manage, and / or maintain the core network device (operation, management, and maintenance of the core network device) and / or is configured to operate, manage, and / or maintain the access network device (operation, management, and maintenance of the access network device). The OAM of the access network device and the OAM of the core network device may be the same or different. This is not limited. For example, the present disclosure includes a first OAM and a second OAM. The first OAM is the operation, management, and maintenance of the core network device, and the second OAM is the operation, management, and maintenance of the access network device. The first OAM and / or the second OAM may include an AI entity. In another example, the present disclosure includes a third OAM, and the third OAM is the operation, management, and maintenance of both the core network device and the access network device.
[0091] Optionally, to accommodate and support AI, an AI entity may be integrated into a terminal or a terminal chip.
[0092] Optionally, in the present disclosure, an AI entity may be a software model, a hardware structure, or a combination of a hardware structure and a software model. This is not limited. An AI entity may have another name and is mainly configured to implement an AI function (or called an AI-related operation). The specific name of an AI entity is not limited in the present disclosure.
[0093] In this disclosure, an AI model is a specific method for implementing an AI function, and an AI model represents a mapping relationship between the input and output of the model. The AI model may be a neural network, a random forest model, a support vector machine model, a decision tree model, or another machine learning model. The AI model may also be called a model for short. The AI-related operations may include at least one of the following: data collection, model training, model validation, model information release, model inference (or called model inference or prediction), or inference result release, etc.
[0094] In this disclosure, the terminal device may be referred to as a terminal for short. The terminal device may communicate with one or more core networks via a radio access network (RAN).
[0095] In the present disclosure, a terminal device may be a device having a wireless transceiver function or a chip that may be disposed within the device. A terminal device may also be referred to as a user equipment (UE), an access terminal, a subscriber unit, a subscriber station, a mobile station, a remote station, a remote terminal, a mobile device, a user terminal, a wireless communication device, a user agent, or a user equipment, etc. A terminal device of the present disclosure may be a mobile phone, a Pad, a computer with a wireless transceiver function, a virtual reality (VR) terminal, an augmented reality (AR) terminal, a wearable device, a vehicle, an unmanned aerial vehicle, a helicopter, an airplane, a ship, a robot, a robot arm, or a smart home device, etc. The terminal device of the present disclosure may be widely used for communication in various scenarios, including, but not limited to, one of the following scenarios: enhanced mobile broadband (eMBB), ultra-reliable low-latency communication (URLLC), device-to-device (D2D), vehicle-to-everything (V2X), machine-type communication (MTC), massive machine-type communication (mMTC), Internet of things (IOT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grid, smart furniture, smart office, smart wearable, smart transportation, or smart city. The specific technology and the specific device form used by the terminal device are not limited by the present disclosure.
[0096] In the present disclosure, a device configured to implement a function of a terminal may be a terminal, or may be a device capable of supporting a terminal to implement its function, such as a chip system, a hardware circuit, a software module, or a combination of a hardware circuit and a software module, etc. The device may be installed in the terminal or may be combined with the terminal in use.
[0097] The access network device may be a base station, a NodeB, an evolved NodeB (eNodeB), a transmission reception point (TRP), a next generation NodeB (gNB) of a fifth generation (5G) mobile communication system, an access network device of an open radio access network (O-RAN or open RAN), a next generation base station of a sixth generation (6G) mobile communication system, a base station of a future mobile communication system, or an access node of a wireless fidelity (Wi-Fi) system, or may be a module or unit that completes some functions of a base station, such as a central unit (CU), a distributed unit (DU), a central unit control plane (CU-CP) module, or a central unit user plane (CU-UP) module. The radio access network device may be a macro base station, a micro base station, or an indoor base station, or may be a relay node, a donor node, or the like. The specific technology used by the access network devices and the specific device configurations are not limited by this disclosure.
[0098] (1) Protocol Layer Structure
[0099] The communication between the access network device and the terminal is compiled in a specific protocol layer structure. This protocol layer structure may include a control plane protocol layer structure and a user plane protocol layer structure. For example, the control plane protocol layer structure may include the functions of protocol layers such as a radio resource control (RRC) layer, a packet data convergence protocol (PDCP) layer, a radio link control (RLC) layer, a media access control (MAC) layer, and a physical layer. For example, the user plane protocol layer structure may include the functions of protocol layers such as a PDCP layer, an RLC layer, a MAC layer, and a physical layer. In a possible implementation, a service data adaptation protocol (SDAP) layer may be further included above the PDCP layer.
[0100] Optionally, the protocol layer structure between the access network device and the terminal may further include an artificial intelligence (AI) layer used to perform transmission of data related to AI functions.
[0101] (2) Central unit (CU) and distributed unit (DU)
[0102] The access network device may include a CU and a DU. A plurality of DUs may be controlled by one CU in a centralized manner. For example, an interface between a CU and a DU may be called an F1 interface. A control plane (CP) interface may be F1-C, and a user plane (UP) interface may be F1-U. The specific names of the interfaces are not limited in this disclosure. The CU and the DU may be divided based on the protocol layers of the wireless network. For example, the functions of the PDCP layer and the protocol layers above the PDCP layer are located in the CU, and the functions of the protocol layers below the PDCP layer (e.g., the RLC layer and the MAC layer) are located in the DU. In another example, the functions of the protocol layers above the PDCP layer are located in the CU, and the functions of the PDCP layer and the protocol layers below the PDCP layer are located in the DU. This is not limited.
[0103] The division of the processing functions of the CU and the DU based on the protocol layer is merely an example, and the processing functions of the CU and the DU may alternatively be divided in another manner. For example, the CU or the DU may be divided into functions having more protocol layers. In another example, the CU or the DU may be further divided into several processing functions having protocol layers. In one design, some functions of the RLC layer and functions of the protocol layers above the RLC layer are set to the CU, and the remaining functions of the RLC layer and functions of the protocol layers below the RLC layer are set to the DU. In another design, the division of the functions of the CU or the DU may alternatively be performed based on the service type or other system requirements. For example, the division may be performed based on latency. Functions whose processing time needs to meet the latency requirement are set to the DU, and functions whose processing time does not need to meet the latency requirement are set to the CU. In another design, the CU may alternatively have one or more functions of the core network. For example, the CU may be located on the network side to facilitate central management. In another design, a radio unit (RU) of the DU is remotely located. Optionally, the RU may have radio frequency capabilities.
[0104] Optionally, the DU and the RU may be differentiated by a physical layer (PHY). For example, the DU may implement the upper layer functions of the PHY layer, and the RU may implement the lower layer functions of the PHY layer. If the PHY layer is used for transmission, the functions of the PHY layer may include at least one of the following functions: cyclic redundancy check (CRC) bit, channel coding, rate matching, scrambling, modulation, layer mapping, precoding, resource mapping, physical antenna mapping, or radio frequency transmission. If the PHY layer is used for reception, the functions of the PHY layer may include at least one of the following: CRC check, channel decoding, rate matching, descrambling, demodulation, layer demapping, channel detection, resource demapping, physical antenna demapping, or radio frequency reception. The upper layer functions of the PHY layer may include some functions of the PHY layer. For example, some of these functions are closer to the MAC layer. The lower layer functions of the PHY layer may include some other functions of the PHY layer. For example, some of these other functions are closer to the radio frequency functions. For example, the upper layer functions of the PHY layer may include CRC code addition, channel coding, rate matching, scrambling, modulation, and layer mapping, and the lower layer functions of the PHY layer may include precoding, resource mapping, physical antenna mapping, and radio frequency transmission. Alternatively, the upper layer functions of the PHY layer may include CRC code addition, channel coding, rate matching, scrambling, modulation, layer mapping, and precoding. The lower layer functions of the PHY layer may include resource mapping, physical antenna mapping, and radio frequency transmission. For example, the upper layer functions of the PHY layer may include CRC checking, channel decoding, rate matching, decoding, demodulation, and layer de-mapping, and the lower layer functions of the PHY layer may include channel detection, resource de-mapping, physical antenna de-mapping, and radio frequency reception.Alternatively, the upper layer functions of the PHY layer may include CRC checking, channel decoding, rate de-matching, decoding, demodulation, layer de-mapping, and channel detection, and the lower layer functions of the PHY layer may include resource de-mapping, physical antenna de-mapping, and radio frequency reception.
[0105] For example, the functions of the CU may be implemented by one entity or by different entities. For example, the functions of the CU may be further divided. In other words, the control plane and the user plane are implemented separately by different entities, namely, a control plane CU entity (i.e., a CU-CP entity) and a user plane CU entity (i.e., a CU-UP entity). The CU-CP entity and the CU-UP entity are coupled to the DU and cooperate to complete the functions of the access network device.
[0106] Optionally, any one of the DU, CU, CU-CP, CU-UP, and RU may be a software module, a hardware structure, or a combination of a software module and a hardware structure. This is not limited. Different entities may exist in different forms. This is not limited. For example, the DU, CU, CU-CP, and CU-UP are software modules, and the RU is a hardware structure. These modules and the methods performed by these modules are also included in the protection scope of the present disclosure.
[0107] In a possible implementation, the access network device includes a CU-CP, a CU-UP, a DU, and a RU. For example, in the present disclosure, the DU, or the DU and the RU, or the CU-CP, the DU, and the RU, or the CU-UP, the DU, and the RU perform the functions of the access network device. This is not limited. The methods performed by these modules are also included in the scope of protection of the present disclosure.
[0108] In the present disclosure, an apparatus configured to implement a function of an access network device may be an access network device, or may be an apparatus capable of supporting an access network device to implement the function, such as a chip system, etc. The apparatus may be installed in the access network device or may be used in combination with the access network device.
[0109] The present disclosure is applicable to the following scenario: an access network device uses an AI model to predict information that needs to be measured by a terminal device, so as to reduce the frequent transmission of downlink signals by the access network device, and reduce the frequent measurement of the information by the terminal device based on the downlink signals, thereby reducing the resource overhead of the information and saving the computing resources of the terminal device.
[0110] For example, a high-frequency band larger than 6 gigahertz (GHz) is introduced into the NR system for data communication. In high-frequency communication, a large number of antennas are usually used for beam forming, so that a beam gain can be obtained to compensate for the limited propagation distance caused by the high-frequency propagation characteristics. High-frequency wireless channels usually have a clear sparsity. In other words, the main energy of the channel is concentrated on a limited number of paths. When there is a line of sight (LOS) path without obstruction between the transmitter and the receiver, the main energy between the receiver and the transmitter is concentrated on the line of sight path. When there is a non-line of sight (NLOS) obstruction between the transmitter and the receiver, the main energy between the receiver and the transmitter is mainly concentrated on the path where the signal can reach once reflected. In general, each path has a different incidence angle and a different exit angle. Therefore, when the transmitter of the high-frequency communication system matches the beam direction with the incidence angle of the main path of the channel and the receiver of the high-frequency communication system matches the beam direction with the exit angle of the main path of the channel, most of the channel transmission energy can be obtained to complete high-quality communication. A beam direction may be considered as one particular data transmission direction.
[0111] A simple beam management solution can be designed. We assume that all candidate beamforming beams are included in one codebook, and each codeword in the codebook can be used to precode a signal, and the precoded signal is propagated in space based on a certain spatial directivity to form a beamforming effect. Thus, one codeword corresponds to one beam. The codewords in the codebook are represented in the form of a complex vector, and the number of elements of the complex vector of each codeword is equal to the number of antennas (or denoted as antenna ports), for example, R contains C codewords, and R=[w1, w2, ..., w C Each codeword w i teeth,
[0112]
number
[0113] where i ranges from 1 to C, C is an integer equal to or greater than 1, and N T is the number of antennas, and the complex number ν j represents the weight of the j-th antenna, and j ranges from 1 to N T is in the range of N T is an integer equal to or greater than 1. i If is intended to precode a signal, then it is the codeword w i This is equivalent to adjusting the amplifiers and / or phase shifters of the corresponding antennas based on the weight v of each antenna in the signal processing unit 100 so that the signal propagation pattern transmitted or received by the antennas is spatially directional, in other words, a beamformed beam is generated.
[0114] The beams corresponding to different code words in the code book may point in different directions to cover as much of the entire propagation space as possible. Each time a signal is transmitted, one suitable code word that matches the direction of the main path of the channel is selected from the code book. The access network device holds its code book, and the terminal device holds its code book. Take the transmission in the downlink direction as an example. When communication is required, the access network device scans the candidate code words in the code book in a specific order. That is, the access network device transmits a reference signal using a beam corresponding to each of the candidate code words in the code book. The terminal device receives this reference signal, and feeds back information about the reference signal corresponding to the received beam whose received signal strength is the highest or exceeds a certain threshold, for example, feeds back an index of the reference signal, or performs feedback on the resource corresponding to the reference signal. The access network device determines an optimal beam based on the feedback information of the terminal device, for example, determines the transmission beam with the highest received signal strength as the optimal beam, and transmits downlink data to the terminal device using the optimal beam. This process is sometimes called beam alignment. However, the air interface overhead is high because the entire codebook needs to be scanned to select the optimal codeword. For a fixed-length slot, a large amount of time resources is used for beam alignment, which constrains the maximum capacity of data transmission. Furthermore, in high-speed mobility scenarios, the channel changes quickly and the channel correlation time is short, making it difficult for the receiver and transmitter to complete effective beam alignment. As a result, communication cannot be established.
[0115] In the present disclosure, beam prediction is performed using machine learning techniques, and the overhead of beam scanning can be reduced. Specifically, for one codebook, the transmitter no longer scans all codewords corresponding to the codebook, but scans some codewords in the codebook according to a certain rule, i.e., selects only some codewords in the codebook to transmit a reference signal. The receiver measures the received reference signal, obtains measurement information, and feeds back the measurement information to the transmitter. The transmitter predicts the optimal codeword index (and beam index) of the entire codebook based on the received measurement information. The prediction function is implemented by an AI model. Since only some codewords in the codebook need to be scanned, the air interface overhead of beam alignment can be reduced. Furthermore, in the present disclosure, the same AI model can be deployed in the access network device and the terminal device, the terminal device can predict the optimal codeword index based on the AI model, and the terminal device can further obtain the actual optimal codeword index by measuring the reference signal. If the predicted optimal codeword index does not match the actual optimal codeword index, the access network device or the terminal device may retain the AI model and check the AI model, as described in more detail below, including, but not limited to, performing training with an initial model or performing update training with a current model.
[0116] Before describing the method of the present disclosure, we first briefly discuss some relevant knowledge about artificial intelligence. Artificial intelligence enables machines to have human intelligence. For example, a machine can simulate some intelligent behavior of a human using computer software and hardware. Machine learning methods or many other methods can be used to implement artificial intelligence. For example, machine learning includes neural networks (NN). Neural networks are a specific implementation of machine learning. According to the general approximation theorem, neural networks have the ability to learn any mapping since, in theory, they can approximate any continuous function. Thus, neural networks can accurately perform abstract modeling of complex high-dimensional problems. The concept of neural networks originates from the neuronal structure of brain tissue. Each neuron performs a weighted summation operation on the input values of that neuron and outputs the result of the weighted summation via an activation function. When the input of a neuron is x=[x0, ..., x n ], and the weights corresponding to the inputs are w = [w0, ..., w n ] and the offset of the weighted sum is b. The form of the activation function can vary. Suppose the activation function of a neuron is y = f(z) = max(0,z), then the output of this neuron is
[0117]
number
[0118] As another example, if the activation function of a neuron is y=f(z)=z, then the output of this neuron is
[0119]
number
[0120] It is. i , x i, and b can take on a variety of possible values, such as fractional numbers, integers (including 0, positive integers, or negative integers, etc.), or complex numbers. The activation functions of different neurons in a neural network may be the same or different.
[0121] A neural network generally includes a multi-layer structure, and each layer may include one or more neurons. Increasing the depth and / or width of a neural network can improve the representation capability of the neural network and provide a more powerful information extraction and abstract modeling capability for complex systems. The depth of a neural network may refer to the number of layers included in the neural network, and the number of neurons included in each layer may be referred to as the layer width. FIG. 2 is a diagram of the layer relationship of a neural network. In one implementation, a neural network includes an input layer and an output layer. After performing neuron processing on the received input, the input layer of the neural network transfers the result to the output layer, and the output layer obtains the output result of the neural network. In another implementation, a neural network includes an input layer, a hidden layer, and an output layer. The input layer of the neural network performs neuron processing on the received input and then transfers the result to an intermediate hidden layer. The hidden layer then transfers the calculation result to the output layer or an adjacent hidden layer. Finally, the output layer obtains the output result of the neural network. A neural network may include one hidden layer, or multiple hidden layers connected in series. This is not limited. In the training process of a neural network, a loss function may be defined. The loss function describes the difference value or difference between the output value of the neural network and the ideal target value. The specific form of the loss function is not limited in this disclosure. The training process of a neural network is a process of adjusting neural network parameters, such as the number and width of layers of the neural network, the weights of neurons, and / or the parameters of the activation functions of neurons, so that the value of the loss function is below a threshold or meets the target requirement. In other words, the difference between the output value of the neural network and the ideal target value is the smallest.
[0122] FIG. 3 is an exemplary diagram of a first application framework of AI in a communication system. In FIG. 3, a data source is used to store training data and inference data. A model training node (model training host) analyzes or trains the training data provided by the data source to obtain an AI model, and places the AI model in a model inference node (model inference host). The AI model represents a mapping relationship between the input and output of the model. Obtaining an AI model by learning with the model training node is equivalent to obtaining a mapping relationship between the input and output of the model by learning with the training data with the model training node. The model inference node uses the AI model to perform inference based on the inference data provided by the data source, and obtains an inference result. This method may also be described as follows: That is, the model inference node inputs inference data into the AI model, and obtains an output using the AI model. The output is the inference result. The inference result may indicate a setting parameter used (acted upon) by an action subject, and / or an operation performed by the action subject. The inference result may be planned in a unified manner by an actor entity and sent to one or more action subjects (e.g., network elements) for action.
[0123] In the present disclosure, the application framework shown in FIG. 3 may be disposed in the network element shown in FIG. 1. For example, the application framework in FIG. 3 may be disposed in the access network device or AI entity in FIG. 1. For example, in the access network device, the model training node may analyze or train training data provided by a data source to obtain a model. The model inference node may perform inference using the model and the inference data provided by the data source to obtain an output of the model. Specifically, the input of the model includes inference data, and the output of the model is an inference result corresponding to the model. For example, the model training node is a CU, and the model inference node is a CU or DU, or the model training node is a DU, and the model inference node is a DU, or the model training node is a quasi-real-time RIC described below, and the model inference node is a quasi-real-time RIC, CU, or DU. For example, the terminal device is regarded as the action subject in FIG. 3. The access network device may send inference data and / or inference results corresponding to the model to the terminal device, and the terminal device may perform a corresponding operation based on the inference data and / or inference results.
[0124] In the following, with reference to Figures 4(a) to 4(d), a network architecture to which the communication solutions provided in the present disclosure may be applied will be described.
[0125] As shown in FIG. 4(a), in a first possible implementation, the access network device includes a near-real-time access network intelligent controller (RAN intelligent controller, RIC) module configured to perform model learning and / or inference. For example, the near-real-time RIC may obtain information about the network side and / or the terminal side from at least one of the CU, DU, and RU, which may be used as training data or inference data. Optionally, the near-real-time RIC may submit inference results to at least one of the CU, DU, and RU. Optionally, the CU and the DU may exchange inference results. Optionally, the DU and the RU may exchange inference results. For example, the near-real-time RIC submits the inference results to the DU, and the DU submits the inference results to the RU. For example, the near-real-time RIC may be configured to train an AI model, and inference is performed using the AI model.
[0126] As shown in FIG. 4(a), in a second possible implementation, a non-real-time RIC (optionally, the non-real-time RIC may be located in an OAM or core network device) other than an access network device is included and used for model learning and inference. For example, the non-real-time RIC may obtain information about the network side and / or the terminal side from at least one of the CU, DU, and RU. This information may be used as training data or inference data, and the inference result may be submitted to at least one of the CU, DU, and RU. Optionally, the CU and DU may exchange the inference result. Optionally, the DU and RU may exchange the inference result. For example, the non-real-time RIC submits the inference result to the DU, and the DU submits the inference result to the RU. For example, the non-real-time RIC is configured to train an AI model, and inference is performed using the AI model.
[0127] As shown in FIG. 4(a), in a third possible implementation, the access network device includes a near-real-time RIC, and a non-real-time RIC other than the access network device (optionally, the non-real-time RIC may be located in an OAM or core network device) is included. As in the second possible implementation, the non-real-time RIC may be configured to perform model learning and / or inference, and / or as in the first possible implementation, the near-real-time RIC may obtain AI model information from the non-real-time RIC, obtain information about the network side and / or the terminal side from at least one of the CU, DU, and RU, and obtain an inference result using the information and the AI model. Optionally, the near-real-time RIC may be submitted the inference result to at least one of the CU, DU, and RU. Optionally, the CU and the DU may exchange the inference result. Optionally, the DU and the RU may exchange the inference result. For example, the near-real-time RIC submits the inference result to the DU, and the DU submits the inference result to the RU. For example, the near-real-time RIC is configured to train model A, and inference is performed using this model A. For example, the non-real-time RIC may be configured to train model B, and inference may be performed using model B. For example, the non-real-time RIC may be configured to train model C, and information about model C may be submitted to the near-real-time RIC, and the near-real-time RIC may perform inference using model C.
[0128] Fig. 4(b) is an exemplary diagram of a network architecture to which the method provided in the present disclosure can be applied. Compared with Fig. 4(a), in Fig. 4(b), the CU is separated into a CU-CP and a CU-UP.
[0129] 4(c) is an exemplary diagram of a network architecture to which the method provided in the present disclosure may be applied. As shown in FIG. 4(c), optionally, the access network device includes one or more AI entities, and the function of the AI entities is similar to that of the near-real-time RIC. Optionally, the OAM includes one or more AI entities, and the function of the AI entities is similar to that of the non-real-time RIC. Optionally, the core network device includes one or more AI entities, and the function of the AI entities is similar to that of the non-real-time RIC. When the OAM and the core network device each include an AI entity, the models obtained by training by the AI entities of the OAM and the core network device are different and / or the models used for inference are different.
[0130] In the present disclosure, the different models include, but are not limited to, at least one of the following differences: differences in the structural parameters of the model (e.g., the number of layers of the model, the connection relationships between the layers, the number of neurons in each layer, the activation functions of the neurons, the weights of the neurons, and / or the offsets of the neurons), the input parameters of the model, or the output parameters of the model.
[0131] FIG. 4(d) is an exemplary diagram of a network architecture to which the method provided in the present disclosure may be applied. Compared with FIG. 4(c), the access network device in FIG. 4(d) is separated into a CU and a DU. Optionally, the CU may include an AI entity, and the function of the AI entity is similar to that of the near real-time RIC. Optionally, the DU may include an AI entity, and the function of the AI entity is similar to that of the near real-time RIC. When the CU and the DU each include an AI entity, the models obtained by training by the AI entities of the CU and the DU are different, and / or the models used for inference are different. Optionally, the CU in FIG. 4(d) may be further separated into a CU-CP and a CU-UP. Optionally, one or more AI models may be disposed in the CU-CP. Optionally, one or more AI models may be disposed in the CU-UP. Optionally, in FIG. 4(c) or FIG. 4(d), the OAM of the access network device and the OAM of the core network device may be disposed separately and independently.
[0132] In the present disclosure, inference using one model may result in one parameter or multiple parameters. The learning processes of different models may be located on different devices or nodes, or may be located on the same device or node. The inference processes of different models may be located on different devices or nodes, or may be located on the same device or node.
[0133] It may be understood that in this disclosure, an AI entity or an access network device, etc. may perform some or all of the steps of this disclosure. These steps or operations are merely examples. The present disclosure may further perform other operations or variations of various operations. Furthermore, the steps may be performed in a different order than the order presented in this disclosure, and not all operations of the present disclosure may be performed.
[0134] In the present disclosure, when "less than" or "greater than" is used for comparison, "less than" can be replaced with "equal to or less than" and "greater than" can be replaced with "equal to or more than".
[0135] In various embodiments of the present disclosure, unless otherwise specified or there is no logical contradiction, the terms and / or descriptions in different examples are consistent and may be cross-referenced, and the technical features in different examples may be combined based on their internal logical relationships to form a new example.
[0136] It can be understood that various numbers in the examples of the present disclosure are used merely for distinction to facilitate description, and are not used to limit the scope of the examples of the present disclosure. The sequential numbers of the above processes do not imply an execution order, and the execution order of the processes shall be determined based on the functions and internal logic of the processes.
[0137] The network architecture and service scenarios described in this disclosure are intended to more clearly illustrate the technical solutions of this disclosure, and are not limitations on the technical solutions provided in this disclosure. Those skilled in the art can understand that with the evolution of network architectures and the emergence of new service scenarios, the technical solutions provided in this disclosure can also be applied to similar technical problems.
[0138] In this disclosure, the uplink and downlink directions are discussed separately, specifically, the procedures shown in Figures 5 to 11 relate to the downlink direction, and the procedure shown in Figure 12 relates to the uplink direction.
[0139] Optionally, before the method provided in the present disclosure is executed, capability exchange and / or model information exchange may be further executed between the terminal device and the access network device. Details may be shown in Figure 5. Figure 5 is a schematic flow chart of capability exchange and / or model information exchange according to the present disclosure. The following steps are included:
[0140] S501: A terminal device reports capability information to an access network device.
[0141] The capability information may include at least one of the following information, but is not limited to: whether the terminal device supports the execution of an artificial intelligence model, the supported artificial intelligence model type (e.g., a convolutional neural network (CNN), a recursive neural network (RNN), or a random forest model), the size of the memory space that can be used by the terminal device to store the artificial intelligence model, the computing capability information of the terminal device, which may be the computing capability for executing the artificial intelligence model, such as information such as the computing speed of the processor and / or the volume of data that can be processed by the processor, energy consumption information of the terminal device, such as the operating power consumption of the chip and / or the battery capacity, and / or hardware information of the terminal device, including, but not limited to, antenna setting information (the number and / or deflection direction of antennas) or the number of radio frequency channels.
[0142] The terminal device may actively report capability information or may report capability information upon receiving a request message from the access network device, which is not a limitation of the present disclosure.
[0143] S501 is an optional operation. For example, if the capability of the terminal device is agreed upon in a protocol, the method shown in FIG. 5 does not need to include S501.
[0144] S502: The access network device sends information about the first prediction model to the terminal device.
[0145] The access network device may transmit information about the first predictive model if it determines that the terminal device supports the execution of the artificial intelligence model. Alternatively, the access network device may not transmit information about the first predictive model. In other words, S502 is an optional operation. For example, the first predictive model may be agreed upon in a protocol, or the access network device and the terminal device download the first predictive model via the same download address.
[0146] In a first possible implementation, the information about the first predictive model may be information including model parameters, model structures, cascade relationships, and activation functions used to describe the first predictive model. The terminal device may determine the first predictive model based on the information about the first predictive model. In a second possible implementation, the information about the first predictive model may alternatively be computer code describing the first predictive model, and the terminal device may compile the computer code to obtain the first predictive model. In a third possible implementation, the information about the first predictive model may be a download address for downloading the first predictive model, and the terminal device may download the first predictive model based on the download address.
[0147] In the present disclosure, the prediction model, such as the first prediction model, the second prediction model, the third prediction model, or the fourth prediction model described below, may be an artificial intelligence model. Specifically, the prediction model may be a trained artificial intelligence model. The prediction model may be configured to predict a downlink signal having an optimal signal quality received by a terminal device. For example, the optimal signal quality may mean that the reference signal received power (RSRP) of the signal is the highest, or that the received signal-to-noise ratio of the signal is the highest, etc. This is not limited.
[0148] For example, FIG. 6 is a diagram of the structure of an artificial intelligence model according to the present disclosure. The type of the artificial intelligence model can be a neural network model, a random forest model, or a linear regression model, etc. This is not limited in the present disclosure. The artificial intelligence model includes a five-layer convolutional network and a two-layer fully connected network. This is merely an example and does not represent any limitation on the quantity of the convolutional layer network and the fully connected layer network, and the connection relationship between the convolutional layer network and the fully connected layer network. The convolutional layer network processes the acquired input and transmits the input to the fully connected layer network. The fully connected layer network finally outputs the result. For example, the input of the artificial intelligence model can be measurement information obtained based on a wireless signal, and the output of the artificial intelligence model is a codeword index or beam index corresponding to a predicted optimal wireless signal. The frequency domain is used as an example. The propagation form of the wireless signal on the channel is Y i =H i X i +N, where H i is the frequency domain channel response used when the i-th beam is transmitted, and X i is the radio signal transmitted using the i-th beam, N is the noise, and Y i is the received signal. i is an integer. The artificial intelligence model is i By using the measurement information of the received signal Y, it is possible to obtain a beam that transmits a wireless signal having the best signal quality or the highest signal strength, i.e., to determine the beam that is the best beam. i The measurement information is Y i The reference signal received power of H i , and / or the received signal Y i Using the time domain as an example, the propagation form of the wireless signal on the channel can be expressed as follows:
[0149]
number
[0150] where h i is the time-domain channel response when the i-th beam is transmitted, and x i is the radio signal transmitted using the i-th beam, n0 is the noise, and y i is the received signal, i is an integer,
[0151]
number
[0152] represents the time domain convolution. The artificial intelligence model is based on multiple received signals y i By using the measurement information of the received signal y, it is possible to obtain a beam that transmits a wireless signal having the best signal quality or the highest signal strength, i.e., to determine the beam that is the best beam. i The measurement information is i The reference signal received power of h i and / or the received signal y i It could be.
[0153] It is assumed that the codebook of the access network device includes C codewords, in other words, the complete codebook includes C codewords, and the C codewords correspond to C beams. In this disclosure, the scan pattern of the access network device includes M codewords in the codebook (M≦C, and M is greater than 0), and g(Y i ) is the received signal Y i In this case, the input format for the prediction model is
[0154]
number
[0155] In other words, the dimension of the input information may be M-dimensional information, the output of the prediction model may be a codeword index or a beam index corresponding to a predicted optimal wireless signal, and the dimension of the output information may be a one-dimensional scalar. Optionally, the method includes:
[0156]
number
[0157] may be used as an input to the prediction model after pre-processing is performed, and the output is a codeword index or beam index corresponding to the predicted optimal wireless signal, and the dimension of the output information may be a one-dimensional scalar. The pre-processing may include at least the following: real and imaginary parts are separated and used separately as inputs, real and imaginary parts are separated and spliced into 2M dimensional information, amplitude and phase are separated and used separately as inputs, amplitude and phase are separated and spliced into 2M dimensional information, and scaling, quantization, filtering, etc. are performed with scale factors. This is not limited to this.
[0158] In the present disclosure, the models trained by the network side, such as the first prediction model, the third prediction model, or the fourth prediction model, may be trained by the access network device, such as trained by the near real-time RIC, CU, or DU of the access network device. Alternatively, a pre-trained prediction model is obtained from a third-party network entity. For example, an AI entity performs the training, and the access network device obtains the trained prediction model from the AI entity. Optionally, the AI entity is located in the OAM or core network device. The specific training process of the prediction model is not limited in the present disclosure. For example, in the first prediction model, the training data required to train the prediction model is measurement information, and the training data is the Y i =H i Xi +N. The channel response H i The 3rd Generation Partnership Project rd The channel model may be generated using a channel model that meets the definition of the Generation Partnership Project (3GPP), may be collected in a real environment, or may be generated by another simulation platform, such as a ray tracking channel simulation platform, which is not a limitation of this disclosure.
[0159] In the present disclosure, multiple prediction models may be trained, with each prediction model corresponding to a scan pattern. For a terminal device, the access network device may configure a prediction model corresponding to the scan pattern based on the scan pattern configured for the terminal device.
[0160] In the present disclosure, the same artificial intelligence model may be located in the access network device and the terminal device.
[0161] In one implementation, the first prediction model is also deployed in the access network device, or an artificial intelligence model having the same parameters is deployed in the access network device and the terminal device. The artificial intelligence model having the same parameters is deployed in the access network device and the terminal device, which may mean that the model parameters, model structure, cascade relationship, activation function, etc. of the artificial intelligence model deployed in the access network device are completely the same as those of the artificial intelligence model deployed in the terminal device.
[0162] In another implementation, information such as model structure of the artificial intelligence model of the access network device may be different from that of the artificial intelligence model of the terminal device. However, when the input information is the same, the output information is also the same or the difference between the output information is less than or equal to a threshold. In other words, when the access network device and the terminal device agree on the same model input data format and input the same measurement information, the same prediction result can be output through each artificial intelligence model.
[0163] After the first prediction model is deployed in the terminal device, the access network device may transmit a downlink signal to the terminal device, so that the terminal device can input the estimated downlink measurement field information obtained by measuring the downlink signal into the first prediction model to determine whether the prediction result of the first prediction model is accurate. The following will be described in detail.
[0164] Figure 7 is a schematic flow diagram of a communication method according to the present disclosure. The procedure is described using an example of an interaction between a terminal device and an access network device. In the procedure shown in Figure 7, the same prediction model may be configured in the terminal device and the access network device, and the access network device may use the prediction model to predict a downlink signal with the highest received signal strength among the downlink signals received by the terminal device. The terminal device may check the prediction result of the configured prediction model, and may send a feedback to the access network device if the prediction result is inaccurate. Specifically, the method includes the following steps:
[0165] Optional S700: The access network device sends the indication information to the terminal device.
[0166] The display information may indicate at least one of the following information:
[0167] (1) Scan pattern information indicating a scan pattern. The scan pattern indicates one or more code words selected from a code book by an access network device and the order of selecting the code words. Since one code word corresponds to one beam, the scan pattern may also indicate one or more beams used by the access network device and the transmission order of those beams. Information such as the code words in the code book used by the access network device, the use order of the code words, the beams used, and the transmission order of the beams may be determined using the scan pattern. Optionally, the indication information may not include the scan pattern information. For example, the scan pattern information may be agreed upon in a protocol or set in another manner. This is not limited in the present disclosure.
[0168] (2) Input and output format information of the prediction model, indicating information such as the dimension of the input information and the dimension of the output information of the prediction model. For example, if the scan pattern indicates M codewords and the output is a codeword index corresponding to an optimal downlink signal, the dimension of the input information is M and the dimension of the output information is 1. Optionally, the display information may not include the input and output format information of the prediction model. For example, the input and output format information of the prediction model may be agreed upon in a protocol or set in another manner. This is not limited in the present disclosure.
[0169] (3) Test mode information indicating whether the terminal device enters a test mode or indicating the mode used by the access network device to transmit downlink signals.
[0170] In the present disclosure, the access network device may transmit downlink signals in two modes, a first mode and a second mode. The first mode may be referred to as a checking mode, and the second mode may be referred to as a non-checking mode. In a possible implementation, when transmitting a downlink signal in the first mode, the access network device may instruct the terminal device to enter the checking mode through the checking mode information, or when transmitting a downlink signal in the second mode, the access network device may instruct the terminal device to enter the non-checking mode through the checking mode information. In another possible implementation, the access network device may further configure a resource for the checking mode or the non-checking mode to make the terminal device enter the checking mode or the non-checking mode. For example, the resource for the checking mode is periodic or aperiodic. When the access network device periodically or aperiodically scans beams corresponding to all codewords in the codebook, the terminal device enters the checking mode. When the access network device scans beams corresponding to some codewords in the codebook, the terminal device enters the non-checking mode.
[0171] In the present disclosure, the access network device includes a, and the codebook includes one or more code words, each code word may be for precoding a signal, and the precoded signal is propagated in space based on a specific spatial directivity to form a beamforming effect, i.e., one code word corresponds to one beam or one data transmission direction.
[0172] In the check mode, all the codewords in the codebook are selected by the access network device, in other words, all the codewords in the codebook are used to transmit downlink signals, and the beams corresponding to all the codewords in the codebook are transmitted.
[0173] In the unchecked mode, the access network device may select some code words from the code book based on the scan pattern, i.e., the code words in the code book indicated by the scan pattern are for transmitting the downlink signal, and other code words are not for transmitting the downlink signal.
[0174] Take as an example that the codebook contains 64 codewords, as shown in Figure 8. In Figure 8, one grid represents one beam, one beam corresponds to one codeword, there are 64 beams in total, and different beams have different directions in space.
[0175] In the inspection mode, the access network device may sequentially transmit 64 downlink signals using 64 beams in the order of horizontal angles and then in the order of vertical angles, where one beam is for transmitting one downlink signal. Correspondingly, the terminal device sequentially receives the 64 downlink signals transmitted using the 64 beams in the same order.
[0176] In FIG. 8, an explanation is given using an example where the scan pattern shows 16 beams corresponding to 16 codewords. The 16 beams are represented by small black grids in FIG. 8. Establishing a coordinate system using the horizontal angle as the x-axis and the vertical angle as the y-axis, the position coordinates of the 16 beams are (1,1), (1,3), (1,5), (1,7), (3,1), (3,3), (3,5), (3,7), (5,1), (5,3), (5,5), (5,7), (7,1), (7,3), (7,5), and (7,7), respectively. In the non-checking mode, the access network device may sequentially transmit downlink signals using one or more beams indicated by the scan pattern in the order of the horizontal angle and then in the order of the vertical angle. Correspondingly, the terminal device sequentially receives the downlink signals in the same order.
[0177] In the above example, the order in which the access network device transmits the beam is merely an example and does not represent a restriction on the order. The access network device and the terminal device may alternatively use another order. This is not limited.
[0178] S701: An access network device transmits X downlink signals to a terminal device, and correspondingly, the terminal device receives X downlink signals from the access network device.
[0179] X may be an integer greater than 1. Optionally, different downlink signals of the X downlink signals may be transmitted in a time division manner and / or different downlink signals of the X downlink signals may be transmitted in a frequency division manner. For example, the access network device transmits different downlink signals at different frequencies using different beams and different antenna panels.
[0180] The type of the downlink signal is not limited. The downlink signal may be a downlink reference signal, for example, a channel state information downlink signal (CSI-RS). Alternatively, the downlink signal may be another type of signal, for example, a synchronization signal and a physical broadcast channel block (SSB). This is not limited in this disclosure. Optionally, the synchronization signal may include a primary synchronization signal (PSS) and a secondary synchronization signal (SSS). The SSB includes a PSS, an SSS, and a PBCH. The SSB may further include a demodulation reference signal (DMRS) of the PBCH.
[0181] For any downlink signal among the X downlink signals, when transmitting the downlink signal, the access network device precodes the downlink signal with one codeword. In other words, the downlink signal corresponds to one beam or one data transmission direction. The terminal device may receive the downlink signal with one codeword. The codeword used by the terminal device to receive the downlink signal may be determined by the terminal device. How to specifically determine the codeword is not limited in this disclosure.
[0182] When the access network device transmits the downlink signal in a check mode, the value of X may be equal to the number of codewords included in the codebook of the access network device. When the access network device transmits the downlink signal in a non-check mode, the value of X may be equal to the number of codewords (or the number of beams) indicated by the scan pattern.
[0183] S702: The terminal device determines X pieces of downlink measurement information based on the X downlink signals.
[0184] The terminal device may obtain one downlink measurement information based on one downlink signal. The downlink measurement information may be an RSRP obtained by measuring the downlink signal, or may be a received value of the downlink signal, or may be other information obtained by measuring the downlink signal.
[0185] If the display information of S700 instructs the terminal device to enter a non-testing mode, S703 and S704 may be executed.
[0186] S703: The terminal device sends X pieces of downlink measurement information to the access network device, and correspondingly, the access network device receives the X pieces of downlink measurement information.
[0187] S704: The access network device determines a first optimal downlink signal based on the X pieces of downlink measurement information.
[0188] The first optimal downlink signal is a downlink signal determined based on an output of a first prediction model, the input of the first prediction model being determined based on the X pieces of downlink measurement information.
[0189] Assume that the codebook of the access network device includes C code words, and the access network device can transmit C beams in different directions through the C code words, where C is an integer greater than 0. Since the access network device transmits downlink signals in an unchecked mode, in this case, X is less than or equal to C.
[0190] In this step, the first optimal downlink signal may refer to the downlink signal obtained by prediction that has the best signal quality received by the terminal device, and is not the downlink signal that has the best signal quality obtained by actual measurement. In other words, the access network device does not actually transmit C downlink signals, but only transmits X downlink signals. The first optimal downlink signal predicted by the access network device through the first prediction model may be one of the X downlink signals, or may not be one of the X downlink signals. The first optimal downlink signal obtained by prediction is one of the C downlink signals.
[0191] For example, referring to the above example of FIG. 8, assume that the X downlink signals transmitted by the access network device are transmitted through 16 beams indicated by the scan pattern. The first optimal beam corresponding to the first optimal downlink signal actually obtained by the access network device through prediction may not be any one of the 16 beams. For example, the predicted first optimal beam may be the beam with the position coordinates (6,5) in FIG. 8, i.e., the beam indicated by the arrow in FIG. 8.
[0192] After determining the first optimal downlink signal, the access network device may transmit downlink data to the terminal device via the first optimal beam corresponding to the first optimal downlink signal.
[0193] S704 may be further understood as follows: the access network device may determine a first optimal codeword (or a first optimal beam) based on X pieces of downlink measurement information, and the access network device may transmit downlink data to the terminal device via the first optimal codeword (or the first optimal beam). Specifically, the access network device may obtain a codeword corresponding to a downlink signal with the best signal quality by model-based prediction.
[0194] According to the above method, the access network device only needs to transmit a few downlink signals, and can obtain the optimal downlink signals through prediction based on downlink measurement information to determine the optimal beam to be used for communication, thereby quickly completing beam alignment with low air interface overhead.
[0195] If the display information of S700 instructs the terminal device to enter the inspection mode, the following procedures may be performed.
[0196] S705: The terminal device determines a first downlink signal and a second downlink signal.
[0197] The first downlink signal and the second downlink signal are downlink signals of the X downlink signals, the first downlink signal being a downlink signal obtained by prediction based on a portion of the X downlink measurement information, and the second downlink signal being a downlink signal determined based on the X downlink measurement information.
[0198] Specifically, the first downlink signal is a downlink signal determined based on the output of the first prediction model, and the input of the first prediction model is determined based on Y downlink measurement information of the X downlink measurement information, where Y is an integer less than or equal to X and greater than 0. Optionally, the Y downlink measurement information may be determined based on a scan pattern. For example, the Y downlink signals corresponding to the Y downlink measurement information are downlink signals transmitted through Y beams indicated by the scan pattern. In this case, the value of Y is equal to the quantity of beams indicated by the scan pattern.
[0199] In the present disclosure, the downlink signal, the codeword, and the beam correspond to each other, one downlink signal corresponds to one codeword, the codeword is for precoding the downlink signal, and the transmission direction of the precoded signal is the direction indicated by the beam. The prediction model obtains the downlink signal by prediction, which can be alternatively replaced with the case where the prediction model obtains the codeword by prediction or obtains the beam by prediction. Similarly, the scan pattern indicates a scanned signal, which can be alternatively replaced with the case where the scan pattern indicates a scanned codeword or a scanned beam.
[0200] The second downlink signal is a downlink signal corresponding to the downlink measurement information having the best corresponding signal quality among the X downlink measurement information. For example, if the downlink measurement information is RSRP, a downlink signal corresponding to the RSRP having the maximum value among the X RSRPs may be used as the second downlink signal.
[0201] The terminal device may determine, based on the second downlink signal, whether the first downlink signal predicted based on the first prediction model is accurate. For details, see the following description.
[0202] S706: When the first downlink signal and the second downlink signal satisfy a first condition, the terminal device transmits first information to the access network device, and in response, the access network device receives the first information from the terminal device.
[0203] If the first downlink signal and the second downlink signal satisfy the first condition, the terminal device may consider that the prediction result of the first prediction model is inaccurate and the first prediction model needs to be updated. In this case, in a possible implementation, the first information may indicate the second downlink signal and Y downlink measurement information, and the above information is for updating the first prediction model. How to specifically update the first prediction model will be described in detail below. In another possible implementation, the first information indicates that the first prediction model needs to be updated.
[0204] The first condition may include any one of the following:
[0205] The first downlink signal and the second downlink signal are not the same downlink signal, or The first downlink signal and the second downlink signal are not the same downlink signal, and the difference information between the first downlink signal and the second downlink signal is greater than a first threshold, which may be set by the access network device, or agreed upon in a protocol, or determined by the terminal device.
[0206] In the present disclosure, in a first implementation, the difference information between the first downlink signal and the second downlink signal is an absolute value of a difference between the first downlink measurement information corresponding to the first downlink signal and the second downlink measurement information corresponding to the second downlink signal. For example, if the value of the first downlink measurement information is R1 and the value of the second downlink measurement information is R2, the difference information may be |R1-R2|.
[0207] In a second implementation, the difference information between the first downlink signal and the second downlink signal is an absolute value of a difference between the first position information corresponding to the first downlink signal and the second position information corresponding to the second downlink signal.
[0208] In the present disclosure, the difference between the first location information and the second location information may be a Euclidean distance between the first location information and the second location information.
[0209] For example, as shown in FIG. 9, the codebook includes 64 code words, and there are 64 beams in total. One beam corresponds to one code word, and there are 64 beams in total. Since one beam points to one transmission direction, the 64 beams can represent 64 transmission directions. It is assumed that the 64 beams are arranged in a grid shape, and the horizontal angle includes 8 rows and the vertical angle includes 8 columns. When viewed from the cross section of the transmission direction, the 64 beams may form 64 grids, as shown in FIG. 9, and one grid in FIG. 9 represents one beam. A coordinate system is established by using the horizontal angle as the x-axis and the vertical angle as the y-axis. In FIG. 9, an example is taken in which the position information of the first downlink signal is (6,6) and the position information of the second downlink signal is (8,8). In this case, the Euclidean distance between the first position information corresponding to the first downlink signal and the second position information corresponding to the second downlink signal may satisfy the following form:
[0210]
number
[0211] The location information corresponding to the downlink signal may be indicated to the terminal device by the access network device or may be pre-agreed upon.
[0212] S707: When the first downlink signal and the second downlink signal satisfy a second condition, the terminal device transmits second information to the access network device, and in response, the access network device receives the second information from the terminal device.
[0213] If the first downlink signal and the second downlink signal satisfy the second condition, the terminal device may consider that the prediction result of the first prediction model is accurate and the first prediction model does not need to be updated. In this case, in a possible implementation, the second information may indicate the second downlink signal. In another possible implementation, the second information may indicate that the first prediction model is reliable or that the first prediction model does not need to be updated.
[0214] The second condition may include any one of the following:
[0215] the first downlink signal and the second downlink signal are the same downlink signal, or The first downlink signal and the second downlink signal are not the same downlink signal, but the differential information between the first downlink signal and the second downlink signal is less than or equal to a first threshold.
[0216] In the present disclosure, S706 and S707 may alternatively be replaced with step A or step B below.
[0217] Step A: The terminal device transmits the first information or the second information to the access network device according to the first downlink signal and the second downlink signal. In response, the access network device receives the first information or the second information from the terminal device.
[0218] The first information may be transmitted if the first downlink signal and the second downlink signal satisfy a first condition, and the second information may be transmitted if the first downlink signal and the second downlink signal satisfy a second condition.
[0219] The first information indicates that the first prediction model needs to be updated, or indicates the second downlink signal and the Y downlink measurement information. The second information indicates that the first prediction model is reliable, or indicates that the first prediction model does not need to be updated, or indicates that the second information does not include the Y downlink measurement information. The second information may further indicate the second downlink signal.
[0220] Step B: The terminal device transmits first information to the access network device according to the first downlink signal and the second downlink signal, or does not transmit the first information. Correspondingly, the access network device detects the first information from the terminal device.
[0221] The description of the first information is the same as that in step A. If the first information is detected, the access network device considers that the first prediction model needs to be updated, and if the first information is not detected, the access network device considers that the first prediction model does not need to be updated.
[0222] In the present disclosure, the first prediction model may be updated by the terminal device, or the first prediction model may be updated by the access network device, which will be described separately below.
[0223] When the terminal device updates the first prediction model, the following procedures may be performed.
[0224] S708: The terminal device trains a first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a second prediction model.
[0225] The terminal device may use the Y downlink measurement information and the second downlink signal as a group of training data to perform update training on the first prediction model. The trained model is referred to as a second prediction model. The terminal device may update the first prediction model to the second prediction model. A specific method of updating the first prediction model to the second prediction model is not limited in the present disclosure and may be, for example, using a backpropagation gradient update algorithm.
[0226] After obtaining the second prediction model, the terminal device may represent the second prediction model in multiple manners.
[0227] S709: The terminal device sends third information to the access network device, and correspondingly, the access network device receives the third information from the terminal device.
[0228] In a possible implementation, the third information indicates a second prediction model. Specifically, the third information may include information including model parameters, model structure, cascade relationships, and activation functions, etc., used to describe the second prediction model. The access network device may determine the second prediction model based on the third information. For example, the second prediction model may include a multi-layer convolutional layer network and a multi-layer fully connected layer network, and each layer may include one or more neurons. Correspondingly, the model parameters may include information such as weights corresponding to each neuron in the second prediction model. The model structure may include the quantity of the convolutional layer network and the quantity of the fully connected layer network in the second prediction model, and the connection relationship between the convolutional layer network and the fully connected layer network, etc.
[0229] The third information may also include computer code describing the second prediction model, and the access network device may perform a compilation based on the computer code to obtain the second prediction model.
[0230] In another possible implementation, the third information indicates information about a second prediction model, and the information about the second prediction model includes variation information between the second prediction model and the first prediction model. Specifically, the third information may include this variation information. The variation information refers to the variation between information including model parameters, model structures, cascade relationships, and activation functions, etc., used to describe the second prediction model and information including model parameters, model structures, cascade relationships, and activation functions, etc., used to describe the first prediction model.
[0231] In another possible implementation, the third information indicates information about the second prediction model, and the information about the second prediction model includes gradient information of the second prediction model. Specifically, the third information may include a gradient calculation value when the first prediction model is updated to the second prediction model using a gradient descent method.
[0232] S710: The access network device updates the first prediction model to the second prediction model based on the third information.
[0233] According to the above method, the terminal device may perform prediction using a first prediction model locally configured. If the terminal device determines that the predicted first downlink signal does not match the second downlink signal determined based on the actual downlink measurement information due to a change in channel environment or a mismatch between the prediction model and the channel environment, the terminal device may calibrate the first prediction model, in other words, retrain the first prediction model to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0234] In the present disclosure, in the first scenario, if the first prediction model is updated, the terminal device indicates the second prediction model to the access network device, i.e., executes S709.
[0235] In the second scenario, the terminal device may determine whether the variation information between the second prediction model and the first prediction model is greater than a second threshold, and transmit the third information if the variation information is greater than or equal to the second threshold, and not transmit the third information if the variation information is less than the second threshold. To avoid loss of generality, this disclosure uses the procedure shown in Figure 10 as an example to describe how the terminal device determines whether to transmit the third information.
[0236] S1001: A terminal device trains a prediction model to be updated to obtain an updated prediction model.
[0237] The initial value of the prediction model to be updated is the first prediction model, and the prediction model to be updated is the prediction model currently arranged in the terminal device. The updated prediction model is obtained by training in S708. Assuming that the prediction model to be updated is the first prediction model, the updated prediction model obtained in this case is the second prediction model.
[0238] Suppose several terminal devices have the same prediction model to be updated. These terminal devices train the prediction model to be updated, and feed back the training gradient information, the updated prediction model, or the variation between the updated prediction model and the prediction model to be updated to the access network device. The access network device generates a new fused prediction model based on the feedback of the terminal devices, and then sends the new fused prediction model to these terminal devices to update the first prediction model.
[0239] S1002: The terminal device determines variation information between the updated prediction model and an initially updated prediction model.
[0240] The initial value of the initially updated prediction model is the first prediction model, and the initially updated prediction model is the prediction model previously indicated to the access network device via the third information.
[0241] The updated forecast model is the second forecast model, and it is assumed herein that the initially updated forecast model is the first forecast model.
[0242] S1003: If the variation information is greater than or equal to the second threshold, the terminal device sends third information to the access network device.
[0243] The second threshold may be set by the access network device, or may be agreed upon in a protocol, or may be determined by the terminal device.
[0244] S1004: If the variation information is less than the second threshold, the terminal device does not transmit the third information.
[0245] When training the prediction model to be updated next time, the terminal device may execute the steps from S1001 to S1004 again.
[0246] Since the fluctuation information is less than the second threshold, it can be considered that the updated prediction model is not significantly different from the prediction model before the update, the performance of the updated prediction model is close to that of the prediction model before the update, and the prediction result of the updated prediction model is close to that of the prediction model before the update. Therefore, the updated prediction model may not be indicated to the access network device.
[0247] With reference to the procedure description shown in Figure 10, in one example, it is assumed that the first prediction model is a prediction model A. The terminal device may perform the following steps.
[0248] Step 1: Train predictive model A to obtain predictive model B.
[0249] Step 2: Determine whether the variation information between prediction model A and prediction model B is greater than or equal to a second threshold. If the variation information between prediction model A and prediction model B is greater than or equal to the second threshold, proceed to step 3; if the variation information between prediction model A and prediction model B is less than the second threshold, proceed to step 4.
[0250] Step 3: Transmit third information, where the third information in this step indicates information about the prediction model B.
[0251] Step 4: Skip sending the third information.
[0252] If the variation information of step 2 is less than the second threshold, the terminal device does not send the third information this time.
[0253] Further, assuming that the terminal device needs to further train the prediction model B to obtain the prediction model C, the terminal device may execute the procedure shown in FIG. 10 again.
[0254] Step 5: Train predictive model B to obtain predictive model C.
[0255] Step 6: Determine whether the variation information between the prediction model C and the prediction model A is greater than or equal to a second threshold. If the variation information between the prediction model C and the prediction model A is greater than or equal to the second threshold, proceed to step 7; if the variation information between the prediction model C and the prediction model A is less than the second threshold, proceed to step 8.
[0256] Since the prediction model C is obtained by training the prediction model B, but the prediction model A is located in the access network device and the terminal device does not present the prediction model A to the access network device, it needs to be determined whether the variation information between the latest prediction model in the terminal device and the prediction model in the access network device is greater than or equal to a second threshold.
[0257] Step 7: Transmit the third information, where the third information in this step indicates information about the prediction model C.
[0258] Step 8: Skip sending the third information.
[0259] According to the above procedure, if the variation information between the updated prediction model and the prediction model previously indicated to the access network device is less than a second threshold, the updated prediction model may not be indicated to the access network device, thereby avoiding frequently indicating the updated prediction model to the access network device, reducing signaling overhead and improving system efficiency.
[0260] In the present disclosure, the access network device may also update the first prediction model. When the access network device updates the first prediction model, the following steps may be performed.
[0261] Step 1: The access network device trains a first prediction model based on Y downlink measurement information and a second downlink signal to obtain a third prediction model.
[0262] The access network device may use the Y downlink measurement information and the second downlink signal as a group of training data to perform update training on the first prediction model. The trained model is a third prediction model. The second prediction model obtained by the terminal device through the training of S708 and the third prediction model obtained by the access network device through the training of step 1 may be the same model or different models. In the present disclosure, for ease of explanation, the second prediction model and the third prediction model are distinguished, but this does not necessarily mean that the two models are different models.
[0263] The access network device may update the first prediction model to a third prediction model.
[0264] After obtaining the third prediction model, the access network device may represent the third prediction model in multiple manners.
[0265] In a first implementation, the access network device may directly indicate the third prediction model to the terminal device, i.e., perform step 2.
[0266] Step 2: The access network device sends the fourth information to the terminal device, and correspondingly, the terminal device receives the fourth information from the access network device.
[0267] The fourth information indicates the third prediction model or indicates information about the third prediction model, and the information about the third prediction model includes variation information between the third prediction model and the first prediction model. For the specific content of the fourth information, please refer to the description of the third information. In this specification, the details will not be described again.
[0268] In the second implementation, the access network device may determine whether the variation information between the third prediction model and the first prediction model is greater than a second threshold, and send the fourth information when the variation information is greater than or equal to the second threshold, and not send the fourth information when the variation information is less than the second threshold. The specific process is similar to the procedure shown in Figure 10, and the details are not repeated in this specification.
[0269] Optionally, the access network device may not transmit the fourth information. If the first downlink signal and the second downlink signal of the terminal device satisfy the first condition, the terminal device and the access network device may separately retrain the first prediction model using the second downlink signal and the Y downlink measurement information. In this manner, the prediction models stored in the two devices can be kept consistent with each other with little signaling overhead.
[0270] According to the above method, when the terminal device determines that the predicted first downlink signal does not match the second downlink signal determined based on the actual downlink measurement information due to a change in the channel environment or a mismatch between the prediction model and the channel environment, the terminal device may feed back the second downlink signal and Y pieces of downlink measurement information to the access network device, so that the access network device may calibrate the first prediction model based on the second downlink signal and the Y pieces of downlink measurement information, i.e., retrain the first prediction model, to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0271] In the present disclosure, the prediction model may alternatively not be located in the terminal device, but only in the access network device. The access network device checks the prediction model based on the information fed back by the terminal device. Specifically, Figure 11 is a schematic flow chart of a communication method according to the present disclosure. The method includes the following steps:
[0272] Optional S1100: The access network device sends the indication information to the terminal device.
[0273] For the specific content of the display information, please refer to the description of S700. In this specification, the details will not be described again.
[0274] S1101: An access network device transmits X downlink signals to a terminal device.
[0275] S1102: The access network device receives X pieces of downlink measurement information from the terminal device.
[0276] The X downlink measurement information is determined based on the X downlink signals.
[0277] S1103: The access network device determines a first downlink signal and a second downlink signal.
[0278] The first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, an input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information, and the second downlink signal is determined based on the X downlink measurement information.
[0279] For the specific content of S1103, please refer to the explanation of S705. In this specification, the details will not be described again.
[0280] S1104: If the first downlink signal and the second downlink signal satisfy a first condition, the access network device trains a first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model.
[0281] For the specific content of S1104, please refer to the explanation of S706. In this specification, the details will not be described again.
[0282] Further, the access network device updates the first prediction model to a third prediction model.
[0283] S1105: If the first downlink signal and the second downlink signal satisfy a second condition, the access network device keeps the first prediction model unchanged.
[0284] For the specific content of S1105, please refer to the explanation of S707. In this specification, the details will not be described again.
[0285] According to the above method, only the first prediction model needs to be disposed in the access network device. When the access network device determines that the predicted first downlink signal does not match the second downlink signal determined based on the actual downlink measurement information due to a change in channel environment or a mismatch between the prediction model and the channel environment, the access network device may calibrate the first prediction model based on the second downlink signal and the Y downlink measurement information, i.e., retrain the first prediction model to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0286] The method provided in the present disclosure is also applicable to signal transmission in the uplink direction. For example, in the uplink direction, when communication is required, the access network device scans the candidate codewords in the codebook in a specific order. That is, the access network device receives the uplink signal from the terminal device using the beam corresponding to each of the candidate codewords in the codebook. The access network device finally selects the beam corresponding to the uplink signal with the highest received signal strength as the optimal beam, and receives the uplink data from the terminal device using this beam.
[0287] In the present disclosure, a prediction model may be deployed in an access network device to predict the uplink signal with the highest received signal strength based on only a small number of uplink signals, thereby avoiding the case where the terminal device needs to transmit a large number of uplink signals to determine the uplink signal with the highest signal strength, thereby reducing resource overhead, reducing the energy consumption of the terminal, and improving communication efficiency.
[0288] 12 is a schematic flow diagram of a communication method according to the present disclosure. The method includes the following steps:
[0289] Optional S1200: The access network device sends mode indication information to the terminal device.
[0290] In the present disclosure, the terminal device may transmit uplink signals in two modes: a third mode and a fourth mode. The third mode may also be referred to as a checking mode, and the fourth mode may also be referred to as a non-checking mode.
[0291] In a possible implementation, the access network device may indicate to the terminal device via mode indication information whether to enter the inspection mode. In another possible implementation, the time period during which the terminal device enters the inspection mode may be agreed upon in a protocol or pre-agreed by the access network device and the terminal device via signaling. The terminal device transmits uplink signals in the inspection mode during this time period and transmits uplink signals in the non-inspection mode during another time period. Alternatively, the time period during which the terminal device enters the non-inspection mode may be agreed upon in a protocol or pre-agreed by the access network device and the terminal device via signaling. The terminal device transmits uplink signals in the non-inspection mode during this time period and transmits uplink signals in the inspection mode during another time period.
[0292] Before the terminal device transmits an uplink signal using the first mode, the access network device may instruct the terminal device to enter a checking mode via the checking mode information. Before the terminal device transmits a downlink signal using the second mode, the access network device may instruct the terminal device to enter a non-checking mode via the checking mode information.
[0293] When the terminal device enters the inspection mode, the terminal device needs to transmit X uplink signals in one transmission period. When the terminal device does not enter the inspection mode, in other words, when the terminal device enters the non-inspection mode, the terminal device needs to transmit Y uplink signals in one transmission period.
[0294] The uplink signal may be an uplink reference signal. The specific signal class is not limited in the present disclosure. X is an integer greater than 1, and Y is an integer less than X and greater than 0. The value of X may be equal to the quantity of codewords included in the codebook of the access network device. For ease of understanding, the present disclosure uses X and Y in both uplink and downlink processes for description. In practice, when the method of the present disclosure is used in both uplink and downlink processes, the value of X in the uplink process may be the same as or different from the value of X in the downlink process. This is not limited. Similarly, the value of Y in the uplink process may be the same as or different from the value of Y in the downlink process. This is not limited. This rule is also applicable to other similar cases, and the details will not be repeated.
[0295] When the terminal device enters into a non-checking mode, the terminal device needs to send Y uplink signals to the access network device in one transmission period. Correspondingly, the access network device determines Y uplink measurement information according to the Y uplink signals, and predicts an optimal uplink signal according to the Y uplink measurement information. The specific process is not described again.
[0296] When the terminal device enters the test mode, the following procedures may be performed.
[0297] S1201: A terminal device transmits X uplink signals to an access network device.
[0298] S1202: The access network device receives X uplink signals from a terminal device, and determines X uplink measurement information based on the X uplink signals.
[0299] The access network device may obtain one uplink measurement information based on one uplink signal. The uplink measurement information may be an RSRP obtained by measuring the uplink signal, or may be the uplink signal, or may be other information obtained by measuring the uplink signal.
[0300] S1203: The access network device determines a first uplink signal and a second uplink signal.
[0301] The first uplink signal and the second uplink signal are uplink signals among the X uplink signals, the first uplink signal is an uplink signal obtained by prediction based on some uplink measurement information among the X uplink measurement information, and the second uplink signal is an uplink signal determined based on the X uplink measurement information.
[0302] Specifically, the first uplink signal is an uplink signal determined based on an output of a fourth prediction model, an input of the fourth prediction model is determined based on Y pieces of uplink measurement information out of the X pieces of uplink measurement information, and the second uplink signal is determined based on the X pieces of uplink measurement information.
[0303] S1204: If the first uplink signal and the second uplink signal satisfy a third condition, the access network device trains a fourth prediction model based on the Y uplink measurement information and the second uplink signal to obtain a fifth prediction model.
[0304] Further, the access network device may update the fourth prediction model to a fifth prediction model.
[0305] The third condition may include any one of the following:
[0306] The first uplink signal and the second uplink signal are not the same uplink signal; and The first uplink signal and the second uplink signal are not the same uplink signal, and the differential information between the first uplink signal and the second uplink signal is greater than a third threshold. The third threshold may be set by the access network device, or may be agreed upon in a protocol, and may be determined by the terminal device.
[0307] In a possible implementation, the difference information between the first uplink signal and the second uplink signal is an absolute value of a difference between the first uplink measurement information corresponding to the first uplink signal and the second uplink measurement information corresponding to the second uplink signal.
[0308] In another possible implementation, the difference information between the first uplink signal and the second uplink signal is an absolute value of the difference between the position information corresponding to the first uplink signal and the position information corresponding to the second uplink signal.
[0309] For the specific content of the difference information, please refer to the explanation of S707. In this specification, the details will not be described again.
[0310] S1205: If the first uplink signal and the second uplink signal satisfy a second condition, the access network device keeps the fourth prediction model unchanged.
[0311] The fourth condition may include any one of the following:
[0312] the first uplink signal and the second uplink signal are the same uplink signal; and The first uplink signal and the second uplink signal are not the same uplink signal, and the differential information between the first uplink signal and the second uplink signal is less than or equal to a third threshold.
[0313] According to the above method, the fourth prediction model is disposed in the access network device. When the access network device determines that the predicted first uplink signal is inconsistent with the second uplink signal determined based on the actual uplink measurement information due to the change of the channel environment or the mismatch between the prediction model and the channel environment, the access network device may retrain the fourth prediction model based on the second uplink signal and the Y uplink measurement information to obtain an updated prediction model, and thus the performance of the prediction model can be continuously improved.
[0314] In the above embodiments provided in the present application, the methods provided in the embodiments of the present application are described from the perspective of interaction between devices. To implement the functions of the methods provided in the embodiments of the present application, the access network device or the terminal device may include a hardware structure and / or a software module, and implement the functions in the form of a hardware structure, a software module, or a combination of a hardware structure and a software module. Whether the functions among the above functions are performed using a hardware structure, a software module, or a combination of a hardware structure and a software module depends on the specific application and design constraints of the technical solution.
[0315] In the embodiment of the present application, the division of modules is an example and is merely a logical division of functions. In actual implementation, other division methods may be used. Furthermore, the functional modules of the embodiments of the present application may be integrated into one processor, or may exist physically independently, or two or more modules may be integrated into one module. The integrated module may be implemented in the form of hardware, or may be implemented in the form of a software functional module.
[0316] Similar to the above concept, as shown in FIG. 13, an embodiment of the present application further provides a communication device 1300 configured to implement the functions of an access network device or a terminal device in the above method. The form of the communication device is not limited, and the communication device may be a hardware structure, a software module, or a combination of a hardware structure and a software module. For example, the device may be a software module or a chip system. In an embodiment of the present application, the chip system may include a chip, or may include a chip and another individual component. The device 1300 may include a processing unit 1301 and a communication unit 1302.
[0317] In embodiments of the present application, the communication unit may also be referred to as a transceiver unit and may include a transmitting unit and / or a receiving unit, respectively configured to perform the transmitting and receiving steps performed by the access network device, the AI entity, or the terminal device in the above method embodiments.
[0318] Hereinafter, the communication device provided in the embodiment of the present application will be described in detail with reference to Figures 13 and 14. It should be understood that the description of the device embodiment corresponds to the description of the method embodiment. Therefore, for contents that are not described in detail, please refer to the method embodiment. For the sake of brevity, the details will not be described again in this specification.
[0319] The communication unit may also be referred to as a transceiver device. The processing unit may also be referred to as a processing module, a processing device, or the like. Optionally, a device configured to implement a receiving function in the communication unit 1302 may also be considered as a receiving unit, and a device configured to implement a transmitting function in the communication unit 1302 may also be considered as a transmitting unit. In other words, the communication unit 1302 includes a receiving unit and a transmitting unit. In some cases, the communication unit may be implemented as a pin, a transceiver machine, a transceiver, or a transceiver circuit, or the like. In some cases, the processing unit may be implemented as a processor or a processing board, or the like. In some cases, the receiving unit may be implemented as a pin, a receiver machine, a receiver, or a receiving circuit, or the like. In some cases, the transmitting unit may be implemented as a pin, a transmitter machine, a transmitter, or a transmitting circuit, or the like.
[0320] When the communication device 1300 performs the function of the terminal device in the procedure shown in FIG. 7 of the above embodiment, The processing unit is configured to determine X pieces of downlink measurement information based on X downlink signals from the access network device, where X is an integer greater than 1; the processing unit is configured to determine a first downlink signal and a second downlink signal, where the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, an input of the first prediction model is determined based on Y pieces of downlink measurement information among the X downlink measurement information, and the second downlink signal is determined based on the X downlink measurement information, where Y is an integer less than or equal to X and greater than 0; The communication unit is configured to send first information to the access network device when the first downlink signal and the second downlink signal satisfy a first condition, where the first information indicates the second downlink signal and the Y pieces of downlink measurement information, or the first information indicates that the first prediction model needs to be updated.
[0321] When the communication device 1300 performs the function of the access network device in the procedure shown in FIG. 7 of the above embodiment, The processing unit is configured to send X downlink signals to the terminal device via the communication unit, where X is an integer greater than 1, and receive first information from the terminal device, where the first information indicates the second downlink signal and Y downlink measurement information among the X downlink measurement information, or the first information indicates that the first prediction model needs to be updated; The second downlink signal is determined based on X pieces of downlink measurement information, and the X pieces of downlink measurement information are determined based on the X pieces of downlink signals, where Y is an integer less than or equal to X and greater than 0.
[0322] When the communication device 1300 performs the function of the access network device in the procedure shown in FIG. 11 of the above embodiment, The communication unit is configured to transmit X downlink signals to a terminal device and receive X downlink measurement information from the terminal device, where the X downlink measurement information is determined based on the X downlink signals, where X is an integer greater than 1; The processing unit is configured to determine a first downlink signal and a second downlink signal, where the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, an input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information, and the second downlink signal is determined based on the X downlink measurement information, where Y is an integer less than or equal to X and greater than 0, and the processing unit is configured to train the first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model if the first downlink signal and the second downlink signal satisfy a first condition, and update the first prediction model to the third prediction model.
[0323] When the communication device 1300 performs the function of the access network device in the procedure shown in FIG. 12 of the above embodiment, The communication unit is configured to receive X uplink signals from a terminal device and determine X uplink measurement information based on the X uplink signals, where X is an integer greater than 1; The processing unit is configured to determine a first uplink signal and a second uplink signal, where the first uplink signal and the second uplink signal are uplink signals among the X uplink signals, the first uplink signal is an uplink signal determined based on an output of a fourth prediction model, an input of the fourth prediction model is determined based on Y pieces of uplink measurement information among the X pieces of uplink measurement information, and the second uplink signal is determined based on the X pieces of uplink measurement information, where Y is an integer less than X and greater than 0, and the processing unit is configured to train the fourth prediction model based on the Y pieces of uplink measurement information and the second uplink signal if the first uplink signal and the second uplink signal satisfy a third condition to obtain a fifth prediction model, and update the fourth prediction model to the fifth prediction model.
[0324] When the communication device 1300 performs the functions of the terminal device in the above embodiment, The processing unit is configured to determine X downlink measurement information based on the X downlink signals from the access network device, where X is an integer greater than 1; The processing unit is configured to determine a first downlink signal and a second downlink signal, where the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of a first prediction model, an input of the first prediction model is determined based on Y downlink measurement information among the X downlink measurement information, and the second downlink signal is determined based on the X downlink measurement information, Y is an integer less than or equal to X and greater than 0; The communication unit is configured to send first information or second information to the access network device based on the first downlink signal and the second downlink signal, where the first information indicates that the first prediction model needs to be updated, or the first information indicates the second downlink signal and Y pieces of downlink measurement information, and the second information indicates that the first prediction model is reliable, or the second information indicates that the first prediction model does not need to be updated, or the second information indicates that the second information does not include the Y pieces of downlink measurement information.
[0325] When the communication device 1300 performs the function of the access network device in the above embodiment, The communication unit is configured to transmit X downlink signals to the terminal device, where X is an integer greater than 1; the communication unit is configured to receive first information or second information from the terminal device, where the first information indicates a second downlink signal and Y downlink measurement information of the X downlink measurement information, or the first information indicates that the first prediction model needs to be updated, the second information indicates a second downlink signal, or the second information indicates that the first prediction model does not need to be updated, the second downlink signal is determined based on the X downlink measurement information, and the X downlink measurement information is determined based on the X downlink signals, and Y is an integer less than or equal to X and greater than 0.
[0326] The above is merely an example. The processing unit 1301 and the communication unit 1302 may further perform other functions. For further detailed description, please refer to the relevant description of the embodiment of the method shown in FIG. 7, FIG. 11, or FIG. 12. Details will not be repeated in this specification.
[0327] Figure 14 is a diagram showing a communication device 1400 according to an embodiment of the present application. The device shown in Figure 14 may be a hardware circuit implementation of the device shown in Figure 13. This communication device is applicable to the above flow chart and performs the functions of a terminal device or an access network device in the above method embodiment. For ease of explanation, Figure 14 shows only the main components of the communication device.
[0328] As shown in FIG. 14, the communication device 1400 includes a processor 1410 and an interface circuit 1420. The processor 1410 and the interface circuit 1420 are coupled to each other. It may be understood that the interface circuit 1420 may be a transceiver, a pin, an interface circuit, or an input / output interface. Optionally, the communication device 1400 may further include a memory 1430 configured to store instructions to be executed by the processor 1410, to store input data required for execution of the instructions by the processor 1410, or to store data generated after the processor 1410 executes the instructions. Optionally, a portion or the entirety of the memory 1430 may be located within the processor 1410.
[0329] When the communications device 1400 is configured to implement the method shown in FIG. 7, FIG. 11 or FIG. 12, the processor 1410 is configured to implement the functions of the processing unit 1301 and the interface circuit 1420 is configured to implement the functions of the communications unit 1302.
[0330] When the communication device is a chip used in a terminal device, the chip in the terminal device implements the functions of the terminal device in the above method embodiment. The chip in the terminal device receives information from another module (e.g., a radio frequency module or an antenna) in the terminal device, where the information is transmitted from the access network device to the terminal device. Alternatively, the chip in the terminal device transmits information to another module (e.g., a radio frequency module or an antenna) in the terminal device, where the information is transmitted from the terminal device to the access network device.
[0331] When the communication device is a chip used in an access network device, the chip in the access network device implements the function of the access network device in the above method embodiment. The chip in the access network device receives information from another module (e.g., a radio frequency module or an antenna) in the access network device, where the information is transmitted from the terminal device to the access network device. Alternatively, the chip in the access network device transmits information to another module (e.g., a radio frequency module or an antenna) in the access network device, where the information is transmitted from the access network device to the terminal device.
[0332] As mentioned above, the method of FIG. 7, FIG. 11, or FIG. 12 takes the exchange between the access network device and the terminal device as an example. If the AI entity is located outside the access network device and is an independent module or network element, the terminal device or the access network device may forward the second downlink signal and the Y downlink measurement information to the AI entity, and the AI entity trains a first prediction model based on the second downlink signal and the Y downlink measurement information. The AI entity may indicate information about the trained second prediction model to the terminal device or the access network device. Other processes are also similar to the procedures of FIG. 7, FIG. 11, or FIG. 12. Therefore, the above description of the access network device may also apply to the AI entity.
[0333] It may be understood that a processor in this disclosure may be a central processing unit, another general-purpose processor, a digital signal processor, an application specific integrated circuit, a field programmable gate array or other programmable logic device, a transistor logic device, a hardware component, or any combination thereof. A general-purpose processor may be a microprocessor or any general processor, etc.
[0334] Memory in this disclosure may be random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, registers, hard disk, removable hard disk, or any other form of storage medium known in the art.
[0335] Those skilled in the art may understand that the present disclosure may be provided as a method, a system, or a computer program product. Thus, the present application may take the form of a hardware-only embodiment, a software-only embodiment, or an embodiment comprising a combination of software and hardware. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk memory, optical memory, etc.) that contains computer-usable program code.
[0336] The present application has been described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the present application. It will be understood that computer program instructions can be used to implement each process and / or each block of the flowcharts and / or block diagrams, and combinations of the processes and / or blocks of the flowcharts and / or block diagrams. These computer program instructions can be provided to a processor of a general purpose computer, a special purpose computer, an embedded computer, or any other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or any other programmable data processing device generate an apparatus that implements the specific functionality of one or more steps in the flowcharts and / or one or more blocks in the block diagrams.
[0337] These computer program instructions may be stored in a computer readable memory that can instruct a computer or any other programmable data processing device to function in a particular manner such that the instructions stored in the computer readable memory produce an article that includes an instruction apparatus that implements a particular function of one or more processes in the flowcharts and / or one or more blocks in the block diagrams.
[0338] It is obvious that a person skilled in the art can make various modifications and variations of the present application without departing from the scope of the present application. The present application is intended to cover these modifications and variations of the present application as long as they fall within the scope of protection defined by the following claims and their equivalent technologies.
Claims
1. 1. A communication method comprising: determining X downlink measurement information based on X downlink signals from the access network device, where X is an integer greater than 1; determining a first downlink signal and a second downlink signal, the first downlink signal and the second downlink signal being downlink signals of the X downlink signals, the first downlink signal being a downlink signal determined based on an output of a first prediction model, an input of the first prediction model being determined based on Y downlink measurement information of the X downlink measurement information, and the second downlink signal being determined based on the X downlink measurement information, Y being an integer less than or equal to X and greater than 0; sending first information to the access network device when the first downlink signal and the second downlink signal satisfy a first condition, the first information indicating the second downlink signal and the Y downlink measurement information, or the first information indicating that the first prediction model needs to be updated; A communication method comprising:
2. 1. A communication method comprising: determining X downlink measurement information based on X downlink signals from the access network device, where X is an integer greater than 1; determining a first downlink signal and a second downlink signal, the first downlink signal and the second downlink signal being downlink signals of the X downlink signals, the first downlink signal being a downlink signal determined based on an output of a first prediction model, an input of the first prediction model being determined based on Y downlink measurement information of the X downlink measurement information, and the second downlink signal being determined based on the X downlink measurement information, Y being an integer less than or equal to X and greater than 0; transmitting first information or second information to the access network device based on the first downlink signal and the second downlink signal, wherein the first information indicates that the first prediction model needs to be updated, or the first information indicates the second downlink signal and the Y downlink measurement information, and the second information indicates that the first prediction model is reliable, or the second information indicates that the first prediction model does not need to be updated, or the second information indicates that the second information does not include the Y downlink measurement information; A communication method comprising:
3. 3. The method of claim 1, further comprising: if the first downlink signal and the second downlink signal satisfy a second condition, transmitting the second information to the access network device, the second information indicating the second downlink signal or the second information indicating that the first prediction model does not need to be updated.
4. The first condition is: A difference information between the first downlink signal and the second downlink signal is greater than a first threshold. The method of claim 1 , 2 or 3, comprising:
5. The second condition is: the first downlink signal and the second downlink signal are the same downlink signal; or The first downlink signal and the second downlink signal are not the same downlink signal, but the difference information between the first downlink signal and the second downlink signal is less than or equal to the first threshold. The method according to claim 3 or 4, comprising:
6. 6. The method of claim 4 or 5, wherein the difference information between the first downlink signal and the second downlink signal comprises an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal.
7. The method comprises: training the first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a second prediction model; transmitting third information to the access network device, the third information being indicative of the second prediction model or indicative of information about the second prediction model; The method of any one of claims 1 to 6, further comprising:
8. The step of transmitting third information to the access network device further comprises: sending the third information to the access network device when variation information between the second prediction model and the first prediction model is equal to or greater than a second threshold; 8. The method of claim 7, comprising:
9. The method of claim 7 or 8, wherein the information about the second predictive model includes the variation information between the second predictive model and the first predictive model.
10. The method comprises: receiving fourth information from the access network device, the fourth information indicating information about a third prediction model; updating the first forecasting model to the third forecasting model; The method of any one of claims 1 to 6, further comprising:
11. The method of claim 10 , wherein the third prediction model is obtained by training the first prediction model based on the Y downlink measurement information and the second downlink signal.
12. The method according to claim 1 , wherein the first prediction model is the same as a prediction model configured in the access network device.
13. The method of any one of claims 1 to 12, wherein the first prediction model is agreed in a protocol or configured by the access network device.
14. 1. A communication method comprising: transmitting X downlink signals to a terminal device, where X is an integer greater than 1; receiving first information from the terminal device, the first information indicating a second downlink signal and Y pieces of downlink measurement information among the X pieces of downlink measurement information, or the first information indicating that a first prediction model needs to be updated; Including, the second downlink signal is determined based on the X downlink measurement information, the X downlink measurement information is determined based on the X downlink signals, and Y is an integer less than or equal to X and greater than 0; Communication methods.
15. 1. A communication method comprising: transmitting X downlink signals to a terminal device, where X is an integer greater than 1; receiving first information or second information from the terminal device, the first information indicating a second downlink signal and Y downlink measurement information among the X downlink measurement information, or the first information indicating that a first prediction model needs to be updated, the second information indicating the second downlink signal, or the second information indicating that the first prediction model does not need to be updated, the second downlink signal being determined based on the X downlink measurement information, and the X downlink measurement information being determined based on the X downlink signals, where Y is an integer less than or equal to X and greater than 0; A communication method comprising:
16. 16. The method of claim 14 or 15, wherein a first downlink signal and the second downlink signal satisfy a first condition, the first downlink signal and the second downlink signal are downlink signals among the X downlink signals, the first downlink signal is a downlink signal determined based on an output of the first prediction model, and an input of the first prediction model is determined based on the Y downlink measurement information among the X downlink measurement information.
17. receiving second information from the terminal device, the second information being indicative of the second downlink signal or the second information being indicative of that the first prediction model does not need to be updated; 17. The method of claim 14, 15, or 16, comprising:
18. 20. The method of claim 17, wherein the first downlink signal and the second downlink signal satisfy a second condition.
19. The first condition is: A difference information between the first downlink signal and the second downlink signal is greater than a first threshold.
17. The method of claim 15 or 16, comprising:
20. The second condition is: the first downlink signal and the second downlink signal are the same downlink signal; or The first downlink signal and the second downlink signal are not the same downlink signal, but the difference information between the first downlink signal and the second downlink signal is less than or equal to a first threshold.
20. The method of claim 18, comprising:
21. 21. The method of claim 19 or 20, wherein the difference information between the first downlink signal and the second downlink signal comprises an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal.
22. The method comprises: training the first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model; transmitting fourth information to the terminal device, the fourth information being indicative of the third prediction model or indicative of information about the third prediction model; 22. The method of any one of claims 14 to 21, further comprising:
23. 23. The method of claim 22, wherein the information about the third predictive model includes variability information between the third predictive model and the first predictive model.
24. The method comprises: receiving third information from the terminal device, the third information indicating information about a second prediction model; updating the first forecasting model to the second forecasting model; 22. The method of any one of claims 14 to 21, further comprising:
25. 25. The method according to any one of claims 14 to 24, wherein the first prediction model is the same as a prediction model configured in the terminal device.
26. 1. A communication method comprising: transmitting X downlink signals to a terminal device and receiving X downlink measurement information from the terminal device, the X downlink measurement information being determined based on the X downlink signals, where X is an integer greater than 1; determining a first downlink signal and a second downlink signal, the first downlink signal and the second downlink signal being downlink signals of the X downlink signals, the first downlink signal being a downlink signal determined based on an output of a first prediction model, an input of the first prediction model being determined based on Y downlink measurement information of the X downlink measurement information, and the second downlink signal being determined based on the X downlink measurement information, Y being an integer less than or equal to X and greater than 0; training the first prediction model based on the Y downlink measurement information and the second downlink signal to obtain a third prediction model when the first downlink signal and the second downlink signal satisfy a first condition; updating the first forecasting model to the third forecasting model; A communication method comprising:
27. 27. The method of claim 26, wherein the first condition includes a difference information between the first downlink signal and the second downlink signal being greater than a first threshold.
28. if the first downlink signal and the second downlink signal satisfy a second condition, the first prediction model is maintained unchanged; 28. The method of claim 26 or 27, wherein the second condition includes that the first downlink signal and the second downlink signal are the same downlink signal, or that the first downlink signal and the second downlink signal are not the same downlink signal and the difference information between the first downlink signal and the second downlink signal is less than or equal to the first threshold.
29. the difference information between the first downlink signal and the second downlink signal comprises an absolute value of a difference between first downlink measurement information corresponding to the first downlink signal and second downlink measurement information corresponding to the second downlink signal; or 30. The method of claim 28, wherein the difference information between the first downlink signal and the second downlink signal comprises an absolute value of a difference between first location information corresponding to the first downlink signal and second location information corresponding to the second downlink signal.
30. 1. A communication method comprising: receiving X uplink signals from a terminal device and determining X uplink measurement information based on the X uplink signals, where X is an integer greater than 1; determining a first uplink signal and a second uplink signal, the first uplink signal and the second uplink signal being uplink signals among the X uplink signals, the first uplink signal being an uplink signal determined based on an output of a fourth prediction model, an input of the fourth prediction model being determined based on Y uplink measurement information among the X uplink measurement information, and the second uplink signal being determined based on the X uplink measurement information, where Y is an integer less than X and greater than 0; training the fourth prediction model based on the Y uplink measurement information and the second uplink signal to obtain a fifth prediction model when the first uplink signal and the second uplink signal satisfy a third condition; updating the fourth forecasting model to the fifth forecasting model; A communication method comprising:
31. 31. The method of claim 30, wherein the third condition includes a difference information between the first uplink signal and the second uplink signal being greater than a third threshold.
32. if the first uplink signal and the second uplink signal satisfy a fourth condition, the fourth prediction model is maintained unchanged; 32. The method of claim 30 or 31, wherein the fourth condition includes that the first uplink signal and the second uplink signal are the same uplink signal, or that the first uplink signal and the second uplink signal are not the same uplink signal but the difference information between the first uplink signal and the second uplink signal is less than or equal to the third threshold.
33. the difference information between the first uplink signal and the second uplink signal comprises an absolute value of a difference between first uplink measurement information corresponding to the first uplink signal and second uplink measurement information corresponding to the second uplink signal; or 33. The method of claim 32, wherein the difference information between the first uplink signal and the second uplink signal comprises an absolute value of a difference between location information corresponding to the first uplink signal and location information corresponding to the second uplink signal.
34. A communication device configured to carry out a method according to any one of claims 1 to 13.
35. A communications device configured to carry out a method according to any one of claims 14 to 33.
36. 11. A communications device comprising: a processor, coupled to a memory; A communications device, wherein the processor is configured to execute computer programs or instructions stored in the memory to enable the communications device to perform a method according to any one of claims 1 to 13.
37. 11. A communications device comprising: a processor, coupled to a memory; A communications device, wherein the processor is configured to execute computer programs or instructions stored in the memory to enable the communications device to perform a method according to any one of claims 14 to 33.
38. A computer readable storage medium storing a computer program or instructions, which when executed on a computer enables the computer to carry out a method according to any one of claims 1 to 13, or enables the computer to carry out a method according to any one of claims 14 to 33.
39. A computer program product comprising instructions which, when executed on a computer, enable the computer to carry out a method according to any one of claims 1 to 13 or enable the computer to carry out a method according to any one of claims 14 to 33.
40. A chip comprising a processor, coupled to a memory and configured to execute computer programs or instructions stored in the memory to enable the chip to perform a method according to any one of claims 1 to 33.
41. A communication system comprising a terminal device and an access network device, the terminal device configured to perform the method according to any one of claims 1 to 13 and the access network device configured to perform the method according to any one of claims 14 to 33.
Citation Information
Patent Citations
Neural Network Formation Configuration Feedback for Wireless Communications
US20210064996A1
User equipment reporting for updating of machine learning algorithms
US20210326726A1
Machine learning model selection in beamformed communications
US20210328630A1
