Electronic device, method and storage medium for wireless communication system

Abstract

The invention relates to an electronic device, a method and a storage medium for a wireless communication system. A first electronic device for a wireless communication system includes at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to cause the first electronic device to inform a second electronic device of data through the at least one processor, the data is notified by one of the following notification modes: a first notification mode configured to transmit or receive the data using an AI model deployed on one side of the first electronic device or the second electronic device; a second notification mode configured to transmit and receive data using a jointly trained pair of AI models deployed on both sides of the first electronic device and the second electronic device; and a third notification mode configured to transmit the data by the first electronic device to the second device without using the AI model for notifying the data at both the first electronic device and the second electronic device.

Description

Technical Field

[0001] The present disclosure generally relates to wireless communication systems, and more particularly to techniques related to data transmission / notification in wireless communication systems. Background Art

[0002] With the development of communication scenarios and communication technologies, data transmission in wireless communication systems may have more flexible requirements in terms of communication metrics such as data volume, latency, reliability, and resource overhead than before. Merely relying on communication modes based on traditional source and / or channel coding techniques may be difficult to meet the flexible requirements of such communication metrics in some communication scenarios in current wireless communication.

[0003] Therefore, it is necessary to expand new communication modes to adapt to more flexible requirements in various communication metrics. Summary of the Invention

[0004] The present disclosure proposes a solution related to data transmission / notification in a wireless communication system. Specifically, the present disclosure provides an electronic device, a method, and a storage medium for a wireless communication system.

[0005] One aspect of the present disclosure relates to a first electronic device for a wireless communication system, including: at least one processor; and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to notify data to a second electronic device, wherein the data is notified through one of the following notification modes: a first notification mode configured to use an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device to send or receive the data; a second notification mode configured to use a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device to send and receive the data; and a third notification mode configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0006] Another aspect of the present disclosure relates to a second electronic device for a wireless communication system, comprising: at least one processor; and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, by means of the at least one processor, cause the second electronic device to obtain data to be notified by a first electronic device, wherein the data is notified by one of the following notification modes: a first notification mode configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device; a second notification mode configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and a third notification mode configured to have the first electronic device send the data to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0007] Another aspect of the present disclosure relates to a method for a first electronic device of a wireless communication system, comprising notifying a second electronic device of data, wherein the data is notified by one of the following notification modes: a first notification mode configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device; a second notification mode configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and a third notification mode configured to have the first electronic device send the data to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0008] Another aspect of the present disclosure relates to a method for a second electronic device of a wireless communication system, comprising obtaining data to be notified by a first electronic device, wherein the data is notified by one of the following notification modes: a first notification mode configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device; a second notification mode configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and a third notification mode configured to have the first electronic device send the data to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0009] Another aspect of the present disclosure relates to a first electronic device for a wireless communication system, comprising: at least one processor; and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to send data as one of a group of sender devices to a second electronic device, the group of sender devices including one or more sender devices, wherein on both sides of each sender device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed, and wherein the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to process the data using a first AI model in a first AI model pair selected from the plurality of AI model pairs and send the processed data to the second electronic device, such that the second electronic device uses a second AI model in the first AI model pair to reconstruct the data based on the processed data received from the first electronic device.

[0010] Another aspect of the present disclosure relates to a second electronic device for a wireless communication system, comprising: at least one processor; and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to receive data from each sender device in a group of sender devices, the group of sender devices including one or more sender devices, wherein on both sides of each sender device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed, and wherein the at least one memory and the computer program code are further configured to, through the at least one processor, cause the second electronic device to receive processed data obtained by processing the data using a first AI model in a corresponding AI model pair selected from the plurality of AI model pairs from each sender device, and use a second AI model in the corresponding AI model pair to reconstruct the data based on the received processed data.

[0011] Another aspect of the present disclosure relates to a method for a first electronic device in a wireless communication system, including the first electronic device as a transmitting device among a group of transmitting devices to transmit data to a second electronic device, where the group of transmitting devices includes one or more transmitting devices. Wherein, on both sides of each transmitting device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed. And, wherein the method further includes processing the data using a first AI model in a first AI model pair selected from the plurality of AI model pairs and transmitting the processed data to the second electronic device, so that the second electronic device uses a second AI model in the first AI model pair to reconstruct the data based on the processed data received from the first electronic device.

[0012] Another aspect of the present disclosure relates to a method for a second electronic device in a wireless communication system, including receiving data from each transmitting device among a group of transmitting devices, where the group of transmitting devices includes one or more transmitting devices. Wherein, on both sides of each transmitting device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed. And, wherein the method further includes receiving, from each transmitting device, processed data obtained by processing the data using a first AI model in a corresponding AI model pair selected from the plurality of AI model pairs, and reconstructing the data using a second AI model in the corresponding AI model pair based on the received processed data.

[0013] Another aspect of the present disclosure relates to a non-transitory computer-readable storage medium storing executable instructions, which when executed implement the method as described in the above aspects.

[0014] Another aspect of the present disclosure relates to a computer program product including executable instructions, which when executed implement the method as described in the above aspects.

[0015] The above summary is provided to summarize some exemplary embodiments to provide a basic understanding of aspects of the subject matter described herein. Therefore, the above features are merely examples and should not be construed as narrowing the scope or spirit of the subject matter described herein in any way. Other features, aspects, and advantages of the subject matter described herein will become apparent from the following detailed description taken in conjunction with the accompanying drawings. Description of the Drawings

[0016] A better understanding of the present disclosure can be obtained when considering the following detailed description of embodiments in conjunction with the accompanying drawings. The same or similar reference numerals are used in the various drawings to denote the same or similar components. The various drawings, together with the following detailed description, are included in this specification and form a part of the specification, and are used to illustrate embodiments of the present disclosure and to explain the principles and advantages of the present disclosure. Among them:

[0017] Figure 1 Schematically shows a conventional notification mode in a wireless communication system;

[0018] Figure 2 Conceptual configuration of an electronic device on the transmitting end side according to a first embodiment of the present disclosure;

[0019] Figure 3 Schematically shows a conceptual operation flow on the transmitting end side according to a first embodiment of the present disclosure;

[0020] Figure 4 Schematically shows a schematic representation of a second notification mode according to a first embodiment of the present disclosure;

[0021] Figure 5 Conceptual configuration of an electronic device on the receiving end side according to a first embodiment of the present disclosure;

[0022] Figure 6 Schematically shows a conceptual operation flow on the receiving end side according to a first embodiment of the present disclosure;

[0023] Figure 7 Schematically shows the interaction between the transmitting end and the receiving end according to a first exemplary implementation of a first embodiment of the present disclosure;

[0024] Figure 8A Schematically shows a schematic representation of a first notification mode according to a first exemplary implementation of a first embodiment of the present disclosure;

[0025] Figure 8B Schematically shows the interaction between the transmitting end and the receiving end using the first notification mode according to a first exemplary implementation of a first embodiment of the present disclosure;

[0026] Figure 9A Schematically shows a schematic representation of a second notification mode according to a first exemplary implementation of a first embodiment of the present disclosure;

[0027] Figure 9B Schematically shows the interaction between the transmitting end and the receiving end using the second notification mode according to a first exemplary implementation of a first embodiment of the present disclosure;

[0028] Figure 10ASchematically shows a schematic representation of a third notification mode according to a first exemplary implementation of the present disclosure;

[0029] Figure 10B Schematically shows the interaction between a sender and a receiver using the third notification mode according to a first exemplary implementation of the first embodiment of the present disclosure;

[0030] Figure 11 Schematically shows the interaction between a sender and a receiver according to a second exemplary implementation of the first embodiment of the present disclosure;

[0031] Figure 12A And Figure 12B Schematically shows an application scenario according to a second embodiment of the present disclosure;

[0032] Figure 13 Schematically shows a conceptual configuration of an electronic device on the sender side according to a second embodiment of the present disclosure;

[0033] Figure 14 Schematically shows a conceptual operation flow on the sender side according to a second embodiment of the present disclosure;

[0034] Figure 15 Schematically shows a conceptual configuration of an electronic device on the receiver side according to a second embodiment of the present disclosure;

[0035] Figure 16 Schematically shows a conceptual operation flow on the receiver side according to a second embodiment of the present disclosure;

[0036] Figure 17 Schematically shows the interaction between a sender and a receiver according to a first exemplary implementation of the second embodiment of the present disclosure;

[0037] Figure 18 Schematically shows the interaction between a sender and a receiver according to a second exemplary implementation of the second embodiment of the present disclosure;

[0038] Figure 19 Schematically shows the interaction between a sender and a receiver during training of an AI model according to a second embodiment of the present disclosure;

[0039] Figure 20 Is a block diagram of an example structure of a personal computer as an information processing device that can be adopted in an embodiment of the present disclosure;

[0040] Figure 21 Is a block diagram of a first example showing a schematic configuration of a gNB to which the technology of the present disclosure can be applied;

[0041] Figure 22Block diagram showing a second example of a schematic configuration of a gNB to which the technology of the present disclosure can be applied;

[0042] Figure 23 Block diagram showing an example of a schematic configuration of a smart phone to which the technology of the present disclosure can be applied; and

[0043] Figure 24 Block diagram showing an example of a schematic configuration of an in-vehicle navigation device to which the technology of the present disclosure can be applied.

[0044] Although the embodiments described in the present disclosure may be susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and are described in detail herein. However, it should be understood that the drawings and the detailed description thereof are not intended to limit the embodiments to the particular forms disclosed, but on the contrary, are intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the claims. Detailed Description of the Invention

[0045] The following describes representative applications of various aspects such as devices and methods according to the present disclosure. The description of these examples is only for adding context and helping to understand the described embodiments. Thus, it is clear to those skilled in the art that the described embodiments may be implemented without some or all of the specific details. In other cases, well-known process steps are not described in detail to avoid unnecessarily obscuring the described embodiments. Other applications are possible, and the solutions of the present disclosure are not limited to these examples.

[0046] Typically, a wireless communication system includes at least a transmitting device (hereinafter simply referred to as the transmitter) and a receiving device (hereinafter simply referred to as the receiver). In the present disclosure, both the transmitter and the receiver can be a control device or a terminal device. For example, the control device can act as a transmitter to notify data to the terminal device, or the terminal device can act as a transmitter to notify data to the control device, or the terminal device can act as a transmitter to notify data to another terminal device acting as a receiver. The control device can provide communication services for one or more terminal devices.

[0047] In the present disclosure, the term "base station" or "control device" has the full breadth of its ordinary meaning and at least includes a wireless communication station that is part of a wireless communication system or radio system to facilitate communication. As an example, the base station can be, for example, an eNB of the 4G communication standard, a gNB of the 5G NR communication standard, a remote radio head, a wireless access point, a drone control tower, or a communication device performing similar functions. In the present disclosure, "base station" and "control device" can be used interchangeably, or the "control device" can be implemented as part of the "base station". The application examples of the base station / terminal device will be described in detail below with reference to the accompanying drawings taking the base station as an example.

[0048] In the present disclosure, the term "terminal device" or "user equipment (UE)" has the full breadth of its ordinary meaning and at least includes a terminal device that is part of a wireless communication system or radio system to facilitate communication. As an example, the terminal device can be, for example, a mobile phone, a laptop, a tablet computer, a vehicle-mounted communication device, a wearable device, a sensor, or other such terminal devices or their components. In the present disclosure, "terminal device" and "user equipment" (hereinafter may be abbreviated as "UE") can be used interchangeably, or the "terminal device" can be implemented as part of the "user equipment".

[0049] In the present disclosure, the term "transmitting end side" / "transmitting end device side" has the full breadth of its ordinary meaning and generally indicates the side that sends or notifies data to the other party. Similarly, the term "terminal device side" / "user equipment side" has the full breadth of its ordinary meaning and can correspondingly indicate the side that receives or obtains data from the other party.

[0050] In the present disclosure, the term "AI model" has the full breadth of its ordinary meaning and generally refers to any applicable machine learning model, deep learning model, rule model, weak artificial intelligence model, strong artificial intelligence model, etc. obtained through training and capable of implementing the functions defined in the present disclosure.

[0051] In a wireless communication system, when the transmitting end needs to notify data to the receiving end, according to the traditional communication mode, such as Figure 1 shown, the transmitting end can encode the data to be notified to the receiving end (for example, source coding and channel coding), then modulate the encoded data, and send the modulated data to the receiving end via, for example, a noisy channel. The receiving end demodulates and decodes the received data in sequence (for example, channel decoding and source decoding) to obtain the data that the transmitting end expects to notify it.

[0052] When some communication scenarios require sending a large amount of data, or when the noise and interference in a wireless communication system are relatively high, using this traditional data transmission method may result in a delay that is difficult to meet the requirements of both communication parties. Or, using this traditional data transmission method may lead to a relatively large resource overhead (e.g., communication resources in terms of time, frequency, space, etc.). Or, in some special cases (e.g., special channel conditions), using this traditional data transmission method may also result in a decrease in data transmission reliability.

[0053] In addition, as introduced in the background art, with the development of communication scenarios and communication technologies, different communication scenarios may have more flexible requirements for communication metrics such as data volume, delay, reliability, and resource overhead than in the past.

[0054] Therefore, it is necessary to expand new communication modes to adapt to more flexible requirements in various communication metrics.

[0055] With the development of wireless communication and artificial intelligence, the present disclosure contemplates introducing AI into a wireless communication system, thereby serving as a new communication mode for data transmission / notification.

[0056] For example, this new communication mode may consider deploying an AI model at the sending end and / or the receiving end. Such an AI model may be configured to perform any one or more of operations such as source coding, channel coding, and modulation on the data, perform any one or more of operations such as source decoding, channel decoding, and demodulation on the data, or perform any additional processing required for the current communication on the data, e.g., perform processing such as compression or decompression on the data, predict the current required data based on other data (e.g., historical data), and so on.

[0057] First Embodiment

[0058] According to a first embodiment of the present disclosure, data can be notified from a sending end to a receiving end by using one of the following three notification modes: a first notification mode, configured to use an artificial intelligence (AI) model deployed on one side of the sending end or the receiving end to send or receive the data; a second notification mode, configured to use a pair of jointly trained AI models deployed on both sides of the sending end and the receiving end to send and receive the data; and a third notification mode, configured to send the data from the sending end to the receiving end without using an AI model for notifying the data at both the sending end and the receiving end.

[0059] The following will be combined with Figures 2 - 11 to elaborate on the first embodiment in detail.

[0060] Structure and Operation Process of the Transmitter According to the First Embodiment of the Present Disclosure

[0061] First, the reference Figure 2 Illustrate the conceptual structure of the electronic device 20 for the sending end according to an embodiment of the present disclosure. Figure 2 The illustrated electronic device 20 may include various units to implement corresponding operations according to the first embodiment of the present disclosure. In this example, the electronic device 20 includes a communication unit 202 and a control unit 204. According to the present disclosure, the electronic device 20 may be a control device or a terminal device acting as a sending end. In one implementation, the electronic device 20 is implemented as the control device or the terminal device acting as the sending end itself or a part thereof, or is implemented as a device for controlling the control device or the terminal device acting as the sending end or otherwise related thereto or a part of such device. Various operations described below in conjunction with the sending end may be implemented by the units 202, 204 of the electronic device 20 or other possible units.

[0062] As Figure 2 shown, the electronic device 20 may include a communication unit 202. The communication unit 202 may be configured to notify data to another electronic device. For example, the data to be notified may be any data that needs to be notified by the sending end to the receiving end according to the corresponding communication scenario, such as, but not limited to, channel state information (CSI), video data, sensor data, etc. More generally, the communication unit 202 may be configured to send signals to other electronic devices or receive signals from other electronic devices (such as, traffic data, control signaling, and any other signals that need to be sent between electronic devices).

[0063] The electronic device 20 may further include a control unit 204. The control unit 204 may be configured to control the communication unit 202 to notify data in one of the following notification modes: a first notification mode, configured to send or receive the data by using an AI model deployed on one side of the electronic device 20 or the receiving end electronic device; a second notification mode, configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the electronic device 20 and the receiving end electronic device; and a third notification mode, configured to send the data from the electronic device 20 to the receiving end electronic device without using an AI model for notifying the data in both the electronic device 20 and the receiving end electronic device. For example, the control unit 204 may be configured to select a notification model and / or an AI model, and control the communication unit 202 to notify the receiving end electronic device of the notification model and / or the AI model to be used. Alternatively, the control unit 204 may be configured to determine the notification model and / or the AI model to be used based on the indication information received via the communication unit 202. More generally, the control unit 204 may be configured to perform any appropriate control on the electronic device to enable it to complete the corresponding operation.

[0064] It should be noted that the above-mentioned individual units are only logical modules divided according to their specific functions, rather than being used to limit the specific implementation methods. For example, they can be implemented in software, hardware, or a combination of software and hardware. The functions of the units disclosed herein can be implemented using circuits or processing circuits. The processing circuit can refer to various implementations of a digital circuit system, an analog circuit system, or a mixed-signal (combination of analog and digital) circuit system that performs functions in a computing system. The processing circuit can include, for example, circuits such as integrated circuits (ICs), application-specific integrated circuits (ASICs), parts or circuits of a single processor core, the entire processor core, a single processor, a programmable hardware device such as a field-programmable gate array (FPGA), and / or a system including multiple processors. A processor is considered a processing circuit or circuit because it includes transistors and other circuits therein.

[0065] In the present disclosure, a circuit, unit, device, or apparatus is hardware that performs or is programmed to perform the described functions. The hardware can be any hardware disclosed herein or otherwise known that is programmed or configured to perform the described functions. When the hardware is a processor that can be considered a type of circuit, the circuit, device, or unit is a combination of hardware and software, and the software is used to configure the hardware and / or the processor. In a hardware implementation, the hardware can be programmed or configured to perform the functions. In a software or software-hardware combination implementation, the software can be used to configure the hardware and / or the processor. In actual implementation, the above-mentioned individual units can be implemented as independent physical entities, or can also be implemented by a single entity (for example, a processor (CPU or DSP, etc.), an integrated circuit, etc.).

[0066] Next, reference will be made to Figure 3 the conceptual operation flow 30 of the transmitting end shown to detail each operation implemented by the electronic device 20 as the transmitting end.

[0067] The operation of the transmitting end starts at S302.

[0068] At S304, the transmitting end notifies the receiving end of data, and the data is notified through one of the following notification modes: a first notification mode, configured to use an artificial intelligence (AI) model deployed on one side of the transmitting end or the receiving end to send or receive the data; a second notification mode, configured to use a pair of jointly trained AI models deployed on both sides of the transmitting end and the receiving end to send and receive the data; and a third notification mode, configured to send the data from the transmitting end to the receiving end without using an AI model for notifying the data at both the transmitting end and the receiving end. For example, the data to be notified can be any data that needs to be notified from the transmitting end to the receiving end according to the corresponding communication scenario, such as but not limited to channel state information (CSI), video data, sensor data, etc.

[0069] In S304, specifically, the sending end can first determine which notification mode to use to notify the receiving end of the data. For example, the sending end can select one of the above three notification modes at least based on communication metrics. The communication metrics can, for example, include at least one or more of the following metrics: the transmission delay on the communication link between the sending end and the receiving end (for example, it can be the actual delay and / or the expected delay), the data transmission accuracy requirement, the communication resource overhead, the transmission rate between the sending end and the receiving end, the queue length of the data that the sending end is prepared to send to the receiving end, and the computing capabilities of the sending end and / or the receiving end. For another example, the sending end can receive information from the receiving end indicating which notification mode to use. In this case, the receiving end can select the notification mode to be used at least based on the above communication metrics.

[0070] For example, in the first and second notification modes, due to the introduction of an AI model for jointly encoding / decoding / compressing / decompressing / predicting the data to be sent and / or received, the data transmission accuracy may be lower than that when using traditional encoding / decoding and modulation / demodulation methods to transmit data under the same channel conditions, that is, lower than the data transmission accuracy of the third notification mode. In addition, since the first notification mode only deploys the AI model on one side of the sending end or the receiving end, some association information between the sending end and the receiving end is lost during the generation of the AI model due to the lack of a process of jointly training the data on both sides of the sending end and the receiving end, and / or because this one-sidedly deployed AI model usually performs some data prediction processing, resulting in prediction errors. Therefore, the data transmission accuracy of the first notification mode may be lower than that of the second notification mode. However, the use of the AI model can reduce the total data processing time, thereby reducing the transmission delay and increasing the transmission rate. After the original data is processed by the AI model, a very small amount of output data for transmission can be generated, thereby saving communication resource (such as time, frequency, or space resources) overhead. Therefore, the sending end or the receiving end can select the most appropriate notification mode to notify the data based on the requirement for data transmission accuracy and the trade-off in terms of transmission delay, transmission rate, and communication resource overhead.

[0071] In particular, in some special scenarios, such as in the case of CSI feedback detailed below, the amount of data to be transmitted between the sending end and the receiving end may affect the requirement for data transmission accuracy. Therefore, the amount of data transmitted between the sending end and the receiving end (for example, the queue length of the data to be sent) can also be considered to determine the requirement for data transmission accuracy, and then the most appropriate notification mode can be selected to notify the data according to the required data transmission accuracy.

[0072] In addition, deploying an AI model may have requirements for the computing capabilities of the sender / receiver. Therefore, some sender / receiver devices with limited computing capabilities may not be suitable for deploying the AI model. Thus, the most appropriate notification mode can also be selected based on the computing capabilities of the sender / receiver to notify the data.

[0073] For another example, the receiver can also select the notification mode to be used. In this case, before notifying the data, the sender can also receive indication information from the receiver indicating the notification mode to be used next. For another example, the notification mode to be used can also be default, for example, based on a prior agreement between the two parties, or based on a predetermined association between a specific communication scenario and the notification mode.

[0074] Preferably, the notification mode to be adopted can be determined dynamically. For example, the notification mode to be adopted can be determined and adjusted in real time based on the real-time changes of the communication metrics described above. For another example, the communication metrics described above can also be determined periodically, and the notification mode to be adopted can be adjusted accordingly.

[0075] Before S304, the sender can also deploy at least one AI model for the first notification mode and / or the second notification mode.

[0076] Specifically, according to the present disclosure, the first notification mode can be configured as any one of the following modes: sub-mode (1) deploying an AI model at the sender, using the deployed AI model by the sender to process the data to be notified to the receiver and sending the processed data to the receiver; sub-mode (2) deploying an AI model at the receiver, using the deployed AI model by the receiver to receive the data to be notified to the receiver from the sender; and sub-mode (3) deploying an AI model at the receiver, using the deployed AI model by the receiver to predict the data based on the historical data related to the data to be notified to the receiver.

[0077] In the first notification mode, the AI model deployed at the sending end or the receiving end can be any suitable AI model for preprocessing the data to be notified or performing additional processing on the received data. For example, the AI model deployed at the sending end can be an AI model that compresses the data to be notified, extracts key information, reduces redundant information, adds additional information related to the scenario, etc. In this case, the input of the AI model deployed at the sending end can be the original data to be notified, and the output can be the processed data after processing the data to be notified. For another example, the AI model deployed at the sending end can also be a model that encodes and modulates the data to be notified in an AI manner. In this case, the input of the AI model deployed at the sending end can be the original data to be notified, and the output can be the modulated signal waiting to be sent over the channel. For example, the AI model deployed at the receiving end can be an AI model that decompresses the data received from the sending end, extracts key information, predicts the complete data to be notified based on the data received from the sending end, etc. In this case, the input of the AI model deployed at the receiving end can be the data received from the sending end, and the output can be the complete data to be notified by the sending end to the receiving end generated after being processed by the AI model. For another example, the AI model deployed at the receiving end can also be a model that decodes and demodulates the data to be notified in an AI manner. In this case, the input of the AI model deployed at the receiving end can be the modulated signal that has been conventionally encoded and modulated received from the sending end, and the output can be the data demodulated and decoded by the AI model.

[0078] In particular, the AI model deployed at the receiving end can also be a prediction model for predicting data. In this case, the input of the AI model deployed at the receiving end can be the historical data related to the data to be predicted and optionally any data for assisting in the prediction, and the output can be the predicted data to be notified by the sending end to the receiving end.

[0079] When the sender determines at S304 to use the first notification mode, the sender may further determine, for example, which of the above three sub-modes to use to notify the data. For example, the sender can determine at least based on the communication scenario and the AI model deployment situation of the sender and the receiver. For example, when only the sender has the ability to deploy AI, or when the involved communication scenario is suitable for using AI to preprocess the data to be notified to the receiver (for example, compression, extraction of key information, addition of additional information related to the scenario, etc.), sub-mode (1) can be used to notify the data. For another example, when only the receiver has the ability to deploy AI, or when the involved communication scenario is suitable for using AI to perform additional processing on the data received from the sender (for example, decompression, extraction of key information, prediction of the complete data to be notified based on the data received from the sender, etc.), sub-mode (2) can be used to notify the data. For still another example, in some special cases (for example, in the scenario of CSI feedback described in detail below), the sender may not even send any current data to be notified to the receiver, but only notify the receiver to use the AI deployed at the receiver to predict the current data to be notified based on the historical data previously received by the receiver from the sender. In these special cases, the sender can determine to use sub-mode (3) to notify the data.

[0080] For another example, the receiver can also be the one to select the sub-mode to be used. In this case, before notifying the data, the sender can also receive indication information from the receiver indicating the specific sub-mode to be used next. For still another example, which sub-mode to use can also be default or preset according to the communication scenario and / or the AI model deployment situation of the sender and the receiver (for example, in previous communications, the sender and the receiver have already understood each other's AI model deployment situations).

[0081] According to the present disclosure, the second notification mode can be configured such that the sender uses the first AI model of a pair of jointly trained AI models to process the data to be notified to the receiver and sends the processed data to the receiver, and the receiver uses the second AI model of the pair of AI models to reconstruct the data based on the processed data received from the sender.

[0082] Figure 4 A schematic representation of the second notification mode is shown. As Figure 4 shown, in this mode, the data x to be notified from the sender to the receiver 1 is first processed by the first AI model deployed at the sender to obtain the processed data z 1 (for example, this data z 1 can be in a complex form). Subsequently, through the transmission of the wireless channel, the data z 1 becomes Due to the presence of noise and interference in the wireless channel, not necessarily the same as z 1 is exactly the same. The received at the receiving end is input into the second AI model. After being processed by this model, the receiving end can obtain the reconstructed data Generally, in the case of using a well-trained AI model pair, the reconstructed data is basically the same as the original data x to be notified 1 is basically the same.

[0083] As Figure 4 shown, the first AI model and the second AI model deployed on both sides of the sending end and the receiving end respectively are a pair of AI models that are jointly trained. In this notification mode, the sending end, the receiving end, and the wireless channel passing therebetween can be regarded as an N-layer neural network as a whole. In this neural network, the wireless channel is also a part of it. The wireless channel can be modeled as a non-training layer in the neural network according to the channel characteristics, that is, a neural network layer with fixed parameters, and its parameters will not change during model training. During the training of the AI model pair, the first AI model at the sending end and the second AI model at the receiving end need to perform joint learning (for example, the training data is paired data, jointly adjusting training parameters, etc.) to obtain the parameters of the entire neural network.

[0084] According to the second notification mode, the jointly trained AI model pair can be particularly applicable to perform paired operations on the data to be notified, such as encoding and decoding, modulation and demodulation, compression and decompression, etc. In particular, the jointly trained AI model pair can be particularly applicable to perform Joint Source-Channel Coding (JSCC) on the data to be notified. As is known to those skilled in the art, the purpose of source coding is to remove redundant information in the data to be transmitted, so as to improve the efficiency of data transmission, while the purpose of channel coding is to add redundant information to the data to be transmitted to achieve functions such as detection and error correction, thereby improving the reliability of data transmission. However, the design ideas of these two types of coding are to remove redundant information and add redundant information respectively, so they are opposite. If the source coding and channel coding are designed separately in the traditional way, it is difficult to obtain the optimal such solution. Therefore, a joint coding scheme called JSCC is proposed. In JSCC, an AI model can be used to jointly design at least the source coding and channel coding, so that the end-to-end transmission performance of the wireless communication system reaches the optimal. In some JSCC schemes, the AI model can also jointly design the source coding, channel coding, and modulation, so that the sending end can use the AI model to output a modulated signal that can be directly transmitted, and the receiving end can use the AI model to directly decode the data to be notified by the sending end from the received signal.

[0085] Returning again to Figure 3 , after the sending end notifies the data to the receiving end using one of the determined notification modes at S304, the operation of the sending end ends at S306.

[0086] It should be noted that Figure 3 The operation steps of the sending end shown are only illustrative. In practice, the operation of the sending end may also include some additional or alternative steps. For example, as mentioned in the above description, before S304, the sending end can deploy at least one AI model for subsequent data notification. For another example, before S304, the sending end can determine which notification mode to use and which AI model to use. In some cases, the sending end can also signal to the receiving end the notification mode and / or AI model to be used for data notification, or the sending end can receive from the receiving end information about the notification mode and / or AI model to be used for data notification.

[0087] Structure and Operation Process of the Receiver According to the First Embodiment of the Present Disclosure

[0088] The above has detailed the exemplary structure and exemplary operations of the sending end according to the present disclosure. Next, the exemplary structure and exemplary operation process of the receiving end device according to the present disclosure will be described in conjunction with Figures 5 - 6 illustrate.

[0089] Figure 5 The electronic device 50 shown may include various units to implement corresponding operations according to the first embodiment of the present disclosure. In this example, the electronic device 50 includes a communication unit 502 and a control unit 504. According to the present disclosure, the electronic device 50 may be a control device or a terminal device acting as a receiving end. In one implementation, the electronic device 50 is implemented as the control device or the terminal device acting as a receiving end itself or a part thereof, or is implemented as a device for controlling the control device or the terminal device acting as a receiving end or otherwise related thereto or a part of such device. Various operations described below in conjunction with the sending end may be implemented by the units 502, 504 of the electronic device 50 or other possible units.

[0090] As Figure 5 shown, the electronic device 50 may include a communication unit 502. The communication unit 502 may be configured to receive data from another electronic device. For example, the received data may be any data notified by the sending end to the receiving end according to the corresponding communication scenario, such as but not limited to channel state information (CSI), video data, sensor data, etc. More generally, the communication unit 502 may be configured to send signals to other electronic devices or receive signals from other electronic devices (such as service data, control signaling, and any other signals that need to be sent between electronic devices).

[0091] The electronic device 50 may further include a control unit 504. The control unit 504 may be configured to control the communication unit 502 to obtain the data to be notified by the sending end electronic device by adopting one of the following notification modes: a first notification mode, configured to use an AI model deployed on one side of the electronic device 50 or the sending end electronic device to send or receive the data; a second notification mode, configured to use a pair of jointly trained AI models deployed on both sides of the electronic device 50 and the sending end electronic device to send and receive the data; and a third notification mode, configured to, in the case where neither the electronic device 50 nor the sending end electronic device uses an AI model for notifying the data, the sending end electronic device sends the data to the electronic device 50. For example, the control unit 504 may be configured to determine the notification model and / or the AI model to be used based on the indication information received via the communication unit 502. Alternatively, the control unit 504 may be configured to select the notification model and / or the AI model, and control the communication unit 502 to notify the sending end electronic device of the notification model and / or the AI model to be adopted. More generally, the control unit 204 may be configured to perform any appropriate control on the electronic device to enable it to complete the corresponding operations.

[0092] It should be noted that the above-mentioned respective units are only logical modules divided according to their specific implemented functions, rather than limiting the specific implementation manners. For example, they can be implemented in software, hardware, or a combination of software and hardware. The functions of the units disclosed herein can be implemented using circuits or processing circuits. Processing circuits can refer to various implementations of digital circuit systems, analog circuit systems, or mixed-signal (a combination of analog and digital) circuit systems that perform functions in a computing system. Processing circuits can include, for example, circuits such as integrated circuits (ICs), application-specific integrated circuits (ASICs), parts or circuits of a single processor core, the entire processor core, a single processor, programmable hardware devices such as field-programmable gate arrays (FPGAs), and / or systems including multiple processors. A processor is considered a processing circuit or a circuit because it includes transistors and other circuits therein.

[0093] In the present disclosure, a circuit, unit, device, or apparatus is hardware that performs or is programmed to perform the functions. The hardware can be any hardware disclosed herein or otherwise known that is programmed or configured to perform the functions. When the hardware is a processor that can be considered a type of circuit, the circuit, device, or unit is a combination of hardware and software, and the software is used to configure the hardware and / or the processor. In the hardware implementation manner, the hardware can be programmed or configured to perform the functions. In the software or software-hardware combination implementation manner, the software can be used to configure the hardware and / or the processor. In actual implementation, the above-mentioned respective units can be implemented as independent physical entities, or can also be implemented by a single entity (for example, a processor (CPU or DSP, etc.), an integrated circuit, etc.).

[0094] Next, reference will be made to Figure 6 the conceptual operation flow 60 of the sending end shown to detail each operation implemented by the electronic device 50 as the receiving end.

[0095] The operation of the receiving end starts at S602.

[0096] At S604, the receiving end obtains the data to be notified by the sending end, and the data is notified through one of the following notification modes: the first notification mode, configured to send or receive the data by using an AI model deployed on one side of the receiving end or the sending end; the second notification mode, configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the receiving end and the sending end; and the third notification mode, configured to send the data from the sending end to the receiving end without using an AI model for notifying the data at both the receiving end and the sending end.

[0097] In S604, specifically, the receiving end can first determine which notification mode to use to obtain the data to be notified by the sending end. For example, the receiving end can receive indication information from the sending end indicating the notification mode to be used next. In this case, as described above with reference to Figure 3 the sending end selects one of the above three notification modes at least based on communication metrics and notifies the receiving end of the selected notification mode. For another example, the notification mode to be used can also be selected by the receiving end. For example, the receiving end can select the notification mode to be used at least based on communication metrics in a similar manner as described above with reference to Figure 3 and notify the sending end of the selected notification mode. The specific method of selecting the notification mode has been described in detail above and will not be elaborated here. For another example, the notification mode to be used can also be default, for example, based on a prior agreement between the two parties or based on a predetermined association between a specific communication scenario and the notification mode.

[0098] The receiving end can also deploy at least one AI model for the first notification mode and / or the second notification mode before S604.

[0099] As described above, the first notification mode can be configured as any of the following modes: Mode (1) deploy an AI model at the sending end, and the sending end uses the deployed AI model to process the data to be notified to the receiving end and send the processed data to the receiving end; Mode (2) deploy an AI model at the receiving end, and the receiving end uses the deployed AI model to receive the data to be notified to the receiving end from the sending end; and (3) deploy an AI model at the receiving end, and the receiving end uses the deployed AI model to predict the data based on historical data related to the data to be notified to the receiving end.

[0100] In the case where the receiving end determines to use the first notification mode at S604, the receiving end can further determine which of the above three sub-modes to use to obtain the data to be notified by the sending end. For example, the receiving end can receive indication information from the sending end indicating which sub-mode to use. As described in detail above, which sub-mode to use can be determined at least according to the communication scenario and the AI model deployment situation of the sending end and the receiving end. For another example, the receiving end can also determine which sub-mode to use in a similar manner by itself and send indication information to the sending end. For another example, which sub-mode to use can also be default or preset according to the communication scenario and / or the AI model deployment situation of the sending end and the receiving end (for example, in a previous communication, the sending end and the receiving end have already understood each other's AI model deployment situations).

[0101] As described in detail above, the second notification mode can be configured such that the sending end processes the data to be notified to the receiving end using the first AI model of a pair of jointly trained AI models and sends the processed data to the receiving end, and the receiving end uses the second AI model of the pair of AI models to reconstruct the data based on the processed data received from the sending end. The AI model to be deployed by the receiving end in the second notification mode has been described above and will not be elaborated here.

[0102] After the receiving end obtains the data to be notified by the sending end using one of the determined notification modes at S604, the operation of the receiving end ends at S606.

[0103] It should be noted that Figure 6 The illustrated operation steps of the receiving end are merely illustrative. In practice, the operation of the receiving end may also include some additional or alternative steps. For example, as mentioned in the above description, before S604, the receiving end may deploy at least one AI model for subsequent use in notifying data. For another example, before S604, the sending end may determine which notification mode to use and which AI model to use. In some cases, the receiving end may also signal to the sending end the notification mode and / or AI model to be used for data notification, or the receiving end may receive from the sending end information about the notification mode and / or AI model to be used for data notification.

[0104] First Example Implementation of the First Embodiment of the Present Disclosure

[0105] Next, reference will be made to Figures 7 - 1 0 to illustrate a first example implementation of applying the solution of the first embodiment to the scenario of CSI feedback.

[0106] In the first example implementation, the data to be notified by the sending end to the receiving end may be channel state information (CSI) feedback. In this first example implementation, for example, the sending end may be a user equipment (UE), and the receiving end may be a base station (BS).

[0107] As Figure 7 shown, first, the sending end and the receiving end may determine the queue length of the communication data to be sent to the receiving end. For example, the sending end determines the queue length of the communication data waiting to be sent to the receiving end. In some cases (e.g., when the receiving end, i.e., the base station, determines which notification mode to use based on this queue length), the sending end may send the determined queue length of the communication data to the receiving end. For example, any applicable channel / signaling / message may be used to send the information indicating the queue length, such as via the physical uplink control channel (PUCCH).

[0108] Next, the sending end and / or the receiving end can determine the notification mode to be used based on this queue length. In this example implementation, either the sending end or the receiving end can make such a determination based on the queue length. For example, after one of the sending end or the receiving end determines the notification mode, that party can send indication information indicating the determined notification mode to the other party. Specifically, for example, after determining the queue length of the communication data to be sent to the receiving end, the sending end can determine which notification mode to use to notify CSI feedback based on this queue length, and send indication information indicating which notification mode to use to notify CSI feedback to the receiving end. Alternatively, the sending end can also send information indicating the queue length to the receiving end and receive indication information indicating which notification mode to use to notify CSI feedback from the receiving end. In the latter case, the indication information is determined based on the queue length. For example, any applicable channel / signaling / message can be used to send the indication information indicating the notification mode, such as via a Physical Uplink Control Channel (PUCCH), Downlink Control Information (DCI), etc.

[0109] The process of determining the notification mode in the first example implementation of the first embodiment is described in detail below.

[0110] In wireless communication, especially for communication scenarios that require low latency and high energy efficiency (such as XR scenarios), scheduling across the application layer, data layer, and physical layer is usually considered to minimize the average queuing delay of data under the constraint of limited average data transmission power consumption. In cross-layer scheduling, there are already scheduling algorithms that consider both latency and power consumption in a trade-off manner. For example, such an algorithm will cause higher power to be used for transmission when the channel condition is poor, and lower power to be used for transmission when the channel condition is good. And when the queue length of the communication data waiting to be sent is short, low power is used for transmission and a coding and modulation scheme with a low transmission rate and a high packet loss tolerance rate is adopted, while when the queue length of the communication data waiting to be sent is long, high power is used for transmission and a coding and modulation scheme with a high transmission rate and a low packet loss tolerance rate is adopted.

[0111] For example, in such cross-layer scheduling, the longer the queue of the communication data waiting to be sent, the more data packets are sent in a transmission time slot. Therefore, in this transmission time slot, it is necessary to more accurately estimate the wireless channel state to ensure the performance of data transmission (for example, the trade-off performance between latency and power consumption). In wireless communication, the channel state is generally estimated by the user equipment through CSI feedback. However, providing a high-precision channel state estimate requires a large amount of information to be fed back by the user equipment, resulting in a large communication overhead. Therefore, it is necessary to reasonably and dynamically adjust the accuracy of channel state information estimation to reduce the communication overhead as much as possible.

[0112] In view of the above, the present disclosure proposes that the longer the queue length of the communication data waiting to be sent, the higher the precision of channel estimation is adopted, while the shorter the queue length, the lower the precision of channel estimation is adopted, so as to control the overhead brought by channel estimation as a whole.

[0113] In this first example implementation of the first embodiment of the present disclosure, it is considered to control the precision of channel estimation by adopting different notification modes to notify CSI feedback.

[0114] Specifically, in this example implementation, in the first notification mode, for example, an AI model for estimating the channel state data (i.e., current CSI feedback) of the current transmission time slot based at least on the channel state data (i.e., historical CSI feedback) for historical transmission time slots can be deployed at the receiving end. Figure 8A And Figure 8B respectively show the first notification mode of this first example implementation and the interaction between the sending end and the receiving end using the first notification mode.

[0115] For example, such an AI model can predict the current CSI feedback only based on historical CSI feedback data. As Figure 8A shown, assuming that the current transmission time slot is n, and the CSI feedback data is represented as then the input of such an AI model deployed at the receiving end can be a series of historical CSI feedback data: wherein, T can represent the number of CSI feedbacks in the historical transmission time slots to be adopted, and the output of the AI model can be the CSI feedback data of the current transmission time slot predicted by the model In this case, in the current transmission time slot, the communication overhead introduced by channel estimation can be 0. This case can also be regarded as the CSI feedback to be notified by the sending end to the receiving end being implicitly notified based on the AI model at the receiving end. For another example, such an AI model can also estimate the current CSI feedback based on historical CSI feedback data and a small amount of data for the current channel state sent by the sending end. In this case, the small amount of data sent by the sending end can be any appropriate data for assisting in predicting the current CSI feedback.

[0116] Since in this first notification mode, the receiving end obtains the CSI feedback based on the prediction of the AI model, the precision of channel estimation is relatively low. Therefore, the first notification mode is applicable to the case where the queue length of the communication data waiting to be sent is short. In other words, when the queue length is less than or equal to the first threshold, the receiving end obtains the CSI feedback based on the first notification mode.

[0117] As Figure 8BAs shown, in the first notification mode, for example, the sending end can send information indicating the queue length of the communication data it determines to send to the receiving end to the receiving end, so that the receiving end can judge by itself to use the first notification mode based on this queue length and enable the corresponding AI model to predict CSI feedback. In this case, the first threshold for judging whether to enable the first notification mode based on the queue length can be predetermined or agreed upon by the sending end and the receiving end in previous communications. Alternatively, after the receiving end judges to use the first notification mode based on the queue length, it can also signal to the sending end that it will use the first notification model to predict CSI feedback next.

[0118] For another example, after determining the queue length, the sending end can also determine to use the first notification mode based on this queue length and explicitly notify the receiving end to use the first notification mode, so that the receiving end enables the corresponding AI model to predict CSI feedback.

[0119] In the case where the AI model deployed at the receiving end requires a small amount of channel state information for auxiliary prediction from the sending end, the receiving end can request this data from the sending end, or, according to a prior agreement or configuration, the sending end can also actively send this data to the receiving end without the request of the receiving end.

[0120] In the first example implementation, in the second notification mode, for example, a pair of AI models jointly trained to determine the channel state data of the current transmission time slot (i.e., the current CSI feedback) can be deployed on both the sending end and the receiving end. Figure 9A and Figure 9B respectively show the second notification mode of this first example implementation and the interaction between the sending end and the receiving end using the second notification mode.

[0121] For example, the first AI model in the pair of AI models can be deployed at the sending end. The first AI model can process the CSI feedback data (for example, perform compression on data of this type of CSI feedback, or perform joint coding including source coding, channel coding, and modulation on the CSI feedback data, or perform both compression and joint coding on the CSI feedback data), generating processed feedback data. The sending end can send the processed feedback data to the receiving end via any appropriate channel / signaling (such as PUCCH). The second AI model in the pair of AI models can be deployed at the receiving end. The second AI model can process the received data (for example, perform decompression on data of this type of CSI feedback, or perform joint decoding including source decoding, channel decoding, and demodulation on the received data, or perform both decompression and joint decoding on the received data), thereby reconstructing the CSI feedback that the sending end aims to notify the receiving end.

[0122] In this example implementation, the pair of AI models deployed at the sending end and the receiving end can be any jointly trained pair of AI models suitable for CSI feedback transmission. For example, such a pair of AI models can perform more complex processing on the CSI feedback data, resulting in less communication overhead for transmitting the processed CSI feedback data and higher reconstruction accuracy. For example, as Figure 9A shown, the first AI model in the pair of AI models deployed at the sending end can take not only the CSI feedback data of the current transmission time slot as input, but also the CSI feedback data of historical transmission time slots as input. For example, assume that the current transmission time slot is n, and the CSI feedback data is represented as h[n]. Then the input of the first AI model deployed at the sending end can be the CSI feedback data of the current transmission time slot together with a series of historical CSI feedback data: h[n-T], h[n-T+1], …, h[n-1], h[n], where T can represent the number of CSI feedbacks in the historical transmission time slots to be adopted. After being processed by the first AI model, the processed data output from the first AI model may be data with a very small data volume. For example, the data volume may be only 2 bits or 4 bits. The second AI model deployed at the receiving end can process the processed data with a very small data volume received from the sending end, so as to reconstruct the CSI feedback that the sending end aims to notify the receiving end of

[0123] In the second notification mode, although the sending end does not directly send the CSI feedback itself to the receiving end, the actual CSI feedback of the current transmission time slot is processed by using the jointly trained AI model pair. Therefore, the accuracy of channel estimation is higher than that in the first notification mode, but lower than that in the third notification mode where the sending end directly sends the actual CSI feedback to the receiving end as described below. Therefore, this second notification mode is applicable to the case where the queue length of the communication data waiting to be sent is moderate. In other words, when the queue length is greater than the first threshold and less than or equal to the second threshold, the sending end uses the second notification mode to notify the receiving end of the CSI feedback.

[0124] As Figure 9B shown, in the second notification mode, first, similar to Figure 8B the first notification mode shown, the sending end can send the receiving end information indicating the queue length of the communication data that it determines to be ready to send to the receiving end, so that the receiving end can, based on this queue length, judge by itself that it needs to use the second notification mode and enable the corresponding AI model to process the CSI feedback. Or, after determining to use the second notification mode based on the queue length, the sending end can explicitly notify the receiving end that it needs to use the second notification mode, so that the receiving end enables the corresponding AI model to process the CSI feedback.

[0125] Subsequently, i.e., after determining to enable the second notification mode, the receiving end can send a Channel State Information Reference Signal (CSI-RS) to the sending end. Next, the sending end can estimate the channel state based on the received CSI-RS, i.e., calculate the CSI feedback, and process the CSI feedback using the first AI model in a pair of jointly trained AI models.

[0126] Subsequently, the sending end can send the processed CSI feedback data to the receiving end. The receiving end can process the received data using the second AI model in the pair of AI models to reconstruct the CSI feedback of the current transmission time slot.

[0127] In a first example implementation, in the third notification mode, any AI models may not be utilized, and the sending end sends the CSI feedback to the receiving end in a traditional manner. Figure 10A and Figure 10B respectively show the third notification mode of this first example implementation and the interaction between the sending end and the receiving end using the first notification mode.

[0128] For example, the receiving end can send CSI-RS to the sending end. The sending end can estimate the channel state based on the received CSI-RS, i.e., calculate the CSI feedback. As Figure 10A shown, assuming the current transmission time slot is n, the CSI feedback calculated by the sending end can be represented as a vector h[n] of length L bits. The sending end can directly send this vector to the receiving end via any appropriate channel / signaling (e.g., PUCCH channel).

[0129] Since in the third notification mode, the actual CSI feedback is directly sent to the receiving end, the accuracy of channel estimation is the highest. Therefore, this third notification mode is applicable to the case where the queue length of the communication data waiting to be sent is relatively long. In other words, when the queue length is greater than the second threshold, the sending end uses the third notification mode to send the CSI feedback to the receiving end.

[0130] As Figure 10B shown, in the third notification mode, first, similar to the first notification mode shown in Figure 8B and the second notification mode shown in Figure 9B , the sending end can send information indicating the queue length of the communication data it determines to be ready to send to the receiving end, so that the receiving end can, based on this queue length, judge by itself to use the second notification mode and enable the corresponding AI model to process the CSI feedback. Or, after determining to use the second notification mode based on the queue length, the sending end can explicitly notify the receiving end to use the second notification mode, so that the receiving end enables the corresponding AI model to process the CSI feedback.

[0131] Subsequently, that is, after determining to enable the third notification mode, the receiving end can send CSI-RS to the sending end. Next, the sending end can estimate the channel state based on the received CSI-RS, that is, calculate the CSI feedback.

[0132] Next, the sending end can directly send the calculated CSI feedback to the receiving end.

[0133] Second Example Implementation of the First Embodiment of the Present Disclosure

[0134] Next, reference will be made to Figure 11 describe a second example implementation of applying the solution of the first embodiment to the scenario of video data transmission.

[0135] In the second example implementation, the data that the sending end needs to notify the receiving end can be video data.

[0136] As Figure 11 shown, first, the sending end and the receiving end can determine the transmission delay for data transmission from the sending end to the receiving end. For example, the sending end can measure this transmission delay. For another example, the receiving end can also measure this transmission delay. In the latter case, the receiving end can send the measured transmission delay to the sending end.

[0137] Next, the sending end and / or the receiving end can determine the notification mode to be used based on the transmission delay. In this example implementation, either the sending end or the receiving end can make this determination based on the transmission delay. For example, when either the sending end or the receiving end determines the notification mode, that party can send indication information indicating the determined notification mode to the other party.

[0138] The determination process of the notification mode in the second example implementation of the first embodiment will be described in detail below.

[0139] In this example implementation, in the first notification mode, for example, the AI model deployed at the sending end can be an AI model that compresses video data, extracts key information (such as extracting key frames or key pixels in a frame), reduces redundant information (such as removing some frames or some pixels in a frame), adds additional information (such as metadata information related to video coding and decoding), etc. For another example, the AI model deployed at the sending end can also be a model that encodes and / or modulates video data in an AI manner. For example, the AI model deployed at the receiving end can be an AI model that decompresses the data received from the sending end and predicts the complete data to be notified based on the data received from the sending end (such as predicting the complete frame sequence or a complete frame based on partial frames or partial pixels in a frame sent by the sending end). For another example, the AI model deployed at the receiving end can also be a model that decodes and / or demodulates the received data in an AI manner.

[0140] In this example implementation, in the second notification mode, the pair of jointly trained AI models deployed on both the sending end and the receiving end can be pairs of AI models that perform paired operations on video data, such as pairs of AI models for encoding and decoding, modulating and demodulating, compressing and decompressing. In particular, in the second notification mode, the pair of jointly trained AI models deployed on both the sending end and the receiving end can be the pair of AI models for JSCC described above.

[0141] In this example implementation, in the third notification mode, the sending end and the receiving end transmit video data according to traditional video coding and decoding techniques.

[0142] In the scenario of video data transmission, traditional video coding and decoding techniques may cause a relatively large transmission delay due to relatively complex calculations. However, traditional video coding and decoding techniques can be lossless, so high data transmission reliability can be guaranteed. In the case where an AI model is deployed on either the sending end or the receiving end to transmit video data, the design of the AI model may not guarantee lossless coding and decoding of video data. Especially in the case where the AI model compresses / decompresses the video data to be transmitted, deletes some information, or makes predictions on at least part of the video data, the receiving end may not be able to receive or reconstruct the original video data. Therefore, deploying an AI model on only one side of the sending end or the receiving end to transmit video data may lead to a reduction in data transmission reliability. However, with the help of an AI model, the video data to be sent can be quickly generated or the video data can be reconstructed. Therefore, with the help of an AI model, the end-to-end transmission delay can be reduced. In addition, in the case where an AI model is only deployed on one side of the sending end or the receiving end, due to the lack of a process of jointly training data on both the sending end and the receiving end, some association information between the sending end and the receiving end is lost when generating the AI model, and / or due to the data prediction processing / redundant information deletion processing introduced by such a one-sidedly deployed AI model, errors may be caused. Therefore, the data transmission accuracy of the first notification model with a one-sidedly deployed AI model may be lower than the data transmission accuracy of the second notification mode with AI models deployed on both the sending end and the receiving end.

[0143] Considering the above factors, in response to the transmission delay being less than or equal to the first threshold, in other words, in response to a relatively small transmission delay, the sending end can use the third notification mode to send video data to the receiving end. In response to the transmission delay being greater than the first threshold and less than or equal to the second threshold, in other words, in response to a medium transmission delay, the sending end can use the second notification mode to notify the receiving end of the video data. In response to the transmission delay being greater than the second threshold, in other words, in response to a relatively large transmission delay, the receiving end can obtain the video data based on the first notification mode. For example, as described above, the receiving end can predict the complete frame sequence and / or all the pixels in a frame based on the partial frames received from the sending end and / or the partial pixels in a frame.

[0144] Reference has been made to Figures 2 - 11 to introduce the first embodiment according to the present disclosure. With the solution of the first embodiment, the notification mode most suitable for the current communication requirements can be selected to notify data from the sending end to the receiving end, so as to better meet the more flexible requirements in various communication metrics. Preferably, the notification mode to be used can be dynamically adjusted during the communication process, so as to more flexibly adapt to the real-time changes in the requirements of communication metrics.

[0145] Second Embodiment

[0146] In the above description of the first embodiment, the second notification mode for notifying data by arranging a pair of jointly trained AI models on both the sending end and the receiving end has been described in detail. In fact, there are many communication scenarios in which a group of sending ends send data to a receiving end. For example, the group of sending ends can include one or more sending end devices. Below, first refer to Figure 12A and Figure 12B to illustrate two exemplary communication scenarios in this case.

[0147] Figure 12A Fig. shows a scenario of live broadcast of a sports event such as a football game. In this scenario, multiple cameras can be arranged around the stadium, and each camera captures a picture at a specific angle. The pictures captured by these cameras can be first sent to the TV station / director's room. Subsequently, the most appropriate picture can be selected from these pictures at different angles and sent to the user side. For example, the picture to be selected can be a picture that the user may be interested in, such as a picture including a key player together with the football, a picture including the confrontation between players, a close-up picture of the coach / audience, and so on.

[0148] Figure 12BA scenario of an intelligent connected vehicle is shown. In this scenario, a vehicle may be equipped with multiple sensors, such as image sensors, lidar (LiDAR), millimeter-wave radar (mmWave radar), etc. These sensors can act as different transmitters and send the information they capture to the vehicle's control device. Subsequently, the vehicle's control device can select appropriate information from this information for operations such as presenting a picture around the vehicle and measuring the distance between the vehicle and other objects.

[0149] In Figure 12A With Figure 12B In the exemplified scenario where a group of transmitters send data to a receiver, the most important data for the receiver may be only the data sent by one transmitter or a few transmitters in this group of transmitters. For example, in Figure 12A In the scenario, the most important data for the receiver may be only the data sent by the camera that captures the most suitable picture for playback to the user side. In Figure 12B In the scenario, the most important data for the receiver may be only the data sent by the sensor that captures the most accurate information according to the current climate, light conditions, etc. If each transmitter in this group of transmitters sends the highest quality data (e.g., for video data, the best picture quality (such as high frame rate (e.g., 60FPS) and high resolution (e.g., 4K)), and for sensor data, the highest numerical accuracy, etc.) to the receiver, it may result in unnecessary resource consumption and / or latency.

[0150] Therefore, it can be considered to deploy different jointly trained pairs of AI models between each transmitter in this group of transmitters and the receiver. These pairs of AI models can be adapted to, for example, generate processed data with different data volumes and / or data precisions by the sender, and be adapted to reconstruct data with different data precisions and / or data qualities by the receiver, so as to be able to flexibly control the data transmission from a group of transmitter devices to a receiver device.

[0151] Specifically, according to the second embodiment of the present disclosure, a group of transmitters can send data to a receiver. This group of transmitter devices can include one or more transmitter devices, and multiple jointly trained pairs of AI models are deployed on both sides of each transmitter and the receiver. Each transmitter can use the first AI model in the first pair of AI models selected from the multiple pairs of AI models to process the data to be sent to the receiver and send the processed data to the receiver, so that the receiver uses the second AI model in the first pair of AI models to reconstruct the data based on the processed data received from the receiver.

[0152] The following will Figures 13 - 19 describe the second embodiment in detail.

[0153] Structure and Operation Process of the Transmitter According to the Second Embodiment of the Present Disclosure

[0154] First, the reference Figure 13 illustrates the conceptual structure of the electronic device 130 for the transmitting end according to an embodiment of the present disclosure. Figure 13 The illustrated electronic device 130 may include various units to implement corresponding operations according to the first embodiment of the present disclosure. In this example, the electronic device 130 includes a communication unit 13002 and a control unit 13004. According to the present disclosure, the electronic device 130 may be a control device or a terminal device acting as the transmitting end. In one implementation, the electronic device 130 is implemented as the control device or the terminal device acting as the transmitting end itself or a part thereof, or is implemented as a device for controlling the control device or the terminal device acting as the transmitting end or otherwise related thereto or a part of such device. Various operations described below in connection with the transmitting end may be implemented by the units 13002, 13004 of the electronic device 130 or other possible units.

[0155] As Figure 13 shown, similar to the first embodiment, the electronic device 130 may include a communication unit 13002. The communication unit 13002 may be configured to send data to another electronic device. For example, the data to be sent may be any data that needs to be notified by the transmitting end to the receiving end according to the corresponding communication scenario, such as but not limited to video data, sensor data, etc. For another example, the communication unit 13002 may also be configured to receive from another electronic device information indicating the priority of the electronic device 130 among a group of electronic devices acting as the transmitting end. For another example, the communication unit 13002 may also be configured to receive from another electronic device feedback information indicating the accuracy of processing and reconstructing data using a certain AI model. More generally, the communication unit 13002 may be configured to send signals to other electronic devices or receive signals from other electronic devices (such as service data, control signaling, and any other signals that need to be sent between electronic devices).

[0156] The electronic device 130 may further include a control unit 13004. The control unit 13004 may be configured to select a first AI model pair from a plurality of AI model pairs, and then control the communication unit 13004 to use the first AI model in the first AI model pair to process data to be sent to another electronic device and send the processed data to the other electronic device, so that the other electronic device uses the second AI model in the first AI model pair to reconstruct the data based on the processed data received from the electronic device 130. For example, the control unit 13004 may be configured to notify another electronic device acting as a receiving end of transmission parameters including at least the ID of the selected AI model pair and an event enabling the model pair. Alternatively, the control unit 13004 may be configured based on transmission parameter information indicating transmission parameters received via the communication unit 13002. More generally, the control unit 13004 may be configured to perform any appropriate control on the electronic device to enable it to complete the corresponding operation.

[0157] It should be noted that the above-mentioned respective units are only logical modules divided according to their specific implemented functions, rather than for restricting specific implementation manners. For example, they can be implemented in a software, hardware, or a combination of software and hardware manner. The functions of the units disclosed herein can be implemented using a circuit or a processing circuit. The processing circuit may refer to various implementations of a digital circuit system, an analog circuit system, or a mixed-signal (a combination of analog and digital) circuit system that performs functions in a computing system. The processing circuit may include, for example, circuits such as an integrated circuit (IC), an application-specific integrated circuit (ASIC), a part or a circuit of a single processor core, an entire processor core, a single processor, a programmable hardware device such as a field-programmable gate array (FPGA), and / or a system including multiple processors. A processor is considered a processing circuit or a circuit because it includes transistors and other circuits therein.

[0158] In the present disclosure, a circuit, a unit, a device, or an apparatus is hardware that executes or is programmed to execute the function. The hardware may be any hardware disclosed herein or otherwise known that is programmed or configured to execute the function. When the hardware is a processor that can be considered a type of circuit, the circuit, the device, or the unit is a combination of hardware and software, and the software is used to configure the hardware and / or the processor. In a hardware implementation manner, the hardware may be programmed or configured to execute the function. In a software or a combination of software and hardware implementation manner, the software may be used to configure the hardware and / or the processor. In actual implementation, the above-mentioned respective units may be implemented as independent physical entities, or may also be implemented by a single entity (for example, a processor (CPU or DSP, etc.), an integrated circuit, etc.).

[0159] Next, reference will be made to Figure 14The conceptual operation flow 140 of the transmitting end shown will be used to elaborate in detail on each operation implemented by the electronic device 130 as the transmitting end.

[0160] The operation of the transmitting end starts at S1402.

[0161] At S1404, the transmitting end determines which pair of AI models among multiple pairs of AI models to use for sending data.

[0162] According to one example implementation, the pair of AI models can be selected based on the priority of the transmitting end among a group of transmitting ends. For example, the priority of the transmitting end can be determined at least based on at least one of the following: the communication scenario between a group of transmitting end devices to which the transmitting end belongs and the receiving end, and the importance of the data respectively sent by each transmitting end in the group of transmitting end devices to the receiving end.

[0163] For example, the priority of each transmitting end in a group of transmitting end devices can be specified in advance. For example, such priority can be pre - specified by the transmitting end or the receiving end based on various factors or pre - negotiated by the transmitting end and the receiving end. For example, these factors can include at least the communication scenario and / or the device attributes of the transmitting end (e.g., the performance of transmitting data, such as the inherent transmission accuracy and the inherent processing speed of the device).

[0164] For another example, the receiving end can determine the difference in the importance between the data sent by each transmitting end based on the data sent by each transmitting end, and then feedback to each transmitting end the priority of that transmitting end determined based on the data importance. For example, the receiving end can use any suitable method to determine the importance of the received data. In particular, in some application scenarios, any appropriate artificial intelligence / machine learning model can be used to determine the importance of different data. For example, in the scenario of video transmission, any appropriate artificial intelligence / machine learning model can be used to determine whether the data sent by a certain transmitting end indicates whether the current picture is a picture that the user is interested in (such as determining whether the current picture includes a key athlete together with football, whether it includes athlete confrontation, etc.). For example, the higher the importance of the data, the higher the priority of the corresponding transmitting end. After determining the priority of the transmitting end, the receiving end can feedback the determined priority to the transmitting end. For example, the receiving end can use any suitable signaling for such feedback, such as via uplink control information (UCI).

[0165] Preferably, the priority of the transmitting end can be determined dynamically. For example, it can be determined or adjusted in real - time or periodically according to the current communication scenario and / or the importance of the data sent by each transmitting end in real - time or periodically.

[0166] According to the implementation of the above example, either the sender or the receiver can determine the pair of AI models to be used based on the priority of the sender. For example, for a sender with a higher priority, a pair of AI models with better performance can be used, while for a sender with a lower priority, a pair of AI models with poorer performance can be used. Specifically, for example, the performance of a pair of AI models can be evaluated according to the amount of data output by the AI model used at the sender in the pair of AI models, and / or the accuracy of the data reconstructed by the AI model used at the receiver and / or the processing speed of the AI model. For example, a pair of high-performance AI models allows a small amount of data to be transmitted between the sender and the receiver, and data that is basically the same as the original data can be reconstructed. For example, for video transmission, a pair of high-performance AI models allows a 60FPS, 4K video frame sequence with a large amount of data input at the sender to be processed into data with a relatively small amount of data at a faster speed, so as to be transmitted using less communication resources and obtain a smaller delay. Moreover, a high-performance AI model also allows a 60FPS, 4K video frame sequence that is basically lossless compared to the original 60FPS, 4K video frame sequence to be reconstructed at the receiver based on the data with a small amount of data. For another example, for sensor data transmission, a pair of high-performance AI models allows the input sensor data to be processed at a faster speed at the sender, and allows sensor data with basically no reduction in accuracy compared to the original sensor data to be reconstructed at the receiver based on the received data. On the contrary, a pair of low-performance AI models may generate a relatively large amount of data at the sender, or may require low-quality data to be input at the sender (for example, video data with a lower resolution or sensor data with a lower accuracy), or may have a lower accuracy in reconstructing data at the receiver, or may also have a slower processing speed.

[0167] It should be noted that it is not necessarily optimal for the sender to uniformly use a pair of high-performance AI models without discrimination. For example, the operation of a high-performance AI model may consume more power, occupy more storage space, or occupy more computing power during its operation. Therefore, when the priority of the sender is low, it may be more optimal to use a low-performance AI model to reduce unnecessary consumption of power, storage, and / or computing power, etc. Therefore, the present disclosure proposes to use high-performance AI models only for senders with higher priorities.

[0168] According to another example implementation, a pair of AI models can be selected based on the accuracy of the reconstructed data fed back by the receiver to the sender. For example, the accuracy includes any one or a combination of the following items: the verification result of the reconstructed data, and the accuracy rate calculated based on the comparison between the data sent by the sender and the reconstructed data.

[0169] Specifically, each sender can send data processed by the first AI model in a pair of AI models to the receiver. After receiving the processed data, the receiver can calculate the reconstruction accuracy of reconstructing the data using the second AI model in the pair of AI models. The data here can be the data that the sender wants to send to the receiver (e.g., video data, sensor data, etc.), or it can be data dedicated to determining the reconstruction accuracy of the data.

[0170] According to one example, the receiver can simply use any appropriate verification method to verify the reconstructed data. Verification methods include, for example, cyclic redundancy check (CRC check), MD5 check, hash check, and so on. For example, the data sequence input to the first AI model by the sender can be expressed as z = (z 1 , z 2 , …, z n ), and the data sequence output from the second AI model at the receiver can be expressed as The receiver can use any appropriate verification method to determine the number of data that pass the verification and the number of data that fail the verification in the data sequence output from the second AI model, and can calculate the ratio of the number of data that pass the verification to the number of data in the data sequence as the reconstruction accuracy. For example, assuming the length of the data sequence is n and the number of data that pass the verification in the data sequence output from the second AI model is n1, then the reconstruction accuracy can be expressed as n1 / n. In this example, no additional data needs to be sent between the sender and the receiver to obtain the reconstruction accuracy, so it is advantageously possible to avoid increasing the large overhead between the sender and the receiver.

[0171] According to another example, the receiver can calculate the comparison value between the data sent by the sender and the reconstructed data, such as the minimum mean square error (MMSE) between the data reconstructed via the AI model and the original data, and use it as the reconstruction accuracy. This method can calculate the reconstruction accuracy more accurately, but it requires the sender to directly send the original data to the receiver without going through the AI model, so it may generate additional communication overhead.

[0172] According to the implementation of the above examples, either the sender or the receiver can determine the pair of AI models to be used based on the reconstruction accuracy. After determining the reconstruction accuracy, the receiver can feedback the determined reconstruction accuracy to the sender. The sender can then determine whether to select this pair of AI models for subsequent data transmission according to whether the accuracy reaches the expected value. For another example, after determining the reconstruction accuracy, the receiver can also determine whether the accuracy meets the expected value by itself, and either feedback information indicating whether to continue using this pair of AI models to the sender, or implicitly indicate to the sender to continue using this pair of AI models by not making any feedback.

[0173] Preferably, the reconstruction accuracy can be determined dynamically. For example, the receiving end can calculate the reconstruction accuracy of the currently used AI model pair in real time or periodically, and thereby adjust the AI model pair to be used in real time or periodically.

[0174] Two exemplary implementations for determining the AI model pair to be used have been described. According to the second embodiment, these two exemplary implementations can be implemented in combination. For example, the AI model pair to be used can first be determined based on the priority of each sending end, and then the AI model pair actually used by each sending end and receiving end can be adjusted according to the reconstruction accuracy, for example, to meet the reconstruction accuracy requirements between each sending end and receiving end. For another example, multiple applicable AI model pairs can first be selected based on the reconstruction accuracy, and then the AI model pair whose performance meets its priority can be selected from these multiple AI model pairs according to the priority of the sending end. Of course, the two exemplary implementations can also be combined in any other appropriate way to select the AI model pair.

[0175] Continuing to refer to Figure 14 , in S1406, the sending end can use the determined AI model to process and send data.

[0176] For example, during the determination of which AI model to use for sending data in S1404, the sending end can directly use the corresponding AI model to send data based on the priority and / or data reconstruction accuracy feedback from the receiving end and the predetermined correspondence between the priority and / or data reconstruction accuracy and the AI model pair. In other words, the sending end can directly use the predetermined corresponding AI model pair to send data according to the priority and / or data reconstruction accuracy without explicitly notifying the receiving end to enable the AI model pair.

[0177] For another example, before S1406, the sending end can also send to the receiving end or receive from the receiving end the transmission parameter information indicating the transmission parameters. The transmission parameter information can at least include the ID of the selected AI model pair and the information indicating the time to enable the AI model pair, so that the indicated AI model pair is enabled to send data at the indicated time. In the case where the receiving end sends the transmission parameter information to the sending end, the transmission parameter information can be sent to the sending end together with the information indicating the priority and / or data reconstruction accuracy of the sending end determined by the receiving end. For example, the transmission parameter information can be sent using any appropriate signaling, such as Radio Resource Control (RRC) signaling, Uplink Control Information (UCI), or Downlink Control Information (DCI).

[0178] The operation of the sending end ends at S1408.

[0179] It should be noted that Figure 14The operation steps of the sending end shown are merely illustrative. In practice, the operations of the sending end may also include some additional or alternative steps. For example, as mentioned in the above description, between S1404 and S1406, the sending end may also send sending parameter information to the receiving end or receive sending parameter information from the receiving end. For another example, during S1406, the sending end may also perform operations for dynamically adjusting the AI model pair to be used with the receiving end, such as the receiving end feeding back updated sending end priority and / or data reconstruction accuracy to the sending end, and transmitting updated sending parameter information between the sending end and the receiving end to adjust the AI model pair and the like.

[0180] Structure and Operation Process of the Receiver According to the Second Embodiment of the Present Disclosure

[0181] The exemplary structure and exemplary operations of the sending end according to the present disclosure have been described in detail above. Next, the exemplary structure and exemplary operation process of the receiving end device according to the present disclosure will be described in conjunction with Figures 15 - 16 illustrate the exemplary structure and exemplary operation process of the receiving end device according to the present disclosure.

[0182] Figure 15 The illustrated electronic device 150 may include various units to implement corresponding operations according to the second embodiment of the present disclosure. In this example, the electronic device 150 includes a communication unit 1502 and a control unit 1504. According to the present disclosure, the electronic device 150 may be a control device or a terminal device acting as the receiving end. In one implementation, the electronic device 150 is implemented as the control device or terminal device acting as the receiving end itself or a part thereof, or is implemented as a device for controlling the control device or terminal device acting as the receiving end or otherwise related thereto or a part of such device. Various operations described below in conjunction with the sending end may be implemented by the units 1502, 1504 of the electronic device 150 or other possible units.

[0183] As Figure 15 shown, similar to the first embodiment, the electronic device 150 may include a communication unit 1502. The communication unit 1502 may be configured to receive data from another electronic device. For example, the received data may be any data notified by the sending end to the receiving end according to the corresponding communication scenario, such as but not limited to video data, sensor data, etc. For another example, the communication unit 1502 may also be configured to send information indicating the priority of the other electronic device in a group of electronic devices acting as the sending end to another electronic device. For another example, the communication unit 1502 may also be configured to send feedback information indicating the accuracy of processing and reconstructing data using a certain AI model pair to another electronic device. More generally, the communication unit 1502 may be configured to send signals to other electronic devices or receive signals from other electronic devices (such as service data, control signaling, and any other signals that need to be sent between electronic devices).

[0184] The electronic device 150 may further include a control unit 504. The control unit 1504 may be configured to select a first AI model pair from multiple AI model pairs, and then control the communication unit 1504 to process the data sent from another electronic device and processed by the first AI model in the first AI model pair using the second AI model in the first AI model pair, so as to reconstruct the data that the other electronic device intends to send to the electronic device 150. For example, the control unit 1504 may be configured to be notified by another electronic device acting as a sending end of at least the ID of the selected AI model pair and the sending parameters of the event enabling the model pair. Alternatively, the control unit 1504 may be configured based on the sending parameter information indicating the sending parameters received via the communication unit 1502. More generally, the control unit 1504 may be configured to perform any appropriate control on the electronic device to enable it to complete the corresponding operations.

[0185] It should be noted that the above-mentioned respective units are only logical modules divided according to their specific functions implemented, rather than used to limit the specific implementation manners. For example, they can be implemented in a software, hardware, or a combination of software and hardware manner. The functions of the units disclosed herein can be implemented using circuits or processing circuits. The processing circuit may refer to various implementations of a digital circuit system, an analog circuit system, or a mixed-signal (combination of analog and digital) circuit system that performs functions in a computing system. The processing circuit may include, for example, circuits such as an integrated circuit (IC), an application-specific integrated circuit (ASIC), a part or circuit of a single processor core, the entire processor core, a single processor, a programmable hardware device such as a field-programmable gate array (FPGA), and / or a system including multiple processors. A processor is considered a processing circuit or circuit because it includes transistors and other circuits therein.

[0186] In the present disclosure, a circuit, unit, device, or apparatus is hardware that performs or is programmed to perform the said functions. The hardware may be any hardware disclosed herein or otherwise known that is programmed or configured to perform the said functions. When the hardware is a processor that can be considered a type of circuit, the circuit, device, or unit is a combination of hardware and software, and the software is used to configure the hardware and / or the processor. In the hardware implementation manner, the hardware may be programmed or configured to perform the functions. In the software or software and hardware combination implementation manner, the software may be used to configure the hardware and / or the processor. In actual implementation, the above-mentioned respective units may be implemented as independent physical entities, or may also be implemented by a single entity (for example, a processor (such as a CPU or DSP, etc.), an integrated circuit, etc.).

[0187] Next, reference will be made to Figure 16The conceptual operation process 160 of the receiving end shown is used to explain in detail each operation implemented by the electronic device 150 as the receiving end.

[0188] The operation of the receiving end starts at S16002.

[0189] At S16004, the receiving end determines which pair of AI models to use to receive data.

[0190] In S16004, as described above, the pair of AI models can be determined based on the priority of the sending end in a group of sending ends, and the priority of the sending end can be pre-specified or feedback from the receiving end to the sending end. In the latter case, the receiving end can determine the difference in the importance between the data sent by each sending end and / or determine the communication scenario between the sending end and the receiving end based on the data sent by each sending end, and then feedback the priority of each sending end determined at least based on the data importance and / or communication scenario to each sending end.

[0191] Also as described above, the pair of AI models can also be determined based on the accuracy of the receiving end in reconstructing data. For example, the receiving end can receive data processed by the first AI model in a pair of AI models from the sending end. After receiving the processed data, the receiving end can calculate the reconstruction accuracy of reconstructing the data using the second AI model in the pair of AI models. As described in detail above, the receiving end can use any appropriate verification method to verify the reconstructed data, and calculate the ratio of the number of data passing the verification to the number of data in the data sequence as the reconstruction accuracy. For another example, the receiving end can also calculate the comparison value between the data reconstructed via the AI model and the original data, such as the minimum mean square error (MMSE), and use it as the reconstruction accuracy.

[0192] Also as described above, the pair of AI models can also be determined based on the combination of the priority of the sending end and the accuracy of reconstructing data.

[0193] As described above, the receiving end can feedback the determined priority of the sending end and / or the accuracy of reconstructing data to the sending end. Subsequently, either the sending end or the receiving end can determine the pair of AI models to be used based on the priority of the sending end and / or the accuracy of reconstructing data, and in cases other than the case where the sending end directly enables the pair of AI models without explicit notification, send sending parameter information indicating the sending parameters to the other party. The sending parameter information can at least include the ID of the selected pair of AI models and information indicating the time to enable the pair of AI models.

[0194] At S16006, the receiving end can use the determined AI model to process the data received from the sending end that has been processed by the sending end's AI model, so as to reconstruct the data that the sending end intends to send to the receiving end.

[0195] The operation of the receiving end ends at S16008.

[0196] It should be noted that Figure 16 The illustrated operation steps of the receiving end are merely illustrative. In practice, the operation of the receiving end may also include some additional or alternative steps. For example, as mentioned in the above description, between S16004 and S16006, the receiving end may also send sending parameter information to the sending end or receive sending parameter information from the sending end. For another example, during S16006, the receiving end may also perform operations for dynamically adjusting the AI model pair to be used with the sending end, such as the receiving end feeding back updated sending end priority and / or data reconstruction accuracy to the sending end, and transmitting updated sending parameter information between the sending end and the receiving end to adjust the AI model pair and the like.

[0197] Reference has been made to Figures 13 - 16 illustrated the details of the sending end and the receiving end according to the second embodiment of the present disclosure. The following refers to Figures 17 - 19 illustrate an exemplary interaction between the sending end and the receiving end according to the second embodiment of the present disclosure.

[0198] Figure 17 Schematically shows the interaction between the sending end and the receiving end according to the first example implementation of the second embodiment of the present disclosure.

[0199] As Figure 17 shown, first, the sending end sends data to the receiving end, that is, in the current communication scenario, the data that the sending end wants to send to the receiving end, such as video data or sensor data and the like. Here, the sending end can use any appropriate method to send the data to the receiving end. For example, the sending end can use traditional methods to perform traditional encoding, modulation and other operations on the data to be sent. Or, it can also use the default AI model pair or the AI model pair determined by both parties before to send the data.

[0200] Subsequently, the receiving end can determine the priority of the sending end based on the data received from the sending end. As described in detail above, the receiving end can receive data from each sending end in a group of sending ends, and compare the importance differences between the received data from different sending ends and / or determine the current communication scenario based on this data. After determining the importance of the data of each sending end and / or the communication scenario, the receiving end can feedback its priority to the corresponding sending end accordingly. For example, a sending end with high data importance can correspond to a higher priority, while a sending end with lower data importance can correspond to a lower priority. For example, the receiving end can use any suitable signaling to perform such feedback, such as via uplink control information (UCI).

[0201] Next, both the sending end and the receiving end can determine the pair of AI models to be used based on the priority of the sending end. For example, for a sending end with a higher priority, a pair of AI models with better performance can be used, while for a sending end with a lower priority, a pair of AI models with poorer performance can be used. For example, after the sending end receives the priority feedback from the receiving end, it can directly use the corresponding AI model to send data based on the predetermined correspondence between the priority and the pair of AI models. Correspondingly, the receiving end can also directly use the corresponding AI model to receive data based on the predetermined correspondence between the priority of the corresponding sending end and the pair of AI models.

[0202] More generally, as Figure 17 shown, either the sending end or the receiving end can determine the pair of AI models to be used based on the priority of the sending end, and send sending parameter information indicating the sending parameters to the other party. The sending parameter information can at least include the ID of the selected pair of AI models and the information indicating the time to enable this pair of AI models, so as to enable the indicated pair of AI models to send data at the indicated time. In the case where the receiving end sends this sending parameter information to the sending end, the sending parameter information can be sent to the sending end together with the priority of the sending end determined by the receiving end. For example, the sending parameter information can be sent using any appropriate signaling, such as RRC signaling, UCI or DCI.

[0203] After determining the pair of AI models / sending parameters, data can be sent from the sending end to the receiving end using the selected pair of AI models (for example, the pair of AI models specified in the sending parameters).

[0204] Figure 18 Schematically shows the interaction between the sending end and the receiving end according to the second exemplary implementation of the second embodiment of the present disclosure.

[0205] First, the sending end can send the data processed by the first AI model in a pair of AI models (e.g., AI model pair x) to the receiving end. The data here can be the data that the sending end wants to send to the receiving end (e.g., video data, sensor data, etc.), or it can be the data dedicated to determining the reconstruction accuracy of the data.

[0206] Next, the receiving end can calculate the reconstruction accuracy of reconstructing the data using the second AI model in the pair of AI models. As described in detail above, the reconstruction accuracy can be determined based on verification, or calculated based on the accuracy rate (e.g., minimum mean square error (MMSE)) calculated by comparing the data reconstructed via the AI model with the original data.

[0207] Next, the receiving end can feedback the calculated accuracy to the sending end.

[0208] Subsequently, both the sending end and the receiving end can determine the AI model pair to be used based on the accuracy at the sending end. For example, the sending end can determine whether to select the AI model pair x for subsequent data transmission according to whether the accuracy feedback by the receiving end reaches the expected value. For another example, after determining the reconstruction accuracy, the receiving end can also determine whether the accuracy meets the expected value by itself, and either feedback information indicating whether to continue using the AI model pair to the sending end, or implicitly indicate to the sending end to continue using the AI model pair by not making any feedback.

[0209] More generally, as Figure 17 shown, either the sending end or the receiving end can determine the AI model pair to be used based on the reconstruction accuracy, and send sending parameter information indicating the sending parameters to the other party. The sending parameter information can at least include the ID of the selected AI model pair and the information indicating the time to enable the AI model pair, so that the indicated AI model pair is enabled to send data at the indicated time. In the case where the receiving end sends the sending parameter information to the sending end, the sending parameter information can be sent to the sending end together with the reconstruction accuracy determined by the receiving end. For example, the sending parameter information can be sent using any appropriate signaling, such as RRC signaling, UCI or DCI.

[0210] After determining the AI model pair / sending parameters, the data can be sent from the sending end to the receiving end using the selected AI model pair (e.g., the AI model pair specified in the sending parameters).

[0211] Figure 19 Schematically shows the interaction between the sending end and the receiving end during the training of the AI model pair according to the second embodiment of the present disclosure.

[0212] According to the present disclosure, during the training of an AI model pair, the AI model parameters can be adjusted based on the accuracy of reconstructing the data that the sending end intends to send to the receiving end as feedback by the receiving end. In particular, this training method is not limited to the second embodiment of the present disclosure and is also applicable to any situation where it is necessary to jointly train an AI model pair between the sending end and the receiving end. For example, this training method is also applicable to the second communication mode according to the first embodiment of the present disclosure.

[0213] As Figure 19 shown, the sending end can send the data processed by the first AI model of a pair of AI models to the receiving end. For example, the data here can be feature data specifically for model training.

[0214] Next, the receiving end can evaluate the AI model. For example, the reconstruction accuracy of reconstructing the data using the second AI model in the pair of AI models can be calculated. Similar to the reconstruction accuracy calculation method described above, the reconstruction accuracy can be determined based on verification, that is, calculated as the ratio of the number of data passing the verification to the number of data in the data sequence. Alternatively, the reconstruction accuracy can also be based on the minimum mean square error (MMSE) calculated between the data reconstructed via the AI model and the original data. The receiving end can evaluate multiple AI models separately, for example, calculate the reconstruction accuracy of multiple AI models. For example, assuming there are m AI models, then the reconstruction accuracies for the 1st to the mth AI models can be represented as the set A = [A 1 , A 2 , …, A m . For example, assuming m = 3, that is, there are 3 AI models, then the reconstruction accuracy can be represented as A = [[0, 0.5], [0.5, 0.8], [0.8, 1]].

[0215] Next, the receiving end can feedback to the sending end the ID of the AI model targeted by the current feedback, the calculated accuracy or an indication of whether the accuracy meets the standard, and the AI model parameters to be updated. For example, the AI model parameters to be updated can be any applicable AI model parameters, such as the number of layers of a neural network, the number of nodes in each layer, the weight values of each node, the learning rate, and so on. For example, the receiving end can use any applicable message, signaling, etc. to make such feedback to the sending end. In particular, when the receiving end is a user equipment and the sending end is a base station, the receiving end can feedback an indication of whether the accuracy meets the standard through an affirmative acknowledgment (ACK) or a negative acknowledgment (NACK), and transmit the specific AI model parameters to be updated in the PUCCH or PUSCH transmitting the ACK / NACK.

[0216] Subsequently, the sending end can update the first AI model (i.e., the AI model for the sending end) in the AI model pair being currently trained according to the feedback information received from the receiving end. For example, adjust the model parameters of the first AI model according to the feedback from the receiving end. At the same time, the receiving end itself can also update the second AI model (i.e., the AI model for the receiving end) in the AI model pair being currently trained. For example, reconfigure the second AI model according to the determined model parameters.

[0217] During the training process, the interactions and operations as Figure 19 shown can be iteratively performed until the desired reconstruction accuracy is achieved.

[0218] It should be noted that in this disclosure, the AI model pair can be jointly trained between any appropriate sending end and receiving end. For example, a wireless channel under various communication conditions can be established between the sending end and the receiving end during model training using any appropriate method, or the jointly trained AI model pair can be widely applicable to wireless channels under various communication conditions through the model training process. The trained AI model pair can be used for any sending end and receiving end. That is to say, the sending end and receiving end for jointly training the AI model pair do not necessarily have to be the sending end and receiving end that use this AI model pair for data transmission. Of course, the sending end and receiving end that use the AI model pair for data transmission can be the sending end and receiving end for jointly training this AI model pair.

[0219] The above has been described in detail with reference to Figures 13 - 19 the second embodiment of the present disclosure. According to the second embodiment of the present disclosure, advantageously, the most appropriate pair of AI models can be selected between each sending end in a group of sending ends and the receiving end for data transmission, so as to flexibly control the data transmission from a group of sending end devices to the receiving end device. For example, using the second embodiment of the present disclosure, it is possible to ensure the accurate and / or timely transmission of important data while avoiding unnecessary resource consumption and / or delay in the transmission of unimportant data.

[0220] Specifically, the first embodiment and the second embodiment can be combined. For example, each sending end in a group of sending ends can notify the receiving end of data according to the first notification mode, the second notification mode, and the third notification mode described in the first embodiment. In the case of determining to use the second notification mode for notification in the manner described in the first embodiment, the corresponding sending end can determine which AI model pair among the multiple deployed AI model pairs to use for notifying data in the manner described in the second embodiment. For example, in the case of combined implementation, each sending end in a group of sending ends can respectively determine the notification mode to be adopted, or the group of sending ends can uniformly adopt one notification mode.

[0221] It should be understood that the machine-executable instructions in the machine-readable storage medium or program product according to the embodiments of the present disclosure can be configured to perform operations corresponding to the above device and method embodiments. When referring to the above device and method embodiments, the embodiments of the machine-readable storage medium or program product are clear to those skilled in the art, and thus will not be described repeatedly. The machine-readable storage medium and program product for carrying or including the above machine-executable instructions also fall within the scope of the present disclosure. Such a storage medium may include, but is not limited to, a floppy disk, an optical disk, a magneto-optical disk, a memory card, a memory stick, and the like.

[0222] In addition, it should be understood that the above series of processes and devices can also be implemented by software and / or firmware. In the case of implementation by software and / or firmware, a program constituting the software is installed from a storage medium or a network into a computer having a dedicated hardware structure, such as Figure 20 the general personal computer 1300 shown. When various programs are installed in this computer, it can execute various functions and the like. Figure 20 is a block diagram showing an example structure of a personal computer as an information processing device that can be adopted in an embodiment of the present disclosure. In one example, the personal computer may correspond to the above-described exemplary transmitter or receiver according to the present disclosure.

[0223] In Figure 20 it, a central processing unit (CPU) 1301 executes various processes according to a program stored in a read-only memory (ROM) 1302 or a program loaded from a storage section 1308 into a random access memory (RAM) 1303. In the RAM 1303, data required when the CPU 1301 executes various processes and the like is also stored as needed.

[0224] The CPU 1301, the ROM 1302, and the RAM 1303 are connected to each other via a bus 1304. An input / output interface 1305 is also connected to the bus 1304.

[0225] The following components are connected to the input / output interface 1305: an input section 1306 including a keyboard, a mouse, etc.; an output section 1307 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1308 including a hard disk, etc.; and a communication section 1309 including a network interface card such as a LAN card, a modem, etc. The communication section 1309 performs communication processing via a network such as the Internet.

[0226] As needed, the driver 1310 is also connected to the input / output interface 1305. A removable medium 1311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is mounted on the driver 1310 as needed, so that a computer program read therefrom is installed in the storage section 1308 as needed.

[0227] In the case where the above-described series of processes are implemented by software, a program constituting the software is installed from a network such as the Internet or a storage medium such as the removable medium 1311.

[0228] Those skilled in the art should understand that such a storage medium is not limited to Figure 20 the removable medium 1311 shown in which a program is stored and distributed separately from the device to provide the program to the user. Examples of the removable medium 1311 include a magnetic disk (including a floppy disk (registered trademark)), an optical disk (including a compact disc read-only memory (CD-ROM) and a digital versatile disc (DVD)), a magneto-optical disk (including a mini disc (MD) (registered trademark)), and a semiconductor memory. Alternatively, the storage medium may be a ROM 1302, a hard disk included in the storage section 1308, etc., in which a program is stored and distributed to the user together with the device including them.

[0229] The technology of the present disclosure can be applied to various products.

[0230] As described above, both the transmitter and the receiver of the present disclosure can be a control device or a terminal device. For example, the electronic devices 20, 50, 130, and 150 according to the embodiments of the present disclosure can be implemented as various control devices / base stations or included in various control devices / base stations, while as Figure 3 、 Figure 6 、 Figure 14 and Figure 16 shown, the methods can also be implemented by various control devices / base stations. For example, the electronic devices 20, 50, 130, and 150 according to the embodiments of the present disclosure can also be implemented as various terminal devices / user equipment or included in various terminal devices / user equipment, while as Figure 3 、 Figure 6 、 Figure 14 and Figure 16 shown, the methods can also be implemented by various terminal devices / user equipment.

[0231] For example, the control device / base station mentioned in the present disclosure can be implemented as any type of base station, such as a gNode B (gNB), such as a macro gNB and a small gNB. A small gNB can be a gNB that covers a cell smaller than a macro cell, such as a pico gNB, a femto gNB, and a home (femto) gNB. Alternatively, the base station can be implemented as any other type of base station, such as a Node B and a Base Transceiver Station (BTS). The base station can include: a main body configured to control wireless communication (also referred to as a base station device); and one or more Remote Radio Heads (RRHs) provided at a location different from the main body. Additionally, various types of terminals described below can operate as a base station by temporarily or semi-persistently performing base station functions.

[0232] For example, the terminal device mentioned in the present disclosure is also referred to as a user equipment in some examples and can be implemented as a mobile terminal (such as a smart phone, a tablet personal computer (PC), a notebook PC, a portable game terminal, a portable / dongle-type mobile router, and a digital camera device) or a vehicle-mounted terminal (such as a car navigation device). The user equipment can also be implemented as a terminal that performs machine-to-machine (M2M) communication (also referred to as a machine type communication (MTC) terminal). In addition, the user equipment can be a wireless communication module (such as an integrated circuit module including a single chip) installed on each of the above terminals.

[0233] The following will refer to Figures 21 to 24 describe examples according to the present disclosure.

[0234] [Examples of Base Stations]

[0235] It should be understood that the term base station in the present disclosure has the full breadth of its ordinary meaning and at least includes a wireless communication station used as part of a wireless communication system or a radio system for facilitating communication. Examples of base stations can be, for example, but not limited to the following: The base station can be one or both of a Base Transceiver Station (BTS) and a Base Station Controller (BSC) in a GSM system, one or both of a Radio Network Controller (RNC) and a Node B in a WCDMA system, an eNB in an LTE and LTE-Advanced system, a gNB that appears in a 5G communication system, an eLTE eNB, etc., or can be a corresponding network node in a future communication system. Some functions of the base station in the present disclosure can also be implemented as an entity that has a control function for communication in D2D, M2M, and V2V communication scenarios, or as an entity that plays a role in spectrum coordination in a cognitive radio communication scenario.

[0236] First Example

[0237] Figure 21 is a block diagram showing a first example of a schematic configuration of a gNB to which the techniques of the present disclosure can be applied. The gNB 1400 includes a plurality of antennas 1410 and a base station device 1420. The base station device 1420 and each antenna 1410 can be connected to each other via RF cables. In one implementation, the gNB 1400 (or the base station device 1420) here can correspond to the above-mentioned electronic device 10 and / or the electronic device 80.

[0238] Each of the antennas 1410 includes a single or multiple antenna elements (such as multiple antenna elements included in a multiple-input multiple-output (MIMO) antenna), and is used to transmit and receive wireless signals for the base station device 1420. As Figure 21 shown, the gNB 1400 can include a plurality of antennas 1410. For example, the plurality of antennas 1410 can be compatible with a plurality of frequency bands used by the gNB 1400.

[0239] The base station device 1420 includes a controller 1421, a memory 1422, a network interface 1423, and a wireless communication interface 1425.

[0240] The controller 1421 can be, for example, a CPU or a DSP, and operates various functions of the higher layers of the base station device 1420. For example, the controller 1421 generates data packets based on the data in the signals processed by the wireless communication interface 1425, and transmits the generated packets via the network interface 1423. The controller 1421 can bundle data from multiple baseband processors to generate bundled packets, and transmit the generated bundled packets. The controller 1421 can have a logical function to execute controls such as radio resource control, radio bearer control, mobility management, admission control, and scheduling. This control can be executed in combination with nearby gNBs or core network nodes. The memory 1422 includes a RAM and a ROM, and stores programs executed by the controller 421 and various types of control data (such as a terminal list, transmission power data, and scheduling data).

[0241] The network interface 1423 is a communication interface for connecting the base station device 1420 to the core network 1424. The controller 1421 can communicate with a core network node or another gNB via the network interface 1423. In this case, the gNB 1400 and the core network node or other gNBs can be connected to each other through logical interfaces (such as the S1 interface and the X2 interface). The network interface 1423 can also be a wired communication interface or a wireless communication interface for a wireless backhaul line. If the network interface 1423 is a wireless communication interface, compared with the frequency band used by the wireless communication interface 1425, the network interface 1923 can use a higher frequency band for wireless communication.

[0242] The wireless communication interface 1425 supports any cellular communication scheme (such as Long-Term Evolution (LTE) and LTE-Advanced), and provides a wireless connection to a terminal in the cell located at the gNB 1400 via the antenna 1410. The wireless communication interface 1425 generally may include, for example, a baseband (BB) processor 1426 and an RF circuit 1427. The BB processor 1426 may perform, for example, encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and perform various types of signal processing of layers (such as L1, Media Access Control (MAC), Radio Link Control (RLC), and Packet Data Convergence Protocol (PDCP)). Instead of the controller 1421, the BB processor 1426 may have a part or all of the above-described logical functions. The BB processor 1426 may be a memory storing a communication control program, or a module including a processor configured to execute the program and related circuits. The update program may change the functions of the BB processor 1426. The module may be a card or blade inserted into a slot of the base station device 1420. Alternatively, the module may also be a chip mounted on the card or blade. Meanwhile, the RF circuit 1427 may include, for example, mixers, filters, and amplifiers, and transmit and receive wireless signals via the antenna 1410. Although Figure 21 an example showing one RF circuit 1427 connected to one antenna 1410 is illustrated, the present disclosure is not limited to this illustration, but one RF circuit 1427 may be connected to multiple antennas 1410 simultaneously.

[0243] As Figure 21 shown, the wireless communication interface 1425 may include multiple BB processors 1426. For example, the multiple BB processors 1426 may be compatible with multiple frequency bands used by the gNB 1400. As Figure 21 shown, the wireless communication interface 1425 may include multiple RF circuits 1427. For example, the multiple RF circuits 1427 may be compatible with multiple antenna elements. Although Figure 21 an example showing that the wireless communication interface 1425 includes multiple BB processors 1426 and multiple RF circuits 1427 is illustrated, the wireless communication interface 1425 may also include a single BB processor 1426 or a single RF circuit 1427.

[0244] Second example

[0245] Figure 22FIG. is a block diagram showing a second example of a schematic configuration of a gNB to which the technology of the present disclosure can be applied. The gNB 1530 includes a plurality of antennas 1540, a base station device 1550, and an RRH 1560. The RRH 1560 and each antenna 1540 can be connected to each other via an RF cable. The base station device 1550 and the RRH 1560 can be connected to each other via a high-speed line such as an optical fiber cable. In one implementation, the gNB 1530 (or the base station device 1550) here can correspond to the above-described electronic devices 50 and / or 100.

[0246] Each of the antennas 1540 includes a single or multiple antenna elements (such as multiple antenna elements included in a MIMO antenna) and is used to transmit and receive wireless signals for the RRH 1560. As Figure 22 shown, the gNB 1530 can include a plurality of antennas 1540. For example, the plurality of antennas 1540 can be compatible with the multiple frequency bands used by the gNB 1530.

[0247] The base station device 1550 includes a controller 1551, a memory 1552, a network interface 1553, a wireless communication interface 1555, and a connection interface 1557. The controller 1551, the memory 1552, and the network interface 1553 are the same as the controller 1421, the memory 1422, and the network interface 1423 described with reference to Figure 21 description.

[0248] The wireless communication interface 1555 supports any cellular communication scheme (such as LTE and LTE-Advanced), and provides wireless communication to terminals located in the sector corresponding to the RRH 1560 via the RRH 1560 and the antenna 1540. The wireless communication interface 1555 generally may include, for example, a BB processor 1556. Except that the BB processor 1556 is connected to the RF circuit 1564 of the RRH 1560 via the connection interface 1557, the BB processor 1556 is the same as the BB processor 1426 described with reference to Figure 21 description. As Figure 22 shown, the wireless communication interface 1555 can include a plurality of BB processors 1556. For example, the plurality of BB processors 1556 can be compatible with the multiple frequency bands used by the gNB 1530. Although Figure 22 an example in which the wireless communication interface 1555 includes a plurality of BB processors 1556 is shown, the wireless communication interface 1555 may also include a single BB processor 1556.

[0249] The connection interface 1557 is an interface for connecting the base station device 1550 (wireless communication interface 1555) to the RRH 1560. The connection interface 1557 can also be a communication module for communication in the above-mentioned high-speed line for connecting the base station device 1550 (wireless communication interface 1555) to the RRH 1560.

[0250] The RRH 1560 includes a connection interface 1561 and a wireless communication interface 1563.

[0251] The connection interface 1561 is an interface for connecting the RRH 1560 (wireless communication interface 1563) to the base station device 1550. The connection interface 1561 can also be a communication module for communication in the above-mentioned high-speed line.

[0252] The wireless communication interface 1563 transmits and receives wireless signals via the antenna 1540. The wireless communication interface 1563 generally can include, for example, an RF circuit 1564. The RF circuit 1564 can include, for example, a mixer, a filter, and an amplifier, and transmits and receives wireless signals via the antenna 1540. Although Figure 22 an example showing one RF circuit 1564 connected to one antenna 1540 is illustrated, the present disclosure is not limited to this illustration, but rather one RF circuit 1564 can be connected to multiple antennas 1540 simultaneously.

[0253] As Figure 22 shown, the wireless communication interface 1563 can include multiple RF circuits 1564. For example, multiple RF circuits 1564 can support multiple antenna elements. Although Figure 22 an example showing that the wireless communication interface 1563 includes multiple RF circuits 1564 is illustrated, the wireless communication interface 1563 can also include a single RF circuit 1564.

[0254] [Example of User Equipment]

[0255] First Example

[0256] Figure 23 is a block diagram showing an example of a schematic configuration of a smart phone 1600 to which the technology of the present disclosure can be applied. The smart phone 1600 includes a processor 1601, a memory 1602, a storage device 1603, an external connection interface 1604, a camera device 1606, a sensor 1607, a microphone 1608, an input device 1609, a display device 1610, a speaker 1611, a wireless communication interface 1612, one or more antenna switches 1615, one or more antennas 1616, a bus 1617, a battery 1618, and an auxiliary controller 1619. In one implementation, the smart phone 1600 (or the processor 1601) here can correspond to the above-mentioned electronic device 50 and / or electronic device 100.

[0257] The processor 1601 can be, for example, a CPU or a system on chip (SoC), and controls the functions of the application layer and other layers of the smartphone 1600. The memory 1602 includes RAM and ROM, and stores data and programs executed by the processor 1601. The storage device 1603 can include storage media such as semiconductor memories and hard disks. The external connection interface 1604 is an interface for connecting external devices (such as memory cards and universal serial bus (USB) devices) to the smartphone 1600.

[0258] The imaging device 1606 includes image sensors (such as charge-coupled devices (CCDs) and complementary metal-oxide-semiconductor (CMOS)), and generates captured images. The sensor 1607 can include a set of sensors such as measurement sensors, gyro sensors, geomagnetic sensors, and acceleration sensors. The microphone 1608 converts the sound input to the smartphone 1600 into an audio signal. The input device 1609 includes, for example, a touch sensor configured to detect touches on the screen of the display device 1610, a keypad, a keyboard, buttons, or switches, and receives operations or information input from the user. The display device 1610 includes a screen (such as a liquid crystal display (LCD) and an organic light-emitting diode (OLED) display), and displays the output images of the smartphone 1600. The speaker 1611 converts the audio signal output from the smartphone 1600 into sound.

[0259] The wireless communication interface 1612 supports any cellular communication scheme (such as LTE and LTE-Advanced), and performs wireless communication. The wireless communication interface 1612 generally can include, for example, a BB processor 1613 and an RF circuit 1619. The BB processor 1613 can perform, for example, encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and performs various types of signal processing for wireless communication. At the same time, the RF circuit 1614 can include, for example, mixers, filters, and amplifiers, and transmits and receives wireless signals via the antenna 1616. The wireless communication interface 1612 can be a single chip module on which the BB processor 1613 and the RF circuit 1614 are integrated. As Figure 23 shown, the wireless communication interface 1612 can include multiple BB processors 1613 and multiple RF circuits 1614. Although Figure 23 an example in which the wireless communication interface 1612 includes multiple BB processors 1613 and multiple RF circuits 1614 is shown, the wireless communication interface 1612 can also include a single BB processor 1613 or a single RF circuit 1614.

[0260] In addition to the cellular communication scheme, the wireless communication interface 1612 may support other types of wireless communication schemes, such as short-range wireless communication schemes, near-field communication schemes, and wireless local area network (LAN) schemes. In this case, the wireless communication interface 1612 may include a BB processor 1613 and an RF circuit 1614 for each wireless communication scheme.

[0261] Each of the antenna switches 1615 switches the connection destination of the antenna 1616 among a plurality of circuits (e.g., circuits for different wireless communication schemes) included in the wireless communication interface 1612.

[0262] Each of the antennas 1616 includes a single or multiple antenna elements (such as the multiple antenna elements included in a MIMO antenna) and is used for the wireless communication interface 1612 to transmit and receive wireless signals. As Figure 23 shown, the smart phone 1600 may include a plurality of antennas 1616. Although Figure 23 an example in which the smart phone 1600 includes a plurality of antennas 1616 is shown, the smart phone 1600 may also include a single antenna 1616.

[0263] In addition, the smart phone 1600 may include an antenna 1616 for each wireless communication scheme. In this case, the antenna switch 1615 may be omitted from the configuration of the smart phone 1600.

[0264] The bus 1617 connects the processor 1601, the memory 1602, the storage device 1603, the external connection interface 1604, the imaging device 1606, the sensor 1607, the microphone 1608, the input device 1609, the display device 1610, the speaker 1611, the wireless communication interface 1612, and the auxiliary controller 1619 to each other. The battery 1618 supplies power to Figure 23 each block of the smart phone 1600 shown via a feeder line, which is partially shown as a dashed line in the figure. The auxiliary controller 1619 operates the minimum necessary functions of the smart phone 1600, for example, in the sleep mode.

[0265] Second Example

[0266] Figure 24FIG. is a block diagram showing an example of a schematic configuration of an in-vehicle navigation device 1720 to which the technology of the present disclosure can be applied. The in-vehicle navigation device 1720 includes a processor 1721, a memory 1722, a Global Positioning System (GPS) module 1724, a sensor 1725, a data interface 1726, a content player 1727, a storage medium interface 1728, an input device 1729, a display device 1730, a speaker 1731, a wireless communication interface 1733, one or more antenna switches 1736, one or more antennas 1737, and a battery 1738. In one implementation, the in-vehicle navigation device 1720 (or the processor 1721) here may correspond to the above-described electronic device 50 and / or electronic device 100.

[0267] The processor 1721 may be, for example, a CPU or an SoC, and controls the navigation function and other functions of the in-vehicle navigation device 1720. The memory 1722 includes a RAM and a ROM, and stores data and programs executed by the processor 1721.

[0268] The GPS module 1724 uses GPS signals received from GPS satellites to measure the position of the in-vehicle navigation device 1720 (such as latitude, longitude, and altitude). The sensor 1725 may include a set of sensors, such as a gyro sensor, a geomagnetic sensor, and an air pressure sensor. The data interface 1726 is connected to, for example, an in-vehicle network 1741 via a terminal (not shown), and acquires data generated by the vehicle (such as vehicle speed data).

[0269] The content player 1727 reproduces content stored in a storage medium (such as a CD and a DVD) inserted into the storage medium interface 1728. The input device 1729 includes, for example, a touch sensor, a button, or a switch configured to detect a touch on the screen of the display device 1730, and receives operations or information input from the user. The display device 1730 includes a screen such as an LCD or an OLED display, and displays images of the navigation function or reproduced content. The speaker 1731 outputs sounds of the navigation function or reproduced content.

[0270] The wireless communication interface 1733 supports any cellular communication scheme (such as LTE and LTE-Advanced), and performs wireless communication. The wireless communication interface 1733 generally may include, for example, a BB processor 1734 and an RF circuit 1735. The BB processor 1734 may perform, for example, encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and perform various types of signal processing for wireless communication. At the same time, the RF circuit 1735 may include, for example, a mixer, a filter, and an amplifier, and transmits and receives wireless signals via the antenna 1737. The wireless communication interface 1733 may also be a single chip module on which the BB processor 1734 and the RF circuit 1735 are integrated. AsFigure 24 As shown, the wireless communication interface 1733 may include a plurality of BB processors 1734 and a plurality of RF circuits 1735. Although Figure 24 an example in which the wireless communication interface 1733 includes a plurality of BB processors 1734 and a plurality of RF circuits 1735 is shown, the wireless communication interface 1733 may also include a single BB processor 1734 or a single RF circuit 1735.

[0271] In addition, in addition to the cellular communication scheme, the wireless communication interface 1733 may support other types of wireless communication schemes, such as short-range wireless communication schemes, near-field communication schemes, and wireless LAN schemes. In this case, for each wireless communication scheme, the wireless communication interface 1733 may include a BB processor 1734 and an RF circuit 1735.

[0272] Each of the antenna switches 1736 switches the connection destination of the antenna 1737 among a plurality of circuits (such as circuits for different wireless communication schemes) included in the wireless communication interface 1733.

[0273] Each of the antennas 1737 includes a single or a plurality of antenna elements (such as a plurality of antenna elements included in a MIMO antenna), and is used for the wireless communication interface 1733 to transmit and receive wireless signals. As Figure 24 shown, the car navigation device 1720 may include a plurality of antennas 1737. Although Figure 24 an example in which the car navigation device 1720 includes a plurality of antennas 1737 is shown, the car navigation device 1720 may also include a single antenna 1737.

[0274] In addition, the car navigation device 1720 may include an antenna 1737 for each wireless communication scheme. In this case, the antenna switch 1736 may be omitted from the configuration of the car navigation device 1720.

[0275] The battery 1738 supplies power to each block of the car navigation device 1720 shown via a feeder line, which is partially shown as a dotted line in the figure. The battery 1738 accumulates the power supplied from the vehicle. Figure 24 shown, the car navigation device 1720 of

[0276] The technology of the present disclosure may also be implemented as an in-vehicle system (or vehicle) 1740 including one or more blocks of the car navigation device 1720, the in-vehicle network 1741, and the vehicle module 1742. The vehicle module 1742 generates vehicle data (such as vehicle speed, engine speed, and fault information), and outputs the generated data to the in-vehicle network 1741.

[0277] The exemplary embodiments of the present disclosure have been described above with reference to the accompanying drawings. However, the present disclosure is of course not limited to the above examples. Those skilled in the art can obtain various changes and modifications within the scope of the appended claims, and it should be understood that these changes and modifications will naturally fall within the technical scope of the present disclosure.

[0278] It should be understood that the machine-executable instructions in the machine-readable storage medium or program product according to the embodiments of the present disclosure can be configured to perform operations corresponding to the above device and method embodiments. When referring to the above device and method embodiments, the embodiments of the machine-readable storage medium or program product are clear to those skilled in the art, and thus will not be described repeatedly. The machine-readable storage medium and program product for carrying or including the above machine-executable instructions also fall within the scope of the present disclosure. Such a storage medium may include, but is not limited to, a floppy disk, an optical disk, a magneto-optical disk, a memory card, a memory stick, and the like.

[0279] In addition, it should be understood that the above series of processes and devices can also be implemented by software and / or firmware. In the case of implementation by software and / or firmware, a corresponding program constituting the corresponding software is stored in the storage medium of the relevant device, and when the program is executed, various functions can be performed.

[0280] For example, multiple functions included in one unit in the above embodiments can be implemented by separate devices. Alternatively, multiple functions implemented by multiple units in the above embodiments can be respectively implemented by separate devices. In addition, one of the above functions can be implemented by multiple units. Needless to say, such a configuration is included in the technical scope of the present disclosure.

[0281] In this specification, the steps described in the flowcharts include not only the processes executed in time series in the described order, but also processes executed in parallel or individually rather than necessarily in time series. In addition, even in the steps of processing in time series, needless to say, the order can also be appropriately changed.

[0282] Although the present disclosure and its advantages have been described in detail, it should be understood that various changes, substitutions, and transformations can be made without departing from the spirit and scope of the present disclosure as defined by the appended claims. Moreover, the term "comprising" in the embodiments of the present disclosure, "including" or any other variant thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or device including a series of elements includes not only those elements but also other elements not explicitly listed, or also includes elements inherent to such a process, method, article, or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article, or device including the said element.

[0283] In addition, the present disclosure may also have the following configurations:

[0284] (1) A first electronic device for a wireless communication system, comprising:

[0285] At least one processor; and

[0286] At least one memory, including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to notify data to a second electronic device, wherein the data is notified through one of the following notification modes:

[0287] A first notification mode, configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device;

[0288] A second notification mode, configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and

[0289] A third notification mode, configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0290] (2) The first electronic device according to (1), wherein

[0291] The first notification mode is configured as any one of the following modes: (1) deploying an AI model at the first electronic device, the first electronic device using the deployed AI model to process the data and sending the processed data to the second electronic device, (2) deploying an AI model at the second electronic device, the second electronic device using the deployed AI model to receive the data from the first electronic device, and (3) deploying an AI model at the second electronic device, the second electronic device using the deployed AI model to predict the data based at least on historical data related to the data;

[0292] The second notification mode is configured as: the first electronic device using the first AI model in the pair of AI models to process the data and sending the processed data to the second electronic device, and the second electronic device using the second AI model in the pair of AI models to reconstruct the data based on the processed data received from the first electronic device.

[0293] (3) The first electronic device according to (1) or (2), wherein

[0294] The notification mode is at least selected based on communication metrics,

[0295] The communication metrics include at least one or more of the following metrics: the transmission delay on the communication link between the first electronic device and the second electronic device, the data transmission accuracy requirement, the communication resource overhead, the transmission rate between the first electronic device and the second electronic device, the queue length of the data that the first electronic device is ready to send to the second electronic device, and the computing power of the first electronic device and / or the second electronic device.

[0296] (4) The first electronic device according to (1) or (2), wherein the data to be notified is channel state information (CSI) feedback.

[0297] (5) The first electronic device according to (4), wherein

[0298] the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to determine the queue length of the communication data ready to be sent to the second electronic device, and

[0299] the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to:

[0300] determine which notification mode to use to notify the CSI feedback based on the queue length, and send indication information indicating which notification mode to use to notify the CSI feedback to the second electronic device, or

[0301] send information indicating the queue length to the second electronic device, so that the second electronic device determines which notification mode to use to notify the CSI feedback based on the queue length.

[0302] (6) The first electronic device according to (5), wherein

[0303] in response to the queue length being less than or equal to a first threshold, cause the second electronic device to obtain the CSI feedback based on a first notification mode,

[0304] in response to the queue length being greater than the first threshold and less than or equal to a second threshold, the first electronic device uses a second notification mode to notify the second electronic device of the CSI feedback, or

[0305] in response to the queue length being greater than the second threshold, the first electronic device uses a third notification mode to send the CSI feedback to the second electronic device.

[0306] (7) The first electronic device according to (5), wherein

[0307] the information indicating the queue length or the indication information is sent through a physical uplink control channel (PUCCH).

[0308] (8) The first electronic device according to (1) or (2), wherein the data to be notified is video data.

[0309] (9) The first electronic device according to (8), wherein

[0310] the at least one memory and the computer program code are further configured to, by means of the at least one processor, cause the first electronic device to determine a transmission delay for transmission from the first electronic device to the second electronic device, and wherein

[0311] in response to the transmission delay being less than or equal to a first threshold, the first electronic device uses a third notification mode to send the video data to the second electronic device,

[0312] in response to the transmission delay being greater than the first threshold and less than or equal to a second threshold, the first electronic device uses a second notification mode to notify the second electronic device of the video data, or

[0313] in response to the transmission delay being greater than the second threshold, the second electronic device is caused to obtain the video data based on a first notification mode.

[0314] (10) The first electronic device according to (1) or (2), wherein

[0315] the at least one memory and the computer program code are configured to, by means of the at least one processor, cause the first electronic device to send data to the second electronic device as one of a group of sending-end devices, the group of sending-end devices including one or more sending-end devices,

[0316] wherein on both sides of each sending-end device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed, and,

[0317] wherein the at least one memory and the computer program code are further configured to, by means of the at least one processor, cause the first electronic device to send the data to the second electronic device using a second notification mode based on a first AI model pair selected from the plurality of AI model pairs.

[0318] (11) A second electronic device for a wireless communication system, comprising:

[0319] at least one processor; and

[0320] at least one memory, including computer program code, wherein the at least one memory and the computer program code are configured to, by means of the at least one processor, cause the second electronic device to obtain data to be notified by the first electronic device, wherein the data is notified by one of the following notification modes:

[0321] The first notification mode is configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device;

[0322] The second notification mode is configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and

[0323] The third notification mode is configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0324] (12) The second electronic device according to (11), wherein

[0325] The first notification mode is configured to be any one of the following modes: (1) Deploy an AI model at the first electronic device, and the first electronic device uses the deployed AI model to process the data and send the processed data to the second electronic device; (2) Deploy an AI model at the second electronic device, and the second electronic device uses the deployed AI model to receive the data from the first electronic device; and (3) Deploy an AI model at the second electronic device, and the second electronic device uses the deployed AI model to predict the data based at least on historical data related to the data;

[0326] The second notification mode is configured to: The first electronic device uses the first AI model in the pair of AI models to process the data and send the processed data to the second electronic device, and the second electronic device uses the second AI model in the pair of AI models to reconstruct the data based on the processed data received from the first electronic device.

[0327] (13) The second electronic device according to (11) or (12), wherein

[0328] The notification mode is at least selected based on communication metrics,

[0329] The communication metrics at least include one or more of the following metrics: transmission delay on the communication link between the first electronic device and the second electronic device, data transmission accuracy requirement, communication resource overhead, transmission rate between the first electronic device and the second electronic device, queue length of the data that the first electronic device is ready to send to the second electronic device, and computing capabilities of the first electronic device and / or the second electronic device.

[0330] (14) The second electronic device according to (11) or (12), wherein the data to be notified is channel state information (CSI) feedback.

[0331] (15) The second electronic device as described in (14), wherein,

[0332] the at least one memory and the computer program code are further configured to, by means of the at least one processor, cause the second electronic device to:

[0333] receive indication information from the first electronic device indicating which notification mode is to be used to notify CSI feedback, wherein the indication information is determined based on the queue length of the communication data that the first electronic device is ready to send to the second electronic device, or

[0334] receive information indicating the queue length of the communication data that the first electronic device is ready to send to the second electronic device from the first electronic device, and determine which notification mode is to be used to notify CSI feedback based on the queue length.

[0335] (16) The second electronic device as described in (15), wherein,

[0336] in response to the queue length being less than or equal to a first threshold, the second electronic device obtains CSI feedback based on a first notification mode,

[0337] in response to the queue length being greater than the first threshold and less than or equal to a second threshold, the second electronic device obtains CSI feedback from the first electronic device based on a second notification mode, or

[0338] in response to the queue length being greater than the second threshold, the second electronic device uses a third notification mode to receive CSI feedback from the first electronic device.

[0339] (17) The second electronic device as described in (15), wherein,

[0340] the information indicating the queue length or the indication information is sent through a physical uplink control channel (PUCCH).

[0341] (18) The second electronic device as described in (11) or (12), wherein the data to be notified is video data.

[0342] (19) The second electronic device as described in (18), wherein,

[0343] the at least one memory and the computer program code are further configured to, by means of the at least one processor, cause the second electronic device to determine the transmission delay for transmission from the first electronic device to the second electronic device, and wherein

[0344] in response to the transmission delay being less than or equal to a first threshold, the second electronic device uses a third notification mode to receive the video data from the first electronic device,

[0345] In response to the transmission delay being greater than a first threshold and less than or equal to a second threshold, the second electronic device obtains the video data from the first electronic device based on a second notification mode, or

[0346] In response to the transmission delay being greater than the second threshold, the second electronic device obtains the video data based on a first notification mode.

[0347] (20) The second electronic device according to (11) or (12), wherein,

[0348] The at least one memory and the computer program code are configured to, via the at least one processor, cause the second electronic device to receive data from each sending device in a group of sending devices including the first electronic device, the group of sending devices including one or more sending devices,

[0349] wherein, on both sides of each sending device and the second electronic device, a pair of jointly trained artificial intelligence (AI) models are deployed, and,

[0350] wherein the at least one memory and the computer program code are further configured to, via the at least one processor, cause the second electronic device to obtain the data from each sending device using a second communication mode based on a corresponding AI model pair selected from the multiple AI model pairs.

[0351] (21) A method for a first electronic device in a wireless communication system, including notifying data to a second electronic device, wherein the data is notified through one of the following notification modes:

[0352] A first notification mode, configured to use an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device to send or receive the data;

[0353] A second notification mode, configured to use a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device to send and receive the data; and

[0354] A third notification mode, configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0355] (22) A method for a second electronic device in a wireless communication system, including obtaining data to be notified by a first electronic device, wherein the data is notified through one of the following notification modes:

[0356] The first notification mode is configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side in the first electronic device or the second electronic device;

[0357] The second notification mode is configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and

[0358] The third notification mode is configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

[0359] (23) A first electronic device for a wireless communication system, comprising:

[0360] At least one processor; and

[0361] At least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to send data to a second electronic device as one of a group of sender devices, the group of sender devices including one or more sender devices,

[0362] wherein a plurality of pairs of jointly trained artificial intelligence (AI) models are deployed on both sides of each sender device and the second electronic device, and,

[0363] wherein the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to process the data by using a first AI model in a first pair of AI models selected from the plurality of pairs of AI models and send the processed data to the second electronic device, such that the second electronic device uses a second AI model in the first pair of AI models to reconstruct the data based on the processed data received from the first electronic device.

[0364] (24) The first electronic device according to (23), wherein,

[0365] The first pair of AI models is selected based on the priority of the first electronic device in the group of sender devices.

[0366] (25) The first electronic device according to (24), wherein,

[0367] The at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to receive information indicating the priority from the second electronic device.

[0368] (26) The first electronic device as described in (24) or (25), wherein

[0369] the priority is determined based on at least one of the following: the communication scenario between the set of sending devices and the second electronic device, and the importance of the data respectively sent by each sending device in the set of sending devices to the second electronic device.

[0370] (27) The first electronic device as described in (23) or (24), wherein

[0371] the first pair of AI models is selected based on the accuracy of reconstructing the data.

[0372] (28) The first electronic device as described in (27), wherein

[0373] the accuracy includes any one or a combination of the following: the verification result of the reconstructed data, and the accuracy rate calculated based on the comparison between the data and the reconstructed data.

[0374] (29) The first electronic device as described in (23) or (24), wherein

[0375] the at least one memory and the computer program code are configured to cause the first electronic device to send, via the at least one processor, transmission parameter information indicating transmission parameters to the second electronic device, or

[0376] the at least one memory and the computer program code are configured to cause the first electronic device to receive, via the at least one processor, transmission parameter information indicating transmission parameters from the second electronic device,

[0377] wherein the transmission parameter information at least includes the ID of the selected pair of AI models and information indicating the time to enable the pair of AI models.

[0378] (30) The first electronic device as described in (29), wherein

[0379] the transmission parameter information is sent or received via one of the following signaling: Radio Resource Control (RRC) signaling, Uplink Control Information (UCI), or Downlink Control Information (DCI).

[0380] (31) The first electronic device as described in (23) or (24), wherein

[0381] During the training of the multiple pairs of AI models, the AI model parameters are adjusted based on the accuracy of reconstructing the data fed back by the second electronic device to the first electronic device.

[0382] (32)A second electronic device for a wireless communication system, comprising:

[0383] At least one processor; and

[0384] At least one memory, including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to receive data from each sending device in a group of sending devices, the group of sending devices including one or more sending devices,

[0385] wherein, on both sides of each sending device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed, and,

[0386] wherein the at least one memory and the computer program code are further configured to, through the at least one processor, cause the second electronic device to receive processed data obtained by processing the data using a first AI model in a corresponding AI model pair selected from the plurality of AI model pairs from each sending device, and reconstruct the data based on the received processed data using a second AI model in the corresponding AI model pair.

[0387] (33)The second electronic device according to (32), wherein,

[0388] the corresponding AI model pair is selected based on the priority of each sending device in the group of sending devices.

[0389] (34)The second electronic device according to (33), wherein,

[0390] the at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to send information indicating the priority to each sending device.

[0391] (35)The second electronic device according to (33) or (34), wherein,

[0392] the priority is determined based on at least one of the following: the communication scenario between the group of sending devices and the second electronic device, and the importance of the data respectively sent by each sending device in the group of sending devices to the second electronic device.

[0393] (36)The second electronic device according to (32) or (33), wherein,

[0394] the at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to feedback the accuracy of reconstructing the data to each sending device, and

[0395] wherein, the corresponding AI model pairs are selected based on the accuracy.

[0396] (37) The second electronic device as described in (36), wherein,

[0397] The accuracy includes any one or a combination of the following: the verification result of the reconstructed data, and the accuracy rate calculated based on the comparison between the data and the reconstructed data.

[0398] (38) The second electronic device as described in (32) or (33), wherein,

[0399] The at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to send transmission parameter information indicating transmission parameters to the first electronic device, or,

[0400] The at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to receive transmission parameter information indicating transmission parameters from the first electronic device,

[0401] wherein, the transmission parameter information at least includes the ID of the selected AI model pair and the information indicating the time to enable the AI model pair.

[0402] (39) The second electronic device as described in (38), wherein,

[0403] The transmission parameter information is sent or received via one of the following signaling: Radio Resource Control (RRC) signaling, Uplink Control Information (UCI), or Downlink Control Information (DCI).

[0404] (40) The second electronic device as described in (32) or (33), wherein,

[0405] The at least one memory and the computer program code are configured to, through the at least one processor, cause the second electronic device to feedback the reconstruction accuracy of the data during the training of the multiple AI model pairs, so that the AI model parameters are dynamically adjusted based on the accuracy.

[0406] (41) A method for a first electronic device in a wireless communication system, including the first electronic device sending data to a second electronic device as one of a group of sending device, the group of sending devices including one or more sending devices,

[0407] wherein, on both sides of each sending device and the second electronic device, multiple jointly trained artificial intelligence (AI) model pairs are deployed, and,

[0408] Wherein, the method further comprises processing the data using a first AI model in a first AI model pair selected from the plurality of AI model pairs, and sending the processed data to a second electronic device, such that the second electronic device uses a second AI model in the first AI model pair to reconstruct the data based on the processed data received from the first electronic device.

[0409] (42) A method for a second electronic device in a wireless communication system, comprising receiving data from each of a group of transmitting devices, the group of transmitting devices including one or more transmitting devices,

[0410] Wherein, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed on both sides of each transmitting device and the second electronic device, and,

[0411] Wherein, the method further comprises receiving processed data obtained by processing the data using a first AI model in a corresponding AI model pair selected from the plurality of AI model pairs from each transmitting device, and reconstructing the data based on the received processed data using a second AI model in the corresponding AI model pair.

[0412] (43) A non-transitory computer-readable storage medium storing executable instructions, the executable instructions, when executed, implementing the method according to any one of (21), (22), (41), and (42).

[0413] (44) A computer program product comprising executable instructions, the executable instructions, when executed, implementing the method according to any one of (21), (22), (41), and (42).

Claims

1. A first electronic device for a wireless communication system, comprising: at least one processor; and at least one memory including computer program code, wherein the at least one memory and the computer program code are configured to, through the at least one processor, cause the first electronic device to notify a second electronic device of data, and the data is notified through one of the following notification modes: A first notification mode, configured to send or receive the data by using an artificial intelligence (AI) model deployed on one side of the first electronic device or the second electronic device; A second notification mode, configured to send and receive the data by using a pair of jointly trained AI models deployed on both sides of the first electronic device and the second electronic device; and A third notification mode, configured to send the data from the first electronic device to the second device without using an AI model for notifying the data at both the first electronic device and the second electronic device.

2. The first electronic device according to claim 1, wherein the first notification mode is configured to be any one of the following modes: (1) an AI model is deployed at the first electronic device, and the first electronic device uses the deployed AI model to process the data and send the processed data to the second electronic device; (2) an AI model is deployed at the second electronic device, and the second electronic device uses the deployed AI model to receive the data from the first electronic device; and (3) an AI model is deployed at the second electronic device, and the second electronic device uses the deployed AI model to predict the data based at least on historical data related to the data; the second notification mode is configured to: the first electronic device uses the first AI model in the pair of AI models to process the data and send the processed data to the second electronic device, and the second electronic device uses the second AI model in the pair of AI models to reconstruct the data based on the processed data received from the first electronic device.

3. The first electronic device according to claim 1 or 2, wherein the notification mode is at least selected based on communication metrics, and the communication metrics at least include one or more of the following metrics: transmission delay on the communication link between the first electronic device and the second electronic device, data transmission accuracy requirement, communication resource overhead, transmission rate between the first electronic device and the second electronic device, queue length of the data that the first electronic device is ready to send to the second electronic device, and computing capabilities of the first electronic device and / or the second electronic device.

4. The first electronic device according to claim 1 or 2, wherein the data to be notified is channel state information (CSI) feedback.

5. The first electronic device according to claim 4, wherein the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to determine the queue length of the communication data ready to be sent to the second electronic device, and the at least one memory and the computer program code are further configured to, through the at least one processor, cause the first electronic device to: Determine which notification mode to use to notify CSI feedback based on the queue length, and send indication information indicating which notification mode to use to notify CSI feedback to a second electronic device, or Send information indicating the queue length to a second electronic device, so that the second electronic device determines which notification mode to use to notify CSI feedback based on the queue length.

6. The first electronic device according to claim 5, wherein, In response to the queue length being less than or equal to a first threshold, cause the second electronic device to obtain CSI feedback based on a first notification mode, In response to the queue length being greater than the first threshold and less than or equal to a second threshold, the first electronic device uses a second notification mode to notify the second electronic device of CSI feedback, or In response to the queue length being greater than the second threshold, the first electronic device uses a third notification mode to send CSI feedback to the second electronic device.

7. The first electronic device according to claim 5, wherein, The information indicating the queue length or the indication information is sent through a physical uplink control channel (PUCCH).

8. The first electronic device according to claim 1 or 2, wherein, The data to be notified is video data.

9. The first electronic device according to claim 8, wherein, The at least one memory and computer program code are further configured to, through the at least one processor, cause the first electronic device to determine a transmission delay for transmission from the first electronic device to the second electronic device, and wherein In response to the transmission delay being less than or equal to a first threshold, the first electronic device uses a third notification mode to send the video data to the second electronic device, In response to the transmission delay being greater than the first threshold and less than or equal to a second threshold, the first electronic device uses a second notification mode to notify the second electronic device of the video data, or In response to the transmission delay being greater than the second threshold, cause the second electronic device to obtain the video data based on a first notification mode.

10. The first electronic device according to claim 1 or 2, wherein, The at least one memory and computer program code are configured to, through the at least one processor, cause the first electronic device to send data to a second electronic device as one of a group of sender devices, the group of sender devices including one or more sender devices, wherein, on both sides of each sender device and the second electronic device, a plurality of jointly trained artificial intelligence (AI) model pairs are deployed, and, wherein, the at least one memory and computer program code are further configured to, through the at least one processor, cause the first electronic device to use a second notification mode to send the data to the second electronic device based on a first AI model pair selected from the plurality of AI model pairs.

Images