Method and apparatus for artificial intelligence or machine learning training

Abstract

Aspects of the present disclosure provide a method and device for artificial intelligence / machine learning (AI / ML) model training in a wireless communication network to overcome technical issues related to the transmission of AI / ML model training data, such as the inability to determine the importance of data due to the same data type. A first device receives, collectively or separately, AI / ML model training assistance information and information related to the transmission of each AI / ML model training data from a second device. The first device determines data status information (DSI) for each AI / ML model training data based on the AI / ML model training assistance information. The first device transmits each AI / ML model training data to the second device based on at least one of the DSI for each AI / ML model training data or the information related to the transmission of each AI / ML model training data.

Description

[Technical Field]

[0001] The present disclosure relates generally to wireless communications and, in particular embodiments, to methods and apparatus for artificial intelligence or machine learning (AI / ML) training. [Background technology]

[0002] Artificial intelligence techniques may be applied to communications, including artificial intelligence or machine learning (AI / ML)-based communications at the physical layer and / or at the medium access control (MAC) layer. For example, at the physical layer, AI / ML-based communications may aim to optimize component design and / or improve algorithm performance. For the MAC layer, AI / ML-based communications may aim to utilize AI / ML capabilities to learn, predict, and / or make decisions to solve complex optimization problems with possible better strategies and / or optimal solutions, for example, to optimize functionality at the MAC layer.

[0003] In some implementations, AI / ML architectures in wireless communication networks may involve multiple nodes, which may be organized in one of two modes: centralized and distributed, both of which may be deployed in access networks, core networks, or edge computing systems or third-party networks. Centralized training and computing architectures are sometimes limited by large communication overhead and strict user data privacy. Distributed training and computing architectures may include several frameworks, such as distributed machine learning and federated learning.

[0004] However, communications in wireless communications systems, including communications related to AI / ML model training at multiple nodes, typically occur over non-ideal channels. Non-ideal conditions, such as electromagnetic interference, signal degradation, phase delay, fading, and other non-idealities, can attenuate and / or distort communication signals or potentially interfere with or degrade the communication capabilities of the system.

[0005] Conventional AI / ML model training processes generally rely on hybrid automatic repeat request (HARQ) feedback and retransmission processes to attempt to ensure that data communicated between devices involved in AI / ML model training is successfully received. However, the communication overhead and delays associated with such retransmissions can be problematic.

[0006] Additionally, the processing power and / or availability of training data for AI / ML training processes can vary significantly among different nodes / devices, meaning that the ability of different nodes to productively participate in an AI / ML model training process can vary significantly. In practice, such imbalances often mean that the training delay of an AI / ML model training process involving multiple nodes / devices, such as a distributed learning or federated learning-based AI / ML model training process, is dominated by the node / device with the largest delay due to communication and / or computational delays.

[0007] Therefore, one way to reduce training delays for an AI / ML model training process may be to minimize communication and / or computational delays. These delays may be reduced by utilizing only critical data, e.g., by transmitting only critical data and / or by performing the AI / ML model training process using only critical data.

[0008] In existing wireless communication systems, such as 5G network systems, the importance of data can be determined based on the Quality of Service (QoS) defined in higher layers. For example, the priority in the 5G QoS Identifier (5QI) can be used to indicate the importance of data. For uplink data scheduling, the priority in the 5QI is mapped to the priority in the logical channels of the MAC layer. When a user equipment (UE) performs processing such as MAC power distribution unit (PDU) multiplexing and assembly, all selected logical channels are served in descending order of priority. In other words, logical channels with higher priorities have higher opportunities for transmission and are therefore considered to have higher importance levels.

[0009] The data importance defined in the higher layer may be associated with a specific data type. Each data type may be considered to have a certain data importance level. In other words, each data type is associated with a respective data importance level, i.e., different data types exhibit different data importance levels. Therefore, the importance of data may be determined based on its data type. However, when the data type is the same, it is unclear how the data importance should be determined. This may be particularly problematic for intelligent communication systems, such as 6G networks, in which AI / ML models are deployed.

[0010] For these and other reasons, new methods and devices are desired so that new AI-enabled applications and processes can be implemented while minimizing the signaling and communication overhead and delays associated with existing AI / ML model training procedures. Summary of the Invention

[0011] Existing artificial intelligence or machine learning (AI / ML) model training processes have limitations. For example, as described above, it is not clear how data importance should be determined when the data type is the same, which can be particularly problematic for intelligent communication systems, such as 6G wireless networks, in which AI / ML models are deployed. In 6G networks, user equipment (UE) may collect AI / ML model training data samples and report the collected samples to a network (e.g., a base station (BS)). The AI / ML model training data samples may include reference signals or measurements from sensors (e.g., cameras). The source of the AI / ML model training data may be stable. For example, a reference signal is periodically transmitted from a BS. However, the content of the AI / ML training data samples may change at different time instances. This means that the importance of an AI / ML training data sample may change at different time instances. For example, the importance of an AI / ML training data sample collected at one time instance may be higher than the importance of another AI / ML training data sample collected at another time instance. However, because each AI / ML training data sample can be the same type of data, the importance of an AI / ML training data sample is not necessarily determined based on data type.

[0012] In addition, in a 6G intelligent communication system, the performance of an AI / ML model may be determined not only based on inference accuracy but also based on transmission overhead and latency. If a UE transmits all collected AI / ML model training data samples regardless of their data importance level, enormous network resources may be required for transmitting the AI / ML training data, thus degrading the overall performance of the AI / ML model and the communication system.

[0013] Existing communication systems (e.g., 5G networks) have some problems with determining or evaluating data importance. For example, in 5G, data importance may be determined or evaluated solely based on the priority in a 5G QoS identifier (5QI) determined at a higher layer. However, the physical layer in 5G cannot determine data importance. Furthermore, as mentioned above, the concept of data importance does not exist for data having the same data type. Without determining data importance, the overall performance of AI / ML models and intelligent communication systems in 6G may be degraded due to, for example, huge transmission overhead.

[0014] Aspects of the present disclosure provide solutions to overcome at least some of the aforementioned limitations, e.g., particular methods and devices for artificial intelligence or machine learning (AI / ML) model training.

[0015] According to a first broad aspect of the present disclosure, provided herein is a method for supporting AI / ML model training in a wireless communications network. The method according to the first broad aspect of the present disclosure may include receiving, by a first device, AI / ML model training assistance information from a second device. The method according to the first broad aspect of the present disclosure may further include determining, by the first device, data state information (DSI) for each AI / ML model training data based on the AI / ML model training assistance information. The method according to the first broad aspect of the present disclosure may further include receiving, by the first device, information related to transmission of the each AI / ML model training data from the second device. The method according to the first broad aspect of the present disclosure may further include transmitting, by the first device, each AI / ML model training data to the second device based on at least one of the DSI for the each AI / ML model training data or the information related to the transmission of the each AI / ML model training data.

[0016] In some embodiments of the method according to the first broad aspect of this disclosure, the respective AI / ML model training data is selectively transmitted, the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, and the processor-executable instructions, when executed, further include processor-executable instructions that cause the processor to determine whether the respective AI / ML model training data should be transmitted to the second device based on the DSI threshold and the DSI of the respective AI / ML model training data.

[0017] In some embodiments of the method according to the first broad aspect of this disclosure, the respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data should be transmitted to the second device, and the processor-executable instructions, when executed, further include processor-executable instructions that cause the processor to transmit at least one of DSI or AI / ML model training dataset information for the respective AI / ML model training data to the second device and receive information from the second device indicating whether the respective AI / ML model training data should be transmitted to the second device.

[0018] In some embodiments of the method according to the first broad aspect of this disclosure, the AI / ML model training dataset information includes at least one of an AI / ML model training dataset size, or DSI variance information of the AI / ML model training dataset.

[0019] In some embodiments of a method according to a first broad aspect of this disclosure, the AI / ML model training dataset information is transmitted using a buffer status report (BSR) or a scheduling request (SR).

[0020] In some embodiments of a method according to the first broad aspect of this disclosure, the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data.

[0021] In some embodiments of a method according to a first broad aspect of the disclosure, each reporting format indicates a respective transmission accuracy, and the respective transmission accuracy level is determined based on the level of DSI of the respective AI / ML model training data.

[0022] In some embodiments of a method according to the first broad aspect of this disclosure, the respective reporting formats indicate settings for transmission of the respective AI / ML model training data.

[0023] In some embodiments of a method according to a first broad aspect of this disclosure, the configuration for the transmission of the respective AI / ML model training data indicates at least one of resources to be used for the transmission of the respective AI / ML model training data or a quantization granularity to be used for the transmission of the respective AI / ML model training data.

[0024] In some embodiments of a method according to the first broad aspect of this disclosure, the respective report format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

[0025] In some embodiments of the method according to the first broad aspect of this disclosure, the relationship between the DSI of each AI / ML model training data and the respective reporting format is set by the second device.

[0026] In some embodiments of the method according to the first broad aspect of this disclosure, the AI / ML model training assistance information includes at least one of information about a reference AI / ML model or at least one reference input data value.

[0027] In some embodiments of the method according to the first broad aspect of the disclosure, the information about the reference AI / ML model includes at least one of a reference AI / ML model type, a reference AI / ML model structure, one or more reference AI / ML model parameters, a reference AI / ML model gradient, a reference AI / ML model activation function, a reference AI / ML model input data type, a reference AI / ML model output data type, a reference AI / ML model input data dimension, or a reference AI / ML model output data dimension.

[0028] In some embodiments of the method according to the first broad aspect of this disclosure, determining the DSI for each of the AI / ML model training data includes inputting the each of the AI / ML model training data to a reference AI / ML model and determining the DSI based on an output of the reference AI / ML model.

[0029] In some embodiments of a method according to the first broad aspect of this disclosure, each AI / ML model training data input to the reference AI / ML model replaces at least one reference input data value.

[0030] In some embodiments of the method according to the first broad aspect of this disclosure, the processor-executable instructions further include processor-executable instructions that, when executed, cause the processor to receive updated AI / ML model training assistance information from the second device.

[0031] In some embodiments of the method according to the first broad aspect of this disclosure, the DSI of the respective AI / ML model training data includes information indicative of at least one of: data uncertainty of the respective AI / ML model training data, data importance of the respective AI / ML model training data, a degree of requirement of the respective AI / ML model training data for AI / ML model training, or data diversity of the respective AI / ML model training data.

[0032] In some embodiments of a method according to the first broad aspect of this disclosure, the data uncertainty of each AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

[0033] In some embodiments of the method according to the first broad aspect of this disclosure, the AI / ML model training is performed at least in part by the second device.

[0034] In some embodiments of a method according to the first broad aspect of this disclosure, the respective AI / ML model training data includes at least one of local AI / ML model training data for a local AI / ML model of the first device or local gradients associated with the local AI / ML model.

[0035] In some embodiments of the method according to the first broad aspect of this disclosure, the first and second devices cooperate for AI / ML model training, and the processor-executable instructions further include processor-executable instructions that, when executed, cause the processor to perform a portion of the AI / ML model training before determining the DSI of the respective AI / ML model training data, the respective AI / ML model training data including respective outputs of the portion of the AI / ML model training.

[0036] According to a second broad aspect of the present disclosure, provided herein is a method for AI / ML model training in a wireless communications network. The method according to the second broad aspect of the present disclosure may include transmitting, by a first device, AI / ML model training assistance information to a second device for use in determining data state information (DSI) for the respective AI / ML model training data. The method according to the second broad aspect of the present disclosure may further include transmitting, by the first device, information related to the transmission of the respective AI / ML model training data to the second device. The method according to the second broad aspect of the present disclosure may further include receiving, by the first device, the respective AI / ML model training data from the second device, wherein the respective AI / ML model training data is transmitted based on at least one of the DSI for the respective AI / ML model training data or the information related to the transmission of the respective AI / ML model training data. The method according to the second broad aspect of the present disclosure may further include performing, by the first device, AI / ML model training using the respective AI / ML model training data.

[0037] In some embodiments of the method according to the second broad aspect of this disclosure, the respective AI / ML model training data is selectively transmitted, the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, and the processor-executable instructions further include processor-executable instructions that, when executed, cause the processor to set the DSI threshold for use in determining whether the respective AI / ML model training data should be transmitted to the first device.

[0038] In some embodiments of the method according to the second broad aspect of the disclosure, the respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data should be transmitted to the first device, and the processor-executable instructions, when executed, cause the processor to receive from the second device at least one of DSI or AI / ML model training dataset information for the respective AI / ML model training data; determine using the at least one of the DSI or AI / ML model training dataset information for the respective AI / ML model training data whether the respective AI / ML model training data should be transmitted to the first device; and transmit to the second device the information indicating whether the respective AI / ML model training data should be transmitted to the first device.

[0039] In some embodiments of the method according to the second broad aspect of this disclosure, the AI / ML model training dataset information includes at least one of an AI / ML model training dataset size or DSI variance information of the AI / ML model training dataset.

[0040] In some embodiments of a method according to the second broad aspect of this disclosure, the AI / ML model training dataset information is transmitted using a buffer status report (BSR) or a scheduling request (SR).

[0041] In some embodiments of a method according to the second broad aspect of this disclosure, the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data.

[0042] In some embodiments of a method according to the second broad aspect of the present disclosure, the respective reporting formats indicate respective transmission accuracy levels, the respective transmission accuracy levels being determined based on the level of DSI of the respective AI / ML model training data.

[0043] In some embodiments of a method according to the second broad aspect of this disclosure, the respective reporting formats indicate settings for transmission of the respective AI / ML model training data.

[0044] In some embodiments of the method according to the second broad aspect of this disclosure, the configuration for the transmission of the respective AI / ML model training data indicates at least one of resources to be used for the transmission of the respective AI / ML model training data or a quantization granularity to be used for the transmission of the respective AI / ML model training data.

[0045] In some embodiments of a method according to the second broad aspect of this disclosure, the respective reporting format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

[0046] In some embodiments of the method according to the second broad aspect of this disclosure, the processor-executable instructions further include processor-executable instructions that, when executed, cause the processor to establish a relationship between the DSI of each AI / ML model training data and the respective reporting format.

[0047] In some embodiments of the method according to the second broad aspect of this disclosure, the AI / ML model training assistance information includes at least one of information about a reference AI / ML model or at least one reference input data value.

[0048] In some embodiments of the method according to the second broad aspect of the disclosure, the information about the reference AI / ML model includes at least one of a reference AI / ML model type, a reference AI / ML model structure, one or more reference AI / ML model parameters, a reference AI / ML model gradient, a reference AI / ML model activation function, a reference AI / ML model input data type, a reference AI / ML model output data type, a reference AI / ML model input data dimension, or a reference AI / ML model output data dimension.

[0049] In some embodiments of the method according to the second broad aspect of this disclosure, the processor-executable instructions further include processor-executable instructions that, when executed, cause the processor to update the AI / ML model training assistance information and transmit the updated AI / ML model training assistance information to the second device.

[0050] In some embodiments of the method according to the second broad aspect of this disclosure, the DSI of the respective AI / ML model training data includes information indicative of at least one of: data uncertainty of the respective AI / ML model training data, data importance of the respective AI / ML model training data, a degree of requirement of the respective AI / ML model training data for AI / ML model training, or data diversity of the respective AI / ML model training data.

[0051] In some embodiments of the method according to the second broad aspect of this disclosure, the DSI for each AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

[0052] In some embodiments of the method according to the second broad aspect of the present disclosure, the respective AI / ML model training data includes at least one of local AI / ML model training data for a local AI / ML model of the second device or local gradients associated with the local AI / ML model.

[0053] In some embodiments of a method according to the second broad aspect of the disclosure, the first and second devices cooperate for AI / ML model training such that the first device performs a portion of the AI / ML model training before determining the DSI of the respective AI / ML model training data, and the respective AI / ML model training data includes a respective output of the portion of the AI / ML model training.

[0054] A corresponding device for carrying out the method is disclosed.

[0055] For example, according to another aspect of the present disclosure, there is provided a device including a processor and a memory storing processor-executable instructions that, when executed, cause the processor to perform a method according to the first broad aspect or the second broad aspect of the present disclosure described above.

[0056] According to another aspect of the present disclosure, there is provided a device including one or more units for implementing any of the method aspects disclosed in this disclosure. The term "unit" is used broadly and may be referred to by any of a variety of names including, for example, module, component, element, means, etc. A unit may be implemented using hardware, software, firmware, or a combination thereof.

[0057] According to some aspects of the present disclosure, the performance of an AI / ML model may be improved and overfitting may be avoided during the training process of the AI / ML model. Furthermore, transmission overhead (e.g., air interface overhead) may be reduced because a smaller amount of AI / ML model training data samples may be transmitted based on data state information (DSI) of the AI / ML model training data. The DSI (e.g., data uncertainty) may be measured at a device (e.g., UE, BS) before the device reports or transmits the AI / ML model training data.

[0058] According to some aspects of the present disclosure, for example, in federated learning, transmission of AI / ML model training data (e.g., local gradients) that does not contribute to the convergence of the global AI / ML model is avoided. Thus, signaling overhead in federated learning may be reduced, AI / ML model performance may be improved, and AI / ML model training may be enhanced.

[0059] According to some aspects of the present disclosure, fast convergence of an AI / ML model may be achieved in two-sided AI / ML model training. Due to the reduced number of transmissions of AI / ML model training datasets, extra signaling overhead may be avoided.

[0060] According to some aspects of the present disclosure, a balance between improved performance of AI / ML models and reduced transmission overhead may be achieved. [Brief explanation of the drawings]

[0061] Reference will now be made, by way of example only, to the accompanying drawings which illustrate exemplary embodiments of the present application.

[0062] [Figure 1] 1 is a simplified schematic diagram of a communication system, according to an example. [Figure 2] FIG. 1 illustrates another example of a communication system. [Figure 3] 1 illustrates an example of an electronic device (ED), a terrestrial transmitting / receiving point (T-TRP), and a non-terrestrial transmitting / receiving point (NT-TRP). [Figure 4] FIG. 1 illustrates exemplary units or modules in a device. [Figure 5] FIG. 1 illustrates four EDs communicating with network devices in a communication system according to an embodiment of the present disclosure. [Figure 6A] FIG. 1 illustrates an example of a neural network having multiple layers of neurons, according to embodiments of the present disclosure. [Figure 6B]FIG. 1 illustrates an example of a neuron that may be used as a building block for a neural network, according to embodiments of the present disclosure. [Figure 7] FIG. 1 illustrates an example of one-sided AI / ML model training at a base station (BS), according to an embodiment of the present disclosure. [Figure 8A] FIG. 1 illustrates an example of a reference AI / ML model, according to an embodiment of the present disclosure. [Figure 8B] FIG. 1 illustrates an example of a reference AI / ML model with reference AI / ML model input data, according to an embodiment of the present disclosure. [Figure 8C] FIG. 1 illustrates an example of measuring data uncertainty using an AI / ML model for channel information, according to an embodiment of the present disclosure. [Figure 9A] FIG. 1 illustrates an example process for reporting AI / ML model training data from a user equipment (UE) to a BS, according to an embodiment of the present disclosure. [Figure 9B] FIG. 1 illustrates an example process for reporting AI / ML model training data from a user equipment (UE) to a BS, according to an embodiment of the present disclosure. [Figure 10] FIG. 1 illustrates an example of one-sided AI / ML model training in a UE, according to an embodiment of the present disclosure. [Figure 11A] FIG. 1 illustrates an example process for reporting AI / ML model training data from a BS to a UE, according to an embodiment of the present disclosure. [Figure 11B] FIG. 1 illustrates an example process for reporting AI / ML model training data from a BS to a UE, according to an embodiment of the present disclosure. [Figure 12A] FIG. 1 illustrates an example of the process of training an AI / ML model in federated learning. [Figure 12B] FIG. 1 illustrates an example process for AI / ML model training in federated learning using data state information (DSI) of AI / ML model training data, according to an embodiment of the present disclosure. [Figure 13] FIG. 1 illustrates an example of two-sided AI / ML model training, according to an embodiment of the present disclosure. [Figure 14]FIG. 1 illustrates an example of AI / ML model training with data transmission precision adaptation, according to an embodiment of the present disclosure. [Figure 15] FIG. 1 is a flow diagram illustrating an example process for AI / ML model training, according to an embodiment of the present disclosure. [Figure 16] FIG. 10 is a flow diagram illustrating another exemplary process for AI / ML model training, according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0063] Similar reference numbers may be used in different figures to indicate similar components.

[0064] In this disclosure, an "AI / ML model" refers to a data-driven algorithm that applies AI / ML techniques to generate a set of outputs based on a set of inputs.

[0065] In this disclosure, "AI / ML model training" refers to the process of training an AI / ML model by learning input-output relationships in a data-driven manner and obtaining a trained AI / ML model for inference.

[0066] In this disclosure, "inference" or "AI / ML inference" refers to the process of using a trained AI / ML model to generate a set of outputs based on a set of inputs.

[0067] In this disclosure, "federated learning / training" refers to a machine learning technique that trains an AI / ML model across multiple distributed edge nodes (e.g., UEs, gNBs), each of which performs local model training using local data samples. This technique requires multiple model exchanges but does not require the exchange of local data samples.

[0068] For purposes of illustration, certain exemplary embodiments will now be described in more detail below in conjunction with the drawings.

[0069] The embodiments described herein represent sufficient information to practice the claimed subject matter and show how to practice such subject matter. Upon reading the following description in light of the accompanying drawings, those skilled in the art will understand the concepts of the claimed subject matter and will recognize applications of these concepts not specifically addressed herein. It is understood that these concepts and applications fall within the scope of this disclosure and the appended claims.

[0070] Furthermore, it should be understood that any module, component, or device disclosed herein that executes instructions may include or otherwise have access to a non-transitory computer / processor-readable storage medium or media for storage of information such as computer / processor-readable instructions, data structures, program modules, and / or other data. A non-exhaustive list of examples of non-transitory computer / processor-readable storage media includes magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, compact disk read-only memory (CD-ROM), digital video disk or digital versatile disk (i.e., DVD), optical disk such as Blu-ray Disc™ or other optical storage, volatile and non-volatile, removable and non-removable media implemented in any manner or technology, random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technology. Any such non-transitory computer / processor storage medium may be part of or accessible or connectable to the device. Computer / processor readable / executable instructions for implementing the applications or modules described herein may be stored by or otherwise retained by such non-transitory computer / processor readable storage media.

[0071] Exemplary Communication Systems and Devices Referring to FIG. 1 , a simplified schematic diagram of a communication system is provided by way of illustrative example and not limitation. The communication system 100 comprises a radio access network 120. The radio access network 120 may be a next-generation (e.g., sixth-generation (6G) or later) radio access network or a legacy (e.g., 5G, 4G, 3G, or 2G) radio access network. One or more communication electrical devices (EDs) 110a-110j (collectively referred to as 110) may be interconnected to each other or connected to one or more network nodes (170a, 170b, collectively referred to as 170) within the radio access network 120. A core network 130 may be part of the communication system and may be dependent on or independent of the radio access technology used in the communication system 100. The communication system 100 also comprises a public switched telephone network (PSTN) 140, the Internet 150, and other networks 160.

[0072] FIG. 2 illustrates an exemplary communication system 100. Generally, the communication system 100 enables multiple wireless or wired elements to communicate data and other content. The purpose of the communication system 100 may be to provide content such as voice, data, video, and / or text via broadcast, multicast, unicast, and the like. The communication system 100 may operate by sharing resources such as carrier spectrum bandwidth among its components. The communication system 100 may include terrestrial and / or non-terrestrial communication systems. The communication system 100 may provide a wide range of communication services and applications (such as earth monitoring, remote sensing, passive sensing and positioning, navigation and tracking, autonomous delivery and mobility, etc.). The communication system 100 may provide a high degree of availability and robustness through the cooperative operation of the terrestrial and non-terrestrial communication systems. For example, integrating a non-terrestrial communication system (or components thereof) into a terrestrial communication system may result in what may be considered a heterogeneous network comprising multiple layers. Compared to traditional communication networks, heterogeneous networks may achieve better overall performance through efficient multi-link joint operation, more flexible function sharing, and faster physical layer link switching between terrestrial and non-terrestrial networks.

[0073] The terrestrial and non-terrestrial communication systems may be considered subsystems of a communication system. In the illustrated example, communication system 100 includes electronic devices (EDs) 110a-110d (collectively referred to as EDs 110), radio access networks (RANs) 120a-120b, a non-terrestrial communication network 120c, a core network 130, a public switched telephone network (PSTN) 140, the Internet 150, and other networks 160. RANs 120a-120b include respective base stations (BSs) 170a-170b, which may be collectively referred to as terrestrial transmission / reception points (T-TRPs) 170a-170b. Non-terrestrial communication network 120c includes access nodes 120c, which may be collectively referred to as non-terrestrial transmission / reception points (NT-TRPs) 172.

[0074] Any ED 110 may alternatively or additionally be configured to interface with, access, or communicate with any other T-TRPs 170a-170b and NT-TRPs 172, the Internet 150, the core network 130, the PSTN 140, other networks 160, or any combination of the foregoing. In some examples, the ED 110a may communicate uplink and / or downlink transmissions via an interface 190a with the T-TRP 170a. In some examples, the EDs 110a, 110b, and 110d may also communicate directly with each other via one or more sidelink air interfaces 190b. In some examples, the ED 110d may communicate uplink and / or downlink transmissions via an interface 190c with the NT-TRP 172.

[0075] Air interfaces 190a and 190b may use similar communication technologies, such as any suitable radio access technology. For example, communication system 100 may implement one or more channel access methods in air interfaces 190a and 190b, such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), or single-carrier FDMA (SC-FDMA). Air interfaces 190a and 190b may utilize other higher-dimensional signal spaces, which may involve combinations of orthogonal and / or non-orthogonal dimensions.

[0076] The air interface 190c may enable communication between the EDs 110d and one or more NT-TRPs 172 via a wireless link or simply a link. In some examples, the link is a dedicated connection for unicast transmission, a connection for broadcast transmission, or a connection between a group of EDs and one or more NT-TRPs for multicast transmission.

[0077] The RANs 120a and 120b communicate with the core network 130 to provide various services, such as voice, data, and other services, to the EDs 110a, 110b, and 110c. The RANs 120a and 120b and / or the core network 130 may be in direct or indirect communication with one or more other RANs (not shown), which may or may not be directly served by the core network 130 and which may or may not employ the same radio access technology as the RAN 120a, RAN 120b, or both. The core network 130 may also serve as a gateway access between (i) the RANs 120a and 120b or the EDs 110a, 110b, and 110c, or both, and (ii) other networks (such as the PSTN 140, the Internet 150, and other networks 160). Additionally, some or all of the EDs 110a, 110b, and 110c may include functionality for communicating with different wireless networks over different wireless links using different wireless technologies and / or protocols. Alternatively (or in addition to) wireless communications, the EDs 110a, 110b, and 110c may communicate with a service provider or switch (not shown) and the Internet 150 via wired communication channels. The PSTN 140 may include a circuit-switched telephone network for providing Plain Old Telephone Service (POTS). The Internet 150 may include a network of computers and / or subnets (intranets) and may incorporate protocols such as Internet Protocol (IP), Transmission Control Protocol (TCP), and User Datagram Protocol (UDP). The EDs 110a, 110b, and 110c may be multimode devices capable of operating according to multiple radio access technologies and may incorporate multiple transceivers necessary to support such.

[0078] 3 shows another example of the ED 110 and the base stations 170a, 170b, and / or 170c. The ED 110 is used to connect people, objects, machines, etc. The ED 110 may be widely used in various scenarios, such as, for example, cellular communication, device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, peer-to-peer (P2P) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), Internet of Things (IoT), virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, telemedicine, smart grid, smart furniture, smart office, smart wearable, smart transportation, smart city, drones, robots, remote sensing, passive sensing, positioning, navigation and tracking, autonomous delivery and mobility, etc.

[0079] Each ED 110 represents any suitable end-user device for wireless operation and may include (or may be referred to as) a device such as a user equipment / device (UE), a wireless transmit / receive unit (WTRU), a mobile station, a fixed or mobile subscriber unit, a cellular telephone, a station (STA), a machine-type communication (MTC) device, a personal digital assistant (PDA), a smartphone, a laptop, a computer, a tablet, a wireless sensor, a consumer electronics device, a smartbook, a vehicle, a car, a truck, a bus, a train, or an IoT device, an industrial device, or an apparatus (e.g., a communication module, a modem, or a chip) in any of the aforementioned devices, among other possibilities. The next-generation ED 110 may be referred to using other terms. The base stations 170a and 170b are T-TRPs and are hereinafter referred to as T-TRP 170. Also, as shown in FIG. 3, the NT-TRP is hereinafter referred to as NT-TRP 172. Each ED110 connected to the T-TRP170 and / or NT-TRP172 may be dynamically or semi-statically turned on (i.e., established, activated, or enabled), turned off (i.e., released, deactivated, or disabled), and / or configured in response to one or more of connection availability and connection need.

[0080] The ED 110 includes a transmitter 201 and a receiver 203 coupled to one or more antennas 204. Only one antenna 204 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 201 and the receiver 203 may be integrated, for example, as a transceiver. The transceiver is configured to modulate data or other content for transmission by at least one antenna 204 or a network interface controller (NIC). The transceiver is also configured to demodulate data or other content received by the at least one antenna 204. Each transceiver includes any suitable structure for generating signals for wireless or wired transmission and / or processing signals received wirelessly or wired. Each antenna 204 includes any suitable structure for transmitting and / or receiving wireless or wired signals.

[0081] The ED 110 includes at least one memory 208. The memory 208 stores instructions and data used, generated, or collected by the ED 110. For example, the memory 208 may store software instructions or modules configured to implement some or all of the functions and / or embodiments described herein and executed by the processing unit 210. Each memory 208 includes any suitable volatile and / or non-volatile storage and retrieval device. Any suitable type of memory may be used, such as random access memory (RAM), read-only memory (ROM), hard disk, optical disk, subscriber identity module (SIM) card, memory stick, secure digital (SD) memory card, on-processor cache, etc.

[0082] ED 110 may further include one or more input / output devices (not shown) or interfaces (such as a wired interface to Internet 150 in FIG. 1). The input / output devices enable interaction with a user or other devices in a network. Each input / output device includes any suitable structure for providing information to or receiving information from a user, such as a speaker, microphone, keypad, keyboard, display, or touch screen, including network interface communications.

[0083] The ED 110 further includes a processor 210 for performing operations including operations related to preparing a transmission for uplink transmission to the NT-TRP 172 and / or the T-TRP 170, operations related to processing a downlink transmission received from the NT-TRP 172 and / or the T-TRP 170, and operations related to processing a sidelink transmission between another ED 110. Processing operations related to preparing a transmission for uplink transmission may include operations such as encoding, modulating, transmit beamforming, and generating symbols for transmission. Processing operations related to processing a downlink transmission may include operations such as receive beamforming, demodulating and decoding received symbols. Depending on the embodiment, the downlink transmission may be received by the receiver 203, possibly using receive beamforming, and the processor 210 may extract signaling from the downlink transmission (e.g., by detecting and / or decoding the signaling). One example of signaling may be a reference signal transmitted by the NT-TRP 172 and / or the T-TRP 170. In some embodiments, processor 276 implements transmit beamforming and / or receive beamforming based on beam direction indications, e.g., beam angle information (BAI), received from T-TRP 170. In some embodiments, processor 210 may perform operations related to network access (e.g., initial access) and / or downlink synchronization, such as operations related to detecting synchronization sequences, decoding and acquiring system information, etc. In some embodiments, processor 210 may perform channel estimation, e.g., using reference signals received from NT-TRP 172 and / or T-TRP 170.

[0084] Although not shown, the processor 210 may form part of the transmitter 201 and / or the receiver 203. Although not shown, the memory 208 may form part of the processor 210.

[0085] The processor 210 and the processing components of the transmitter 201 and receiver 203 may each be implemented by the same or different one or more processors configured to execute instructions stored in a memory (e.g., in memory 208). Alternatively, some or all of the processor 210 and the processing components of the transmitter 201 and receiver 203 may be implemented using special purpose circuitry, such as a programmed field programmable gate array (FPGA), a graphical processing unit (GPU), or an application specific integrated circuit (ASIC).

[0086] In some implementations, the T-TRP 170 may be known by other names such as a base station, base transceiver station (BTS), radio base station, network node, network device, network-side device, transceiver node, Node B, evolved Node B (eNodeB or eNB), Home eNodeB, next-generation Node B (gNB), transmission point (TP), site controller, access point (AP), or wireless router, relay station, remote radio head, terrestrial node, terrestrial network device, or terrestrial base station, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), positioning node, etc. The T-TRP 170 may be a macro BS, pico BS, relay node, donor node, etc., or a combination thereof. The T-TRP 170 may refer to a counterfeit device or apparatus (e.g., a communication module, modem, or chip) within the aforementioned devices.

[0087] In some embodiments, parts of the T-TRP 170 may be distributed. For example, some of the modules of the T-TRP 170 may be located remotely from the equipment housing the antenna of the T-TRP 170 and may be coupled to the equipment housing the antenna via a communications link (not shown), sometimes known as fronthaul, such as a Common Public Radio Interface (CPRI). Thus, in some embodiments, the term T-TRP 170 may refer to a network-side module that performs processing operations such as determining the location of the ED 110, resource allocation (scheduling), message generation, and encoding / decoding, and that is not necessarily part of the equipment housing the antenna of the T-TRP 170. The modules may also be coupled to other T-TRPs. In some embodiments, the T-TRP 170 may actually be multiple T-TRPs operating together to serve the ED 110, for example, through coordinated multipoint transmission.

[0088] The T-TRP 170 includes at least one transmitter 252 and at least one receiver 254 coupled to one or more antennas 256. Only one antenna 256 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 252 and receiver 254 may be integrated as a transceiver. The T-TRP 170 further includes a processor 260 for performing operations including operations related to preparing a transmission for downlink transmission to the ED 110, processing uplink transmissions received from the ED 110, preparing a transmission for backhaul transmission to the NT-TRP 172, and processing transmissions received via the backhaul from the NT-TRP 172. Processing operations related to preparing a transmission for downlink or backhaul transmission may include operations such as encoding, modulation, precoding (e.g., MIMO precoding), transmit beamforming, and generating symbols for transmission. Processing operations related to processing transmissions received in the uplink or over the backhaul may include operations such as receive beamforming and demodulating and decoding received symbols. The processor 260 may also perform operations related to network access (e.g., initial access) and / or downlink synchronization, such as generating synchronization signal block (SSB) content and generating system information. In some embodiments, the processor 260 also generates a beam direction indication, e.g., a BAI, that may be scheduled for transmission by the scheduler 253. The processor 260 performs other network-side processing operations described herein, such as determining the location of the ED 110 and determining where to deploy the NT-TRP 172. In some embodiments, the processor 260 may generate signaling, for example, to configure one or more parameters of the ED 110 and / or one or more parameters of the NT-TRP 172. Any signaling generated by the processor 260 is sent by the transmitter 252.It should be noted that "signaling" as used herein may alternatively be referred to as control signaling. Dynamic signaling may be transmitted in a control channel, e.g., a Physical Downlink Control Channel (PDCCH), and static or semi-static higher layer signaling may be included in packets transmitted in a data channel, e.g., a Physical Downlink Shared Channel (PDSCH).

[0089] The scheduler 253 may be coupled to the processor 260. The scheduler 253 may be included within or operate separately from the T-TRP 170, which may schedule uplink, downlink, and / or backhaul transmissions, including issuing scheduling grants and / or configuring non-scheduled (“configured grants”) resources. The T-TRP 170 further includes a memory 258 for storing information and data. The memory 258 stores instructions and data used, generated, or collected by the T-TRP 170. For example, the memory 258 may store software instructions or modules configured to implement some or all of the functions and / or embodiments described herein and executed by the processor 260.

[0090] Although not shown, the processor 260 may form part of the transmitter 252 and / or the receiver 254. Also, although not shown, the processor 260 may implement the scheduler 253. Although not shown, the memory 258 may form part of the processor 260.

[0091] The processor 260, the scheduler 253, and the processing components of the transmitter 252 and the receiver 254 may each be implemented by the same or different one or more processors configured to execute instructions stored in a memory, for example, the memory 258. Alternatively, some or all of the processing components of the processor 260, the scheduler 253, and the transmitter 252 and the receiver 254 may be implemented using dedicated circuitry, such as an FPGA, a GPU, or an ASIC.

[0092] Although the NT-TRP 172 is shown as a drone by way of example only, the NT-TRP 172 may be implemented in any suitable non-terrestrial form. The NT-TRP 172 may also be known by other names, such as a non-terrestrial node, a non-terrestrial network device, or a non-terrestrial base station, in some implementations. The NT-TRP 172 includes a transmitter 272 and a receiver 274 coupled to one or more antennas 280. Only one antenna 280 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 272 and the receiver 274 may be integrated as a transceiver. The NT-TRP 172 further includes a processor 276 for performing operations, including operations related to preparing transmissions for downlink transmission to the ED 110, processing uplink transmissions received from the ED 110, preparing transmissions for backhaul transmission to the T-TRP 170, and processing transmissions received via the backhaul from the T-TRP 170. Processing operations related to preparing a transmission for downlink or backhaul transmission may include operations such as encoding, modulation, precoding (e.g., MIMO precoding), transmit beamforming, and generating symbols for transmission. Processing operations related to processing a transmission received in the uplink or via the backhaul may include operations such as receive beamforming, and demodulating and decoding received symbols. In some embodiments, the processor 276 implements transmit beamforming and / or receive beamforming based on beam direction information (e.g., BAI) received from the T-TRP 170. In some embodiments, the processor 276 may generate signaling, for example, to configure one or more parameters of the ED 110. In some embodiments, the NT-TRP 172 implements physical layer processing but does not implement higher layer functions, such as functions at the medium access control (MAC) or radio link control (RLC) layer. This is by way of example only; more generally, the NT-TRP 172 may implement higher layer functions in addition to physical layer processing.

[0093] The NT-TRP 172 further includes a memory 278 for storing information and data. Although not shown, the processor 276 may form part of the transmitter 272 and / or the receiver 274. Although not shown, the memory 278 may form part of the processor 276.

[0094] The processor 276 and the processing components of the transmitter 272 and receiver 274 may each be implemented by the same or different one or more processors configured to execute instructions stored in a memory, e.g., memory 278. Alternatively, some or all of the processing components of the processor 276 and the transmitter 272 and receiver 274 may be implemented using dedicated circuitry, such as a programmed FPGA, GPU, or ASIC. In some embodiments, the NT-TRP 172 may actually be multiple NT-TRPs operating together to service the ED 110, e.g., through coordinated multipoint transmission.

[0095] It should be noted that "TRP" as used herein can refer to T-TRP or NT-TRP.

[0096] T-TRP170, NT-TRP172, and / or ED110 may contain other components, which have been omitted for clarity.

[0097] One or more steps of the method of the embodiments provided herein may be performed by a corresponding unit or module according to FIG. 4. FIG. 4 illustrates units or modules within a device, such as within the ED 110, the T-TRP 170, or the NT-TRP 172. For example, a signal may be transmitted by a transmitting unit or a transmitting module. For example, a signal may be transmitted by a transmitting unit or a transmitting module. A signal may be received by a receiving unit or a receiving module. A signal may be processed by a processing unit or a processing module. Other steps may be performed by an artificial intelligence (AI) or machine learning (ML) module. Each unit or module may be implemented using hardware, one or more components or devices executing software, or a combination thereof. For example, one or more of the units or modules may be an integrated circuit, such as a programmed FPGA, a GPU, or an ASIC. Where modules are implemented using software, for example, for execution by a processor, it will be understood that they may be retrieved by the processor, in single or multiple instances, individually or together for processing, in whole or in part, as needed, and that the modules themselves may include instructions for further deployment and instantiation.

[0098] Additional details regarding ED110, T-TRP170, and NT-TRP172 are known to those skilled in the art, and therefore, these details are omitted here.

[0099] Control signaling is described herein in some embodiments. Control signaling may sometimes be alternatively referred to as signaling, or control information, or configuration information, or configuration. In some cases, control signaling may be dynamically indicated at the physical layer, for example, in a control channel. An example of dynamically indicated control signaling is information sent in physical layer control signaling, for example, downlink control information (DCI). Control signaling may sometimes be alternatively indicated semi-statically, for example, in RRC signaling or MAC control element (CE). A dynamic indication may be an indication in a lower layer, for example, physical layer / Layer 1 signaling (e.g., in DCI), rather than in a higher layer (e.g., not in RRC signaling or MAC CE). A semi-static indication may be an indication in semi-static signaling. As used herein, semi-static signaling may refer to signaling that is not dynamic, e.g., higher layer signaling, RRC signaling, and / or MAC CE. As used herein, dynamic signaling may refer to signaling that is dynamic, e.g., physical layer control signaling sent at the physical layer, such as DCI.

[0100] An air interface generally includes several components and associated parameters that collectively specify how transmissions should be sent and / or received over a wireless communication link between two or more communication devices. For example, an air interface may include one or more components that define a waveform, frame structure, multiple access scheme, protocol, coding scheme, and / or modulation scheme for carrying information (e.g., data) over the wireless communication link. The wireless communication link may support a link between a radio access network and user equipment (e.g., a “Uu” link), and / or the wireless communication link may support a device-to-device link (e.g., a “sidelink”), such as between two user equipments, and / or the wireless communication link may support a link between a non-terrestrial (NT) communication network and user equipment (UE). The following are some examples of such components: Waveform components may specify the shape and form of the signal being transmitted. Waveform options may include orthogonal multiple access waveforms and non-orthogonal multiple access waveforms. Non-limiting examples of such waveform options include orthogonal frequency division multiplexing (OFDM), filtered OFDM (f-OFDM), time-windowed OFDM, filter bank multicarrier (FBMC), universal filtered multicarrier (UFMC), generalized frequency division multiplexing (GFDM), wavelet packet modulation (WPM), faster-than-Nyquist (FTN) waveforms, and low peak-to-average power ratio waveforms (low-PAPR WFs). The frame structure component may specify the configuration of a frame or group of frames. The frame structure component may indicate one or more of the time, frequency, pilot signature, code, or other parameters of the frame or group of frames. Further details of the frame structure are described below. The multiple access scheme component may specify multiple access technique options, including techniques that define how communication devices share a common physical channel, such as time division multiple access (TDMA), frequency division multiple access (FDMA), code division multiple access (CDMA), single-carrier frequency division multiple access (SC-FDMA), low-density signature multi-carrier code division multiple access (LDS-MC-CDMA), non-orthogonal multiple access (NOMA), pattern division multiple access (PDMA), lattice division multiple access (LPMA), resource spread multiple access (RSMA), sparse code multiple access (SCMA), etc. Additionally, the multiple access technique options may include scheduled access versus unscheduled access, also known as grant-free access, e.g., via dedicated channel resources (e.g., not shared among multiple communication devices), non-orthogonal multiple access versus orthogonal multiple access, contention-based shared channel resources versus non-contention-based shared channel resources, and cognitive radio-based access. A Hybrid Automatic Repeat Request (HARQ) protocol component may specify how transmissions and / or retransmissions should occur. Non-limiting examples of transmission and / or retransmission mechanism options include those that specify a scheduled data pipe size, a signaling mechanism for transmissions and / or retransmissions, and a retransmission mechanism. The coding and modulation component may specify how the information being transmitted may be coded / decoded and modulated / demodulated for transmission / reception purposes. Coding may refer to methods of error detection and forward error correction. Non-limiting examples of coding options include turbo trellis codes, turbo product codes, fountain codes, low-density parity-check codes, and polar codes. Modulation may simply refer to the constellation (e.g., including the modulation technique and order), or more specifically, to various types of advanced modulation methods such as hierarchical modulation and low-PAPR modulation.

[0101] In some embodiments, the air interface may be a “one-size-fits-all concept.” For example, components within the air interface cannot be changed or adapted once the air interface is defined. In some implementations, only limited parameters or modes of the air interface, such as cyclic prefix (CP) length or multiple-input multiple-output (MIMO) mode, may be configured. In some embodiments, the air interface design may provide a unified or flexible framework for supporting sub- and above-6 GHz frequency (e.g., mmWave) bands for both licensed and unlicensed access. As an example, the flexibility of the configurable air interface provided by scalable numerology and symbol duration may enable transmission parameter optimization for different spectrum bands and for different services / devices. As another example, the unified air interface may be self-contained in the frequency domain, and a frequency-domain self-contained design may support more flexible radio access network (RAN) slicing through channel resource sharing between different services in both frequency and time.

[0102] Frame structure The frame structure is a feature of the wireless communication physical layer that defines the time-domain signal transmission structure, e.g., to enable timing reference and timing alignment of basic time-domain transmission units. Wireless communications between communication devices may occur over time-frequency resources governed by the frame structure. The frame structure may alternatively be referred to as a radio frame structure.

[0103] Depending on the frame structure and / or the configuration of frames within the frame structure, frequency division duplex (FDD) and / or time division duplex (TDD) and / or full duplex (FD) communication may be possible. FDD communication is when transmissions occur in different directions (e.g., uplink vs. downlink) in different frequency bands. TDD communication is when transmissions in different directions (e.g., uplink vs. downlink) occur for different durations. FD communication is when transmission and reception occur on the same time-frequency resource, i.e., a device can both transmit and receive on the same frequency resource simultaneously in time.

[0104] An example of a frame structure is that in Long Term Evolution (LTE) with the following specifications: each frame is 10 ms in duration, each frame has 10 subframes, each of 1 ms in duration, each subframe contains two slots, each of 0.5 ms in duration, each slot is for the transmission of seven OFDM symbols (assuming a normal CP), each OFDM symbol has a symbol duration and a specific bandwidth (or partial bandwidth or bandwidth segment) related to the number of subcarriers and subcarrier spacing, the frame structure is based on OFDM waveform parameters such as subcarrier spacing and CP length (where the CP has options of fixed length or limited length), and the switching gap between uplink and downlink in TDD must be an integer number of OFDM symbol durations.

[0105] Another example of a frame structure is that in New Radio (NR), which has the following specifications: Multiple subcarrier spacings are supported, each corresponding to a different numerology; the frame structure depends on the numerology, but in all cases, the frame length is set to 10 ms, consisting of 10 subframes of 1 ms each; a slot is defined as 14 OFDM symbols; and the slot length depends on the numerology. For example, the NR frame structure for normal CP 15 kHz subcarrier spacing ("numerology 1") differs from the NR frame structure for normal CP 30 kHz subcarrier spacing ("numerology 2"). For 15 kHz subcarrier spacing, the slot length is 1 ms; for 30 kHz subcarrier spacing, the slot length is 0.5 ms. The NR frame structure may have greater flexibility than the LTE frame structure.

[0106] Another example of a frame structure is an exemplary flexible frame structure, e.g., for use in 6G networks and beyond. In a flexible frame structure, a symbol block may be defined as the smallest duration that can be scheduled in the flexible frame structure. A symbol block may be a unit of transmission having an optional redundant portion (e.g., a CP portion) and an information (e.g., data) portion. An OFDM symbol is an example of a symbol block. A symbol block may alternatively be referred to as a symbol. Embodiments of a flexible frame structure include different parameters that may be configurable, e.g., frame length, subframe length, symbol block length, etc. A non-exhaustive list of configurable parameters possible in some embodiments of a flexible frame structure includes the following: (1) Frame: The frame length does not need to be limited to 10 ms and may be configurable and vary over time. In some embodiments, each frame includes one or more downlink synchronization channels and / or one or more downlink broadcast channels, and each synchronization channel and / or broadcast channel may be transmitted in a different direction by different beamforming. The frame length may be two or more possible values ​​and may be configured based on the application scenario. For example, an autonomous vehicle may require relatively fast initial access, in which case the frame length may be set as 5 ms for an autonomous vehicle application. As another example, a smart meter on a home may not require fast initial access, in which case the frame length may be set as 20 ms for the smart meter application. (2) Subframe Duration: Subframes may or may not be defined in a flexible frame structure depending on the implementation. For example, a frame may be defined to include slots but not subframes. In frames where subframes are defined, e.g., for time domain alignment, the duration of the subframes may be configurable. For example, a subframe may be configured to have a length of 0.1 ms, 0.2 ms, 0.5 ms, 1 ms, 2 ms, 5 ms, etc. In some embodiments, if a subframe is not needed in a particular scenario, the subframe length may be defined to be the same as the frame length or may not be defined. (3) Slot Configuration: Slots may or may not be defined in a flexible frame structure, depending on the implementation. In frames in which slots are defined, the slot definition (e.g., in duration and / or number of symbol blocks) may be configurable. In one embodiment, the slot configuration is common to all UEs or a group of UEs. In this case, slot configuration information may be transmitted to the UE in a broadcast channel or a common control channel. In other embodiments, the slot configuration may be UE-specific, in which case slot configuration information may be transmitted in a UE-specific control channel. In some embodiments, slot configuration signaling may be transmitted together with frame configuration signaling and / or subframe configuration signaling. In other embodiments, slot configuration may be transmitted independently from frame configuration signaling and / or subframe configuration signaling. In general, slot configuration may be system-wide, base station-wide, UE group-wide, or UE-specific. (4) Subcarrier Spacing (SCS): SCS is one parameter of a scalable numerology that may allow the SCS to range from 15 KHz to 480 KHz in some cases. The SCS may vary with the frequency of the spectrum and / or the maximum UE speed to minimize the effects of Doppler shift and phase noise. In some examples, there may be separate transmit and receive frames, and the SCS of the symbols in the receive frame structure may be configured independently of the SCS of the symbols in the transmit frame structure. The SCS in the receive frames may be different from the SCS in the transmit frames. In some examples, the SCS of each transmit frame may be half the SCS of each receive frame. If the SCS between the receive and transmit frames is different, for example, if a more flexible symbol duration is implemented using an inverse discrete Fourier transform (IDFT) instead of a fast Fourier transform (FFT), the difference does not necessarily need to be scaled by a factor of two. Additional example frame structures may be used with different SCSs. (5) Flexible Transmission Duration of Basic Transmission Unit: A basic transmission unit may be a symbol block (alternatively called a symbol) that generally includes a redundant portion (referred to as a CP) and an information (e.g., data) portion, although in some embodiments, the CP may be omitted from the symbol block. The CP length may be flexible and configurable. The CP length may be fixed within a frame or may be flexible within a frame, and the CP length may change dynamically, in some cases, for each frame, group of frames, subframe, slot, or scheduling. The information (e.g., data) portion may be flexible and configurable. Another possible parameter related to a symbol block that may be defined is the ratio of the CP duration to the information (e.g., data) duration. In some embodiments, the symbol block length may be adjusted according to channel conditions (e.g., multipath delay, Doppler), and / or latency requirements, and / or available duration. As another example, the symbol block length may be adjusted to fit the available duration within a frame. (6) Flexible Switching Gap: A frame may include both a downlink portion for downlink transmission from the base station and an uplink portion for uplink transmission from the UE. A gap may exist between each uplink and downlink portion, which is called a switching gap. The switching gap length (duration) may be configurable. The switching gap duration may be fixed within a frame or flexible within a frame, and the switching gap duration may possibly change dynamically from frame to frame, or from group of frames, or from subframe to subframe, or from slot to slot, or from scheduling to scheduling.

[0107] Cell / Carrier / Bandwidth Portion (BWP) / Occupied Bandwidth A device, such as a base station, may provide coverage across a cell. Wireless communication with a device may occur over one or more carrier frequencies. A carrier frequency is referred to as a carrier. A carrier may alternatively be referred to as a component carrier (CC). A carrier may be characterized by its bandwidth and a reference frequency, e.g., the center or lowest or highest frequency of the carrier. A carrier may be on a licensed or unlicensed spectrum. Wireless communication with a device may also or alternatively occur over one or more bandwidth portions (BWPs). For example, a carrier may have one or more BWPs. More generally, wireless communication with a device may occur over a spectrum. A spectrum may include one or more carriers and / or one or more BWPs.

[0108] A cell may include one or more downlink resources and optionally one or more uplink resources, or a cell may include one or more uplink resources and optionally one or more downlink resources, or a cell may include both one or more downlink resources and one or more uplink resources. As an example, a cell may include only one downlink carrier / BWP, or only one uplink carrier / BWP, or multiple downlink carriers / BWPs, or multiple uplink carriers / BWPs, or one downlink carrier / BWP and one uplink carrier / BWP, or one downlink carrier / BWP and multiple uplink carriers / BWPs, or multiple downlink carriers / BWPs and one uplink carrier / BWP. In some embodiments, a cell may alternatively or additionally include one or more sidelink resources, including sidelink transmission and reception resources.

[0109] A BWP is a set of contiguous or non-contiguous frequency subcarriers on a carrier, which may have one or more carriers, or a set of contiguous or non-contiguous frequency subcarriers on multiple carriers, or a set of non-contiguous or contiguous frequency subcarriers.

[0110] In some embodiments, a carrier may have one or more BWPs; for example, a carrier may have a 20 MHz bandwidth and consist of one BWP, or a carrier may have an 80 MHz bandwidth and consist of two adjacent contiguous BWPs, etc. In other embodiments, a BWP may have one or more carriers; for example, a BWP may have a 40 MHz bandwidth and consist of two adjacent contiguous carriers, each carrier having a 20 MHz bandwidth. In some embodiments, a BWP may include discontinuous spectral resources consisting of discontinuous carriers, where a first carrier of the discontinuous carriers may be in the mmW band, a second carrier may be in a low band (such as the 2 GHz band), a third carrier (if present) may be in the THz band, and a fourth carrier (if present) may be in the visible light band. The resources in one carrier belonging to a BWP may be contiguous or discontinuous. In some embodiments, a BWP has discontinuous spectral resources on one carrier.

[0111] Wireless communications may occur over an occupied bandwidth, which may be defined as the width of a frequency band below a lower frequency limit and above an upper frequency limit such that the average power emitted is equal to a specified fraction β / 2 of the total average transmitted power, e.g., β / 2=0.5%, respectively.

[0112] The carrier, BWP, or occupied bandwidth may be signaled dynamically by a network device (e.g., a base station), e.g., in physical layer control signaling such as downlink control information (DCI), or semi-statically, e.g., in radio resource control (RRC) signaling or medium access control (MAC) layer, or may be pre-defined based on an application scenario, or may be determined by the UE as a function of other parameters known by the UE, or may be fixed, e.g., by a standard.

[0113] Artificial Intelligence (AI) and / or Machine Learning (ML) The number of new devices in future wireless networks is expected to increase exponentially, and the capabilities of the devices are expected to become increasingly diverse. Also, many new applications and use cases are expected to emerge with more diverse quality of service requirements than those of 5G applications / use cases. These will result in new key performance indicators (KPIs) for future wireless networks (e.g., 6G networks) that may be extremely challenging. AI techniques, such as ML techniques (e.g., deep learning), have been introduced into telecommunications applications with the aim of improving system performance and efficiency.

[0114] Furthermore, advances continue to be made in antennas and bandwidth capabilities, potentially enabling more and / or better communications over wireless links. Additionally, advances continue to be made in the areas of computer architecture and computing power, for example, with the introduction of general-purpose graphics processing units (GP-GPUs). Next-generation communication devices may have more computing power and / or communication capabilities than previous generations, which may enable AI to be employed to implement air interface components. Future generations of networks may also have access to more accurate and / or new information (compared to previous networks) that may form the basis of input to AI models, such as the physical speed / velocity at which the device is traveling, the device's link budget, the device's channel conditions, one or more device capabilities and / or service types to be supported, sensing information, and / or positioning information. To obtain sensing information, the TRP may transmit a signal to a target object (e.g., a suspected UE), and based on the signal's reflection, the TRP or another network device calculates the angle (for beamforming for the device), the device's distance from the TRP, and / or Doppler shift information. Positioning information, sometimes referred to as location determination, may be obtained in a variety of ways, such as by positioning reports from the UE (e.g., reporting the GPS coordinates of the UE), by using positioning reference signals (PRS), by using sensing as described above, by tracking and / or predicting the location of the device, etc.

[0115] AI technologies (including ML technologies) may be applied to communications, including AI-based communications at the physical layer and / or MAC layer. In the case of the physical layer, AI communications may aim to optimize component design and / or improve algorithm performance. For example, AI may be applied in connection with implementing channel coding, channel modeling, channel estimation, channel decoding, modulation, demodulation, MIMO, waveforms, multiple access, physical layer element parameter optimization and update, beamforming, tracking, sensing, and / or positioning, etc. In the case of the MAC layer, AI-based communications may aim to utilize AI capabilities to learn, predict, and / or make decisions to solve complex optimization problems with possible better strategies and / or optimal solutions, for example, to optimize functions in the MAC layer. For example, AI may be applied to implement intelligent TRP management, intelligent beam management, intelligent channel resource allocation, intelligent power control, intelligent spectrum utilization, intelligent MCS, intelligent HARQ strategies, and / or intelligent transmit / receive mode adaptation, etc.

[0116] In some embodiments, the AI ​​architecture may involve multiple nodes, which may be organized in one of two modes: centralized and distributed, both of which may be deployed within the access network, core network, or edge computing system or third-party network. Centralized training and computing architectures are sometimes limited by high communication overhead and strict user data privacy. Distributed training and computing architectures may include several frameworks, such as distributed machine learning and federated learning. In some embodiments, the AI ​​architecture may include an intelligent controller that can run as a single agent or multiple agents based on joint or personal optimization. New protocols and signaling mechanisms are desired so that corresponding interface links can be personalized with customized parameters to meet specific requirements, while minimizing signaling overhead and maximizing overall system spectral efficiency through personalized AI techniques.

[0117] In some embodiments herein, new protocols and signaling mechanisms are provided for operating between and switching between different operating modes for AI training, including between a training mode and a normal operating mode, as well as for measurement and feedback to accommodate different possible measurements and information that may need to be fed back, depending on the implementation.

[0118] AI training 1 and 2, embodiments of the present disclosure may be used to implement AI training involving two or more communication devices in communication system 100. For example, FIG. 5 illustrates four EDs communicating with network device 452 in communication system 100, according to one embodiment. Each of the four EDs is depicted as a respective different UE, hereinafter referred to as UEs 402, 404, 406, and 408. However, EDs do not necessarily have to be UEs.

[0119] The network device 452 is part of a network (e.g., the radio access network 120). The network device 452 may be deployed in an access network, a core network, or an edge computing system or a third-party network, depending on the implementation. The network device 452 may be (or may be part of) a T-TRP or a server. In one example, the network device 452 may be (or may be implemented within) the T-TRP 170 or the NT-TRP 172. In another example, the network device 452 may be a T-TRP controller and / or an NT-TRP controller that can manage the T-TRP 170 or the NT-TRP 172. In some embodiments, the components of the network device 452 may be distributed. The UEs 402, 404, 406, and 408 may communicate directly with the network device 452, for example, if the network device 452 is part of a T-TRP that serves the UEs 402, 404, 406, and 408. Alternatively, the UEs 402, 404, 406, and 408 may communicate with the network device 452 via one or more intermediate components, such as, for example, a T-TRP and / or an NT-TRP. For example, the network device 452 may transmit and / or receive information (e.g., control signaling, data, training sequences, etc.) to and from one or more of the UEs 402, 404, 406, and 408 via backhaul links and wireless channels interposed between the network device 452 and the UEs 402, 404, 406, and 408.

[0120] Each UE 402, 404, 406, and 408 includes a respective processor 210, memory 208, transmitter 201, receiver 203, and one or more antennas 204 (or alternatively, a panel), as described above. For simplicity, only the processor 210, memory 208, transmitter 201, receiver 203, and antenna 204 for UE 402 are shown, although the other UEs 404, 406, and 408 also include the same respective components.

[0121] For each UE 402, 404, 406, and 408, the communication link between that UE and a respective TRP in the network is an air interface. The air interface generally includes several components and associated parameters that collectively specify how transmissions should be sent and / or received over the wireless medium.

[0122] The processor 210 of the UE in FIG. 5 implements one or more air interface components on the UE side. The air interface components configure and / or implement transmission and / or reception over the air interface. Examples of air interface components are described herein. For example, air interface components such as a channel encoder (or decoder) that implements a coding component of the air interface for the UE, and / or a modulator (or demodulator) that implements a modulation component of the air interface for the UE, and / or a waveform generator that implements a waveform component of the air interface for the UE may be in the physical layer. The air interface components may be in or part of a higher layer, such as the MAC layer, for example, a module that implements channel estimation / tracking and / or a module that implements a retransmission protocol (e.g., implements a HARQ protocol component of the air interface for the UE). The processor 210 also directly performs (or controls the UE to perform) the UE-side operations described herein.

[0123] The network device 452 includes a processor 454, a memory 456, and an input / output device 458. The processor 454 implements or instructs other network devices (e.g., a T-TRP) to implement one or more of the network-side air interface components. The air interface components may be implemented differently on the network side for one UE than for another UE. The processor 454 directly performs (or controls network components to perform) the network-side operations described herein.

[0124] Processor 454 may be implemented by the same or different one or more processors configured to execute instructions stored in a memory (e.g., in memory 456). Alternatively, some or all of processor 454 may be implemented using dedicated circuitry such as a programmed FPGA, GPU, or ASIC. Memory 456 may be implemented by volatile and / or non-volatile storage. Any suitable type of memory may be used, such as RAM, ROM, a hard disk, an optical disk, an on-processor cache, etc.

[0125] The input / output devices 458 enable interaction with other devices by receiving (input) and sending (output) information. In some embodiments, the input / output devices 458 may be implemented by a transmitter and / or receiver (or transceiver), and / or one or more interfaces (e.g., a wired interface to an internal network, the Internet, etc.). In some implementations, the input / output devices 458 may be implemented by a network interface, which may in some cases be implemented as a network interface card (NIC), and / or a computer port (e.g., a physical outlet to which a plug or cable connects), and / or a network socket, etc., depending on the implementation.

[0126] Network device 452 and UE 402 are capable of implementing one or more AI-enabled processes. In particular, in the embodiment of FIG. 5, network device 452 and UE 402 include ML modules 410 and 460, respectively. ML module 410 is implemented by processor 210 of UE 402, and ML module 460 is implemented by processor 454 of network device 452; thus, in FIG. 5, ML module 410 is shown within processor 210, and ML module 460 is shown with processor 454. ML modules 410 and 460 execute one or more AI / ML algorithms to perform one or more AI-enabled processes, e.g., AI-enabled link adaptation, for example, to optimize the communications link between the network and UE 402.

[0127] The ML modules 410 and 460 may be implemented using an AI model. The term AI model may refer to a computer algorithm configured to accept defined input data and output defined inference data, and the algorithm's parameters (e.g., weights) may be updated and optimized through training (e.g., using a training dataset or using real-world collected data). The AI ​​model may be implemented using one or more neural networks (e.g., including deep neural networks (DNNs), recurrent neural networks (RNNs), convolutional neural networks (CNNs), and combinations thereof) and various neural network architectures (e.g., autoencoders, generative adversarial networks, etc.). Various techniques may be used to train an AI model to update and optimize its parameters. For example, backpropagation is a common technique for training DNNs, in which a loss function is calculated between the inference data generated by the DNN and some target output (e.g., ground truth data). The gradient of the loss function is calculated with respect to the DNN's parameters, and the calculated gradient is used to update the parameters (e.g., using a gradient descent algorithm) with the aim of minimizing the loss function.

[0128] In some embodiments, the AI ​​model encompasses a neural network used in machine learning. A neural network is composed of multiple computational units (sometimes called neurons) arranged in one or more layers. The process of receiving input at the input layer and generating output at the output layer is sometimes called forward propagation. In forward propagation, each layer receives input (which may have any suitable data format, such as a vector, matrix, or multidimensional array) and performs computations to generate output (which may have different dimensions than the input). The computations performed by a layer typically involve applying (e.g., multiplying) a set of weights (also called coefficients) to the input. Except for the first layer (i.e., input layer) of a neural network, the input to each layer is the output of the previous layer. A neural network may include one or more layers between the first layer (i.e., input layer) and the last layer (i.e., output layer), which may be called internal or hidden layers. For example, FIG. 6A shows an example of a neural network 600 including an input layer, an output layer, and two hidden layers. In this example, it can be seen that the outputs of each of the three neurons in the input layer of neural network 600 are included in the input vectors to each of the three neurons in the first hidden layer. Similarly, the outputs of each of the three neurons in the first hidden layer are included in the input vectors to each of the three neurons in the second hidden layer, and the outputs of each of the three neurons in the second hidden layer are included in the input vectors to each of the two neurons in the output layer. As mentioned above, the basic computational unit in a neural network is the neuron, as shown at 650 in FIG. 6A. FIG. 6B shows an example of a neuron 650 that can be used as a building block for neural network 600. As shown in FIG. 6B, in this example, neuron 650 takes a vector x as input and performs a dot product with an associated vector of weights w. The neuron's final output z is the result of the activation function f() on the dot product. Various neural networks can be designed with various architectures (e.g., various numbers of layers with various functions performed by each layer).

[0129] Neural networks are trained to optimize the neural network's parameters (e.g., weights). This optimization is performed in an automated manner and is sometimes referred to as machine learning. Training a neural network involves forward propagating input data samples to generate output values ​​(also called predicted or inferred output values) and comparing the generated output values ​​with known or desired target values ​​(e.g., ground truth values). A loss function is defined to quantitatively represent the difference between the generated output values ​​and the target values, and the goal of training a neural network is to minimize the loss function. Backpropagation is an algorithm for training neural networks. Backpropagation is used to adjust (also called update) the values ​​of parameters (e.g., weights) in a neural network so that the calculated loss function is reduced. Backpropagation involves calculating the gradient of the loss function with respect to the parameters being optimized, and a gradient algorithm (e.g., gradient descent) is used to update the parameters to reduce the loss function. Backpropagation is performed iteratively so that the loss function converges or is minimized over several iterations. After training conditions are met (e.g., the loss function converges or a predetermined number of training iterations are performed), the neural network is considered trained. The trained neural network may be deployed (or run) to generate output data inferred from input data. In some embodiments, training of the neural network may be ongoing even after the neural network is deployed, so that the neural network's parameters may be iteratively updated with the latest training data.

[0130] Referring again to FIG. 5 , in some embodiments, the UE 402 and the network device 452 may exchange information for training purposes. The information exchanged between the UE 402 and the network device 452 may be implementation-specific and may not have a human-understandable meaning (e.g., it may be intermediate data generated during the execution of an ML algorithm). Additionally or alternatively, the exchanged information may not be predefined by a standard; for example, bits may be exchanged, but the bits may not be associated with a predefined meaning. In some embodiments, the network device 452 may provide or indicate to the UE 402 one or more parameters to be used in the ML module 410 implemented in the UE 402. As an example, the network device 452 may send or indicate updated neural network weights to be implemented in the neural network executed by the ML module 410 on the UE side to attempt to optimize one or more aspects of the modulation and / or coding used for communications between the UE 402 and the T-TRP or NT-TRP.

[0131] In some embodiments, the UE 402 may implement the AI ​​itself, e.g., perform learning, while in other embodiments, the UE 402 may not perform the learning itself, but may be able to operate in conjunction with a network-side AI implementation, e.g., by receiving configuration from the network for an AI model (e.g., a neural network or other ML algorithm) implemented by the ML module 410 and / or by assisting other devices (e.g., network devices or other AI-enabled UEs) to train the AI ​​model (e.g., a neural network or other ML algorithm) by providing requested measurements or observations. For example, in some embodiments, the UE 402 may not itself implement learning or training, but the UE 402 may receive trained configuration information for an ML model determined by the network device 452 and execute the model.

[0132] 5 assumes network-side AI / ML capabilities, there may be cases where the network itself does not perform the training / learning, and instead the UE may perform the learning / training itself, possibly using dedicated training signals sent from the network. In other embodiments, end-to-end (E2E) learning may be implemented by the UE and network device 452.

[0133] Using AI, various processes such as link adaptation may be AI-enabled, for example, by implementing the AI ​​models described above. Some examples of possible AI / ML training processes and over-the-air information exchange procedures between devices during the training phase to facilitate AI-enabled processes according to embodiments of the present disclosure are described below.

[0134] Referring again to FIG. 5 , in the case of wireless federated learning (FL), the network device 452 may initialize a global AI / ML model implemented by the ML module 460, sample a group of UEs, such as the four UEs 402, 404, 406, and 408 shown in FIG. 5 , and broadcast the global AI / ML model parameters to the UEs. Each of the UEs 402, 404, 406, and 408 may then initialize its local AI / ML model using the global AI / ML model parameters and update (train) its local AI / ML model using its own data. Each of the UEs 402, 404, 406, and 408 may then report the parameters of its updated local AI / ML model to the network device 452. The network device 452 may then aggregate the updated parameters reported from the UEs 402, 404, 406, and 408 and update the global AI / ML model. The foregoing procedure is one iteration of the FL-based AI / ML model training procedure. The network device 452 and the UEs 402, 404, 406, and 408 perform multiple iterations until the AI / ML model has converged sufficiently to meet one or more training goals / criteria and the AI / ML model is finalized.

[0135] Aspects of the present disclosure provide solutions to overcome at least some of the aforementioned limitations, e.g., specific methods and devices for artificial intelligence or machine learning (AI / ML) model training. The methods and devices disclosed in this disclosure may overcome technical issues associated with transmitting AI / ML model training data, such as the inability to determine the relative importance of data of the same data type.

[0136] As described above, one way to reduce delays for the AI / ML model training process may be to minimize communication delays and / or computational delays. These delays may be reduced by transmitting each AI / ML model training data based on data importance. Data importance, for example, the importance of each AI / ML model training data, may be measured or determined based on data state information (DSI), such as data uncertainty. For example, if the uncertainty level of the data is relatively high, the data is more important (i.e., of higher importance) to the AI / ML model training process.

[0137] The DSI (e.g., data uncertainty) may be determined in a device, such as, but not limited to, a user equipment (UE) or a base station (BS), based on the AI / ML model training assistance information. The AI / ML model training assistance information may include at least one of information about a reference AI / ML model or at least one reference input data value. For example, the information about the reference AI / ML model may include a reference AI / ML model type, a reference AI / ML model structure, one or more reference AI / ML model parameters, a reference AI / ML model gradient, a reference AI / ML model activation function, a reference AI / ML model input data type, a reference AI / ML model output data type, a reference AI / ML model input data dimension, and / or a reference AI / ML model output data dimension. The reference input data value may be a respective reference value used as input data for the AI / ML model. In some cases, the AI / ML model training assistance information may not include at least one reference input data value, for example, when only one input data for the AI / ML model exists or when a device (e.g., a UE) has all inputs for the AI / ML model.

[0138] In some embodiments, a DSI threshold (e.g., an uncertainty threshold) may be used to determine whether each AI / ML model training data should be transmitted. For example, only AI / ML model training data with an uncertainty level higher than the uncertainty threshold may be reported for AI / ML model training.

[0139] In some embodiments, the AI / ML model training dataset information may be used to determine whether the respective AI / ML model training data should be transmitted. The AI / ML model training dataset information may include AI / ML model training dataset size and / or DSI distribution information of the AI / ML model training dataset.

[0140] A device (e.g., a UE, a BS) may determine a DSI for each AI / ML model training data based on the AI / ML model training assistance information. The device may then transmit the AI / ML model training data to another device (e.g., a UE or a BS on which the AI / ML model training is performed) based on the DSI for the AI / ML model training data and / or information related to the transmission of the AI / ML model training data.

[0141] In some embodiments, a device (e.g., a UE, a BS) may determine the DSI of each AI / ML model training data in federated learning. For example, a UE may determine the data uncertainty of its local AI / ML model training data (local AI / ML model training data at the UE) before uploading / transmitting local gradients associated with the local AI / ML model to a base station.

[0142] In some embodiments using double-sided AI / ML model training, a device (e.g., UE, BS) may determine the DSI of each AI / ML model training data after at least a portion of the AI / ML model training has occurred. For example, there are embodiments in which the UE includes an encoder and a reference decoder, the BS includes a decoder, and both the UE and BS perform the AI / ML model training. In these embodiments, the UE determines the data uncertainty of the AI / ML model training data after the AI / ML model training in the encoder and reference decoder has finished. The respective AI / ML model training data may include the respective outputs of the portion of the AI / ML model training (e.g., the respective outputs of the encoder and reference decoder in the UE).

[0143] In some embodiments, the AI / ML model training data may be transmitted based on the DSI of each AI / ML model training data. The AI / ML model training data may be transmitted according to respective reporting formats that may indicate the respective transmission accuracy. In one example, a UE may transmit AI / ML model training data with high accuracy and a high data uncertainty level to a BS, despite the high transmission overhead this may incur. On the other hand, the UE may transmit AI model training data with low accuracy and a low uncertainty level to a BS to reduce overhead. The relationship between the DSI of each AI / ML model training data and the respective transmission accuracy may be set by the BS in this case.

[0144] One-sided AI / ML model training in BS According to some embodiments, AI / ML model training data may be sent based on the importance of the AI / ML model training data. The importance of the AI / ML model training data may be measured or determined based on data status information (DSI) of the AI / ML model training data. The DSI may include data uncertainty.

[0145] Taking data uncertainty as an example of DSI, AI / ML model training data may be reported or transmitted to another device if the data uncertainty level of the AI / ML model training data is high. AI / ML model training data with a high data uncertainty level may be considered more important than AI / ML model training data with a low data uncertainty level because the AI / ML model training data with a high data uncertainty level may provide more (i.e., more useful) information for AI / ML model training. Furthermore, when AI / ML model training is performed using AI / ML model training data with a high data uncertainty level, the overfitting phenomenon is less likely to occur. In contrast, AI / ML model training data with a low data uncertainty level may be considered less important because such AI / ML model training data may not contribute sufficiently to the convergence of the AI / ML model (e.g., may contribute little or no to faster convergence of the AI / ML model).

[0146] In some embodiments, AI / ML model training and AI / ML inference may be performed at the BS. For the AI / ML model training and AI / ML inference procedures, the BS may receive AI / ML model training data from one or more UEs. The AI / ML model training data may be generated by each UE. The AI / ML model training may be one-sided AI / ML model training because the training is performed only at the BS side.

[0147] FIG. 7 illustrates an example of one-sided AI / ML model training at a BS 720 in a wireless network 700, in accordance with an embodiment of the present disclosure. Referring to FIG. 7, there are multiple UEs 710 and BSs 720 in the network 700. Each UE 710 is communicatively and operably connected to the BS 720. Each UE 710 may transmit AI / ML model training data to the BS 720 based on the importance of the respective AI / ML model training data. In some embodiments, each UE 710 may transmit AI / ML model training data to the BS 720 in a selective manner based on the importance of the respective AI / ML model training data. The importance of each AI / ML model training data may be determined based on the DSI of the respective AI / ML model training data, e.g., the data uncertainty of the respective AI / ML model training.

[0148] To determine the DSI for each AI / ML model training data, each UE 710 may need some assistance information. In some embodiments, this assistance information may be AI / ML model training assistance information provided by the BS 720. Each UE 710 may receive the AI / ML model training assistance information from the BS 720. In some embodiments, the UE 710 may receive the AI / ML model training assistance information periodically. However, in some embodiments, the UE 710 may receive the AI / ML model training assistance information aperiodically. In other words, the BS 720 may not need to transmit the AI / ML model training assistance information before the UE 710 determines the DSI for each AI / ML model training data. The UE 710 may use the received AI / ML model training assistance information to determine the DSI for each AI / ML model training data. The AI / ML model training assistance information may include at least one of information about a reference AI / ML model (or query AI / ML model) or at least one reference input data value. The information about the reference AI / ML model may include at least one of the following: Reference AI / ML model types (e.g., convolutional neural networks (CNNs), recurrent neural networks (RNNs), deep neural networks (DNNs)), Reference AI / ML model structure (e.g., number of layers, number of neurons in each layer), One or more reference AI / ML model parameters (e.g., weights, coefficients), Reference AI / ML model gradients, Reference AI / ML model activation functions (e.g., sigmoid, rectified linear unit (ReLU), exponential linear unit (ELu), softmax), Reference AI / ML model input data types, Reference AI / ML model output data type, Reference AI / ML model input data dimensions, or Reference AI / ML model output data dimension.

[0149] To determine the DSI of each AI / ML model training data, a UE (e.g., UE 710) may input the respective AI / ML model training data into a reference AI / ML model. The UE may determine the DSI of each AI / ML model training data based on the output of the reference AI / ML model that inputted the respective AI / ML model training data.

[0150] Once the DSI for each AI / ML model training data is determined, the UE 710 may report or transmit the respective AI / ML model data to the BS 720. The UE 710 may transmit the respective AI / ML model data based on the DSI for the respective AI / ML model training data and / or information related to the transmission of the respective AI / ML model training data. For example, the respective AI / ML model training data may be selectively transmitted (e.g., only some of the AI / ML model training data is transmitted, while other AI / ML model training data is not transmitted) based on the DSI for the respective AI / ML model training data and / or information related to the transmission of the respective AI / ML model training data. Information related to the transmission of the respective AI / ML model training data is described below or elsewhere herein.

[0151] As described above, the AI / ML model training support information that can be used to determine the DSI for each AI / ML model training data can include information about a reference AI / ML model. Figure 8A shows an example of a reference AI / ML model.

[0152] The reference AI / ML model 800 shown in FIG. 8A may be implemented using a DNN. That is, the type of the reference AI / ML model 800 may be a DNN. The reference AI / ML model input data dimension (i.e., the dimension of the input data of the reference AI / ML model 800) is M, and therefore, M pieces of input data (i.e., input 1 to input 2) are input. M ) exists. The reference AI / ML model output data dimension (i.e., the dimension of the output data of the reference AI / ML model 800) is N, and therefore, there are N pieces of output data (i.e., y1 to y N ) exist. The reference AI / ML model 800 may include L hidden layers (i.e., the number of hidden layers in the reference AI / ML model 800 is L), and each hidden layer includes K neurons (i.e., the number of neurons in each hidden layer is K). The reference AI / ML model 800 may include one or more reference AI / ML model (weight) parameters (e.g., w 11 , w 1K , wM1 , w MK ) The reference AI / ML model parameters may be determined based on the AI / ML model type (e.g., DNN in this case) and / or the reference AI / ML model structure. The reference AI / ML model activation function may be a predetermined function indicated using a preset function index.

[0153] In some embodiments, information regarding the reference AI / ML model may be transmitted from a BS (e.g., BS 720) to a UE (e.g., UE 710) using radio resource control (RRC), medium access control (MAC) control element (MAC-CE), or downlink control information (DCI). Information regarding the reference AI / ML model may be transmitted using broadcast signaling, unicast signaling, or groupcast signaling. In some embodiments, information regarding the reference AI / ML model may be specific to a UE or a group of UEs / devices. In such cases, information regarding the reference AI / ML model may be transmitted using unicast signaling or groupcast signaling.

[0154] As mentioned above, the AI / ML model training assistance information that may be used for determining the DSI of each input / ML model training data may also include at least one reference input data value or at least one reference AI / ML model input data. There may be two ways for a BS (e.g., BS 720) to transmit the reference AI / ML model input data.

[0155] The first method is for a BS (e.g., BS 720) to transmit reference values ​​for all input data of a reference AI / ML model. The UE (e.g., UE 710) may replace some of the received reference values ​​with local AI / ML model training data, as shown in FIG. 8B, which illustrates an example of a reference AI / ML model with reference AI / ML model input data. In particular, the UE that determines the DSI of each AI / ML model training data may use the UE i(e.g., one of the UEs 710), i may replace the i-th reference value with its local AI / ML model training data and then calculate the output of the reference AI / ML model. In some embodiments, to simplify this process, the UE i The reference AI / ML model input data to be replaced by UE may be indicated by some predetermined value. For example, the reference AI / ML model input data to be replaced may be filled with 0 or 1. i may determine the location of the reference AI / ML model input data to replace among all reference data based on the UE index and / or the reference AI / ML model input data dimension (i.e., the dimension of the entire reference AI / ML model input data).

[0156] The second method is for the BS (e.g., BS 720) to transmit reference values ​​for only a portion of the reference AI / ML model input data. The UE (e.g., UE 710) may replace the absent reference values ​​(i.e., input data not filled with reference values) with its local AI / ML model training data. In other words, the UE (e.g., UE 710) may add its local AI / ML model training data as input data for the reference AI / ML model when the input data is not available. The UE i may determine the location of the reference AI / ML model input data to add based on the UE index and / or the reference AI / ML model input data dimension (i.e., the dimension of the entire reference AI / ML model input data).

[0157] In some embodiments, the DSI of each AI / ML model training data may include information indicating at least one of: data uncertainty of the each AI / ML model training data, data importance of the each AI / ML model training data, degree of requirement of the each AI / ML model training data for AI / ML model training, or data diversity of the each AI / ML model training data.

[0158] The data uncertainty of each AI / ML model training data may be determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

[0159] In some embodiments, where the DSI of each AI / ML model training data includes information indicating the data uncertainty of each AI / ML model training data, the data uncertainty of each AI / ML model training data may be determined based on entropy. In such cases, the output data of each AI / ML model training data may be a probability function. For example, in FIG. 8B , where the AI / ML model 820 is designed to solve some classification problem, the input data is x i (i.e., input i ) and P(x i ) is the input data x i is the probability of belonging to category i. The entropy of the output of the AI / ML model 820 can be expressed in equation (1) as follows:

[0160]

number

[0161] According to the definition of entropy in equation (1) above, the larger the value of H(x), the higher the data uncertainty.

[0162] In some embodiments in which the DSI for each AI / ML model training data includes information indicating the data uncertainty of the respective AI / ML model training data, the data uncertainty for each AI / ML model training data may be determined based on a minimum confidence. For example, the AI / ML model (e.g., AI / ML model 820) may not assign a specific class to each AI / ML model training data because it is not certain of the class membership. Therefore, the AI / ML model (e.g., AI / ML model 820) may be trained to select the most informative and uncertain AI / ML model training data samples. In other words, AI / ML model training data with lower confidence in prediction is considered to be AI / ML model training data with higher data uncertainty. A margin sampling method may be used when selecting AI / ML model training data with uncertain predictions (e.g., for which it is difficult to predict the class to which the AI / ML model training data belongs). Using the margin sampling method, the AI / ML model training data with the smallest distance from the hyperplane may be the desired AI / ML model training data (e.g., the most uncertain data).

[0163] In some embodiments where the DSI for each AI / ML model training data includes information indicating the data uncertainty for each AI / ML model training data, the data uncertainty for each AI / ML model training data may be determined based on the generalization error (i.e., out-of-sample error). For supervised learning applications in machine learning and statistical learning theory, the generalization error may be used to determine how accurately an algorithm can predict output values ​​for unprecedented (i.e., never-before-seen) or uncertain data.

[0164] 8C illustrates an example of measuring data uncertainty using an AI / ML model 840 for channel information according to an embodiment of the present disclosure. In the example illustrated in FIG. 8C, data uncertainty is measured based on entropy, and the AI / ML model is implemented using a DNN. The input data of the AI / ML model 840 is the channel information, and the output data of the AI / ML model 840 is the probability of each MCS index. The AI / ML model 840 aims to provide the probability of each MCS index using the channel information as input data.

[0165] UE i (not shown in Figure 8C) is the UE that determines the data uncertainty of each AI / ML model training data. i data i1 is the input of AI / ML model 840 i When this is considered, the output probabilities of MCS5 and MCS7 are 95% and 5%, respectively. i Data i2 is input to AI / ML model 840 i If we consider the data i1 and data i2 as the output probabilities of MCS5 and MCS7, the output probabilities of MCS5 and MCS7 are 51% and 49%, respectively. The output probabilities of other MCS indices are negligible (i.e., close to zero) for both data i1 and data i2, and therefore, the output probabilities of other MCS indices can be ignored in this example. According to equation (1) shown above, the entropy of data i2 is greater than the entropy of data i1, so data i2 provides higher data uncertainty than data i2. Therefore, the UE i may report or transmit data i2 to the BS to enhance the training of the corresponding AI / ML model at the BS (e.g., improve AI / ML model training performance).

[0166] According to some embodiments, after the DSI of the AI / ML model training data is determined, for example, using one or more methods described above or elsewhere in this disclosure, a UE (e.g., UE 710) or a BS (e.g., BS 720) may determine whether the respective AI / ML model training data should be transmitted to the BS (e.g., BS 720). As shown in Figures 9A and 9B, there may be at least two methods for reporting the respective AI / ML model training data.

[0167] FIG. 9A shows a first method for reporting AI / ML model training data in a wireless network 700. In the first method, when the BS 720 transmits AI / ML model training assistance information, the BS 720 may also transmit information related to the transmission of the respective AI / ML model training data. For example, the AI / ML model training assistance information and the information related to the transmission of the respective AI / ML model training data may be transmitted together (collectively). The information related to the transmission of the respective AI / ML model training data may include a DSI threshold. In some embodiments, the DSI of the AI / ML model training data includes information indicating the data uncertainty of the respective AI / ML model training data, the DSI threshold may be a data uncertainty threshold. The BS 720 may set or pre-set a data uncertainty threshold for the UE 710 (e.g., each UE). For each AI / ML model training data, if the data uncertainty of the AI / ML model training data is higher than the data uncertainty threshold received from the BS 720, the UE 710 may transmit the AI / ML model training data to the BS 720 for AI / ML model training.

[0168] In some embodiments, data uncertainty may be quantified in the form of an uncertainty level (e.g., an uncertainty level ranging from level 0 to level N, where N is a positive integer). In this case, the uncertainty threshold may also be quantified in the form of an uncertainty level ranging from level 0 to level N (e.g., the uncertainty threshold is level 3).

[0169] In a first method of reporting AI / ML model training data, it may be the UE 710 that determines whether each AI / ML model training data should be transmitted or reported to the BS 720. As shown above in the example using an uncertainty threshold, the UE 710 may determine whether each AI / ML model training data should be transmitted or reported to the BS 720 based on the DSI threshold and the DSI of the each AI / ML model training data.

[0170] 9B illustrates a second method for reporting AI / ML model training data in wireless network 700. In the second method, the BS 720 may transmit AI / ML model training assistance information without information related to the transmission of the respective AI / ML model training data. After receiving the AI / ML model training assistance information, the UE 710 may transmit at least one of DSI for the respective AI / ML model training data or AI / ML model training dataset information. In some embodiments, the DSI for the respective AI / ML model training data may include information indicating data uncertainty (e.g., uncertainty value, uncertainty level) for the respective AI / ML model training data.

[0171] When the BS 720 receives the DSI of each AI / ML model training data (e.g., information indicating data uncertainty of each AI / ML model training data), the BS 720 may determine whether the respective AI / ML model training data should be transmitted from the UE 710. The BS 720 may then transmit information related to the transmission of the respective AI / ML model training data. In this case, the information related to the transmission of the respective AI / ML model training data may include information indicating whether the respective AI / ML model training data should be transmitted (reported) to the BS 720. For example, for each AI / ML model training data, if the BS 720 determines that the DSI (e.g., data uncertainty) of the AI / ML model training data meets the requirements of the AI / ML model and that the AI / ML model training data should be reported later, the BS 720 may transmit an authorization flag to indicate whether the UE 710 is permitted to transmit the AI / ML model training data. For example, if the authorization flag is set to “1,” the UE 710 is allowed to report the AI / ML model training data. Otherwise, the UE 710 is not permitted to report the AI / ML model training data. In some embodiments, the BS 720 may further indicate dynamic uplink transmission resources to be used for transmitting / reporting the AI / ML model training data. In some embodiments, the transmission resources used for transmitting / reporting the AI / ML model training data are pre-configured.

[0172] As indicated above, in the second method of reporting AI / ML model training data, it may be the BS 720 that determines whether each AI / ML model training data should be reported. The determination is made using the DSI of each AI / ML model training data (e.g., the data uncertainty of each AI / ML model training data) and indicated to the UE 710 using, for example, a permission flag.

[0173] In some embodiments, the BS 720 may update the AI / ML model training assistance information during the AI / ML model training procedure. In some cases, the BS (e.g., the BS 720) may transmit updated AI / ML model training assistance information to the UE (e.g., the UE 710) after the UE determines the DSI of the respective AI / ML model training data. The updated AI / ML model training assistance information may include at least one of information about an updated reference AI / ML model (e.g., updated reference AI / ML model parameters), updated reference input data values, or an updated DSI threshold (e.g., an updated uncertainty threshold). In this way, the evolution of the AI / ML model can be appropriately adapted.

[0174] To reduce transmission overhead, especially when only some of the reference AI / ML model parameters are changed, only updated reference AI / ML model parameters may be transmitted to the UE. In some embodiments, if a large portion of the AI / ML model has changed (e.g., the data uncertainty of a large portion of the AI / ML model training data is higher than a preset uncertainty threshold), the entire reference AI / ML model may be transmitted to the UE.

[0175] In some embodiments, a BS (e.g., BS 720) may use some information received from a UE (e.g., UE 710) to determine whether each AI / ML model training data should be transmitted to the BS. For example, the UE may transmit AI / ML model training dataset information to the BS, which may include at least one of an AI / ML model training dataset or DSI variance information for the AI / ML model training dataset. The DSI variance information for the AI / ML model training dataset may include a cumulative distribution function (CDF) and / or a probability density function (PDF). The AI / ML model training dataset information from each UE may be transmitted along with the corresponding DSI (e.g., data uncertainty level) for each AI / ML model training data. After receiving at least one of the DSI or AI / ML model training dataset information for each AI / ML model training data, the BS may determine whether the each AI / ML model training data should be reported. The BS may transmit information (e.g., a permission flag) to the UE indicating whether the each AI / ML model training data should be transmitted.

[0176] In some embodiments, the AI / ML model training dataset information may be transmitted using a buffer status report (BSR). For example, the BSR may carry the AI / ML model training dataset information by adding one or more extra fields therein. To indicate the CDF or PDF, each AI / ML model training data value may be quantified into N data levels (N is a positive integer). The probabilities of the N data levels may be indicated according to the ascending order of the N data levels.

[0177] In some embodiments, the transmission of the AI / ML model training dataset information may be (implicitly) associated with a scheduling request (SR). The relationship between the SR resource and the AI / ML model training dataset information may be configured or pre-configured by the BS. In this way, the BS may obtain the AI / ML model training dataset information when it receives an SR on the corresponding resource.

[0178] One-sided AI / ML model training in UE In some embodiments, AI / ML model training and AI / ML inference may be performed at the UE. For the AI / ML model training and AI / ML inference procedures, the UE may receive AI / ML model training data from the BS. The AI / ML model training data may be generated by the BS. This AI / ML model training may also be one-sided AI / ML model training because the training is performed only at the UE side.

[0179] FIG. 10 illustrates an example of one-sided AI / ML model training at a UE 710 in a wireless network 700, in accordance with an embodiment of the present disclosure. Referring to FIG. 10, in the network 700, there is a UE 710 and a BS 720 that are operatively connected and capable of communicating with each other. The BS 720 may transmit AI / ML model training data to the UE 710 based on the importance of the respective AI / ML model training data. In some embodiments, the BS 720 may transmit AI / ML model training data to the UE 710 in a selective manner based on the importance of the respective AI / ML model training data. As indicated above or elsewhere in this disclosure, the importance of each AI / ML model training data may be determined based on the DSI of the respective AI / ML model training data, e.g., the data uncertainty of the respective AI / ML model training.

[0180] 7, in the example shown in FIG. 10, both AI / ML model training and AI / ML inference are performed in the UE 710. The UE 710 may transmit AI / ML model training assistance information to the BS 720 and receive AI / ML model training data from the BS 720. For example, the BS 720 may use a sounding reference signal transmitted from the UE 710 to the BS 720 to acquire uplink (UL) channel information that may not be directly acquired at the UE 710. The BS 720 may transmit the acquired UL channel information to the UE 710 as AI / ML model training data when the AI / ML model is deployed in the UE 710.

[0181] In some embodiments of one-sided AI / ML model training at a UE (e.g., UE 710), a BS (e.g., BS 720) may determine a DSI (e.g., data uncertainty) for each AI / ML model training data. When each AI / ML model training data is generated at the BS side, the UE may transmit associated AI / ML model training assistance information to the BS using, for example, RRC, MAC-CE, DCI, broadcast signaling, unicast signaling, and / or groupcast signaling. The AI / ML model training assistance information may include at least one of information about a reference AI / ML model or at least one reference input data value. The BS may determine a DSI for each AI / ML model training data and transmit the AI / ML model training data to the UE based on the DSI for each AI / ML model training data.

[0182] The method for determining the DSI for each AI / ML model may be substantially the same as the method for the embodiment of one-sided AI / ML model training at the BS (e.g., the example described above and shown in Figures 9A and 9B), except that the roles of the BS and UE are swapped.

[0183] In some embodiments, the BS may selectively transmit the respective AI / ML model training data. For example, the BS may determine whether to report or transmit the respective AI / ML model training data based on a DSI threshold (e.g., an uncertainty threshold) received from the UE and the DSI (e.g., data uncertainty) of the respective AI / ML model training data. The manner in which the BS determines whether to transmit the AI / ML model training data to the UE may be substantially similar to the manner in which the UE determines in the embodiment of one-sided AI / ML model training at the BS (e.g., the example shown in FIGS. 9A and 9B), except that the roles of the BS and the UE are swapped.

[0184] In some embodiments, the BS may determine whether to report or transmit each AI / ML model training data based on information related to the transmission of the respective AI / ML model training data received from the UE. The BS may transmit at least one of DSI for the respective AI / ML model training data or AI / ML model training dataset information. The UE may then determine whether the respective AI / ML model training data should be transmitted to the UE and then transmit information indicating whether the respective AI / ML model training data should be transmitted to the BS.

[0185] In some embodiments, the UE may send a request for AI / ML model training data, for example, when the UE needs some AI / ML model training data from the BS. After receiving the AI / ML model training data request, the BS may transmit the AI / ML model training data as described above or elsewhere in this disclosure.

[0186] The method for transmitting the respective AI / ML model training data may be substantially the same as the method for the embodiment of one-sided AI / ML model training at the BS (e.g., the example shown above and in FIGS. 9A and 9B), except that the roles of the BS and UE are swapped. An example illustrating an exemplary process for reporting the respective AI / ML model training data from the BS 720 to the UE 710 in the wireless network 700 is shown in FIGS. 11A and 11B.

[0187] AI / ML model training in federated learning According to some embodiments, AI / ML model training and AI / ML inference may be performed using federated learning techniques. Federated learning, also known as collaborative learning, is a machine learning technique that trains algorithms across multiple distributed edge devices or servers. Each distributed edge device or server may maintain local data samples but may not exchange them with other devices or servers. Federated learning techniques are the opposite of traditional centralized machine learning techniques in that, while in traditional centralized machine learning techniques, all local data sets are uploaded to one server, in federated learning techniques, local data samples are not shared.

[0188] In a wireless federated learning-based (FL-based) AI training process, a network node / device / nodes initialize a global AI model, sample a group of user devices, and broadcast the global AI model parameters to the user devices. Each user device then initializes its local AI model using the global AI model parameters and updates (trains) its local AI model using its own data. Each user device may then report the parameters of its updated local AI model to the network device, which then aggregates the updated parameters reported by the user devices and updates the global AI model. The foregoing procedure is one iteration of a conventional FL-based AI model training procedure. The network devices and participating user devices typically perform multiple iterations until the AI ​​model has converged sufficiently to meet one or more training goals / criteria and the AI ​​model is finalized.

[0189] FIG. 12A shows an example of a process for AI / ML model training in federated learning within a wireless network 700. During federated learning, AI / ML model training may be performed cooperatively or jointly by multiple client devices (e.g., UEs) and one central server device (e.g., a BS). Federated learning may be performed, for example, as described below and shown in FIG. 12A. As shown in FIG. 12A, each of the UEs 710 may train its local AI / ML model using local AI / ML training data. Upon completion of training, each UE 710 may update the local gradients associated with its local AI / ML model. Each UE 710 may then transmit the updated local gradients to the BS 720 (e.g., upload them to a server). After receiving the local gradients from each UE 710, the BS 720 may aggregate all of the received local gradients and generate one or more global gradients associated with a global AI / ML model. The BS 720 may transmit the global gradients to each UE 710 (e.g., download them to each UE). The above procedure can be repeated until the global AI / ML model converges.

[0190] In the federated learning procedure, there may be frequent exchange of data (e.g., transmission of local and global gradients) between the client (e.g., UE 710) and the server (e.g., BS 720), which may result in inordinate transmission overhead.

[0191] To keep transmission overhead within an acceptable level, the transmission of gradients for AI / ML model training should be controlled. To control the transmission of gradients, a client device (e.g., UE 710, BS 720) may transmit local gradients associated with a local AI / ML model to a central server based on the DSI of the respective AI / ML model training data. Here, the respective AI / ML model training data may include at least one of local AI / ML model training data or local gradients associated with the local AI / ML model of the client device's local AI / ML model.

[0192] For example, the UE 710 may determine the DSI (e.g., data uncertainty) of its local AI / ML model training data for its local AI / ML model. If the DSI of the local AI / ML model training data is lower than a DSI threshold (e.g., the data uncertainty of the local AI / ML model training data is lower than an uncertainty threshold, meaning the local AI / ML model is relatively stable), the UE 710 may not transmit local gradients associated with the local AI / ML model (and / or local AI / ML model training data) because the local gradients do not contribute to AI / ML model training (e.g., convergence of the global AI / ML model). In such a case, the UE 710 may skip transmitting (e.g., uploading) local gradients associated with the UE 710's local AI / ML model in the current iteration.

[0193] FIG. 12B illustrates an example process for AI / ML model training in federated learning using data state information (DSI) of AI / ML model training data in a wireless network 700.

[0194] As shown in FIG. 12B , UEs 710a and 710b may train their local AI / ML models using their local AI / ML training data. Once training is complete, each of UEs 710a and 710b may update local gradients associated with its local AI / ML model. Each of UEs 710a and 710b may determine the DSI of its respective AI / ML model in the manner described above or elsewhere in this disclosure. In FIG. 12B , the DSI is data uncertainty, and the data uncertainty of UE 710a's local AI / ML model training data is higher than a set or preset uncertainty threshold. Therefore, UE 710a may transmit the local gradients associated with its local AI / ML model to BS 720. On the other hand, the data uncertainty of UE 710b's local AI / ML model training data is lower than a set or preset uncertainty threshold. Therefore, UE 710b may be prohibited from transmitting the local gradients associated with its local AI / ML model to BS 720. Optionally, the UE 710b may instead send an indication to inform the BS 720 that there are no local AI / ML model updates in this iteration. The remaining procedures may be substantially similar to those described above and shown in FIG. 12A.

[0195] Determining the DSI for each of the AI / ML model training data (e.g., local AI / ML model training data) during federated learning may be performed in the same manner as described above (e.g., the embodiment of one-sided AI / ML model training at the BS) or elsewhere in this disclosure.

[0196] Two-sided AI / ML model training According to some embodiments, AI / ML model training and AI / ML inference may be performed in both the UE and the BS. In other words, the BS and the UE may cooperate to perform the AI / ML model training and AI / ML inference procedures. An example of a double-sided AI / ML model is an autoencoder for channel state information (CSI) compression, as shown in FIG. 13.

[0197] 13 , original channel data 1311 (e.g., original channel state information (CSI)) may be generated at a UE 1310. The original channel data 1311 may be conveyed to a CSI encoder 1312 to obtain compressed channel data 1313 (e.g., compressed CSI). Due to the compression procedure, some or all of the compressed channel data 1313 may be transmitted to a BS 1320 with lower signaling overhead. After receiving the compressed channel data 1321 (e.g., compressed CSI), the BS 1320 may reconstruct the original channel data (i.e., reconstructed channel data 1323) using a decoder 1322. The compressed channel data 1321 input to the decoder 1322 may be the same as the compressed channel data 1313 output from the encoder 1312.

[0198] If the UE 1310 trains the encoder 1312 and the BS 1320 trains the decoder 1322 (i.e., train separately), the UE 1310 may use the reference decoder 1314 to reconstruct the original channel data and thus generate the reconstructed channel data 1315. In this way, divergence between the output of the encoder 1312 (i.e., compressed channel data 1313) and the input of the decoder 1322 (i.e., compressed channel data 1321) may be prevented. The reference decoder 1314 may be configured or predefined by the BS 1320. In some embodiments, information related to the reference decoder 1314 may be forwarded to the UE 1310. In some embodiments, information related to the reference decoder 1314 or the reference decoder 1314 may be considered AI / ML model training assistance information.

[0199] After the reconstructed channel data 1315 is generated, the UE 1310 receives the output of the encoder 1312 (i.e., the compressed channel data 1313 or V mid ) and the output of the reference decoder 1314 (i.e., the reconstructed channel data 1315 or V out) to the BS 1320. The BS 1320 may use the received dataset to train a decoder 1322, where V mid and V out is used as labeled data for training the decoder 1322 in the BS 1320.

[0200] In some embodiments, to support faster AI / ML model training (e.g., faster convergence of the AI / ML model) and avoid extra signaling overhead, the DSI (e.g., data uncertainty) of each AI / ML model training data may be used as one constraining factor when enhancing the AI / ML model training on both sides. For example, the UE 1310 may determine the data uncertainty of a dataset including the output of the encoder 1312 and the output of the reference decoder 1314 after training is completed in the encoder 1312 and the reference decoder 1314. If the data uncertainty level of the dataset is high, e.g., higher than a preset uncertainty threshold, the UE 1310 may determine the data uncertainty of the output of the encoder 1312 (i.e., compressed channel data 1313 or V mid ) and the reconstructed channel data 1315 (i.e., V out ) to the BS 1320. In contrast, if the data uncertainty level of the data set is low, e.g., lower than a preset uncertainty threshold, the UE 1310 may not be allowed to transmit the data set to the BS 1320. The UE 1310 may update one or more parameters of the encoder 1312 to generate a new output (i.e., new compressed channel data 1313 or new V mid ) and reassess the data uncertainty of the dataset.

[0201] Determining the DSI for each of the AI / ML model training data (e.g., local AI / ML model training data) for the two-sided AI / ML model training embodiment may be performed in the same manner as described above (e.g., the one-sided AI / ML model training embodiment at the BS) or elsewhere in this disclosure.

[0202] AI / ML model training data transmission based on reporting format As described above, in some embodiments, the respective AI / ML model training data may be selectively transmitted (e.g., some AI / ML model training data is transmitted, but other AI / ML model training data is not transmitted). For example, the respective AI / ML model training data may be transmitted only when the data uncertainty of the AI / ML model training data is greater than a preset uncertainty threshold. Compared to these embodiments, in some other embodiments, all of the respective AI / ML model training data may be transmitted regardless of the DSI (e.g., data uncertainty) of the respective AI / ML model training data. By transmitting all of the AI / ML model training data, more information may be provided for AI / ML model training, thereby potentially achieving faster convergence of the AI / ML model. However, such data transmission results in higher transmission overhead due to the transmission of less important AI / ML model training data (e.g., AI / ML model training data with lower data uncertainty).

[0203] To achieve a balance between improving AI / ML model performance and reducing transmission overhead, the AI / ML model training data may be transmitted according to respective reporting formats. In some embodiments, each reporting format may indicate a respective transmission accuracy, and the respective level of transmission accuracy may be determined based on the level of DSI of the respective AI / ML model training data. In some embodiments, each reporting format may indicate a setting for transmission of the respective AI / ML model training data.

[0204] As described above, in some embodiments, each reporting format may indicate a respective transmission accuracy. In other words, the AI / ML model training data may be transmitted based on the respective transmission accuracy indicated in each reporting format. If some of the AI / ML model training data has high data uncertainty (i.e., the AI / ML model training data contributes more to AI / ML model training), the transmission of the AI / ML model training data may be performed with high accuracy to ensure successful data transmission. On the other hand, if some of the AI / ML model training data has low data uncertainty (i.e., the AI / ML model training data provides limited information and therefore contributes little to AI / ML model training), the transmission of the AI / ML model training data may be performed with low accuracy to reduce transmission overhead.

[0205] In some embodiments, when the respective transmission accuracy is higher than a predetermined value (e.g., high accuracy), the respective AI / ML model training data is transmitted in its entirety. In other words, the complete AI / ML model training data may be transmitted, or the raw AI / ML model training data may be transmitted without preprocessing. For example, when the AI / ML model training data is channel information, complete channel information including all real and imaginary values ​​may be transmitted when the transmission accuracy is high. On the other hand, when the respective transmission accuracy is lower than a predetermined value (e.g., low accuracy), only a portion of the respective AI / ML model training data or information extracted from the respective AI / ML model training data is transmitted. For example, when the AI / ML model training data is channel information, conventional CSI such as channel quality information (CQI), rank indicator (RI), layer indicator (LI), reference signal received power (RSRP), and precoding matrix indicator (PMI) may be transmitted instead of the complete channel information when the transmission accuracy is low.

[0206] In some embodiments, each reporting format may indicate whether each AI / ML model training data includes channel state information (CSI) or raw channel information. In the case of raw channel information, each reporting format may indicate, for example, whether to transmit channel information for six subcarriers in a resource block (RB) or three subcarriers in an RB.

[0207] As described above, in some embodiments, each reporting format may indicate a configuration for transmitting the respective AI / ML model training data. In other words, the AI / ML model training data may be transmitted based on the configuration for transmitting the respective AI / ML model training data indicated in the respective reporting format. The configuration for transmitting the respective AI / ML model training data may indicate at least one of resources used for transmitting the respective AI / ML model training data or a quantization granularity used for transmitting the respective AI / ML model training data.

[0208] In some embodiments, when the respective transmission accuracy is higher than a predetermined value (e.g., high accuracy), the respective AI / ML model training data may be transmitted using more subcarriers per resource block (RB) or more bits per RB than when the respective transmission accuracy is lower than the predetermined value (e.g., low accuracy). In other words, when the transmission accuracy is high, more transmission resources may be allocated or a lower MCS value may be used for transmitting the respective AI / ML model training data (i.e., a larger number of subcarriers, RBs, subbands, symbols, and / or minislots for transmission). On the other hand, when the transmission accuracy is low, fewer transmission resources may be allocated or a higher MCS value may be used for transmitting the respective AI / ML model training data (i.e., a smaller number of subcarriers, RBs, subbands, symbols, and / or minislots for transmission). For example, for high transmission accuracy, six subcarriers per RB may be used for transmitting the AI / ML model training data, and the resource utilization density is 1 / 2. However, due to low transmission accuracy, only two subcarriers per RB may be used to transmit AI / ML model training data, resulting in a resource utilization density of 1 / 6.

[0209] In some embodiments, when the respective transmission precision is higher than a predetermined value (e.g., high precision), the respective AI / ML model training data may be transmitted using a finer quantization granularity (e.g., using more bits to represent the raw data). On the other hand, when the respective transmission precision is lower than a predetermined value (e.g., low precision), the respective AI / ML model training data may be transmitted using a coarser quantization granularity (e.g., using fewer bits to represent the raw data). For example, when the transmission precision is high, 8 bits (e.g., 4 bits for the real part and 4 bits for the imaginary part) may be used to represent one complex value for transmission of the AI / ML model training data. When the transmission precision is low, only 4 bits (e.g., 2 bits for the real part and 2 bits for the imaginary part) may be used to represent one complex value for transmission of the AI / ML model training data.

[0210] In some embodiments, when the respective transmission accuracy is higher than a predetermined value (e.g., high accuracy), the AI / ML model training data may be transmitted in its entirety using more transmission resources (e.g., a larger number of subcarriers per RB or a smaller number of bits per RB) and / or a finer quantization granularity. On the other hand, when the respective transmission accuracy is lower than a predetermined value (e.g., low accuracy), only a portion of the respective AI / ML model training data or information extracted from the respective AI / ML model training data may be transmitted using fewer transmission resources (e.g., a smaller number of subcarriers per RB or a larger number of bits per RB) and / or a coarser quantization granularity.

[0211] In some embodiments, the relationship between the DSI of each AI / ML model training data and the respective reporting format may be set by the BS or device on which the AI / ML training is performed (e.g., the UE on which one-sided AI / ML model training is performed).

[0212] For example, the BS may set a relationship between data uncertainty for the AI / ML model training data and the corresponding data transmission accuracy. The BS may set a mapping table for the data uncertainty for the AI / ML model training data and the corresponding data transmission accuracy, as shown in Tables 1-4 below. Tables 1-4 show how each data uncertainty level can be mapped to a data transmission accuracy level. In Tables 1-4, the data uncertainty levels range from 1 to 8.

[0213] Table 1 shows the mapping between data uncertainty levels and data transmission accuracy levels with full / partial data.

[0214] [Table 1]

[0215] Table 2 shows the mapping between data uncertainty levels and data transmission accuracy levels by resource granularity.

[0216] [Table 2]

[0217] Table 3 shows the mapping between data uncertainty levels and data transmission precision levels through quantization.

[0218] [Table 3]

[0219] Table 4 is a combination of Tables 1, 2, and 3 and shows the mapping between data uncertainty levels and data transmission precision levels by full / partial data, resource granularity, and quantization.

[0220] [Table 4]

[0221] FIG. 14 illustrates an example of AI / ML model training with data transmission precision adaptation, according to an embodiment of the present disclosure. The AI / ML model 1400 illustrated in FIG. 14 may be implemented using a DNN. In other words, the type of the AI / ML model 1400 may be a DNN. The AI / ML model input data dimension (i.e., the dimension of the input data of the AI / ML model 1400) is M, and therefore, M pieces of input data (i.e., input 1 to input 2) are input. M ) exists. The AI / ML model output data dimension (i.e., the dimension of the output data for the AI / ML model 1400) is 1, and therefore there is one output data (i.e., the optimal MCS in slot n+k). The AI / ML model 1400 may include L hidden layers (i.e., the number of hidden layers in the AI / ML model 1400 is L), and each hidden layer includes K neurons (i.e., the number of neurons in each hidden layer is K). The purpose of the AI / ML model 1400 is to input historical channel information in slot n (e.g., input i ) to predict the optimal MCS in slot n+k. To provide more information to support AI / ML model training, all of the historical channel information may be allowed to be transmitted. However, to reduce transmission overhead, each AI / ML model training data may be transmitted based on its respective transmission accuracy indicated in its respective reporting format. In other words, different transmission accuracy may be applied to the transmission of each AI / ML model training data to reduce transmission overhead.

[0222] Determining the DSI for each of the AI / ML model training data (e.g., local AI / ML model training data) for the two-sided AI / ML model training embodiment may be performed in the same manner as described above (e.g., the one-sided AI / ML model training embodiment at the BS) or elsewhere in this disclosure.

[0223] 15 is a flow diagram illustrating an example process for AI / ML model training in a wireless communication network, according to an embodiment of the present disclosure. Referring to FIG. 15, in some embodiments, the first device may be a UE and the second device may be a BS. In some other embodiments, the first device may be a BS and the second device may be a UE. In some other embodiments, the first device and the second device may be UEs. In some other embodiments, the first device and the second device may be BSs.

[0224] At step 1510, the first device may receive from the second device the AI / ML model training assistance information and information related to the transmission of the respective AI / ML model training data. In some embodiments, the AI / ML model training assistance information and the information related to the transmission of the respective AI / ML model training data are transmitted together, e.g., using one DCI message or one sidelink control information (SCI) message. In some embodiments, the AI / ML model training assistance information and the information related to the transmission of the respective AI / ML model training data are separate. For example, the AI / ML model training assistance information is carried in one DCI or SCI message, and the information related to the transmission of the respective AI / ML model training data is carried in another DCI or SCI message.

[0225] The AI / ML model training assistance information may include at least one of information about a reference AI / ML model or at least one reference input data value. The information about the reference AI / ML model may include at least one of a reference AI / ML model type, a reference AI / ML model structure, one or more reference AI / ML model parameters, a reference AI / ML model gradient, a reference AI / ML model activation function, a reference AI / ML model input data type, a reference AI / ML model output data type, a reference AI / ML model input data dimension, or a reference AI / ML model output data dimension. In some embodiments, the AI / ML model training assistance information may be updated by the second device. In such a case, the second device may send the updated AI / ML model training assistance information to the first device.

[0226] The information associated with the transmission of each AI / ML model training data may include a DSI threshold that may be set by the second device for use in determining whether the each AI / ML model training data should be transmitted to the second device.

[0227] In some embodiments in which the first and second devices cooperate for AI / ML model training, the first device may perform a portion of the AI / ML model training before determining the DSI of the respective AI / ML model training data in step 1520, and the respective AI / ML model training data may include respective outputs of the portion of the AI / ML model training.

[0228] In step 1520, the first device may determine a DSI for each AI / ML model training data based on the AI / ML model training assistance information.

[0229] In some embodiments, the DSI for each AI / ML model training data may include information indicating at least one of the following: data uncertainty for each AI / ML model training data, data importance for each AI / ML model training data, a degree of requirement for each AI / ML model training data for AI / ML model training, or data diversity for each AI / ML model training data. The data uncertainty for each AI / ML model training data may be determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

[0230] In some embodiments, determining the DSI for each AI / ML model training data may include inputting the each AI / ML model training data to a reference AI / ML model and determining the DSI based on an output of the reference AI / ML model. Each AI / ML model training data input to the reference AI / ML model may replace at least one reference input data value.

[0231] In step 1530, the first device may determine whether the respective AI / ML model training data should be transmitted to the second device based on the DSI threshold and the DSI of the respective AI / ML model training data.

[0232] At step 1540, the first device may transmit the respective AI / ML model training data to the second device based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data.

[0233] In some embodiments, the respective AI / ML model training data may be transmitted according to respective reporting formats determined based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate respective transmission accuracy, and the respective level of transmission accuracy may be determined based on the level of the DSI of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate a configuration for the transmission of the respective AI / ML model training data. The configuration for the transmission of the respective AI / ML model training data may indicate at least one of resources used for the transmission of the respective AI / ML model training data or a quantization granularity used for the transmission of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information. In some embodiments, the second device may configure a relationship between the DSI of the respective AI / ML model training data and the respective reporting formats.

[0234] In some embodiments, the respective AI / ML model training data may include at least one of local AI / ML model training data or local gradients associated with the local AI / ML model of the first device.

[0235] The second device may perform AI / ML model training using the respective AI / ML model training data at step 1550. As noted above, in some embodiments, the AI / ML model training is also performed in part by the first device.

[0236] 16 is a flow diagram illustrating another example process for AI / ML model training in a wireless communication network, according to an embodiment of the present disclosure. Referring to FIG. 16, in some embodiments, the first device may be a UE and the second device may be a BS. In some other embodiments, the first device may be a BS and the second device may be a UE. In some other embodiments, the first device and the second device may be UEs. In some other embodiments, the first device and the second device may be BSs.

[0237] In step 1610, the first device may receive AI / ML model training assistance information from the second device. The AI / ML model training assistance information may include at least one of information about a reference AI / ML model or at least one reference input data value. The information about the reference AI / ML model may include at least one of a reference AI / ML model type, a reference AI / ML model structure, one or more reference AI / ML model parameters, a reference AI / ML model gradient, a reference AI / ML model activation function, a reference AI / ML model input data type, a reference AI / ML model output data type, a reference AI / ML model input data dimension, or a reference AI / ML model output data dimension. In some embodiments, the AI / ML model training assistance information may be updated by the second device. In such a case, the second device may send the updated AI / ML model training assistance information to the first device.

[0238] In some embodiments in which the first and second devices cooperate for AI / ML model training, before determining the DSI of the respective AI / ML model training data in step 1620, the first device may perform a portion of the AI / ML model training, and the respective AI / ML model training data may include respective outputs of the portions of the AI / ML model training.

[0239] In step 1620, the first device may determine a DSI for each AI / ML model training data based on the AI / ML model training assistance information.

[0240] In some embodiments, the DSI for each AI / ML model training data may include information indicating at least one of the following: data uncertainty for each AI / ML model training data, data importance for each AI / ML model training data, a degree of requirement for each AI / ML model training data for AI / ML model training, or data diversity for each AI / ML model training data. The data uncertainty for each AI / ML model training data may be determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

[0241] In some embodiments, determining the DSI for each AI / ML model training data may include inputting the each AI / ML model training data to a reference AI / ML model and determining the DSI based on an output of the reference AI / ML model. Each AI / ML model training data input to the reference AI / ML model may replace at least one reference input data value.

[0242] In step 1630, the first device may transmit to the second device at least one of DSI or AI / ML model training dataset information for the respective AI / ML model training data. In some embodiments, the AI / ML model training dataset information may include at least one of an AI / ML model training dataset size or DSI distribution information for the AI / ML model training dataset. In some embodiments, the AI / ML model training dataset information may be transmitted using a buffer status report (BSR) or a scheduling request (SR). Step 1630 may be an optional step.

[0243] In step 1640, the second device may use at least one of the DSI of the respective AI / ML model training data or the AI / ML model training dataset information to determine whether the respective AI / ML model training data should be transmitted to the second device. Step 1640 may be an optional step.

[0244] At step 1650, the second device may transmit information related to the transmission of the respective AI / ML model training data to the first device. The information related to the transmission of the respective AI / ML model training data may include information indicating whether the respective AI / ML model training data should be transmitted to the second device. Thus, at step 1650, the second device may transmit to the first device information indicating whether the respective AI / ML model training data should be transmitted to the second device.

[0245] In step 1660, the first device may transmit the respective AI / ML model training data to the second device based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data.

[0246] In some embodiments, the respective AI / ML model training data may be transmitted according to respective reporting formats determined based on at least one of the DSI of the respective AI / ML model training data or information related to the transmission of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate respective transmission accuracy, and the respective level of transmission accuracy may be determined based on the level of the DSI of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate a configuration for the transmission of the respective AI / ML model training data. The configuration for the transmission of the respective AI / ML model training data may indicate at least one of resources used for the transmission of the respective AI / ML model training data or a quantization granularity used for the transmission of the respective AI / ML model training data. In some embodiments, the respective reporting formats may indicate whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information. In some embodiments, the second device may configure a relationship between the DSI of the respective AI / ML model training data and the respective reporting formats.

[0247] In some embodiments, the respective AI / ML model training data may include at least one of local AI / ML model training data or local gradients associated with the local AI / ML model of the first device.

[0248] The second device may perform AI / ML model training using the respective AI / ML model training data at step 1670. As mentioned above, in some embodiments, the AI / ML model training is also performed in part by the first device.

[0249] The embodiments described above are in the context of a UE communicating with a BS. However, more generally, devices that wirelessly communicate with each other via time-frequency resources do not necessarily need to be one or more UEs communicating with a BS. For example, two or more UEs may wirelessly communicate with each other via a sidelink using device-to-device (D2D) communication. As another example, two network devices (e.g., a terrestrial base station and a non-terrestrial base station such as a drone) may wirelessly communicate with each other via a backhaul link. The embodiments are not limited to uplink and / or downlink communication. For example, in the above embodiments, the BS may be replaced by another device, such as a node in the network or a UE. The uplink / downlink communication may instead be sidelink communication. Thus, as mentioned above, the first device may be a UE or a network device (e.g., a BS), and the second device may be a UE or a network device (e.g., a BS).

[0250] According to some aspects of the present disclosure, the performance of an AI / ML model may be improved and overfitting may be avoided during the training process of the AI / ML model. Furthermore, transmission overhead (e.g., air interface overhead) may be reduced because a smaller amount of AI / ML model training data samples may be transmitted based on data state information (DSI) of the AI / ML model training data. The DSI (e.g., data uncertainty) may be measured at a device (e.g., UE, BS) before the device reports or transmits the AI / ML model training data.

[0251] According to some aspects of the present disclosure, for example, in federated learning, transmission of AI / ML model training data (e.g., local gradients) that does not contribute to the convergence of the global AI / ML model is avoided. Thus, signaling overhead in federated learning may be reduced, AI / ML model performance may be improved, and AI / ML model training may be enhanced.

[0252] According to some aspects of the present disclosure, fast convergence of AI / ML models can be achieved in two-sided AI / ML model training. Due to the reduced number of transmissions of AI / ML model training datasets, extra signaling overhead can be avoided.

[0253] According to some aspects of the present disclosure, a balance between improved performance of AI / ML models and reduced transmission overhead may be achieved.

[0254] Examples of devices (e.g., EDs or UEs and TRPs or network devices) for performing the various methods described herein are also disclosed.

[0255] For example, a (first) device may include a memory for storing processor-executable instructions and a processor for executing the processor-executable instructions. When the processor executes the processor-executable instructions, the processor may be caused to perform one or more method steps of the devices described herein, e.g., with respect to Figures 7-16. For example, the processor may implement operations consistent with an operating mode, e.g., performing necessary measurements and generating content from those measurements, preparing uplink transmissions, processing downlink transmissions (e.g., encoding, decoding, etc.), and configuring and / or commanding transmit / receive on RF chains and antennas, as configured for that operating mode, thereby causing the device to communicate over the air interface in that operating mode.

[0256] It should be noted that the phrase "at least one of A or B" as used herein is interchangeable with the phrase "A and / or B." This refers to a list from which A or B, or both A and B, may be selected. Similarly, "at least one of A, B, or C" as used herein is interchangeable with "A and / or B and / or C" or "A, B, and / or C." This refers to a list from which A or B or C, or both A and B, or both A and C, or both B and C, or all of A, B, and C may be selected. The same principle applies to longer lists having the same format.

[0257] While the present invention has been described with reference to specific features and embodiments thereof, various modifications and combinations can be made without departing from the invention. Accordingly, the description and drawings should be considered merely as illustrative of some embodiments of the invention as defined by the appended claims, and it is intended to embrace any and all modifications, variations, combinations, or equivalents that fall within the scope of the invention. Thus, while the invention and its advantages have been described in detail, various changes, substitutions, and alterations can be made therein without departing from the present disclosure as defined by the appended claims. Moreover, the scope of this application is not intended to be limited to the particular embodiments of the processes, machines, manufacture, compositions, means, methods, and steps described herein. As those skilled in the art will readily appreciate from this disclosure, existing or later-developed processes, machines, manufactures, compositions, means, methods, or steps that perform substantially the same function or achieve substantially the same results as the corresponding embodiments described herein can be utilized in accordance with the present disclosure. Accordingly, it is intended that the appended claims include within their scope such processes, machines, manufactures, compositions, means, methods, or steps.

[0258] Additionally, any module, component, or device illustrated herein that executes instructions may include or otherwise have access to a non-transitory computer / processor-readable storage medium or media for storage of information such as computer / processor-readable instructions, data structures, program modules, and / or other data. A non-exhaustive list of examples of non-transitory computer / processor-readable storage media includes magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, optical disks such as compact disk read-only memory (CD-ROM), digital video disks or digital versatile disks (DVD), Blu-ray Disc™, or other optical storage, volatile and non-volatile, removable and non-removable media implemented in any manner or technology, random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technology. Any such non-transitory computer / processor storage medium may be part of or accessible or connectable to the device. Any applications or modules described herein may be implemented using computer / processor readable / executable instructions, which may be stored or otherwise maintained by such non-transitory computer / processor readable storage media. Acronym Definition AI artificial intelligence LTE Long Term Evolution NR new radio BWP Bandwidth Portion BS base station CA Carrier Aggregation CC Component Carrier CG Cell Group CSI Channel State Information CSI-RS Channel State Information Reference Signal DNN Deep Neural Network DC Dual Connectivity DCI Downlink Control Information DL Downlink DL-SCH Downlink Shared Channel E-UTRA NR dual connectivity with MCG using EN-DC E-UTRA and SCG using NR gNB Next Generation (or 5G) Base Station HARQ-ACK Hybrid Automatic Repeat Request Acknowledgment MCG Master Cell Group MCS modulation and coding scheme MAC-CE Medium Access Control - Control Element PBCH Physical Broadcast Channel PCell Primary Cell PDCCH Physical Downlink Control Channel PDSCH Physical Downlink Shared Channel PRACH Physical Random Access Channel PRG Physical Resource Block Group PSCell Primary SCG cell PSS primary synchronization signal PUCCH Physical Uplink Control Channel PUSCH Physical Uplink Shared Channel RACH Random Access Channel RAPID Random Access Preamble Identity RB Resource Block RE Resource Element RRM Radio Resource Management RMSI Residual System Information RS reference signal RSRP reference signal received power RRC Radio Resource Control SCG Secondary Cell Group SCI Sidelink Control Information SFN System Frame Number SL Side Link SCell Secondary Cell SPS Semi-Persistent Scheduling SR Scheduling Request SRI SRS Resource Indicator SRS Sounding Reference Signal SSS Secondary Synchronization Signal SSB sync signal block SUL complementary uplink TA Timing Advance TAG Timing Advance Group TUE Target UE UCI Uplink Control Information UE User Equipment UL Uplink UL-SCH Uplink Shared Channel

Claims

1. 1. A method for supporting artificial intelligence or machine learning (AI / ML) model training in a wireless communications network, the method comprising: receiving, by the first device, AI / ML model training assistance information from the second device; determining, by the first device, data status information (DSI) for each AI / ML model training data based on the AI / ML model training assistance information; receiving, by the first device, information from the second device related to the transmission of the respective AI / ML model training data; transmitting, by the first device, the respective AI / ML model training data to the second device based on at least one of the DSI of the respective AI / ML model training data or the information related to the transmission of the respective AI / ML model training data; A method comprising:

2. The respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, the method comprising: determining, by the first device, whether the respective AI / ML model training data should be transmitted to the second device based on the DSI threshold and the DSI of the respective AI / ML model training data; The method of claim 1 further comprising:

3. The respective AI / ML model training data is selectively transmitted, and the information related to the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data should be transmitted to the second device, and the method further comprises: transmitting, by the first device, at least one of the DSI or AI / ML model training dataset information of the respective AI / ML model training data to the second device; receiving, by the first device, the information from the second device indicating whether the respective AI / ML model training data should be transmitted to the second device; The method of claim 1 further comprising:

4. The AI / ML model training dataset information includes: AI / ML model training dataset size, or DSI distribution information for AI / ML model training datasets 4. The method of claim 3, comprising at least one of:

5. The method of claim 3 or 4, wherein the AI / ML model training dataset information is transmitted using a Buffer Status Report (BSR) or a Scheduling Request (SR).

6. 2. The method of claim 1 , wherein the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or the information related to transmission of the respective AI / ML model training data.

7. 7. The method of claim 6, wherein the respective reporting formats indicate respective transmission accuracy levels, the respective transmission accuracy levels being determined based on the level of the DSI of the respective AI / ML model training data.

8. 8. The method of claim 6 or 7, wherein the respective report formats indicate settings for the transmission of the respective AI / ML model training data.

9. The configuration for the transmission of the respective AI / ML model training data includes: resources used for the transmission of the respective AI / ML model training data; or the quantization granularity used for the transmission of the respective AI / ML model training data; 9. The method of claim 8, wherein the method exhibits at least one of:

10. 10. The method of any one of claims 6 to 9, wherein the respective report format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

11. 11. The method of claim 6, wherein the relationship between the DSI of the respective AI / ML model training data and the respective reporting format is set by the second device.

12. The method of any one of claims 1 to 11, wherein the AI / ML model training support information comprises at least one of information about a reference AI / ML model or at least one reference input data value.

13. The information regarding the reference AI / ML model is: Reference AI / ML model type, Reference AI / ML model structure, one or more reference AI / ML model parameters; Reference AI / ML model gradient, Reference AI / ML model activation function, See AI / ML model input data types, See AI / ML model output data types, Reference AI / ML model input data dimension, or Reference AI / ML model output data dimension 13. The method of claim 12, comprising at least one of:

14. The step of determining the DSI for each AI / ML model training data includes: inputting the respective AI / ML model training data into the reference AI / ML model; determining the DSI based on the output of the reference AI / ML model; 14. The method of claim 12 or 13, comprising:

15. The method of claim 14 , wherein the respective AI / ML model training data input to the reference AI / ML model replaces the at least one reference input data value.

16. The method comprises: receiving, by the first device, updated AI / ML model training support information from the second device; 16. The method of any one of claims 1 to 15, further comprising:

17. The DSI of each of the AI / ML model training data is: data uncertainty of said respective AI / ML model training data; the data importance of each of the AI / ML model training data; the extent of the requirement of said respective AI / ML model training data for said AI / ML model training; or Data diversity of the respective AI / ML model training data 17. The method of claim 1, further comprising information indicative of at least one of:

18. 18. The method of claim 17, wherein the data uncertainty of the respective AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

19. 19. The method of any one of claims 1 to 18, wherein the AI / ML model training is performed at least in part by the second device.

20. 20. The method of claim 1, wherein the respective AI / ML model training data comprises at least one of local AI / ML model training data for a local AI / ML model of the first device, or local gradients associated with the local AI / ML model.

21. The first device and the second device cooperate for the AI / ML model training, and the method includes: performing, by the first device, a portion of the AI / ML model training prior to determining the DSI of each AI / ML model training data, the respective AI / ML model training data including a respective output of the portion of the AI / ML model training.

21. The method of any one of claims 1 to 20, further comprising:

22. 1. A device configured to support artificial intelligence or machine learning (AI / ML) model training in a wireless communications network, the device comprising: a processor; a memory for storing processor-executable instructions; the processor-executable instructions, when executed, cause the processor to: receiving AI / ML model training assistance information from a second device; determining data status information (DSI) for each AI / ML model training data based on the AI / ML model training support information; receiving information related to the transmission of the respective AI / ML model training data from the second device; transmitting the respective AI / ML model training data to the second device based on at least one of the DSI of the respective AI / ML model training data or the information related to the transmission of the respective AI / ML model training data; A device that causes

23. The respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, and the processor-executable instructions, when executed, cause the processor to: determining whether the respective AI / ML model training data should be transmitted to the second device based on the DSI threshold and the DSI of the respective AI / ML model training data; 23. The device of claim 22, further comprising processor-executable instructions.

24. the respective AI / ML model training data is selectively transmitted, and the information related to the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data is to be transmitted to the second device, and the processor-executable instructions, when executed, cause the processor to: transmitting at least one of the DSI or AI / ML model training dataset information of the respective AI / ML model training data to the second device; receiving the information from the second device indicating whether the respective AI / ML model training data should be transmitted to the second device; 23. The device of claim 22, further comprising processor-executable instructions to:

25. The AI / ML model training dataset information includes: AI / ML model training dataset size, or DSI distribution information for AI / ML model training datasets 25. The device of claim 24, comprising at least one of:

26. 26. The device of claim 24 or 25, wherein the AI / ML model training dataset information is transmitted using a buffer status report (BSR) or a scheduling request (SR).

27. 23. The device of claim 22, wherein the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or the information related to transmission of the respective AI / ML model training data.

28. 28. The device of claim 27, wherein the respective report formats indicate respective transmission accuracy levels, the respective transmission accuracy levels being determined based on the level of the DSI of the respective AI / ML model training data.

29. 29. The device of claim 27 or 28, wherein the respective report formats indicate settings for the transmission of the respective AI / ML model training data.

30. The configuration for the transmission of the respective AI / ML model training data includes: resources used for the transmission of the respective AI / ML model training data; or the quantization granularity used for the transmission of the respective AI / ML model training data; 30. The device of claim 29, exhibiting at least one of:

31. 31. The device of any one of claims 27 to 30, wherein the respective report format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

32. 32. The device of any one of claims 27 to 31, wherein the relationship between the DSI of the respective AI / ML model training data and the respective reporting format is set by the second device.

33. 33. The device of any one of claims 22 to 32, wherein the AI / ML model training assistance information comprises at least one of information relating to a reference AI / ML model or at least one reference input data value.

34. The information regarding the reference AI / ML model is: Reference AI / ML model type, Reference AI / ML model structure, one or more reference AI / ML model parameters; Reference AI / ML model gradient, Reference AI / ML model activation function, See AI / ML model input data types, See AI / ML model output data types, Reference AI / ML model input data dimension, or Reference AI / ML model output data dimension 34. The device of claim 33, comprising at least one of:

35. Determining the DSI of each AI / ML model training data includes: inputting the respective AI / ML model training data into the reference AI / ML model; determining the DSI based on the output of the reference AI / ML model; 35. The device of claim 33 or 34, comprising:

36. 36. The device of claim 35, wherein the respective AI / ML model training data input to the reference AI / ML model replaces the at least one reference input data value.

37. The processor-executable instructions, when executed, cause the processor to: receiving updated AI / ML model training support information from the second device; 37. A device according to any one of claims 22 to 36, further comprising processor-executable instructions.

38. The DSI of each of the AI / ML model training data is: data uncertainty of said respective AI / ML model training data; the data importance of each of the AI / ML model training data; the extent of the requirement of said respective AI / ML model training data for said AI / ML model training; or Data diversity of the respective AI / ML model training data 38. A device according to any one of claims 22 to 37, comprising information indicative of at least one of:

39. 39. The device of claim 38, wherein the data uncertainty of the respective AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

40. 40. The device of any one of claims 22 to 39, wherein the AI / ML model training is performed at least in part by the second device.

41. 41. The device of claim 22, wherein the respective AI / ML model training data includes at least one of local AI / ML model training data for a local AI / ML model of the device, or local gradients associated with the local AI / ML model.

42. The device and the second device cooperate for the AI / ML model training, and the processor-executable instructions, when executed, cause the processor to: performing a portion of the AI / ML model training before determining the DSI for each AI / ML model training data, the respective AI / ML model training data including a respective output of the portion of the AI / ML model training; 42. The device of any one of claims 22 to 41, further comprising processor-executable instructions to cause:

43. A device comprising one or more units for carrying out the method according to any one of claims 1 to 21.

44. 1. A method for artificial intelligence or machine learning (AI / ML) model training in a wireless communications network, the method comprising: transmitting, by the first device, AI / ML model training assistance information to a second device for use in determining data status information (DSI) for each AI / ML model training data; transmitting, by the first device, information related to the transmission of the respective AI / ML model training data to the second device; receiving, by the first device, from the second device, the respective AI / ML model training data, the respective AI / ML model training data being transmitted based on at least one of the DSI of the respective AI / ML model training data or the information related to the transmission of the respective AI / ML model training data; performing, by the first device, the AI / ML model training using the respective AI / ML model training data; A method comprising:

45. The respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, the method comprising: setting, by the first device, the DSI threshold for use in determining whether the respective AI / ML model training data should be transmitted to the first device.

45. The method of claim 44, further comprising:

46. The respective AI / ML model training data is selectively transmitted, and the information related to the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data should be transmitted to the first device, and the method further comprises: receiving, by the first device, from the second device, at least one of the DSI of the respective AI / ML model training data or AI / ML model training dataset information; determining, by the first device, whether the respective AI / ML model training data should be transmitted to the first device using at least one of the DSI of the respective AI / ML model training data or the AI / ML model training dataset information; transmitting, by the first device, the information indicating whether the respective AI / ML model training data should be transmitted to the first device to the second device; 45. The method of claim 44, further comprising:

47. The AI / ML model training dataset information includes: AI / ML model training dataset size, or DSI distribution information for AI / ML model training datasets 47. The method of claim 46, comprising at least one of:

48. 48. The method of claim 46 or 47, wherein the AI / ML model training dataset information is transmitted using a Buffer Status Report (BSR) or a Scheduling Request (SR).

49. 45. The method of claim 44, wherein the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or the information related to transmission of the respective AI / ML model training data.

50. 50. The method of claim 49, wherein the respective reporting formats indicate respective transmission accuracy levels, the respective transmission accuracy levels being determined based on the level of the DSI of the respective AI / ML model training data.

51. 51. The method of claim 49 or 50, wherein the respective report formats indicate settings for the transmission of the respective AI / ML model training data.

52. The configuration for the transmission of the respective AI / ML model training data includes: resources used for the transmission of the respective AI / ML model training data; or the quantization granularity used for the transmission of the respective AI / ML model training data; 52. The method of claim 51, wherein the method exhibits at least one of:

53. 53. The method of any one of claims 49 to 52, wherein the respective report format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

54. The method comprises: establishing, by the first device, a relationship between the DSI of the respective AI / ML model training data and the respective reporting format.

54. The method of any one of claims 49 to 53, further comprising:

55. 55. The method of any one of claims 44 to 54, wherein the AI / ML model training assistance information comprises at least one of information relating to a reference AI / ML model or at least one reference input data value.

56. The information regarding the reference AI / ML model is: Reference AI / ML model type, Reference AI / ML model structure, one or more reference AI / ML model parameters; Reference AI / ML model gradient, Reference AI / ML model activation function, See AI / ML model input data types, See AI / ML model output data types, Reference AI / ML model input data dimension, or Reference AI / ML model output data dimension 56. The method of claim 55, comprising at least one of:

57. The method comprises: updating, by the first device, the AI / ML model training support information; transmitting, by the first device, the updated AI / ML model training assistance information to the second device; 57. The method of any one of claims 44 to 56, further comprising:

58. The DSI of each of the AI / ML model training data is: data uncertainty of said respective AI / ML model training data; the data importance of each of the AI / ML model training data; the extent of the requirement of said respective AI / ML model training data for said AI / ML model training; or Data diversity of the respective AI / ML model training data 58. A method according to any one of claims 44 to 57, including information indicative of at least one of:

59. 59. The method of any one of claims 44 to 58, wherein the DSI of the respective AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

60. 60. The method of any one of claims 44 to 59, wherein the respective AI / ML model training data includes at least one of local AI / ML model training data for a local AI / ML model of the second device, or local gradients associated with the local AI / ML model.

61. 61. The method of any one of claims 44 to 60, wherein the first device and the second device cooperate for the AI / ML model training such that the first device performs a portion of the AI / ML model training before determining the DSI of respective AI / ML model training data, and the respective AI / ML model training data includes respective outputs of the portion of the AI / ML model training.

62. 1. A device configured for artificial intelligence or machine learning (AI / ML) model training in a wireless communications network, the device comprising: a processor; a memory for storing processor-executable instructions; the processor-executable instructions, when executed, cause the processor to: transmitting the AI / ML model training assistance information to a second device for use in determining data status information (DSI) for each AI / ML model training data; transmitting information related to the transmission of the respective AI / ML model training data to the second device; receiving the respective AI / ML model training data from the second device, wherein the respective AI / ML model training data is transmitted based on at least one of the DSI of the respective AI / ML model training data or the information related to the transmission of the respective AI / ML model training data; performing said AI / ML model training using said respective AI / ML model training data; A device that causes

63. The respective AI / ML model training data is selectively transmitted, and the information associated with the transmission of the respective AI / ML model training data includes a DSI threshold, and the processor-executable instructions, when executed, cause the processor to: setting the DSI threshold for use in determining whether the respective AI / ML model training data should be transmitted to the first device; 63. The device of claim 62, further comprising processor-executable instructions.

64. the respective AI / ML model training data is selectively transmitted, the information related to the transmission of the respective AI / ML model training data includes information indicating whether the respective AI / ML model training data is to be transmitted to the device, and the processor-executable instructions, when executed, cause the processor to: receiving at least one of the DSI of the respective AI / ML model training data or AI / ML model training dataset information from the second device; determining whether the respective AI / ML model training data should be transmitted to the device using at least one of the DSI of the respective AI / ML model training data or the AI / ML model training dataset information; transmitting the information to the second device indicating whether the respective AI / ML model training data should be transmitted to the device; 63. The device of claim 62, further comprising processor-executable instructions to cause:

65. The AI / ML model training dataset information includes: AI / ML model training dataset size, or DSI distribution information for AI / ML model training datasets 65. The device of claim 64, comprising at least one of:

66. 66. The device of claim 64 or 65, wherein the AI / ML model training dataset information is transmitted using a buffer status report (BSR) or a scheduling request (SR).

67. 63. The device of claim 62, wherein the respective AI / ML model training data is transmitted according to a respective reporting format determined based on at least one of the DSI of the respective AI / ML model training data or the information related to transmission of the respective AI / ML model training data.

68. 68. The device of claim 67, wherein the respective report formats indicate respective transmission accuracy, the respective transmission accuracy levels being determined based on the level of the DSI of the respective AI / ML model training data.

69. 69. The device of claim 67 or 68, wherein the respective report formats indicate settings for the transmission of the respective AI / ML model training data.

70. The configuration for the transmission of the respective AI / ML model training data includes: resources used for the transmission of the respective AI / ML model training data; or the quantization granularity used for the transmission of the respective AI / ML model training data; 70. The device of claim 69, exhibiting at least one of:

71. 71. The device of any one of claims 67 to 70, wherein the respective report format indicates whether the respective AI / ML model training data includes channel state information (CSI) or raw channel information.

72. The processor-executable instructions, when executed, cause the processor to: establishing a relationship between the DSI of each of the AI / ML model training data and each of the reporting formats; 72. A device according to any one of claims 67 to 71, further comprising processor-executable instructions.

73. 73. The device of any one of claims 62 to 72, wherein the AI / ML model training assistance information comprises at least one of information relating to a reference AI / ML model or at least one reference input data value.

74. The information regarding the reference AI / ML model is: Reference AI / ML model type, Reference AI / ML model structure, one or more reference AI / ML model parameters; Reference AI / ML model gradient, Reference AI / ML model activation function, See AI / ML model input data types, See AI / ML model output data types, Reference AI / ML model input data dimension, or Reference AI / ML model output data dimension 74. The device of claim 73, comprising at least one of:

75. The processor-executable instructions, when executed, cause the processor to: updating the AI / ML model training support information; transmitting the updated AI / ML model training assistance information to the second device; 75. The device of any one of claims 62 to 74, further comprising processor-executable instructions to cause:

76. The DSI of each of the AI / ML model training data is: data uncertainty of said respective AI / ML model training data; the data importance of each of the AI / ML model training data; the extent of the requirement of said respective AI / ML model training data for said AI / ML model training; or Data diversity of the respective AI / ML model training data 76. A device according to any one of claims 62 to 75, comprising information indicative of at least one of:

77. 77. The device of any one of claims 62 to 76, wherein the DSI of the respective AI / ML model training data is determined based on at least one of entropy, minimum confidence, margin sampling, or generalization error.

78. 78. The device of any one of claims 62 to 77, wherein the respective AI / ML model training data includes at least one of local AI / ML model training data for a local AI / ML model of the second device, or local gradients associated with the local AI / ML model.

79. 79. The device of any one of claims 62 to 78, wherein the device and the second device cooperate for the AI / ML model training such that the device performs a portion of the AI / ML model training before determining the DSI of each AI / ML model training data, and the each AI / ML model training data includes a respective output of the portion of the AI / ML model training.

80. 62. A device comprising one or more units for carrying out the method of any one of claims 44 to 61.

Images