Communication method and apparatus, and storage medium and program product

By establishing DPD and DPoD models using OFDM modulated training signals in 6G communication systems and utilizing artificial intelligence technology to compensate for the nonlinear distortion of radio frequency devices, the problem of insufficient coverage of 6G technology was solved, achieving wireless coverage capabilities comparable to 5G.

WO2026232971A1PCT designated stage Publication Date: 2026-11-12ZTE CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/CN2026/083347
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-09
Filing Date
2026-03-13
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

The lack of standardization in existing distortion model determination methods makes it difficult to achieve enhanced coverage capabilities for 6G technology, especially when using mid-to-high frequency bands, where the coverage capability is more than 10dB lower than that of 5G.

Method used

The distortion model is determined by using OFDM modulated training signals in the communication system. This includes sending and receiving training signals at the transmitter and receiver to establish DPD and DPoD models, and using artificial intelligence technology to train the models to compensate for the nonlinear distortion of radio frequency devices.

Benefits of technology

The linear operating point of the radio frequency devices was improved, the coverage capability of 6G technology was enhanced, the problem of insufficient coverage capability was solved, and wireless coverage capability comparable to 5G was achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2026083347_12112026_PF_FP_ABST
    Figure CN2026083347_12112026_PF_FP_ABST
Patent Text Reader

Abstract

Provided are a communication method and apparatus, and a storage medium and a program product, which relate to the technical field of communications. The method is applied to a first node. The method comprises: receiving a first training signal sent by a second node, the first training signal being used by a first node to determine a distortion model, and the first training signal being an OFDM-modulated training signal; and determining the distortion model on the basis of the first training signal.
Need to check novelty before this filing date? Find Prior Art

Description

Communication method, apparatus, storage medium and program product

[0001] The present disclosure claims priority to Chinese Patent Application No. 202510599670.8, filed on May 9, 2025, the entire contents of which are incorporated herein by reference. TECHNICAL FIELD

[0002] The present disclosure relates to the technical field of communication, and in particular to a communication method, apparatus, storage medium and program product. BACKGROUND

[0003] Future mobile communication systems, such as the 6th generation mobile networks (6G) communication system, are gradually becoming the focus of deployment. The vision, demand, technical indicators, research direction, and the like of the future mobile communication system are becoming increasingly clear.

[0004] The 5th generation mobile networks (5G) has determined to use FR1 and FR2 as the main radio frequency frequency band. For the frequency band that can be used by 6G, in addition to the frequency band used by 5G, some other new frequency bands are being discussed, and here the 7 GHz frequency band, such as 6425-7125 MHz, has been determined as a frequency band that can be used by 6G technology in most regions. It is required that 6G and 5G network devices are co-located, and the wireless coverage capability of 6G is comparable to that of 5G. Since the medium-high frequency band used by 6G technology and the medium-low frequency band used by 5G technology differ greatly, the maximum path loss (MPL) supported by 6G technology may exceed the MPL supported by 5G technology by more than 10 dB, so the coverage capability of 6G technology needs to be enhanced. Through the distortion compensation method of the radio frequency device, the linear operating point of the radio frequency device can be improved, which is an important technical exploration direction for improving coverage capability.

[0005] The distortion compensation method includes applying a distortion model for distortion compensation, however, the current distortion model determination method does not have a unified standard, is non-standardized, and is not easy to realize the standardization of the distortion model. SUMMARY

[0006] In a first aspect, a communication method is provided, applied to a first node, comprising:

[0007] receive a first training signal sent by the second node, the first training signal being used for the first node to determine the distortion model, the first training signal being a training signal after orthogonal frequency-division multiplexing (OFDM) modulation;

[0008] determine the distortion model based on the first training signal.

[0009] In a second aspect, a communication method is provided, and applied to a second node, including:

[0010] send a first training signal to a first node, the first training signal being used for the first node to determine a distortion model, the first training signal being a training signal after OFDM modulation.

[0011] In a third aspect, a communication method is provided, and applied to a first node, including:

[0012] send a third training signal to a second node, the third training signal being used for the second node to determine a distortion model, the third training signal being a training signal after OFDM modulation.

[0013] In a fourth aspect, a communication method is provided, and applied to a second node, including:

[0014] receive a third training signal sent by a first node, the third training signal being used for the second node to determine a distortion model, the third training signal being a training signal after OFDM modulation.

[0015] determine the distortion model based on the third training signal.

[0016] In a fifth aspect, a communication apparatus is provided, and applied to a first node, including:

[0017] a receiving unit, configured to receive a first training signal sent by a second node, the first training signal being used for the first node to determine a distortion model, the first training signal being a training signal after OFDM modulation;

[0018] a processing unit, configured to determine the distortion model based on the first training signal.

[0019] In a sixth aspect, a communication apparatus is provided, and applied to a second node, including:

[0020] a sending unit, configured to send a first training signal to a first node, the first training signal being used for the first node to determine a distortion model, the first training signal being a training signal after OFDM modulation.

[0021] In a seventh aspect, a communication apparatus is provided, and applied to a first node, including:

[0022] The transmitting unit is used to send a third training signal to the second node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation.

[0023] Eighthly, a communication device is provided for use in a second node, comprising:

[0024] The receiving unit is used to receive the third training signal sent by the first node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation.

[0025] The processing unit is used to determine the distortion model based on the third training signal.

[0026] A ninth aspect provides a communication device, comprising: a processor and a memory; the memory storing processor-executable instructions; when the processor is configured to execute the instructions, causing the communication device to implement any of the methods provided in any of the first to fourth aspects described above.

[0027] A tenth aspect provides a computer-readable storage medium, comprising a non-transitory computer-readable storage medium storing computer instructions that, when executed on a computer, cause the computer to perform any of the methods provided in any of the first to fourth aspects.

[0028] Eleventhly, a computer program product containing computer instructions is provided, which, when executed on a computer, causes the computer to perform any of the methods provided in any of the first to fourth aspects. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in this disclosure, the accompanying drawings used in some embodiments of this disclosure will be briefly described below. Obviously, the drawings described below are merely drawings of some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings.

[0030] Figure 1 is a location diagram of an AI-based DPD and DPoD implementation according to an embodiment of the present disclosure.

[0031] Figure 2 is a model training diagram of a DPD / DPoD model provided according to an embodiment of the present disclosure.

[0032] Figure 3 is a model inference diagram of a DPD model and a DPoD model provided according to an embodiment of the present disclosure.

[0033] Figure 4 is a structural diagram of a communication system provided according to an embodiment of the present disclosure.

[0034] Figure 5 is a flowchart of a communication method provided according to an embodiment of the present disclosure.

[0035] Figure 6 is a diagram showing the segmentation of a training sequence mapped to each OFDM symbol according to an embodiment of the present disclosure.

[0036] Figure 7 is a diagram showing the segmentation of a training sequence mapped to multiple OFDM symbols according to an embodiment of the present disclosure.

[0037] Figure 8 is a long sequence mapping diagram formed by repeated mapping of training sequences according to an embodiment of the present disclosure.

[0038] Figure 9 is a frequency hopping diagram based on subband training signals provided according to an embodiment of the present disclosure.

[0039] Figure 10 is a sideband-priority frequency hopping diagram provided according to an embodiment of the present disclosure.

[0040] Figure 11 is a diagram showing the binding relationship between a dataset and an associated ID according to an embodiment of this disclosure.

[0041] Figure 12 is a diagram of a model monitoring method for a DPD model and a DPoD model provided according to an embodiment of the present disclosure.

[0042] Figure 13 is a flowchart of another communication method provided according to an embodiment of the present disclosure.

[0043] Figure 14 is a flowchart of another communication method provided according to an embodiment of the present disclosure.

[0044] Figure 15 is a flowchart of another communication method provided according to an embodiment of the present disclosure.

[0045] Figure 16 is a block diagram of a communication device provided according to an embodiment of the present disclosure.

[0046] Figure 17 is a block diagram of another communication device provided according to an embodiment of the present disclosure.

[0047] Figure 18 is a block diagram of another communication device provided according to an embodiment of the present disclosure.

[0048] Figure 19 is a block diagram of another communication device provided according to an embodiment of the present disclosure.

[0049] Figure 20 is a block diagram of another communication device provided according to an embodiment of the present disclosure. Detailed Implementation

[0050] To enable those skilled in the art to better understand the technical solutions of the embodiments of this disclosure, the technical solutions of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0051] Unless the context otherwise requires, throughout the specification and claims, the term "comprise" and other forms such as the third-person singular "comprises" and the present participle "comprising" are interpreted as open-ended and encompassing, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiments," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples.

[0052] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this disclosure, unless otherwise stated, "a plurality of" means two or more.

[0053] In this disclosure, the terms "exemplarily" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplarily" or "for example" in this disclosure should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the terms "exemplarily" or "for example" is intended to present the relevant concepts in a specific manner.

[0054] In addition, the use of “based on” implies openness and inclusivity, because processes, steps, calculations or other actions “based on” one or more of the stated conditions or values ​​may in practice be based on additional conditions or values ​​beyond those stated.

[0055] The integration of mobile communication systems with artificial intelligence (AI) and machine learning (ML) is a future trend. For example, the combination of AI / ML technology and mobile communication systems has been attempted in 5G and 5G-Advanced (5G-A). In the future design of 6G mobile communication systems, AI / ML will be a key technology inherent in mobile communication systems.

[0056] Artificial intelligence (AI) encompasses self-learning devices, components, software, and modules such as machine learning, deep learning, reinforcement learning, transfer learning, deep reinforcement learning, and meta-learning. In some cases, AI is implemented through AI networks (or neural networks). A neural network consists of multiple layers, each containing at least one node. Typically, a neural network includes an input layer, an output layer, and at least one hidden layer. Each layer of the neural network uses, but is not limited to, at least one of the following: fully connected layers, dense layers, convolutional layers, transposed convolutional layers, directly connected layers, activation functions, normalization layers, and pooling layers. In other cases, each layer of the neural network may include a sub-neural network, such as a residual network block (or ResNet block), a dense network (DenseNet block), or a recurrent neural network (RNN). AI networks include the neural network model and / or the corresponding neural network parameters. Here, the neural network model can be simply referred to as the network model, and the neural network parameters can be simply referred to as network parameters. A network model defines the architecture of a neural network, including the number of layers, the size of each layer, the activation function, the connections, the convolutional kernels and strides, and the convolution type (e.g., 1D convolution, 2D convolution, 3D convolution, hollow convolution, transposed convolution, separable convolution, grouped convolution, expanded convolution, etc.). Network parameters are the weights and / or biases of each layer in the network model and their values. A network model can correspond to multiple sets of different neural network parameter values ​​to adapt to different scenarios. The values ​​of network parameters can be obtained through offline training and / or online training, such as training the neural network model by inputting at least one sample and label. A neural network model can correspond to multiple different neural network parameter values.

[0057] AI / ML (Artificial Intelligence / Machine Learning) is a promising direction for enhancing mobile communication systems. Introducing AI / ML technologies into mobile communication systems, such as 5G, 5G-A, and 6G, can improve system efficiency. For example, AI / ML inference and prediction can reduce reference signal overhead, reduce channel state information feedback overhead, or improve terminal positioning accuracy. Its benefits have been preliminarily verified or are currently being verified.

[0058] For communication systems employing AI / ML technology, in all parts of this disclosure, an AI / ML model is narrowly defined as: a mathematical algorithm that applies machine learning techniques to generate a set of outputs based on a set of inputs. It may include metadata, which includes relevant information about the training model and the applicable operating environment. The AI / ML model can be managed, stored, and transmitted as an artifact, which may be a container, image, or proprietary file format.

[0059] In a broad sense, an AI / ML "model" is a general term used to describe a device's ability to perform processing methods, functions, features, or groups of features. A "model" can be a function, functional module, processing method, information processing method, implementation method, functional group, configuration, configuration set, dataset (e.g., for model training), or data-driven algorithm, etc.

[0060] In all parts of this disclosure, the terms algorithm, function, algorithm function, or function algorithm may be used to describe the model, algorithm, function, algorithm function, or function algorithm, unless otherwise specified, they have the same meaning and connotation.

[0061] In all parts of this disclosure, when AI and ML are described individually, AI and ML are used interchangeably. AI / ML is equivalent to AI and ML used alone.

[0062] Digital pre-distortion (DPD) and digital post-distortion (DPoD) are signal processing techniques used in wireless communication systems to correct distortions introduced by nonlinear devices (typically power amplifiers, PAs) or nonlinear channels. These techniques are crucial for ensuring high signal quality and spectral efficiency, especially in modern communication systems such as 5G New Radio (NR) and 6G.

[0063] DPD technology can improve the efficiency of power amplifiers and save power. Using DPD on the base station side can effectively save energy and achieve low-carbon goals.

[0064] Uplink coverage is a bottleneck for future 6G deployment. When terminals also adopt DPD technology, the maximum power reduction (MPR) can be reduced, the efficiency of the power amplifier can be improved and power can be saved. More importantly, power space can be freed up to increase the uplink coverage distance.

[0065] DPD is a distortion compensation technology applied at the transmitting end, while DPoD is a distortion compensation technology applied at the receiving end. Both can be used at the base station or terminal side.

[0066] DPD (Distributed Power Processing) is a preprocessing technique applied to the input signal before it is fed into the power amplifier. It counteracts the nonlinearity and distortion introduced by the power amplifier by intentionally distorting the input signal in the opposite direction of the desired distortion. DPD is used in transmitters. Its algorithm works as follows:

[0067] Establish a model of the nonlinear behavior of PA. This model can be obtained through mathematical derivation, simulation, experimental data summarization, or model training.

[0068] This model is used to predistort the input signal so that after the nonlinear effect of PA, the output signal becomes linear and distortion-free.

[0069] The process is adaptive, meaning that the DPD system needs to continuously update the model based on the feedback from the PA output.

[0070] DPoD is a post-processing technique applied to received signals after they have been disturbed by nonlinear channel interference or nonlinear distortion of the transceiver's radio frequency devices. Used in the receiver, DPoD corrects various distortions introduced during transmission or reception. The working principle of the DPoD algorithm is as follows:

[0071] DPoD requires knowledge of distortion characteristics (e.g., PA or channel model). Distortion characteristics can be obtained through mathematical derivation, simulation, experimental data summarization, or model training.

[0072] After signal transmission and distortion, DPoD applies a correction algorithm to recover the original signal as closely as possible.

[0073] It is typically used in receivers to compensate for distortion introduced in the uplink or downlink.

[0074] Because the distortion characteristics change in real time due to the nonlinear channel or the radio frequency devices of the transceiver, continuous retraining is required to obtain the real-time distortion characteristics. When the model generalizes sufficiently, the frequency of retraining can be greatly reduced.

[0075] The model can be a simple or complex mathematical model, or an artificial intelligence model. AI and ML-based DPD and DPoD can more accurately compensate for PA nonlinearity and distortion or loss caused by wireless channels.

[0076] Figure 1 is a location diagram of an AI-based DPD and DPoD implementation according to an embodiment of the present disclosure. Figure 1 shows the implementation location of the DPD and DPoD devices. The model in the example is based on artificial intelligence.

[0077] For DPD (Distributed Power Distribution), if we consider using AI-based model training to obtain a model of PA (Power Amplifier) ​​nonlinear behavior, or for DPoD (Distributed Power Distribution), if we consider using AI-based model training to obtain distortion characteristics (e.g., PA or channel model), a specific training signal is first required. In a communication system, the training signal can be a specific training sequence. This sequence enters the PA or wireless channel, where it is introduced with radio frequency distortion by the PA or disturbed by the wireless channel, resulting in a distorted output signal. By comparing this signal with the original training sequence, a mathematical expression of the distortion or disturbance can be obtained. This mathematical expression of distortion can be used to obtain a mathematical model for inverse compensation through mathematical analysis and induction, such as a mathematical model expressed using Volterra series, which is suitable for describing nonlinear systems with memory effects.

[0078] Alternatively, an AI model can be trained to form a DPD or DPoD model with inverse compensation based on the AI ​​model (expressed in terms of model structure, macro parameters, etc.). The input to the AI-based DPD or DPoD model is a distorted signal formed by passing a large number of training sequences through a PA or wireless channel, and the original training sequence is a label used for comparison. After long-term training and convergence, a stable AI-based DPD or DPoD model can be obtained for inverse compensation of distortion. Figure 2 is a model training diagram of a DPD / DPoD model provided according to an embodiment of this disclosure. Figure 2 shows the input for training the AI-based DPD or DPoD model, which can be trained based on the distorted training signal and the original training signal as input.

[0079] Figure 2 illustrates how to train an AI-based DPD or DPoD model. Whether it's a DPD or DPoD model, the device running the model uses the known original training signal as the label as input, and the distorted training signal as the other input. After training convergence, an AI-based DPD or DPoD model can be obtained, which can be described using model structure, model macro parameters, etc.

[0080] Figure 3 is a model inference diagram of a DPD model and a DPoD model provided according to an embodiment of the present disclosure. Figure 3 shows the input and output of AI-based DPD model or DPoD model inference.

[0081] The model inference of DPD and DPoD in Figure 3 differs. DPD model inference primarily performs pre-compensation / pre-distortion on the input signal before it is fed into the power amplifier (PA), so its input is the original communication signal, and its output is the pre-distorted communication signal. DPoD model inference, on the other hand, primarily performs post-processing of the distorted signal to compensate for the distortion; therefore, its input is the distorted communication signal, and its output is either the compensated communication signal or the distorted communication signal.

[0082] The above has introduced DPD and DPoD. The following will describe a communication method provided by an embodiment of this disclosure in conjunction with the accompanying drawings.

[0083] The technical solutions provided in this disclosure can be applied to various mobile communication systems, such as 5G NR mobile communication systems, future mobile communication systems (e.g., 6G communication systems), or multiple communication convergence systems, etc. This disclosure does not limit them.

[0084] In this disclosure, the mobile communication system (including but not limited to 3G, 4G, 5G, and future mobile communication systems such as 6G) may include network-side devices (e.g., including but not limited to base stations) and receiving-side devices (e.g., including but not limited to terminals). It should be understood that, in this example, for instance, in the downlink, the first communication node (also referred to as the first communication node device, the first node) may be a base station-side device, and the second communication node (also referred to as the second communication node device, the second node) may be a terminal-side device. In some examples, such as in the uplink, the first communication node may also be a terminal-side device, and the second communication node may also be a base station-side device. In some examples, such as in device-to-device communication between the two communication nodes, both the first and second communication nodes may be base stations or terminals. Therefore, whether the first and second nodes are base stations or terminals needs to be determined based on the context.

[0085] Figure 4 is a structural diagram of a communication system provided according to an embodiment of the present disclosure. As shown in Figure 4, the communication system includes, but is not limited to, a first node 110 and a second node 120. Here, the first node 110 and the second node 120 can transmit and receive wireless signals and perform related interactions.

[0086] In a wireless communication scenario, the first node 110 and the second node 120 communicate via a wireless channel. For example, the first node 110 may be a terminal, and the second node 120 a base station; the terminal and the base station communicate via a wireless channel. Alternatively, the first node 110 may be a terminal, and the second node 120 a wireless router; the wireless router and the terminal communicate via a wireless channel. Another example is that the first node 110 may be a first base station, and the second node 120 a second base station; the first base station and the second base station communicate via a wireless channel. Yet another example is that the first node 110 may be a first terminal, and the second node 120 a second terminal; the first terminal and the second terminal communicate via a wireless channel. Finally, the first node 110 may be a repeater, and the second node 120 a base station; the base station and the repeater communicate via a wireless channel. Finally, the first node 110 may be a terminal, and the second node 120 a repeater; the repeater and the terminal communicate via a wireless channel. For example, node 110 is a first repeater, and node 120 is a second repeater; the first repeater and the second repeater communicate via a wireless channel. Alternatively, node 110 can be a base station, and node 120 a satellite; the satellite and the base station communicate via a wireless channel. Another example: node 110 can be a satellite, and node 120 a base station; the base station and the satellite communicate via a wireless channel. Yet another example: node 110 can be a terminal, and node 120 a satellite; the satellite and the terminal communicate via a wireless channel. Again, node 110 can be a satellite, and node 120 a terminal; the terminal and the satellite communicate via a wireless channel. Finally, node 110 can be ground equipment, and node 120 can be an aircraft; the aircraft and the ground equipment communicate via a wireless channel. Finally, node 110 can be a first aircraft, and node 120 a second aircraft; the first aircraft and the second aircraft communicate via a wireless channel.

[0087] Unless otherwise specified, the terms "first" node, "second" node, "first" method, "second" method, "first" approach, "second" approach, "first" matrix, "second" matrix, "first" part, and "second" part in this disclosure are used only for descriptive distinction and do not represent a sequential or chronological order.

[0088] In this disclosure, the base station can be a base station in Long Term Evolution (LTE), Long Term Evolution Advanced (LTEA), or an evolved Node B (eNB or eNodeB), a base station device in a 5G network, or a base station in a future mobile communication system (such as 6G). The base station can include various macro base stations, micro base stations, femtocell base stations, wireless remote extensions, reconfigurable intelligent surfaces (RISS), routers, wireless fidelity (WIFI) devices, or various network-side devices such as primary cells and secondary cells.

[0089] In this disclosure, a terminal is a device with wireless transceiver capabilities that can be deployed on land, including indoors or outdoors; on water (such as on ships); and in the air (such as on airplanes, balloons, and satellites). The terminal can be a mobile phone, tablet computer, computer with wireless transceiver capabilities, virtual reality (VR) terminal, augmented reality (AR) terminal, wireless terminal in industrial control, wireless terminal in self-driving, wireless terminal in remote medical care, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, etc. The embodiments of this disclosure do not limit the application scenarios. A terminal may also be referred to as a user, user equipment, access terminal, UE unit, UE station, mobile station, mobile station, remote station, remote terminal, mobile device, UE terminal, wireless communication device, UE agent, or UE device, etc. The embodiments of this disclosure are not limited to these terms.

[0090] Figure 4 is an exemplary structural diagram. The number of devices included in the communication system shown in Figure 4 is not limited; for example, the number of first nodes and second nodes is not limited. Furthermore, in addition to the devices shown in Figure 4, the communication system shown in Figure 4 may include other devices, which are not limited.

[0091] Next, as shown in Figure 5, this embodiment of the present disclosure provides a communication method applied to a first node, which may be the first node 110 shown in Figure 4 above. The method may include the following steps:

[0092] S101, Receive the first training signal sent by the second node.

[0093] Here, the first training signal is used to determine the distortion model, and the first training signal is the training signal after OFDM modulation.

[0094] The second node can be the second node 120 shown in Figure 4 above. For ease of description, the following embodiments will be described using the first node as the terminal and the second node as the base station.

[0095] In some embodiments, the distortion model is used for distortion processing of communication signals, or for compensating distortion processing of communication signals.

[0096] In some embodiments, the distortion model is a DPoD model. That is, the second node sends a first training signal to the first node, and the first training signal is used by the first node to determine the DPoD model.

[0097] The generation method of the first training signal is described below.

[0098] Future 6G will still use OFDM for signal transmission. OFDM is not only a modulation technique but also possesses the characteristics of multiplexing. OFDM divides a broadband signal into multiple narrowband subcarriers, each of which can be modulated independently, effectively combating frequency-selective fading and inter-symbol interference. The subcarriers are orthogonal, meaning they can overlap in the spectrum without interfering with each other, thus improving spectral efficiency. Cyclic prefix (CP)-OFDM and discrete fourier transform-spread (DFT-S)-OFDM, derived from OFDM, retain the characteristics of OFDM and are more suitable for equipment implementation.

[0099] The first training signal used for distortion compensation can be generated based on a training sequence. The training sequence can be modulated based on OFDM and mapped to a certain bandwidth (see the frequency domain usage below). The mapping of the training sequence is not limited to a single OFDM symbol and can be extended to multiple symbols in the time domain (see the time domain usage below). That is, in some embodiments, the first training signal is determined based on the training sequence, which is modulated based on OFDM, mapped to a configured bandwidth in the frequency domain, and mapped to one or more OFDM symbols in the time domain before transmission.

[0100] The training sequence is generated based on pseudo-random (PN) sequences. Pseudo-random sequences possess the characteristics of Gaussian white noise or uniform distribution, which can simulate random interference in actual communication. For the DPD model, the symbols formed by OFDM modulation of the pseudo-random sequence are fed into the power amplifier (PA), generating a distorted signal. For the DPoD model, the symbols formed by OFDM modulation of the pseudo-random sequence are processed at the transmitter, transmitted through the air channel, and received at the receiver. The receiver receives a distorted signal caused by both the transmission process and the air channel.

[0101] The pseudo-random sequence to be generated has a finite length of M bits. Several generation methods are as follows:

[0102] The pseudo-random sequence is mainly implemented based on a linear feedback shift register (LFSR), generating the Gold sequence by combining two m-sequences. The m-sequences, generated by the LFSR, possess the maximum length characteristic and excellent periodicity. The Gold sequence is generated by combining two different m-sequences through an XOR operation, enhancing randomness and robustness against interference. One possible generation method is shown in the formula:

[0103] c(n) is the final output pseudo-random sequence, n = 0, 1, ..., M-1, classified as a Gold sequence, whose sequence content has only two possibilities: {0, 1}. i (n) is the m-sequence input for generating pseudo-randomness, X i (n) can be initialized independently as i changes.

[0104] Since actual transmitted signals need to undergo different modulation methods, such as π / 2-binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 16QAM, and higher-order QAM modulation, for the generalization considerations of AI models, pseudo-random sequences also need to traverse all possible modulation methods, called modulation symbols. For example, when c(n) undergoes QPSK modulation, the pseudo-random sequence transforms into:

[0105] The use of other non-QPSK modulation methods is similar and will not be elaborated here.

[0106] Based on this, in some embodiments, the training sequence includes any one of the following: an unmodulated pseudo-random sequence, a modulated pseudo-random sequence, a sequence of all zeros, or a sequence of all ones, where modulation includes at least one of the following:

[0107] π / 2-BPSK, BPSK, QPSK, and N-order QAM, where N is an integer greater than or equal to 16. In other words, the first training signal can be a modulated training signal, and the modulation method includes at least one of the above.

[0108] Besides generating pseudo-random sequences to produce training signals, training signals can also be two-tone or multi-tone signals. Two-tone or multi-tone signals are superimposed signals of two or more sinusoidal waves of different frequencies, which can be used for intermodulation distortion analysis. By testing the DPD's ability to suppress intermodulation distortion (IMD), it is beneficial to accurately quantify harmonic suppression capabilities. Since OFDM modulation is inherently a superposition of multiple sinusoidal waves of different frequencies, the generation of multi-tone signals can still be based on OFDM modulation, but there is no modulated source; the modulation information is empty, meaning the OFDM subcarriers are in an information-unloaded state. Alternatively, multi-tone signals can be generated by OFDM modulation using all-zero or all-one information.

[0109] Based on this, in some embodiments, the first training signal is a two-tone or multi-tone signal based on OFDM modulation, and the modulation information on the OFDM subcarrier is empty.

[0110] In some embodiments, the training signal can be defined as a reference signal for training the model. That is, the first training signal is a reference signal used to train the model.

[0111] The following explains how to use the training signal in the time domain.

[0112] When the training sequence is M bits, and a single OFDM symbol cannot fully carry the M-bit training sequence within the bandwidth allocated to the training signal, the long training sequence needs to be segmented and carried across multiple OFDM symbols. To avoid distortion caused by boundary effects resulting from truncating long sequences, segmentation requires the introduction of overlapping regions. The length of the overlapping region within each segment must be no less than the memory depth MM of the device (e.g., power amplifier PA). Common power amplifier modeling requires consideration of significant memory effects. Possible segmentation methods include the following:

[0113] I: Each segment is mapped to each OFDM symbol, suitable for scenarios where the memory depth is within one symbol time.

[0114] As shown in Figure 6, a training sequence of length M bits is divided into N segments, each segment is mapped to an OFDM symbol, and the first P elements in the sequence mapped by the next OFDM symbol are the same as the last P elements in the sequence mapped by the previous OFDM symbol, where P is greater than or equal to the memory depth MM.

[0115] The specific training sequence, with a length of M = 50 bits, is divided into 3 segments, each mapped to an OFDM symbol. The first P = 6 elements of the mapped sequence of the next OFDM symbol are the same as the last 6 elements of the mapped sequence of the previous OFDM symbol. The RE positions in the extra unpadded mapped training sequence of the last symbol are either NULL or filled with meaningless padding bits.

[0116] II: Each segment is mapped to multiple OFDM symbols, which is suitable for scenarios where memory depth requires multiple symbol times.

[0117] As shown in Figure 7, a training sequence of length M bits is divided into N segments, each segment is mapped to a set of multiple OFDM symbols. The first P elements in the sequence mapped by each set of OFDM symbols are the same as the last P elements in the sequence mapped by the previous set of OFDM symbols. P is greater than or equal to the memory depth MM. The mapping of P elements requires one or more OFDM symbols.

[0118] The specific training sequence of length M = 110 bits is divided into two segments, each mapped to a set of three OFDM symbols. The first P = 26 elements of the mapped sequence of the next OFDM symbol set are the same as the last 26 elements of the mapped sequence of the previous OFDM symbol set. In the last set, the RE positions of one or more symbols with extra unpadded mapped training sequences are either NULL or filled with meaningless padding bits.

[0119] Two typical implementation methods can be summarized as follows: this segmented and mapped training sequence processing method is called segmented overlapping sampling, where each segment is connected end-to-end and has overlapping regions. The time length of the OFDM symbols in the overlapping region after segmented mapping should not be shorter than the memory depth. "Connected end-to-end" refers to the concatenation of sequence bits, which is not limited by the time domain.

[0120] Besides segmenting and mapping long sequences while preserving overlapping regions, a method of repeatedly mapping sequences can be used to form even longer sequences. This method also preserves overlapping regions, and the overlapping regions are the entire sequence. The process of repeatedly mapping training sequences to form long sequence mappings can be illustrated in Figure 8.

[0121] Specifically, a training sequence of length M = 24 bits is mapped to a set of two OFDM symbols, and this mapping is repeated multiple times. In Figure 8, after repetition, a total of three 72-bit training sequences are mapped to three sets of six symbols. This naturally forms an overlapping region, meaning the overlapping region contains the complete 24-bit sequence, and the time length of the two OFDM symbols in the overlapping region is not shorter than the memory depth. Each set of OFDM symbols can have its resource element (RE) positions in the extra unfilled mapped training sequences set to NULL or filled with meaningless padding bits.

[0122] The above embodiments summarize the time-domain usage of training signals:

[0123] Method 1: Long training sequences are segmented and mapped onto multiple OFDM symbols. Each segment is contiguous and its elements overlap. Each segment may be mapped to a single OFDM symbol or multiple OFDM symbols. The time length of the OFDM symbols containing the overlapping elements after segmentation is no less than the memory depth. The RE position within each or every group of OFDM symbols that fails to map the training sequence is either NULL or padded with padding bits.

[0124] In other words, in some embodiments, the first training signal is determined based on the training sequence, and the training sequence is segmented to obtain multiple segmented training sequences, which are respectively mapped to one or more OFDM symbols for transmission.

[0125] In some embodiments, the training sequences of each segment are connected end-to-end, and the training sequences of adjacent segments have overlapping elements in the connected region. For example, the first sequence is 12345, and the second sequence is 45678, not 23678 (this is not an overlap when the sequences are connected end-to-end).

[0126] In some embodiments, the time length of the overlapping element or the time length of the OFDM symbol containing the overlapping element is greater than or equal to the memory depth.

[0127] In some embodiments, the resource element positions within OFDM symbols that are not mapped to training sequences in the time domain are empty, or are padding bits.

[0128] Method 2: After the sequence is mapped to one or more OFDM symbols, the mapping is repeated one or more times. Each or each group of OFDM symbols may have its RE position set to NULL or filled with padding bits in the unpadded mapped training sequence.

[0129] In other words, in some embodiments, the first training signal is determined based on a training sequence, which is repeatedly mapped to multiple OFDM symbol groups, each OFDM symbol group including at least one OFDM symbol, and the resource element RE position in the OFDM symbol in each OFDM symbol group that is not mapped to the training sequence is empty or is a padding bit.

[0130] The following explains how to use the frequency domain of the training signal.

[0131] Training signals generated based on OFDM modulation training sequences, being subcarrier-based signals, facilitate convenient allocation of frequency domain resources. Since DPD or DPoD model training demands high generalization, the training signal needs to traverse various possible bandwidths, from the lowest bandwidth unit (RE) or physical resource block (PRB) to the highest bandwidth unit (system bandwidth). Besides the need for flexible bandwidth allocation, the training signal also needs to be flexibly transmitted on certain subbands, and even frequency hopping may be required. Frequency hopping can effectively save bandwidth overhead while achieving generalization. Figure 9 illustrates a frequency hopping technique based on subband training signals.

[0132] In Figure 9, the bandwidth of the generated training signal covers one subband. To traverse the entire bandwidth as much as possible, a maximum of four frequency hopping operations are used, with a frequency hopping time granularity of one OFDM symbol.

[0133] Training signal frequency hopping can use FH start and FH offset The two parameters are described in a more general way. FH start It is the starting point of frequency hopping, generally FH start =PRB0 or ​​RE0. FH offset It refers to the frequency hopping interval, under normal circumstances. Here, Bandwidth is the system bandwidth, frequency band bandwidth, or maximum usable bandwidth, and N is the maximum number of frequency hoppings in a single cycle. In Figure 9, the maximum number of frequency hoppings is 4, i.e., N = 4. Taking Figure 9 as an example, with each frequency hopping, the frequency offset of the training signal increases by one FH. offset .

[0134] In most cases, Bandwidth cannot be used by FH. offset Divisibility means that training signals may not be able to be transmitted at the beginning or end of the full bandwidth. The sidebands of the RF system are special relative to the center of the frequency band and need to be considered separately. In order to traverse the beginning or end of the full bandwidth, a frequency hopping method that prioritizes sideband protection can be adopted, as shown in Figure 10, which is a frequency hopping diagram with sideband priority.

[0135] Still using FHstart and FH offset Two parameters. FH start It is the starting point of frequency hopping, FH start =PRB0 or ​​RE0. FH offset It is the frequency hopping interval. Here, Bandwidth is the system bandwidth, frequency band bandwidth, or maximum usable bandwidth, and N is the maximum number of frequency hoppings within a single cycle. In Figure 10, the maximum number of frequency hoppings is 4, i.e., N = 4. Taking Figure 10 as an example, frequency hopping starts from the lowest frequency subband 0, then jumps to the highest frequency subband 2, then to the second lowest frequency subband 3, and then to the second highest frequency subband 4, completing one cycle. The frequency hopping pattern follows the same logic when the number of frequency hoppings increases to a larger even number. The starting frequency of each frequency hopping subband can be expressed as a formula:

[0136] Subband0: 0;

[0137] Subband1: Bandwidth-FH offset ;

[0138] Subband2: FH offset ;

[0139] Subband3: Bandwidth - 2 * FH offset

[0140] This pattern of traversing the sidebands first and then successively traversing the intermediate subbands is not unique. It's also possible to consider traversing the lower half of the bandwidth first, and then the upper half. Subband0 can also start frequency hopping from the beginning of any subband, not limited to PRB=0 or RE=0.

[0141] The above examples summarize the frequency domain usage of training signals:

[0142] The frequency domain resource usage of the training signal can be configured from the lowest bandwidth unit (RE or PRB) to the highest system bandwidth.

[0143] Configure the training signal frequency hopping mode, with the starting point of the frequency hopping being FH. start The frequency hopping interval is FH offset , Here, Bandwidth is the system bandwidth, frequency band bandwidth, or maximum usable bandwidth, and N is the maximum number of frequency hopping cycles within a single period. The frequency hopping mode adopts a sideband-first approach, traversing the sidebands first, and then traversing the intermediate subbands in sequence.

[0144] In conjunction with the above description, in some embodiments, the frequency domain resource unit occupied by the first training signal includes at least one of the following: RE, PRB, and system bandwidth.

[0145] In some embodiments, the starting point of the frequency hopping of the first training signal includes the first PRB or the first RE. Here, the first PRB can be PRB0, and the first RE can be RE0.

[0146] In some embodiments, the frequency hopping interval of the first training signal is determined based on a first bandwidth and a first number of frequency hopping, wherein the first bandwidth is any one of the system bandwidth, frequency band bandwidth, or the maximum available bandwidth, and the first number of frequency hopping includes the maximum number of frequency hopping in a single cycle.

[0147] Here, the first bandwidth can be the Bandwidth mentioned above, and the first frequency hopping number can be N mentioned above.

[0148] In some embodiments, the frequency hopping mode of the first training signal includes a sideband priority mode, which involves traversing the sidebands first and then traversing the intermediate subbands in sequence.

[0149] The following explains how to use the modulation domain, waveform domain, power domain, and phase domain of the training signal.

[0150] In addition to the resource configuration in the time-frequency domain and code domain, the training signal formed based on OFDM modulation training sequence also needs to be configured or selected accordingly in terms of modulation method, waveform, power and phase.

[0151] Considering the generalization requirements of model training, modulation methods, waveforms, power, and phases need to be traversed to generate corresponding training signals.

[0152] Modulation mode domain: The modulation mode of the training signal should include at least one of π / 2-BPSK, BPSK, QPSK, 16QAM, 64QAM and higher order QAM modulation.

[0153] Waveform domain: The waveform of the training signal should include at least one of CP-OFDM and DFT-s OFDM.

[0154] Power Domain: The generated training signal should contain mixed signals of varying amplitudes, covering the nonlinear response of the power amplifier from the linear region to the saturation region. Example: Input signal amplitude is sampled at logarithmic intervals from 0.1V to 1V to generate a stepped training sequence.

[0155] Phase domain: The generated training signal should contain mixed signals of different phases to cover possible phase drift.

[0156] The above examples summarize the usage of modulation methods, waveforms, power, and phase of training signals:

[0157] Modulation mode domain: The modulation mode of the training signal should include at least one of π / 2-BPSK, BPSK, QPSK, 16QAM, 64QAM and higher order QAM modulation.

[0158] Waveform domain: The waveform of the training signal should include at least one of CP-OFDM and DFT-s OFDM.

[0159] Power domain: The generated training signal should contain mixed signals of different amplitudes, covering the nonlinear response of the power amplifier from the linear region to the saturation region.

[0160] Phase domain: The generated training signal should contain a mixture of signals with different phases.

[0161] In conjunction with the above description, in some embodiments, the waveform of the first training signal includes at least one of the following:

[0162] CP-OFDM, DFT-s OFDM.

[0163] In some embodiments, the first training signal comprises a mixed signal with different amplitudes and / or different phases.

[0164] S102. Determine the distortion model based on the first training signal.

[0165] As an example, a distorted model is determined based on a first training signal and a second training signal, where the second training signal is the undistorted training signal. For instance, a dataset is generated based on the first and second training signals, and then the untrained distorted model is trained using the dataset to obtain the trained distorted model. The second training signal can be the original training signal shown in Figure 2 above. The first training signal has become distorted after transmission through the channel; therefore, the first training signal can be understood as a distorted training signal or a distorted training signal.

[0166] Taking the DPoD model with distortion as an example, for a DPoD model processed on the terminal side, signal distortion is mainly caused by the base station or wireless channel, and the training signal (or sequence) comes from the base station. Network nodes or base stations need to configure the terminal with the training signal (or sequence) sent by the base station. When the terminal receives a specific training signal (or sequence), it can trigger DPoD model calculation or training based on that specific training signal (or sequence). The training signal (or sequence) can also be predefined in the protocol. If the training signal (or sequence) is narrowly interpreted as a code domain signal, the network node or base station still needs to configure auxiliary signaling associated with the training signal (or sequence), such as configuring the time resources, frequency resources, modulation method, waveform, power, and phase of the training signal (or sequence) transmission.

[0167] When the second training signal is configured for the second node, the method further includes:

[0168] Receive first configuration information sent by the second node, the first configuration information being used to configure the second training signal; determine the second training signal based on the first configuration information.

[0169] In some embodiments, the first configuration information is also used to configure at least one of the following associated with the second training signal: training sequence, time resources, frequency resources, modulation scheme, waveform, power, and phase.

[0170] In some embodiments, the distortion model is a DPD model.

[0171] Taking the distortion model as an example of the DPD model, the training of the DPD model on the terminal side is self-sponsored and self-received. The self-sponsored and self-received process includes: the terminal sending an undistorted second training signal; the second training signal being distorted after passing through a power amplifier, becoming a distorted first training signal; and the first node obtaining the distorted first training signal from the second node. Self-sponsored and self-received does not involve base station behavior, but the resources for transmitting and / or receiving the training signal need to be configured by the second node. Based on this, in some embodiments, the first node receives auxiliary signaling sent by the second node. The auxiliary signaling is used to indicate the configuration for transmitting and / or receiving the training signal. The configuration for transmitting and / or receiving the training signal includes at least one of the following: training sequence, time resources, frequency resources, modulation method, waveform, power, and phase.

[0172] Based on the above description, the first node receiving the first training signal sent by the first node can be configured to receive the first training signal based on the auxiliary signaling indication.

[0173] The first node sends the second training signal, which can be configured based on the auxiliary signaling instruction to send the training signal.

[0174] For DPD models processed on the terminal side, signal distortion is caused by the terminal itself. Considering the possibility that the air interface resources used by the mobile communication system terminal may conflict with other terminals, the training signal (or sequence) still needs to be configured to the terminal by the network node or base station. The configuration information (i.e., the first configuration information) includes the code domain resources, time resources, frequency resources, modulation method, waveform, power, phase, etc. of the second training signal (or training sequence).

[0175] In other words, when the DPD model is trained at the first node and the second training signal is configured for the second node, the process of training the DPD model at the first node includes the following steps:

[0176] X1. Obtain the first training signal and the second training signal.

[0177] Here, the first training signal is a distorted / already distorted training signal, and the second training signal is an undistorted training signal.

[0178] As an example, obtaining the first training signal and the second training signal includes:

[0179] The system receives first configuration information sent by the second node, determines a second training signal based on the first configuration information, and sends the second training signal from the first node and receives the first training signal from the first node.

[0180] After the first node sends the second training signal, the second training signal is distorted after passing through the power amplifier and becomes the first training signal, so that the first node receives the first training signal from the first node.

[0181] X2. Determine the DPD model based on the first training signal and the second training signal.

[0182] The above description of the first training signal can also be applied to the first training signal involved in the training of the DPD model on the first node side, and will not be repeated here.

[0183] Besides being configured by the network or base station, the time-domain and frequency-domain resources used for training signals can also be restricted resources by default. Restricted resources refer to resources where normal communication is limited. For example, when a base station or terminal enters a power-saving mode for communication functions, time-frequency domain resources used for power saving through sleep mode can be used by default for transmitting and receiving training signals; or time-frequency domain resources used by the terminal during its sleep period in discontinuous transmission (DTX) / discontinuous reception (DRX). Here, the time-frequency domain resources used for training signals are resources that are not directly configured or notified.

[0184] The above can be summarized as follows: network nodes or base stations configure the terminal with training signals sent by the base station or by the terminal, and the distortion model for compensating for signal distortion is determined based on the training signals.

[0185] Network nodes or base stations configure auxiliary signaling associated with training signals to terminals. The auxiliary signaling includes information such as code domain resources, time resources, frequency resources, modulation method, waveform, power, and phase of the training signal. Alternatively, when auxiliary signaling is unavailable, the training signal uses restricted resources by default. Restricted resources refer to limited communication resources.

[0186] The distortion model is an artificial intelligence-based model, which can be a DPD model with pre-distortion compensation or a DPoD model with post-distortion compensation.

[0187] The inputs for training the above model are the distorted training signal (i.e., the first training signal) and the known training signal (i.e., the second training signal).

[0188] The DPD model's inference input is the communication signal / training signal, and its output is the pre-distorted communication signal / training signal; the DPoD model's inference input is the already distorted communication signal / training signal, and its output is the distorted communication signal / training signal.

[0189] Based on the embodiment shown in Figure 5, since the first training signal sent by the second node to the first node for determining the distortion model is a training signal after OFDM modulation, the distortion model determined based on the first training signal is normalizable, thus realizing the normalization of the distortion model.

[0190] After determining the distortion model, the first node can perform pre-distortion processing or distortion compensation processing on the communication signal based on the distortion model, which can reduce the power consumption of the first node and improve energy saving.

[0191] The following explains the statistics and reporting of historical model usage.

[0192] In some embodiments, after determining the distortion model, the first node can report model information to the second node, so that the second node can determine the first node's use of the distortion model based on the model information.

[0193] The use of AI models generated through training is not static. Due to changes in the usage environment, the specific parameters of internal and external devices, and even the geographical location and usage duration of the model, network devices or terminals need to retrain and update the AI ​​model. Continuously updated models do not mean that the replaced models have lost their value. Remembering historical models and their corresponding device parameters, usage environment, geographical location, usage duration, and other auxiliary information that may affect model replacement, forming a historical model library, can accelerate the convergence speed of model retraining and improve model accuracy. Where storage space allows, auxiliary information can also include the training data packages that formed the historical models.

[0194] Specifically, regarding AI models for nonlinear distortion compensation, since the model is closely related to the RF parameters, it is entirely possible that multiple different AI models will exist depending on factors such as different system bandwidths, different numbers of antenna ports, different device temperatures, and different operating points of RF devices. When the terminal or base station is in power-saving mode, it is very likely that situations such as switching the bandwidth used or switching the number of antenna ports will trigger the use of different AI models.

[0195] Especially when a terminal uses an AI model, regardless of whether the model is trained by a network and transmitted to the terminal, or trained internally by the terminal itself, it must report relevant data about the model it has used historically, including but not limited to the model framework, parameters, and corresponding auxiliary information. At a minimum, it must report the model ID. The model ID is a uniformly defined identifier for each model in the model management function.

[0196] The reporting of historical models can be triggered when initial access is initiated, or when the base station requests the terminal to report, or when the terminal actively reports, such as when the terminal reports its capabilities along with historical model information.

[0197] In addition to reporting historical models, for errors in model usage, such as when a terminal updates its own model through a third party, causing the terminal to be unable to support the use of the new model to complete the functions of the old model under the current base station, the terminal needs to actively report the relevant error information to the network.

[0198] The above embodiments summarize the methods for reporting the historical usage of the model.

[0199] The terminal reports its historical model information to the network device, including but not limited to model ID, model framework, model parameters and corresponding auxiliary information, such as device parameters, usage environment, geographical location, usage duration and training data packets.

[0200] Reporting triggering methods: initial access, base station requests terminal to report, terminal actively reports.

[0201] Error information used by the model, as a special type of model usage history, is also included in the reported information.

[0202] Based on this, in some embodiments, after determining the distortion model, the first node sends model information to the second node. The model information is the model information of the models previously used by the first node, and the model information includes at least one of the following:

[0203] Model identifiers, model framework, model parameters, and corresponding auxiliary information. Historically used models include historically used distorted models.

[0204] In some embodiments, the auxiliary information corresponding to the model includes at least one of the following:

[0205] The model's corresponding device parameters, usage environment, geographical location, usage duration, and dataset. The dataset is the dataset used to determine the model.

[0206] In some embodiments, the auxiliary information corresponding to the model also includes error information, which is error information generated during the use of the model.

[0207] As an example, after the initial access, model information is sent to the second node.

[0208] As another example, the system receives a request message from a second node, which requests model information; in response to the request message, it sends the model information to the second node. This example corresponds to the base station requesting the terminal to report model information as described above.

[0209] The following section introduces the binding of model IDs and dataset IDs.

[0210] The model includes a model identification (ID) (used for model management). A single model ID may have different associated IDs depending on the different datasets used in model training, ensuring consistency between the inference and training environments. The dataset ID identifies the training data and is primarily used for training management. A binding relationship can be established between dataset IDs and associated IDs. In some cases, the base station may require the terminal to train the model on specific datasets. Based on the binding relationship with the datasets, the base station will assign an associated ID to the terminal after the model is trained. The binding relationship is shown in Figure 11.

[0211] The basic process of network-side control binding relationship is as follows:

[0212] Network nodes or base stations configure the binding relationships between multiple different datasets and model association IDs under the same model ID; the network or base station configures the datasets used for terminal training; the network or base station assigns the associated ID to the model trained by the terminal based on the binding relationship.

[0213] The basic process of terminal-side control binding relationship is as follows:

[0214] The terminal generates a training dataset and a binding relationship with associated IDs, and reports the binding relationship to the network or base station.

[0215] Based on this, taking the network-side control binding relationship as an example, the first node receives the association identifier sent by the second node. The association identifier is associated with the dataset used to determine the distortion model. The dataset used is determined based on the first training signal. The association relationship between the dataset and the association identifier is established. In other words, the first training signal generates data (the first training signal is received and processed to form data, and the second training signal is also data). A large amount of data is collected to form a dataset, which can be used for model training.

[0216] Taking the terminal-side control binding relationship as an example, the first node sends the first instruction information to the second node. The first instruction information is used to indicate the association relationship between the dataset used to determine the distortion model and the association identifier. The dataset used is determined based on the first training signal, and the association identifier is the association identifier associated with the dataset used to determine the distortion model.

[0217] The following explains the monitoring of the distortion model.

[0218] Due to variations in the usage environment and the specific parameters of internal and external devices, network nodes or terminals using DPD or DPoD distortion compensation models may experience performance degradation. The method for detecting model performance is called model monitoring, and the module that implements model monitoring is called the model monitoring module. Model monitoring for DPD or DPoD is still based on comparing the original training signal with the deformed training signal.

[0219] Figure 12 shows a model monitoring method for a DPD model and a DPoD model according to an embodiment of this disclosure, illustrating typical model monitoring methods and modules for DPD and DPoD models.

[0220] In the DPD model monitoring method, the comparator inputs are the original training signal and the data processed by the power amplifier (PA). The PA-processed data may require a series of sampling and down-frequency reduction operations, similar to a receiver, to restore the PA output to a signal equivalent to the original training signal before comparison. The comparator output can be varied, including the bit error rate (BER) of the training signal. If the BER is used as an evaluation metric, a BER threshold needs to be configured to compare whether the BER meets the performance requirements.

[0221] In the DPoD model monitoring method, the comparator inputs are the compensated training signal and the original training signal. The comparator output can vary, and can be the bit error rate (BER) of the training signal, etc. If the BER is used as an evaluation metric, a BER threshold needs to be configured to compare whether the BER meets the performance requirements.

[0222] Based on this, in some embodiments, after determining the distorted model, the first node determines the model monitoring result of the distorted model based on the first training signal, the second training signal, and the distorted model. The second training signal is the undistorted training signal, i.e., the original training signal. Based on the model monitoring result, it is determined whether the performance of the distorted model meets the performance requirements.

[0223] Taking the distorted model as an example of the DPD model, based on the first training signal, the second training signal, and the distorted model, the model monitoring results of the distorted model are determined, including:

[0224] The second training signal is input into the DPD model to obtain the pre-distorted training signal;

[0225] The pre-distorted training signal is input to the power amplifier to obtain the processed training signal;

[0226] Based on the second training signal and the processed training signal, the model monitoring results of the DPD model are obtained;

[0227] Taking the distorted model as an example of the DPoD model, based on the first training signal, the second training signal, and the distorted model, the model monitoring results of the distorted model are determined, including:

[0228] The first training signal (i.e. the distorted training signal) is input into the DPoD model to obtain the compensated training signal;

[0229] Based on the compensated training signal and the second training signal, the model monitoring results of the DPoD model are determined.

[0230] Taking the model monitoring results, including the bit error rate, as an example, based on the model monitoring results, determine whether the performance of the distorted model meets the performance requirements, including:

[0231] Based on the bit error rate and bit error rate threshold, it is determined whether the performance of the distortion model meets the performance requirements. The bit error rate threshold is either predefined or configured for the second node.

[0232] If the bit error rate is less than or equal to the bit error rate threshold, the performance of the distortion model is determined to meet the performance requirements; if the bit error rate is greater than the bit error rate threshold, the performance of the distortion model is determined to not meet the performance requirements.

[0233] The following explains the request and notification regarding the number of antennas.

[0234] One of the purposes of DPD and DPoD is to assist terminal or network equipment in improving power amplifier efficiency and achieving energy saving. Equipment energy saving is greatly affected by the number of transceiver antennas used at the time; more transceiver antennas consume more power. Flexible adjustment of the number of transceiver antennas can achieve energy saving while meeting transmission performance requirements.

[0235] Due to limitations in battery power and other factors, terminals have limited energy supply and need to adapt the number of their transceiver antennas to the data transmission scenarios as much as possible. For example, in scenarios such as silent, sleep, low power, or no data transmission, the number can be reduced to a single antenna. Conversely, when large amounts of data are transmitted, the maximum number of transceiver antennas can be configured. The terminal UE's request to adjust the number of antennas can be expressed by reporting or requesting the base station. The base station responds to the terminal UE's report request and configures an appropriate number of transceiver antennas.

[0236] One reporting method could be a schedule request (SR) signaling based on Media Access Control (MAC) layer signaling, which includes the required number of downlink (DL) / uplink (UL) antennas. Alternatively, the required number of DL / UL antennas could be reported using the slower Radio Resource Control (RRC) layer signaling.

[0237] To conserve energy, base stations need to flexibly configure the number of their transmit and receive antennas at appropriate times and in appropriate scenarios. Terminals need to be aware of the base station's transmit and receive antenna count in a timely manner. An efficient notification method is for the base station to broadcast the number of transmit and receive antennas used to the UE (User Equipment). This broadcast message can be based on system messages. When the broadcast message can also be based on on-demand system messages, the base station can also respond to a single terminal's request and notify the terminal of its transmit and receive antenna count. For example, if the base station's maximum transmit and receive antenna capacity is 64T / 64R (T: transmit, R: receive), the base station can flexibly configure its transmit and receive antennas to 32T / 32R, 16T / 16R, or 8T / 8R, and notify the terminal via broadcast.

[0238] Summary of key points of the above embodiments:

[0239] 1. The base station broadcasts the number of its transmit / receive antennas to multiple terminals, or the base station responds to a terminal's request and notifies the terminal of the number of its transmit / receive antennas.

[0240] Second: The terminal reports the number of its transmit / receive antennas to the base station and receives the number of transmit / receive antennas configured by the base station.

[0241] Based on this, as an example, the first node receives the number of transmit / receive antennas sent by the second node.

[0242] In some embodiments, before the first node receives the number of transmit / receive antennas sent by the second node, the first node sends a request message to the second node to request the number of transmit / receive antennas.

[0243] As another example, the first node sends a request to the second node to request the number of transmit / receive antennas. Then, the first node receives configuration information from the second node to configure the number of transmit / receive antennas for itself. The first node then configures the number of transmit / receive antennas based on this configuration information.

[0244] Based on the methods shown in the two examples above, the power consumption of the first and second nodes can be reduced, thereby improving energy efficiency.

[0245] The above embodiments illustrate a communication method provided by the present disclosure from the perspective of a first node. In some embodiments, as shown in FIG13, the present disclosure also provides another communication method, which is applied to a second node and may include the following steps:

[0246] S201, Send the first training signal to the first node.

[0247] Here, the first training signal is used to determine the distortion model at the first node, and the first training signal is the training signal after OFDM modulation.

[0248] In some embodiments, the distortion model is a DPoD model.

[0249] In some embodiments, the second node sends first configuration information to the first node. The first configuration information is used to configure a second training signal. The second training signal is an undistorted training signal. The first node can determine the second training signal based on the first configuration information, and then determine the distorted model based on the undistorted second training signal and the distorted first training signal.

[0250] In some embodiments, the first configuration information is also used to configure at least one of the following associated with the second training signal: training sequence, time resources, frequency resources, modulation scheme, waveform, power, and phase.

[0251] In some embodiments, the first training signal is determined based on a training sequence, which is modulated based on OFDM, mapped to a configuration bandwidth in the frequency domain, and mapped to one or more OFDM symbols in the time domain before transmission.

[0252] In some embodiments, the training sequence includes any one of the following: an unmodulated pseudo-random sequence, a modulated pseudo-random sequence, a sequence of all zeros, or a sequence of all ones, where modulation includes at least one of the following:

[0253] π / 2-phase shift keying (BPSK), BPSK, quadrature phase shift keying (QPSK), and N-order quadrature amplitude modulation (QAM), where N is an integer greater than or equal to 16.

[0254] In some embodiments, the first training signal is a two-tone or multi-tone signal based on OFDM modulation, and the modulation information on the OFDM subcarrier is empty.

[0255] In some embodiments, the first training signal is a reference signal used to train the model.

[0256] In some embodiments, the first training signal is determined based on a training sequence, and the training sequence is segmented to obtain multiple segmented training sequences, which are respectively mapped onto one or more OFDM symbols for transmission.

[0257] In some embodiments, the training sequences of each segment are connected end to end, and the training sequences of adjacent segments have overlapping elements in the region where they are connected end to end.

[0258] In some embodiments, the time length of the overlapping element or the time length of the OFDM symbol containing the overlapping element is greater than or equal to the memory depth.

[0259] In some embodiments, the first training signal is determined based on a training sequence, which is repeatedly mapped to multiple OFDM symbol groups. Each OFDM symbol group includes at least one OFDM symbol. In each OFDM symbol group, the resource element RE position within the OFDM symbol that is not mapped to the training sequence is empty or is a padding bit.

[0260] In some embodiments, the frequency domain resource unit occupied by the first training signal includes at least one of the following: RE, physical resource block (PRB), and system bandwidth.

[0261] In some embodiments, the starting point for frequency hopping of the first training signal includes the first PRB or the first RE.

[0262] In some embodiments, the frequency hopping interval of the first training signal is determined based on a first bandwidth and a first number of frequency hopping, wherein the first bandwidth is any one of the system bandwidth, frequency band bandwidth, or the maximum available bandwidth, and the first number of frequency hopping includes the maximum number of frequency hopping in a single cycle.

[0263] In some embodiments, the frequency hopping mode of the first training signal includes a sideband priority mode, which involves traversing the sidebands first and then traversing the intermediate subbands in sequence.

[0264] In some embodiments, the waveform of the first training signal includes at least one of the following:

[0265] Cyclic prefix OFDM, Discrete Fourier transform extended DFT-s OFDM.

[0266] In some embodiments, the first training signal comprises a mixed signal with different amplitudes and / or different phases.

[0267] In some embodiments, the second node sends auxiliary signaling to the first node. The auxiliary signaling is used to indicate the configuration for transmitting training signals and / or the configuration for receiving training signals. The configuration for transmitting training signals and / or the configuration for receiving training signals include at least one of the following:

[0268] Training sequence, time resources, frequency resources, modulation method, waveform, power, phase.

[0269] In some embodiments, the second node receives model information sent by the first node. The model information is model information of models historically used by the first node, and the model information includes at least one of the following:

[0270] Model identifier, model framework, model parameters, and corresponding auxiliary information.

[0271] In some embodiments, the auxiliary information corresponding to the model includes at least one of the following:

[0272] The model includes the device parameters, usage environment, geographical location, usage duration, and dataset.

[0273] In some embodiments, the auxiliary information corresponding to the model also includes error information, which is error information generated during the use of the model.

[0274] As an example, receiving model information sent by the first node includes: receiving model information sent by the first node in the initial access indication.

[0275] As another example, receiving model information sent by the first node includes: sending request information to the first node, the request information being used to request model information; and receiving model information sent by the first node.

[0276] In some embodiments, the second node sends an association identifier to the first node, the association identifier being an association identifier associated with the dataset used to determine the distortion model; the dataset is determined based on the first training signal;

[0277] In some embodiments, the second node receives first indication information sent by the first node and sends the first indication information to the second node. The first indication information is used to indicate the association relationship between the dataset used to determine the distortion model and the association identifier. The dataset is determined based on the first training signal, and the association identifier is an association identifier associated with the dataset used to determine the distortion model.

[0278] The descriptions of the information in step S201 can be found in the corresponding descriptions of the process of determining the distortion model for the first node, and will not be repeated here.

[0279] The above embodiments are illustrated using the example of a first node determining a distortion model. The following embodiments are illustrated using a second node determining a distortion model. In some embodiments, as shown in FIG14, this disclosure also provides a communication method applied to a first node, which may include the following steps:

[0280] S301, Send the third training signal to the second node.

[0281] Here, the third training signal is used to determine the distortion model at the second node, and the third training signal is a training signal modulated by OFDM.

[0282] As an example, the distortion model determined by the second node is the DPoD model.

[0283] For DPoD models processed at the base station, signal distortion is mainly caused by the terminal or wireless channel, and the training signal (or sequence) originates from the terminal. Network nodes or base stations need to configure the training signal (or sequence) sent by the terminal. When the base station receives a specific training signal (or sequence), it can trigger DPoD model calculation or training based on that specific training signal (or sequence). The training signal (or sequence) can also be predefined in the protocol, but this is less flexible than network configuration. If the training signal (or sequence) is narrowly interpreted as a code domain signal, the network or base station still needs to configure auxiliary signaling associated with the training signal (or sequence), such as configuring the time resources, frequency resources, modulation scheme, waveform, power, and phase of the training signal (or sequence) transmission.

[0284] In other words, in some embodiments, the first node receives auxiliary signaling sent by the second node, generates a third training signal based on the auxiliary signaling, and sends the third training signal to the second node based on the auxiliary signaling. For a description of the auxiliary signaling, please refer to the description of the auxiliary signaling in step S101 above, which will not be repeated here.

[0285] For DPD models processed at the base station, signal distortion is caused by the base station itself, while the training signal (or sequence) is configured by the network or base station itself. Since the time resources, frequency resources, modulation method, waveform, power, phase, and other elements of the training signal (or sequence) transmission are all handled by the base station scheduling, there is no need to notify the terminal through an interface.

[0286] In other words, the DPD model trained on the second node can include the following steps:

[0287] V1. Obtain the first training signal and the second training signal.

[0288] The first training data is a distorted / already distorted training signal, and the second training signal is an undistorted training signal.

[0289] In this example, both the distorted first training signal and the undistorted second training signal are generated by the second node itself.

[0290] V2. Based on the first training signal and the second training signal, determine the DPD model.

[0291] The above description of the first training signal also applies to the third training signal; simply replace the first training signal with the third training signal. For example, in some embodiments, the third training signal is determined based on a training sequence, which is modulated using OFDM, mapped to a configured bandwidth in the frequency domain, and mapped to one or more OFDM symbols in the time domain before transmission.

[0292] In some embodiments, the training sequence includes any one of the following: an unmodulated pseudo-random sequence, a modulated pseudo-random sequence, a sequence of all zeros, or a sequence of all ones, where modulation includes at least one of the following:

[0293] π / 2-phase shift keying (BPSK), BPSK, quadrature phase shift keying (QPSK), and N-order quadrature amplitude modulation (QAM), where N is an integer greater than or equal to 16.

[0294] In some embodiments, the third training signal is a two-tone or multi-tone signal based on OFDM modulation, and the modulation information on the OFDM subcarrier is empty.

[0295] In some embodiments, the third training signal is a reference signal used to train the model.

[0296] In some embodiments, the third training signal is determined based on the training sequence, and the training sequence is segmented to obtain multiple segmented training sequences, which are respectively mapped onto one or more OFDM symbols for transmission.

[0297] In some embodiments, the training sequences of each segment are connected end to end, and the training sequences of adjacent segments have overlapping elements in the region where they are connected end to end.

[0298] In some embodiments, the time length of the overlapping element or the time length of the OFDM symbol containing the overlapping element is greater than or equal to the memory depth.

[0299] In some embodiments, the third training signal is determined based on a training sequence that is repeatedly mapped to multiple OFDM symbol groups. Each OFDM symbol group includes at least one OFDM symbol. In each OFDM symbol group, the resource element RE position within the OFDM symbol that is not mapped to the training sequence is empty or is a padding bit.

[0300] In some embodiments, the frequency domain resource unit occupied by the third training signal includes at least one of the following: RE, physical resource block (PRB), and system bandwidth.

[0301] In some embodiments, the starting point for frequency hopping of the third training signal includes the first PRB or the first RE.

[0302] In some embodiments, the frequency hopping interval of the third training signal is determined based on a first bandwidth and a first number of frequency hopping, wherein the first bandwidth is any one of the system bandwidth, frequency band bandwidth, or the maximum available bandwidth, and the first number of frequency hopping includes the maximum number of frequency hopping in a single cycle.

[0303] In some embodiments, the frequency hopping mode of the third training signal includes a sideband priority mode, which involves traversing the sidebands first and then traversing the intermediate subbands in sequence.

[0304] In some embodiments, the waveform of the third training signal includes at least one of the following:

[0305] Cyclic prefix OFDM, Discrete Fourier transform extended DFT-s OFDM.

[0306] In some embodiments, the third training signal comprises a mixture of signals with different amplitudes and / or different phases.

[0307] The first node sends a third training signal modulated by OFDM to the second node so that the second node can determine the distortion model based on the third training signal, thereby achieving the standardization of the distortion model.

[0308] In some embodiments, the second node sends second configuration information to the first node. The second configuration information is used to configure a fourth training signal, which is an undistorted training signal. Accordingly, the first node receives the second configuration information sent by the second node and determines the fourth training signal based on the second configuration information.

[0309] Taking the DPD model trained by the first node as an example, the first node can determine the DPD model based on the third and fourth training signals.

[0310] The descriptions of the fourth training signal and the second configuration information can be found in the descriptions of the second training signal and the first configuration information described above, and will not be repeated here.

[0311] The descriptions of the embodiments involving the first node sending model information to the second node, the first node receiving the association identifier sent by the second node, and the first node sending the first indication information to the second node, as described in the above introduction of the first node determining the distortion model, can also be applied to the embodiments shown in Figure 14 and the embodiments shown in Figure 15 below, and will not be repeated here.

[0312] In some embodiments, as shown in FIG15, this disclosure also provides another communication method, which is applied to a second node and may include the following steps:

[0313] S401, Receive the third training signal sent by the first node.

[0314] Here, the third training signal is used to determine the distortion model at the second node, and the third training signal is the training signal after OFDM modulation.

[0315] The description of the third training signal can be found in the corresponding description in step S301, and will not be repeated here.

[0316] S402. Determine the distortion model based on the third training signal.

[0317] Here, the distortion model is the DPoD model.

[0318] As an example, the second node determines the distortion model based on the second and third training signals. For instance, a dataset is generated based on the second and third training signals, and then the distortion model is determined based on the dataset.

[0319] In some embodiments, the second node determines the model monitoring result of the distorted model based on the second training signal, the third training signal, and the distorted model, wherein the second training signal is the undistorted training signal; and determines whether the performance of the distorted model meets the performance requirements based on the model monitoring result.

[0320] The description of the second node determining the model monitoring results can be referred to the description of the first node determining the model monitoring results, and will not be repeated here.

[0321] In some embodiments, the second node sends second configuration information to the first node. The second configuration information is used to configure a fourth training signal, which is an undistorted training signal. The first node can determine a distorted model, such as a DPD model, based on the undistorted fourth training signal.

[0322] The descriptions of the fourth training signal and the second configuration information can be found in the descriptions of the second training signal and the first configuration information described above, and will not be repeated here.

[0323] The foregoing primarily describes the solution provided in this disclosure from the perspective of the interaction between various nodes. Each node, such as the first node or the second node, includes corresponding hardware structures and / or software modules to perform the aforementioned functions. Those skilled in the art should readily recognize that, based on the algorithmic steps of the examples described in conjunction with the embodiments disclosed herein, this disclosure can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0324] This disclosure embodiment can divide the first node or the second node into functional modules according to the above method embodiment. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one functional module. The integrated module can be implemented in hardware or software. The module division in this disclosure embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods. The following description uses the example of dividing each functional module according to each function.

[0325] Figure 16 is a block diagram of a communication device according to an embodiment of the present disclosure. As shown in Figure 16, the communication device 50 includes a receiving unit 501 and a processing unit 502. In some embodiments, the communication device 50 further includes a transmitting unit 503.

[0326] The communication device 50 can be the first node or a chip within the first node. When the communication device 50 is used to implement the functions of the first node in the above embodiments, each unit is specifically used to implement the following functions.

[0327] The receiving unit 501 is used to receive the first training signal sent by the second node. The first training signal is used by the first node to determine the distortion model. The first training signal is a training signal after being modulated by orthogonal frequency division multiplexing (OFDM).

[0328] Processing unit 502 is used to determine the distortion model based on the first training signal.

[0329] In some embodiments, the processing unit 502 is specifically configured to determine a distorted model based on a first training signal and a second training signal, wherein the second training signal is an undistorted training signal.

[0330] In some embodiments, the receiving unit 501 is further configured to receive first configuration information sent by the second node, the first configuration information being used to configure the second training signal;

[0331] The processing unit 502 is also used to determine the second training signal based on the first configuration information.

[0332] In some embodiments, the receiving unit 501 is further configured to receive auxiliary signaling sent by the second node, the auxiliary signaling being used to indicate the configuration for sending training signals and / or the configuration for receiving training signals, the configuration for sending training signals and / or the configuration for receiving training signals including at least one of the following:

[0333] Training sequence, time resources, frequency resources, modulation method, waveform, power, phase.

[0334] In some embodiments, the sending unit 503 is further configured to send model information to the second node, wherein the model information is model information of a model historically used by the first node, and the model information includes at least one of the following:

[0335] Model identifier, model framework, model parameters, and corresponding auxiliary information.

[0336] In some embodiments, the sending unit 503 is specifically used to send model information to the second node after initial access.

[0337] In some embodiments, the receiving unit 501 is further configured to receive request information sent by the second node, the request information being used to request model information;

[0338] The sending unit 503 is specifically used to send model information to the second node in response to the request information.

[0339] In some embodiments, the receiving unit 501 is further configured to receive an association identifier sent by the second node, wherein the association identifier is an association identifier associated with the dataset used to determine the distortion model; the dataset is determined based on the first training signal.

[0340] The processing unit 502 is also used to establish the association between the dataset and the associated identifier.

[0341] In some embodiments, the sending unit 503 is further configured to send first indication information to the second node. The first indication information is used to indicate the association relationship between the dataset used to determine the distortion model and the association identifier. The dataset is determined based on the first training signal, and the association identifier is an association identifier associated with the dataset used to determine the distortion model.

[0342] In some embodiments, the processing unit 502 is further configured to determine the model monitoring result of the distorted model based on the first training signal, the second training signal, and the distorted model, wherein the second training signal is an undistorted training signal; and to determine whether the performance of the distorted model meets the performance requirements based on the model monitoring result.

[0343] In some embodiments, the processing unit 502 is specifically used to determine whether the performance of the distortion model meets the performance requirements based on the bit error rate and the bit error rate threshold, wherein the bit error rate threshold is predefined or configured for the second node.

[0344] Figure 17 is a block diagram of another communication device provided according to an embodiment of the present disclosure. As shown in Figure 17, the communication device 60 includes a transmitting unit 601.

[0345] The communication device 60 can be the second node or a chip within the second node. When the communication device 60 is used to implement the functions of the second node in the above embodiments, each unit is specifically used to implement the following functions.

[0346] The transmitting unit 601 is used to transmit a first training signal to the first node. The first training signal is used by the first node to determine the distortion model. The first training signal is a training signal after OFDM modulation.

[0347] In some embodiments, the sending unit 601 is further configured to send first configuration information to the first node, the first configuration information being used to configure a second training signal, the second training signal being an undistorted training signal.

[0348] In some embodiments, the sending unit 601 is further configured to send auxiliary signaling to the first node.

[0349] Figure 18 is a block diagram of another communication device provided according to an embodiment of the present disclosure. As shown in Figure 18, the communication device 70 includes a transmitting unit 701 and a receiving unit 702.

[0350] The communication device 70 can be the first node or a chip within the first node. When the communication device 70 is used to implement the functions of the first node in the above embodiments, each unit is specifically used to implement the following functions.

[0351] The transmitting unit 701 is used to send a third training signal to the second node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation.

[0352] The receiving unit 702 is used to receive the second configuration information sent by the second node. The second configuration information is used to configure the fourth training signal, which is an undistorted training signal.

[0353] Figure 19 is a block diagram of another communication device provided according to an embodiment of the present disclosure. As shown in Figure 19, the communication device 80 includes a receiving unit 801 and a processing unit 802.

[0354] The communication device 80 can be the second node or a chip within the second node. When the communication device 80 is used to implement the functions of the second node in the above embodiments, each unit is specifically used to implement the following functions.

[0355] The receiving unit 801 is used to receive the third training signal sent by the first node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation.

[0356] Processing unit 802 is used to determine the distortion model based on the third training signal.

[0357] The units in Figures 16 to 19 can also be called modules; for example, a transmitting unit can be called a transmitting module. Furthermore, in the embodiments shown in Figures 16 to 19, the names of the units may not be those shown in the figures; for example, a transmitting unit can also be called a communication unit, and a receiving unit can also be called a communication unit.

[0358] If the units in Figures 16 to 19 are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this disclosure, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this disclosure. Storage media for storing computer software products include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0359] In the case where any of the communication devices 50 to 80 described above implements the functions of the integrated module in hardware, a block diagram of another communication device is provided according to an embodiment of this disclosure. As shown in FIG20, the communication device 90 includes: a processor 902, a communication interface 903, and a bus 904. In some embodiments, the communication device 90 may further include a memory 901.

[0360] Processor 902 may implement or execute various exemplary logic blocks, modules, and circuits described in connection with embodiments of this disclosure. Processor 902 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in connection with embodiments of this disclosure. Processor 902 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a digital signal processor (DSP), and a microprocessor.

[0361] The communication interface 903 is used to connect to other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0362] The memory 901 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0363] In some embodiments, the memory 901 may exist independently of the processor 902. The memory 901 may be connected to the processor 902 via a bus 904 and may be used to store instructions or program code. When the processor 902 calls and executes the instructions or program code stored in the memory 901, it can implement the communication method provided in the embodiments of this disclosure.

[0364] In some embodiments, the memory 901 may also be integrated with the processor 902.

[0365] Bus 904 can be an extended industry standard architecture (EISA) bus, etc. Bus 904 can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in Figure 20, but this does not mean that there is only one bus or one type of bus.

[0366] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the first node or the second node can be divided into different functional modules to complete all or part of the functions described above.

[0367] This disclosure also provides a computer-readable storage medium, including a non-transitory computer-readable storage medium on which computer instructions are stored. All or part of the processes in the above method embodiments can be executed by computer instructions instructing related hardware. These computer instructions can be stored in the aforementioned non-transitory computer-readable storage medium. When executed, the computer instructions can include the processes described in the above method embodiments. The aforementioned computer-readable storage medium can also be an external storage device for the aforementioned first node or second node, such as a pluggable hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the aforementioned first node or second node. Further, the aforementioned computer-readable storage medium can also include both internal storage units of the aforementioned first node or second node and external storage devices. The aforementioned computer-readable storage medium is used to store the aforementioned computer program and other programs and data required by the aforementioned first node or second node. The aforementioned computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0368] This disclosure also provides a computer program product comprising a computer program that, when run on a computer, causes the computer to perform any of the communication methods provided in the above embodiments.

[0369] Although this disclosure has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings, the disclosure, and the appended claims in carrying out the claimed disclosure. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce a good effect.

[0370] Although this disclosure has been described in conjunction with specific features and embodiments, it will be apparent that various modifications and combinations can be made therein without departing from the spirit and scope of this disclosure. Accordingly, this specification and drawings are merely exemplary illustrations of the disclosure as defined by the appended claims and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this disclosure. It is obvious that those skilled in the art can make various alterations and modifications to this disclosure without departing from its spirit and scope. Thus, this disclosure is also intended to include any such modifications and modifications that fall within the scope of the claims of this disclosure and their equivalents.

[0371] The above are merely specific embodiments of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any changes or substitutions within the technical scope disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. A communication method, wherein, Applied to the first node, the method includes: The first training signal sent by the second node is received. The first training signal is used by the first node to determine the distortion model. The first training signal is a training signal after being modulated by orthogonal frequency division multiplexing (OFDM). The distortion model is determined based on the first training signal.

2. The method according to claim 1, wherein, The distortion model is the Digital Post-Distortion (DPoD) model.

3. The method according to claim 1 or 2, wherein, Determining the distortion model based on the first training signal includes: The distortion model is determined based on the first training signal and the second training signal, wherein the second training signal is the undistorted training signal.

4. The method according to claim 3, wherein, The second training signal is either predefined or configured for the second node.

5. The method according to claim 4, wherein, When the second training signal is configured for the second node, the method further includes: Receive first configuration information sent by the second node, the first configuration information being used to configure the second training signal; Based on the first configuration information, the second training signal is determined.

6. The method according to claim 5, wherein, The first configuration information is also used to configure at least one of the following related to the second training signal: Training sequence, time resources, frequency resources, modulation method, waveform, power, phase.

7. The method according to any one of claims 1 to 6, wherein, The first training signal is determined based on a training sequence, which is modulated by OFDM, mapped to a configuration bandwidth in the frequency domain, and mapped to one or more OFDM symbols in the time domain before transmission.

8. The method according to claim 7, wherein, The training sequence includes any one of the following: an unmodulated pseudo-random sequence, a modulated pseudo-random sequence, a sequence of all zeros, or a sequence of all ones, wherein the modulation includes at least one of the following: π / 2-phase shift keying (BPSK), BPSK, quadrature phase shift keying (QPSK), and N-order quadrature amplitude modulation (QAM), where N is an integer greater than or equal to 16.

9. The method according to any one of claims 1 to 8, wherein, The first training signal is a two-tone or multi-tone signal based on OFDM modulation, and the modulation information on the subcarrier of the OFDM is empty.

10. The method according to any one of claims 1 to 9, wherein, The first training signal is a reference signal used to train the model.

11. The method according to any one of claims 1 to 10, wherein, The first training signal is determined based on the training sequence, and the training sequence is segmented into multiple segments, which are then mapped onto one or more OFDM symbols for transmission.

12. The method according to claim 11, wherein, The training sequences of each segment are connected end to end, and there are overlapping elements in the connected regions of adjacent segments.

13. The method according to claim 12, wherein, The time length of the overlapping element or the time length of the OFDM symbol containing the overlapping element is greater than or equal to the memory depth.

14. The method according to claim 11, wherein, For OFDM symbols containing unmapped training sequences in the time domain, the resource element positions are either empty or filled bits.

15. The method according to any one of claims 1 to 14, wherein, The first training signal is determined based on a training sequence, which is repeatedly mapped to multiple OFDM symbol groups. Each OFDM symbol group includes at least one OFDM symbol. In each OFDM symbol group, the resource element RE position within the OFDM symbol that is not mapped to the training sequence is empty or is a padding bit.

16. The method according to any one of claims 1 to 15, wherein, The frequency domain resource unit occupied by the first training signal includes at least one of the following: RE, physical resource block (PRB), and system bandwidth.

17. The method according to any one of claims 1 to 16, wherein, The starting point for frequency hopping of the first training signal includes the first PRB or the first RE.

18. The method according to any one of claims 1 to 17, wherein, The frequency hopping interval of the first training signal is determined based on the first bandwidth and the first number of frequency hopping. The first bandwidth is any one of the system bandwidth, frequency band bandwidth, or the maximum usable bandwidth. The first number of frequency hopping includes the maximum number of frequency hopping in a single cycle.

19. The method according to any one of claims 1 to 18, wherein, The frequency hopping mode of the first training signal includes a sideband priority mode, which involves traversing the sidebands first and then traversing the intermediate subbands in sequence.

20. The method according to any one of claims 1 to 19, wherein, The waveform of the first training signal includes at least one of the following: Cyclic prefix OFDM, Discrete Fourier transform extended DFT-s OFDM.

21. The method according to any one of claims 1 to 20, wherein, The first training signal includes mixed signals with different amplitudes and / or different phases.

22. The method according to any one of claims 1 to 21, wherein, The method further includes: The system receives auxiliary signaling sent by the second node, the auxiliary signaling being used to indicate the configuration for transmitting training signals and / or receiving training signals, the configuration for transmitting training signals and / or receiving training signals including at least one of the following: Training sequence, time resources, frequency resources, modulation method, waveform, power, phase.

23. The method according to any one of claims 1 to 22, wherein, The method further includes: Send model information to the second node, wherein the model information is model information of models historically used by the first node, and the model information includes at least one of the following: Model identifier, model framework, model parameters, and corresponding auxiliary information.

24. The method according to claim 23, wherein, The auxiliary information corresponding to the model includes at least one of the following: The model includes the device parameters, usage environment, geographical location, usage duration, and dataset.

25. The method according to claim 24, wherein, The auxiliary information corresponding to the model also includes error information, which is the error information generated during the use of the model.

26. The method according to any one of claims 23 to 25, wherein, Sending model information to the second node includes: After initial access, the model information is sent to the second node.

27. The method according to any one of claims 23 to 25, wherein, Sending model information to the second node includes: Receive request information sent by the second node, the request information being used to request the model information; In response to the request information, the model information is sent to the second node.

28. The method according to any one of claims 1 to 27, wherein, The method further includes: The system receives an association identifier sent by the second node, the association identifier being an association identifier associated with the dataset used to determine the distortion model; the dataset is determined based on the first training signal. Establish the association between the dataset and the associated identifier.

29. The method according to any one of claims 1 to 28, wherein, The method further includes: Send a first indication message to the second node. The first indication message is used to indicate the association relationship between the dataset used to determine the distortion model and the association identifier. The dataset is determined based on the first training signal, and the association identifier is an association identifier associated with the dataset used to determine the distortion model.

30. The method according to any one of claims 1 to 29, wherein, The method further includes: Based on the first training signal, the second training signal, and the distortion model, the model monitoring result of the distortion model is determined, wherein the second training signal is the undistorted training signal; Based on the monitoring results of the model, determine whether the performance of the distortion model meets the performance requirements.

31. The method according to claim 30, wherein, The model monitoring results include the bit error rate. Determining whether the performance of the distortion model meets the performance requirements based on the model monitoring results includes: Based on the bit error rate and bit error rate threshold, it is determined whether the performance of the distortion model meets the performance requirements. The bit error rate threshold is either predefined or configured for the second node.

32. A communication method, wherein, Applied to the second node, the method includes: A first training signal is sent to the first node. The first training signal is used by the first node to determine the distortion model. The first training signal is a training signal after OFDM modulation.

33. The method according to claim 32, wherein, The method further includes: Send first configuration information to the first node. The first configuration information is used to configure a second training signal, which is an undistorted training signal.

34. A communication method, wherein, Applied to the first node, the method includes: A third training signal is sent to the second node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation.

35. The method according to claim 34, wherein, The method further includes: The system receives second configuration information sent by the second node, which is used to configure a fourth training signal, wherein the fourth training signal is an undistorted training signal.

36. A communication method, wherein, Applied to the second node, the method includes: The second node receives a third training signal sent by the first node. The third training signal is used by the second node to determine the distortion model. The third training signal is a training signal after OFDM modulation. The distortion model is determined based on the third training signal.

37. The method of claim 36, wherein, The distortion model is the DPoD model.

38. A communication device, wherein, include: Memory and processor; Memory and processor are coupled; The memory is used to store instructions that can be executed by the processor; When the processor executes the instructions, it performs the method as described in any one of claims 1 to 37.

39. A computer-readable storage medium, wherein, The computer-readable storage medium includes a non-transitory computer-readable storage medium on which computer instructions are stored, which, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 37.

40. A computer program product, wherein, The computer program product includes computer instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 37.