Method executed by network node and network node

Abstract

The embodiment of the invention provides a method executed by a network node and the network node, and relates to the field of artificial intelligence. The method comprises the following steps: executing frequency domain channel estimation based on a decorrelation signal about a demodulation reference signal (DMRS) symbol received from user equipment, and obtaining a first channel estimation value of a channel used by the user equipment; determining a channel type of the channel; and when the channel type of the channel is a high-speed mobile channel, adjusting the frequency offset of the channel, and obtaining a second channel estimation value based on the adjusted frequency offset and the first channel estimation value. Optionally, the method performed by the electronic device may be performed using an artificial intelligence model.

Description

Technical Field

[0001] The present application relates to the field of mobile communication technology, and in particular, to a method and network node for channel estimation of a high speed train (HST) cell of a virtual radio access network (vRAN) system. Background Art

[0002] High-speed mobile communication is a scenario that current communication products must support, and its coverage and quality are an important challenge for current communication products. Accurate channel estimation is the key to ensuring the performance of high-speed mobile scenarios. However, due to the particularity of high-speed mobile channels, accurate frequency offset estimation is a big challenge for channel estimation in high-speed mobile scenarios.

[0003] Existing channel estimation methods cannot accurately estimate the frequency deviation range of high-speed mobile channels. Incorrect frequency deviation will lead to reduced cell capacity, high user access failure rate and poor user experience.

[0004] How to accurately estimate the frequency offset for a high-speed mobile channel, and then better perform channel estimation to meet communication needs, is a technical problem that those skilled in the art have been working hard to study. Summary of the invention

[0005] In order to at least solve the above problems existing in the prior art, the present invention provides a method executed by a network node and a network node.

[0006] According to a first aspect of an embodiment of the present application, a method performed by a network node is provided, comprising: performing frequency domain channel estimation based on a decorrelated signal of a demodulation reference signal DMRS symbol received from a user equipment to obtain a first channel estimation value of a channel used by the user equipment; determining the channel type of the channel; and when the channel type of the channel is a high-speed mobile channel, adjusting the frequency offset of the channel, and obtaining a second channel estimation value based on the adjusted frequency offset and the first channel estimation value.

[0007] Optionally, the step of performing frequency domain channel estimation based on a decorrelated signal of a demodulation reference signal DMRS symbol received from a user equipment to obtain a first channel estimation value of a channel used by the user equipment includes: performing time offset estimation based on the decorrelated signal to obtain a time offset regarding the channel; and performing time offset compensation on the decorrelated signal based on the time offset to obtain the first channel estimation value.

[0008] Optionally, the step of determining the channel type of the channel includes: determining the channel type of the channel through a neural network based on the decorrelation signal.

[0009] Optionally, the neural network is one of a support vector machine, a convolutional neural network, a recurrent neural network, a long short-term memory network, a Transformer network, and a multi-layer perceptron mixer.

[0010] Optionally, the step of determining the channel type of the channel through a neural network based on the decorrelation signal includes: obtaining at least one feature of a phase characteristic, an amplitude characteristic, a radial quantity characteristic of the channel, and a signal-to-noise ratio of the channel according to the decorrelation signal; and determining the channel type of the channel through a neural network based on the at least one feature.

[0011] Optionally, the step of obtaining at least one of a phase characteristic, an amplitude characteristic, a path quantity characteristic of the channel and a signal-to-noise ratio of the channel according to the decorrelation signal includes: determining the signal-to-noise ratio of the channel according to the decorrelation signal; and / or obtaining a first signal matching the frequency domain resource dimension of the neural network by matching the dimension of the decorrelation signal with the frequency domain resource dimension of the neural network, and extracting, based on the first signal, a phase characteristic and an amplitude characteristic of the first signal on each resource element and at least one of the path quantity characteristics of the channel.

[0012] Optionally, the step of determining the channel type of the channel through a neural network based on the at least one feature includes: determining the channel type of the channel through a neural network based on the at least one feature and a first signal.

[0013] Optionally, the step of obtaining a first signal matching the frequency domain resource dimension of the neural network by matching the dimension of the decorrelation signal with the frequency domain resource dimension of the neural network includes: if the dimension of the decorrelation signal is equal to the frequency domain resource dimension of the neural network, determining the decorrelation signal as the first signal; if the dimension of the decorrelation signal is smaller than the frequency domain resource dimension of the neural network, expanding the dimension of the decorrelation signal to the frequency domain resource dimension of the neural network based on the time offset of the channel to obtain the first signal; if the dimension of the decorrelation signal is larger than the frequency domain resource dimension of the neural network, intercepting a portion of the signal from the decorrelation signal based on the frequency domain resource dimension of the neural network as the first signal.

[0014] Optionally, the step of extracting the path quantity characteristics of the channel according to the first signal includes: determining the amplitude of the time domain impulse response of the channel according to the first signal; and extracting the path quantity characteristics of the channel according to the number of times the amplitude of the time domain impulse response exceeds a predetermined threshold.

[0015] Optionally, the step of adjusting the frequency offset of the channel includes: determining a frequency offset interval in the frequency offset that needs to be adjusted; and adjusting the frequency offset in the frequency offset interval.

[0016] Optionally, the step of determining the frequency deviation interval that needs to be adjusted in the frequency deviation includes: determining the frequency deviation interval that needs to be adjusted in the frequency deviation by performing edge detection on the frequency deviation.

[0017] Optionally, the step of determining a frequency deviation interval that needs to be adjusted in the frequency deviation further includes: performing filtering and / or sliding averaging on the frequency deviation before the edge detection.

[0018] Optionally, by performing edge detection on the frequency deviation, the step of determining the frequency deviation interval that needs to be adjusted in the frequency deviation includes: calculating the slope of the frequency deviation; determining the positive jump edge and the negative jump edge of the frequency deviation according to the slope of the frequency deviation; and determining the frequency deviation between adjacent positive jump edges and negative jump edges as the frequency deviation interval that needs to be adjusted.

[0019] Optionally, the step of adjusting the frequency offset of the frequency offset interval includes: reducing the positive frequency offset in the frequency offset interval by a first predetermined value; and / or increasing the negative frequency offset in the frequency offset interval by a first predetermined value.

[0020] According to a second aspect of an embodiment of the present application, a network node is provided, comprising: a transceiver for sending and receiving signals; and a processor coupled to the transceiver and configured to execute the method performed by the network node as described above.

[0021] According to a third aspect of an embodiment of the present application, a computer-readable storage medium storing instructions is provided, wherein, when the instructions are executed by at least one processor, the at least one processor is prompted to execute the method performed by the network node as described above.

[0022] The beneficial effects brought about by the technical solutions provided by the embodiments of the present application will be described below in conjunction with specific optional embodiments, or can be learned from the description of the embodiments, or can be known through the implementation of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly and easily illustrate and understand the technical solutions in the embodiments of the present application, the following is a brief introduction to the drawings required for use in describing the embodiments of the present application.

[0024] Figure 1 is a diagram showing the positions of DMRS symbols.

[0025] Figure 2 is a schematic diagram showing an HST cell.

[0026] Figure 3is a schematic diagram showing a typical Doppler frequency offset (FO) curve of a high-speed mobile channel.

[0027] Figure 4 is a schematic diagram showing a frequency deviation curve based on DMRS symbol estimation.

[0028] Figure 5 is a schematic diagram showing a Doppler frequency deviation curve of a non-high-speed mobile channel (ie, a normal channel).

[0029] Fig. 6A is a flow chart illustrating a method performed by a network node according to an exemplary embodiment of the present application.

[0030] Figure 6B is a schematic diagram illustrating a channel estimation process according to an exemplary embodiment of the present application.

[0031] Fig. 7A is a flowchart illustrating a process of determining a channel type of a channel through a neural network based on a decorrelated signal according to an exemplary embodiment of the present application.

[0032] Figure 7B is a schematic diagram illustrating a feature extraction process according to an exemplary embodiment of the present application.

[0033] Figure 8 is a schematic diagram illustrating a frequency domain adaptation process according to an exemplary embodiment of the present application.

[0034] Fig. 9 2 is a diagram illustrating time domain impulse responses and characteristics of a high-speed mobile channel and a non-high-speed mobile channel according to an exemplary embodiment of the present application.

[0035] Fig.10 is a schematic diagram illustrating a channel type determiner based on a support vector machine according to an exemplary embodiment of the present application.

[0036] Fig.11A is a flow chart illustrating a process of adjusting frequency offset according to an exemplary embodiment of the present application.

[0037] Fig. 11B is a schematic diagram illustrating a process of adjusting frequency offset (ie, adaptive adjustment) according to an exemplary embodiment of the present application.

[0038] Fig.12 is a schematic diagram showing an estimated frequency deviation curve under a high-speed mobile channel.

[0039] Fig.13 is a schematic diagram showing a curve of frequency deviation in which noise is suppressed according to an exemplary embodiment of the present application.

[0040] Fig.14 is a schematic diagram showing a slope curve of a frequency deviation curve according to an exemplary embodiment of the present application.

[0041] Fig.15 is a schematic diagram showing a Marker curve according to an exemplary embodiment of the present application.

[0042] Fig.16 is a schematic diagram showing the division of a frequency deviation curve according to an exemplary embodiment of the present application.

[0043] Fig.17 It is a schematic diagram showing the process, result and comparison of adaptively adjusting a frequency deviation curve according to an exemplary embodiment of the present application.

[0044] Fig.18 is a diagram of a system illustrating a method performed by a network node to which an exemplary embodiment of the present application is applied.

[0045] Fig.19 is a block diagram illustrating a network node according to an exemplary embodiment of the present application.

[0046] Fig. 20 A schematic structural diagram of an electronic device applicable to an exemplary embodiment of the present application is shown in FIG. DETAILED DESCRIPTION

[0047] The following description with reference to the accompanying drawings is provided to facilitate a comprehensive understanding of the various embodiments of the present disclosure as defined by the claims and their equivalents. This description includes various specific details to facilitate understanding but should be considered as exemplary only. Therefore, one of ordinary skill in the art will recognize that various changes and modifications can be made to the various embodiments described herein without departing from the scope and spirit of the present disclosure. In addition, for the sake of clarity and conciseness, descriptions of well-known functions and structures may be omitted.

[0048] The terms and expressions used in the following specification and claims are not limited to their dictionary meanings, but are merely used by the inventor to enable a clear and consistent understanding of the present disclosure. Therefore, it should be apparent to those skilled in the art that the following description of various embodiments of the present disclosure is provided for illustration purposes only and not for the purpose of limiting the present disclosure as defined in the appended claims and their equivalents.

[0049] It should be understood that the singular forms "a", "an", and "the" may also include plural references unless the context clearly indicates otherwise. Thus, for example, reference to "a component surface" includes reference to one or more such surfaces. When we refer to an element as being "connected" or "coupled" to another element, the one element may be directly connected or coupled to the other element, or the one element and the other element may establish a connection relationship through an intermediate element. In addition, "connected" or "coupled" as used herein may include wireless connection or wireless coupling.

[0050] The term "include" or "may include" refers to the presence of the corresponding disclosed functions, operations or components that can be used in various embodiments of the present disclosure, rather than limiting the presence of one or more additional functions, operations or features. In addition, the term "include" or "have" may be interpreted as indicating certain characteristics, numbers, steps, operations, constituent elements, components or combinations thereof, but should not be interpreted as excluding the possibility of the presence of one or more other characteristics, numbers, steps, operations, constituent elements, components or combinations thereof.

[0051] The term "or" used in various embodiments of the present disclosure includes any of the listed terms and all combinations thereof. For example, "A or B" may include A, may include B, or may include both A and B. When describing multiple (two or more) items, if the relationship between the multiple items is not clearly defined, the multiple items may refer to one, multiple, or all of the multiple items. For example, the description of "parameter A includes A1, A2, A3" may be implemented as parameter A including A1 or A2 or A3, or may be implemented as parameter A including at least two of the three items A1, A2, and A3.

[0052] Unless defined differently, all terms (including technical terms or scientific terms) used in the present disclosure have the same meanings as understood by those skilled in the art described in the present disclosure. Common terms as defined in dictionaries are interpreted as having meanings consistent with the context in the relevant technical field, and should not be interpreted ideally or overly formally unless clearly defined in the present disclosure.

[0053] At least some functions of the device or electronic device provided in the embodiments of the present disclosure can be implemented by an AI model, such as at least one module among multiple modules of the device or electronic device can be implemented by an AI model. Functions associated with AI can be performed by non-volatile memory, volatile memory and processor.

[0054] The processor may include one or more processors. In this case, the one or more processors may be general-purpose processors, such as a central processing unit (CPU), an application processor (AP), etc., or pure graphics processing units, such as a graphics processing unit (GPU), a visual processing unit (VPU), and / or an AI-specific processor, such as a neural processing unit (NPU).

[0055] The one or more processors control the processing of input data according to predefined operating rules or artificial intelligence (AI) models stored in non-volatile memory and volatile memory. The predefined operating rules or artificial intelligence models are provided by training or learning.

[0056] Here, providing by learning means obtaining a predefined operating rule or an AI model with desired characteristics by applying a learning algorithm to a plurality of learning data. The learning can be performed in the device or electronic device itself in which the AI ​​according to the embodiment is executed, and / or can be implemented by a separate server / system.

[0057] The AI ​​model may include multiple neural network layers. Each layer has multiple weight values, and each layer performs neural network calculations by calculating between the input data of the layer (such as the calculation results of the previous layer and / or the input data of the AI ​​model) and the multiple weight values ​​of the current layer. Examples of neural networks include, but are not limited to, convolutional neural networks (CNNs), deep neural networks (DNNs), recurrent neural networks (RNNs), restricted Boltzmann machines (RBMs), deep belief networks (DBNs), bidirectional recurrent deep neural networks (BRDNNs), generative adversarial networks (GANs), and deep Q networks.

[0058] A learning algorithm is a method of using a plurality of learning data to train a predetermined target device (e.g., a robot) to enable, allow, or control the target device to make a determination or prediction. Examples of the learning algorithm include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0059] According to the present disclosure, at least one step in the method performed by the network node, such as determining the channel type, can be implemented using an artificial intelligence model. The processor of the electronic device can perform a preprocessing operation on the data to convert it into a form suitable for use as an input to the artificial intelligence model. The artificial intelligence model can be obtained through training. Here, "obtained through training" means that a predefined operating rule or artificial intelligence model configured to perform a desired feature (or purpose) is obtained by training a basic artificial intelligence model with multiple training data through a training algorithm.

[0060] Different from the architecture of the traditional Radio Access Network (RAN), in the architecture of the Open Radio Access Network (ORAN), the Cyclic Prefix (CP) removal module is performed on the O-RAN Radio Unit (O-RU) side. Therefore, the data on the O-RAN Distributed Unit (O-DU) side does not contain the CP part, so channel estimation can only be achieved based on the Demodulation Reference Signal (DMRS) symbol. However, when the number and position of DMRS symbols are fixed, the frequency offset range estimated based on DMRS symbols is limited. For example, in the Long Term Evolution (LTE) system, the position of the DMRS symbols is the 3rd symbol and the 10th symbol, such as Figure 1 As shown, according to the characteristics of the frame structure, the range of frequency offset estimation is [-1000Hz, 1000Hz].

[0061] Considering the relative position of the user equipment to the HST and the relative distance to the base station, the channels used by the user equipment in the HST cell can be divided into high-speed mobile channels and non-high-speed mobile channels (also called ordinary channels). Figure 2 As shown in , the HST cell includes both user equipment located on the HST train using high-speed mobile channels and user equipment that moves slowly or remains stationary near the base station and uses non-high-speed mobile channels.

[0062] Figure 3 The figure shows a typical Doppler frequency offset (FO) curve of a high-speed mobile channel, where the frequency offset corresponding to the blue frame part Case 1 exceeds 1000Hz, which corresponds to the situation where the user equipment is located on a high-speed train and the user equipment is far away from the base station. The channel used by the user equipment is a high-speed mobile channel, and the channel has a large frequency offset at this time. If the frequency offset corresponding to the blue frame part Case 1 is estimated using DMRS symbols, the following is obtained: Figure 4 The frequency deviation represented by the blue curve in the figure is between -600Hz and -1000Hz and between 600Hz and 1000Hz, which is obviously different from Figure 3 Compared with the frequency deviation curve shown in Figure 1, there is a large error. In other words, the frequency deviation range of the high-speed mobile channel exceeds the maximum frequency deviation range that can be estimated by the existing channel estimation method. The wrong frequency deviation will lead to reduced cell capacity, high user access failure rate and poor experience. Figure 3In the figure, the frequency deviation corresponding to the red frame Case 2 is less than 1000Hz, which corresponds to the situation where the user equipment is located on a high-speed train and the user equipment is near the base station. Although the channel used by the user equipment is a high-speed mobile channel, the channel has a small frequency deviation at this time. If the frequency deviation corresponding to the red frame Case 2 is estimated using DMRS symbols, the following is obtained: Figure 4 The frequency deviation represented by the red curve in the figure is between -1000Hz and +1000Hz. Figure 3 The red frame part in Case 2 corresponds to the same frequency deviation.

[0063] Figure 5 The Doppler frequency deviation curve of a non-high-speed mobile channel (i.e., a normal channel) is shown, where the frequency deviation corresponding to Case 3 in the red frame is less than 1000 Hz, which corresponds to the case where the user equipment moves slowly or is stationary. If the DMRS symbol is used to estimate this, the frequency deviation curve can be accurately obtained.

[0064] Based on this, the present application proposes a method executed by a network node. When only DMRS symbols can be used for channel estimation, an AI model is used to determine whether the channel used by the user equipment is a high-speed mobile channel or a normal channel. When the channel used by the user equipment is a high-speed mobile channel, the frequency deviation interval that needs to be adjusted in the frequency deviation of the channel is determined, and the frequency deviation is adjusted for the frequency deviation interval to obtain the adjusted accurate frequency deviation, and the adjusted accurate frequency deviation is used to perform frequency deviation compensation to obtain the final channel estimation value, thereby improving the cell throughput; when the channel used by the user equipment is a normal channel, the frequency deviation of the channel is directly used to perform frequency deviation compensation to obtain the final channel estimation value.

[0065] The following describes several optional embodiments to illustrate the technical solutions of the embodiments of the present disclosure and the technical effects produced by the technical solutions of the present disclosure. It should be noted that the following embodiments can refer to, draw on or combine with each other, and the same terms, similar features and similar implementation steps in different embodiments will not be described repeatedly.

[0066] Fig. 6A is a flow chart illustrating a method performed by a network node according to an exemplary embodiment of the present application. Figure 6Bis a schematic diagram showing a channel estimation process according to an exemplary embodiment of the present application. The network node may be a base station, but the present application is not limited thereto. The network node may also be a network server, which may receive various information from the base station, and determine the first channel estimation value of the channel used by the user equipment according to the various information received, determine the channel type of the channel, adjust the frequency offset of the channel, and obtain the second channel estimation value based on the adjusted frequency offset and the first channel estimation value, and then send the obtained second channel estimation value to the base station. The network node may be composed of one entity or multiple sub-entities. When the network node is composed of multiple sub-entities, different functions may be performed by multiple sub-entities respectively, each sub-entity may have a corresponding name, and the connection between the sub-entities may be a wired connection or a wireless connection, which is not specifically limited in the present application. For ease of description, in the following description, the network node is taken as an example of a base station.

[0067] like Fig. 6A As shown in FIG. 5 , in step S610 , frequency domain channel estimation is performed based on a decorrelated signal of a demodulation reference signal DMRS symbol received from a user equipment to obtain a first channel estimation value of a channel used by the user equipment.

[0068] Specifically, the decorrelated signal is obtained by decorrelating the demodulation reference signal DMRS symbol received from the user equipment. Figure 6B As shown in FIG. 1 , the base station inputs a signal received from the user equipment into the DMRS decorrelator, wherein the signal may be received from the user equipment through a high-speed mobile channel and has a signal as shown in FIG. Figure 3 The signal with the frequency deviation shown in FIG. 1 may also be a signal with the frequency deviation shown in FIG. 1 , which is received from the user equipment through a normal channel. Figure 5 Thereafter, the base station performs conjugate multiplication on the DMRS symbol in the received signal and the DMRS symbol generated locally by the base station, thereby obtaining a decorrelated signal. For example, decorrelation may be performed according to the following equation (1):

[0069] H Decorr =q*conj(dmrs seq ) (1)

[0070] Where q represents the DMRS symbol in the received signal, DMRS seq Indicates the DMRS symbol generated locally by the base station, H Decorr represents the decorrelated signal (i.e., the result of decorrelation), and conj represents the conjugate operation.

[0071] After obtaining the relevant signal, such as Figure 6BAs shown in , first, the frequency domain channel estimator may perform time offset estimation based on the decorrelated signal to obtain the time offset of the channel used by the user equipment. For example, the frequency domain channel estimator may use the decorrelated signal H obtained according to equation (1) above according to the following equation (2): Decorr Perform time offset estimation to obtain the channel time offset θ (also known as the average channel time offset):

[0072]

[0073] Among them, arctan means calculating the phase angle corresponding to the complex number by inverse tangent operation, k means the number of the subcarrier, l means the position of a DMRS symbol, N RB Indicates the number of allocated resource blocks (RBs).

[0074] After obtaining the time offset of the channel, the frequency domain channel estimator may perform time offset compensation on the decorrelated signal based on the obtained time offset to obtain a first channel estimation value of the channel. For example, the frequency domain channel estimator may perform time offset compensation on the decorrelated signal H based on the time offset θ obtained according to equation (2) above. Decorr Perform time offset compensation to obtain the channel response of frequency domain channel estimation, that is, the first channel estimation value H of the channel FD .

[0075] In step S620, the channel type of the channel used by the user equipment is determined.

[0076] In an exemplary embodiment of the present application, step S620 may include determining the channel type of the channel by a neural network based on the decorrelated signal. Fig. 7A This is described in detail.

[0077] Fig. 7A is a flowchart illustrating a process of determining a channel type of a channel through a neural network based on a decorrelated signal according to an exemplary embodiment of the present application.

[0078] In step S710, at least one of a phase characteristic, an amplitude characteristic, a radial quantity characteristic of the channel, and a signal-to-noise ratio of the channel is obtained according to the decorrelated signal.

[0079] A signal-to-noise ratio of the channel is determined according to the decorrelated signal.

[0080] In an exemplary embodiment of the present application, in the process of performing frequency domain channel estimation based on the decorrelated signal, Figure 6B The frequency domain channel estimator in determines the signal-to-noise ratio of the channel. For example, the frequency domain channel estimator can determine the signal-to-noise ratio s of the channel according to the following equation (3):

[0081] s=Power / I (3)

[0082] Wherein, Power represents the useful signal power measured by the base station for the channel used by the user equipment, and I represents the noise power measured by the base station for the channel.

[0083] The first signal matching the dimension of the neural network can be obtained by matching the dimension of the decorrelation signal with the dimension of the frequency domain resources of the neural network.

[0084] Specifically, since the schedulable frequency domain resources in the communication system are dynamically allocated, the dimension of the decorrelation signal will change accordingly. The present application matches the dimension of the decorrelation signal with the dimension of the frequency domain resources of the neural network so that the dimension of the decorrelation signal is consistent with the dimension of the frequency domain resources of the neural network, such as Figure 6B As shown in , the channel type determiner can perform a frequency domain adaptation operation between the dimension of the decorrelated signal and the frequency domain resource dimension of the neural network, thereby reducing the complexity of the neural network and facilitating the input to the neural network.

[0085] In an exemplary embodiment of the present application, the frequency domain resource dimension of the neural network may be set to a predetermined value K, which may be determined based on statistical data, for example, may be set to 10, but the present application is not limited thereto, and the predetermined value K may be set accordingly based on statistical data according to different neural networks. In this case, the first signal matching the frequency domain resource dimension K of the neural network may be determined by comparing the dimension of the decorrelated signal with the frequency domain resource dimension K of the neural network, wherein the first signal may also be referred to as a frequency domain adaptation signal.

[0086] In an exemplary embodiment of the present application, if the dimension of the decorrelation signal is equal to the frequency domain resource dimension K of the neural network, the decorrelation signal is determined as a first signal H that matches the frequency domain resource dimension K of the neural network, and the first signal H can be subsequently input into the neural network for determining the channel type of the channel used by the user equipment.

[0087] If the dimension of the decorrelated signal is smaller than the frequency domain resource dimension K of the neural network, the dimension of the decorrelated signal is expanded to the frequency domain resource dimension K of the neural network based on the time offset of the channel determined in step S610, thereby obtaining a first signal H that matches the frequency domain resource dimension K of the neural network.

[0088] Since each symbol in the decorrelated signal corresponds to a resource element (RE) in the frequency domain, that is, each RE contains a symbol in the decorrelated signal, the dimension of the decorrelated signal can be expanded by expanding the allocated REs corresponding to the decorrelated signal. Figure 8As shown in (a), according to the time offset θ, the allocated RE corresponding to the decorrelation signal is expanded to the frequency domain resource dimension K of the neural network. In other words, the allocated RE corresponding to the decorrelation signal is interpolated according to the time offset, and accordingly, the decorrelation signal after dimension expansion, i.e., the first signal H, can be obtained.

[0089] For example, the extended frequency domain resources can be expressed as:

[0090] x nm =x m *e jd*θ (4)

[0091] Among them, x m represents the mth RE allocated to the user equipment; x nm represents the mth RE in the nth extended RB; d represents x m to x nm The frequency domain distance; θ represents the time offset of the frequency domain channel estimation, that is, the time offset obtained by equation (2) above. In addition, Figure 8 In (a), d1 represents x 1 to x n1 The frequency domain distance.

[0092] For example, assuming that the number of REs in the frequency domain (ie, REs corresponding to decorrelated signals) allocated to the user equipment is 12 REs (ie, 1 RB), that is, If the frequency domain resource dimension K of the neural network is 24 REs (i.e., 2 RBs), it is necessary to expand the REs in the frequency domain allocated to the user equipment, and the expanded frequency domain resources can be expressed as:

[0093]

[0094] Since the frequency domain resources (ie, the allocated REs) are extended and corresponding symbols exist on the extended REs, an extended decorrelated signal, ie, the first signal H, can be obtained accordingly.

[0095] In addition, if the dimension of the decorrelated signal is greater than the frequency domain resource dimension K of the neural network, a portion of the signal is intercepted from the decorrelated signal based on the frequency domain resource dimension of the neural network as the first signal H. Due to the correspondence between the symbols in the decorrelated signal and the RE, the dimension of the decorrelated signal can be reduced by intercepting the RE with the same frequency domain resource dimension K of the neural network from the allocated RE corresponding to the decorrelated signal. In an exemplary embodiment of the present application, Figure 8As shown in (b), REs of the same frequency domain resource dimension K as the neural network, for example, the middle part, can be cut out from the allocated REs corresponding to the decorrelated signal, and accordingly, multiple symbols corresponding to the middle part of the RE in the decorrelated signal can be obtained as the first signal H.

[0096] For example, assuming that the number of REs allocated to the user equipment is 36 REs (ie, 3 RBs), that is, If the frequency domain resource dimension K of the neural network is 12 REs (i.e., 1 RB), it is necessary to reduce the REs allocated to the user equipment, that is, to intercept the REs in the middle part. At this time, the intercepted REs can be expressed as:

[0097]

[0098] Since the frequency domain resources (ie, the allocated REs) are reduced and corresponding symbols exist on the reduced REs, a reduced decorrelated signal, ie, the first signal H, can be obtained accordingly.

[0099] In summary, through the above process, a first signal H matching the frequency domain resource dimension K of the neural network can be obtained.

[0100] Thereafter, at least one of the phase feature and the amplitude feature of the first signal on each RE and the radial quantity feature of the channel is extracted according to the first signal. Figure 6B As shown in , in the channel type determiner, a feature extraction operation is performed on the first signal H obtained by the frequency domain adaptation operation to obtain at least one of a phase feature, an amplitude feature, and a radial quantity feature.

[0101] Specifically, the step of extracting the channel's path quantity characteristics according to the first signal includes: determining the amplitude of the channel's time domain impulse response according to the first signal; and determining the channel's path quantity characteristics according to the number of times the amplitude of the time domain impulse response exceeds a first predetermined threshold.

[0102] In an exemplary embodiment of the present application, the amplitude of the time domain impulse response of the channel can be determined according to the following equation (7), where IDFT represents Inverse Discrete Fourier Transform, || represents a modulo operation, n represents the number of Fourier transform points, and the sequence value of h(n) represents the amplitude of the time domain impulse response of the channel.

[0103] h(n) = |IDFT(H,n)| (7)

[0104] In addition, if Fig. 9As shown in, for non-high-speed mobile channels (i.e., ordinary channels), usually due to the abundance of scatterers in the cell and the existence of multipath in the propagation path, the number of time domain impulse response amplitudes exceeding the first predetermined threshold is relatively large, that is, h(n) shows multiple peaks. However, for high-speed mobile channels, since the base station is set up around, for example, railway tracks, the signal propagation environment is open, and the base station and the user equipment are usually strong line of sight (LOS) path channels, the number of time domain impulse response amplitudes exceeding the first predetermined threshold is relatively small, that is, h(n) shows a single peak. Therefore, if the amplitude of a time domain impulse response exceeds the first predetermined threshold, it can be determined that there is a valid path. By determining the number of times the amplitude of the time domain impulse response h(n) exceeds the first predetermined threshold, the path quantity characteristics of the channel can be determined. For example, the path quantity characteristics of the channel can be determined according to the following equation (8), for example, Figure 7B As shown in , wherein sum represents a summation operation, TH represents a first predetermined threshold value, and the first predetermined threshold value is configurable.

[0105] path=sum(h(n) >TH) (8)

[0106] In an exemplary embodiment of the present application, the amplitude feature A and the phase feature of the first signal on each corresponding RE can be extracted according to the following equations (9) and (10):

[0107] A=|H| (9)

[0108]

[0109] Among them, arctan represents the inverse tangent operation, for example Figure 7B as shown in .

[0110] For example, suppose that the first signal matching the frequency domain resource dimension K of the neural network is obtained through the previous matching operation In this case, after performing an inverse Fourier transform IDFT on H using equation (7), according to equation (8) above, the path quantity feature Path can be expressed as the number of modulus values ​​after IDFT greater than the first predetermined threshold TH, and according to equation (9) above, the amplitude feature A can be expressed as Phase characteristics It can be expressed as

[0111] Through the above process, the phase characteristics and amplitude characteristics of the first signal on each RE and the path quantity characteristics of the channel can be extracted, such as Figure 7B as shown in .

[0112] Return to reference Fig. 7AIn step S720, based on at least one of the phase characteristic, the amplitude characteristic, the channel diameter quantity characteristic, and the signal-to-noise ratio of the channel, the channel type of the channel is determined through a neural network.

[0113] In an exemplary embodiment of the present application, when using a neural network to determine the channel type of a channel, in addition to inputting at least one feature obtained in step S710 into the neural network, the first signal obtained by the matching operation described above is also input into the neural network, and then the neural network predicts the channel type of the channel based on the at least one input feature and the first signal.

[0114] Specifically, if Figure 7B As shown in , the phase features extracted by the feature extraction operation At least one of the amplitude feature A and the path quantity feature path, and the noise ratio s of the channel obtained by the frequency domain channel estimator and the first signal H are input into the neural network for prediction, so that the channel type of the channel can be predicted. For example, when the output of the neural network is 1, it can be set to indicate that the channel is a high-speed mobile channel, and when the output of the neural network is 0, it can be set to indicate that the channel is a non-high-speed mobile channel (i.e., a normal channel). In the present application, the neural network can be a support vector machine (Support Vector Machine, SVM), a convolutional neural network (Convolutional Neural Networks, CNN), a recurrent neural network (Recurrent Neural Network, RNN), a long short-term memory network (Long Short-Term Memory, LSTM), a multilayer perceptron mixer (Multilayer Perceptron Mixer, MLP-Mixer), a Transformer network, etc., but the present application is not limited thereto. In the present application, the signal-to-noise ratio s is introduced as an indicator of the parameters of the neural network. According to the change of the signal-to-noise ratio s, the parameters of the neural network can be dynamically adjusted, thereby enhancing the robustness to the noise environment.

[0115] For example, the process of determining the channel type of a channel is described by taking a neural network as a support vector machine. Fig.10As shown, first, the phase feature, amplitude feature, channel diameter quantity feature, channel signal-to-noise ratio and the first signal H obtained by the above matching operation are normalized, and the above normalized data are input into the embedding layer. Then, the embedding layer fuses the input data and inputs the fusion result into the support vector machine. After that, the support vector machine determines the channel type of the channel according to the received fusion result. For example, when the output of the support vector machine is 1, it indicates that the channel type of the channel is a high-speed mobile channel. When the output of the support vector machine is 0, it indicates that the channel type of the channel is a non-high-speed mobile channel, that is, a normal channel.

[0116] In addition, the support vector machine used above can be a support vector machine obtained through offline training, such as Fig.10 As shown in , the support vector machine may be provided with a separation hyperplane and a classification decision function, wherein the kernel function in the classification decision function may adopt a Gaussian kernel function as shown in the following equation (11):

[0117]

[0118] Where x' and z represent two input data points; ||x'-z|| represents the distance between the two data points x' and z in Gaussian space; σ represents the parameter of the Gaussian kernel function, which is used to control the density of the Gaussian distribution. In addition, when training the support vector machine, the cross entropy function of the channel type shown in the following equation (12) is used as the loss function:

[0119]

[0120] in, represents the probability that the predicted sample of the neural network (i.e., support vector machine) is a positive example, and y represents the sample label. According to the loss function indicated by the above equation (12), after multiple rounds of iterative training, the trained neural network can be obtained.

[0121] Return to reference Fig. 6A In step S630, it is determined whether the channel type of the channel is a high-speed mobile channel. When the channel type of the channel is a high-speed mobile channel, the process proceeds to step S640.

[0122] In step S640, the frequency offset of the channel is adjusted, that is, the following is performed: Figure 6B The frequency deviation adaptive adjustment shown in .

[0123] like Figure 6B As shown in , before executing step S640, it is necessary to first perform frequency offset estimation, that is, estimate the frequency offset of the channel.

[0124] Specifically, the first channel estimation value H obtained by frequency domain channel estimation can be calculated according to the following equation (13):FD The data at the two DMRS symbol positions are conjugate multiplied and accumulated to obtain the accumulated value FO est :

[0125]

[0126] Wherein, l1 and l2 represent the positions of two DMRS symbols. For example, for the example shown in FIG. 1 , l1 is 3 and l2 is 10; N RB , k and conj have the same meanings as those in equation (2) above.

[0127] Then, the accumulated value FO est Find phase On this basis, the channel frequency offset fo is estimated according to the following equation (14):

[0128]

[0129] Wherein, Δt represents the time interval between two DMRS symbols.

[0130] In the present application, the step of estimating the frequency offset of the channel may be performed after step S630 and before step S640, or may be performed before step S630, and the present application does not make any specific limitation on this. The process of adjusting the frequency offset of the channel is described in detail below.

[0131] Specifically, the step of adjusting the frequency offset may include: determining a frequency offset interval that needs to be adjusted in the frequency offset; and adjusting the frequency offset in the frequency offset interval. Fig.11A and Fig. 11B The process of determining the frequency deviation interval that needs to be adjusted in the frequency deviation is specifically described.

[0132] Fig.11A is a flow chart illustrating a process of adjusting frequency offset according to an exemplary embodiment of the present application. Fig. 11B is a schematic diagram illustrating a process of adjusting frequency offset (ie, adaptive adjustment) according to an exemplary embodiment of the present application.

[0133] In step S1110, filtering and / or sliding averaging are performed on the frequency offset to obtain a frequency offset with suppressed noise.

[0134] like Fig. 11B As shown in , when the channel type of the channel used by the user equipment is a high-speed mobile channel, after the frequency offset estimation is performed, a noise suppression operation can be performed. Specifically, due to the influence of noise, the frequency offset curve estimated before step S640 has many burrs, and there are multiple oscillations at the frequency point position of ±1000 Hz, such as Fig.12 As shown, Fig.12The vertical axis represents the frequency offset estimated based on the two DMRSs. In order to reduce the influence of noise, the frequency offset may be first filtered by the following equation (15) to obtain a filtered frequency offset.

[0135] Filter_fo(t)=(1-α)*fo(t)+α*fo(t-1) (15)

[0136] Among them, α represents the filter parameter, fo(t) represents the estimated frequency deviation at the current time t, fo(t-1) represents the estimated frequency deviation at the previous time t-1, and Filter_fo(t) represents the filtered frequency deviation at the current time t. Fig.13 (a) shows the frequency deviation after filtering. Then, a sliding average is performed on the filtered frequency deviation according to the following equation (16) to obtain a frequency deviation with suppressed noise.

[0137]

[0138] Among them, N win is the window length of the sliding average, and MA_fo(t) represents the frequency deviation at which the noise at the current time t is suppressed. Fig.13 (b) shows the frequency deviation after sliding average, that is, the frequency deviation with noise suppressed.

[0139] For example, for HST scenario 3, the maximum Doppler frequency deviation is 1340 Hz. In this case, the filter parameter α in equation (15) can be set to, for example, 0.8, and the sliding average window length N can be set to win If t is set to 10, the filtered frequency deviation Filter_fo(t) at the current time t is expressed as the following equation (17). Accordingly, the frequency deviation with suppressed noise can be expressed as the following equation (18):

[0140] Filter_fo(t)=0.2*fo(t)+0.8*fo(t-1) (17)

[0141]

[0142] In step S1120, edge detection is performed on the frequency offset to determine a frequency offset interval that needs to be adjusted.

[0143] like Fig. 11B As shown in , after the noise suppression operation is performed, the edge detection operation is performed. Specifically, first, the slope of the frequency deviation in which the noise is suppressed is calculated, that is, the slope of the curve of the frequency deviation in which the noise is suppressed is calculated, for example, the slope slop(t) of the curve of the frequency deviation is calculated according to the following equation (19):

[0144]

[0145] Here, Δt1 is a fixed value, for example, may be set to 1, but the present application is not limited thereto. Fig.14 is a diagram illustrating a slope curve of a curve of frequency deviation according to an exemplary embodiment of the present application.

[0146] Then, the positive transition edge and the negative transition edge of the frequency deviation are determined according to the slope of the frequency deviation.

[0147] Specifically, from Fig.14 It can be seen that the slope curve is periodic, and each period contains two positive pulses and two negative pulses, where the positive pulse represents the positive transition edge of the frequency deviation curve, and the negative pulse represents the negative transition edge of the frequency deviation curve. The following is a detailed description of the process of determining the positive transition edge and the negative transition edge. Set the initial value of the marker Marker to 0. If the following conditions (1) and (2) are met, it is considered that the first positive transition edge is detected:

[0148] (1) The Marker at the previous time t-1 is 0, that is, Marker(t-1)=0;

[0149] (2) The absolute value slop(t) of the slope of the frequency deviation curve at the current time t is greater than the second predetermined threshold slopt_TH, that is, slop(t)>slop_TH, where slopt_TH can be set to 200, for example, but the present application is not limited thereto.

[0150] When the first positive transition edge is detected (for example, when Fig.14 When the first positive transition edge is detected at t0 in the figure, the Marker at the current time t is set to 1. Then the frequency deviation curve slope is detected to be less than 0. When the frequency deviation curve slope is detected to be less than 0 (for example, when Fig.14 When the slope of the frequency deviation curve is less than 0 at t1 in the figure, the second positive transition edge of the frequency deviation curve is detected. Before the second positive transition edge is detected, the value of Marker remains at 1. If the following conditions (3) and (4) are met, the second positive transition edge is considered to be detected:

[0151] (3) The Marker at the previous time t-1 is 1, that is, Marker(t-1)=1;

[0152] (4) The absolute value slop(t) of the slope of the frequency deviation curve at the current time t is greater than the second predetermined threshold slopt_TH, that is, slop(t)>slop_TH.

[0153] When the second positive edge is detected (for example, when Fig.14When the second positive transition edge is detected at t2 in the above example, the Marker at the current time t is set to 0. Then, the first negative transition edge is detected. The detection method of the negative transition edge is the same as that of the positive transition edge. That is, if the above conditions (1) and (2) are met, it is considered that the first negative transition edge is detected.

[0154] When the first negative transition edge is detected (for example, when Fig.14 When the first negative transition edge is detected at t3 in the figure, the Marker of the current time t is set to 1. Then the frequency deviation curve slope is detected to be greater than 0. When the frequency deviation curve slope is greater than 0 (for example, when Fig.14 When the slope of the frequency deviation curve is greater than 0 at t4 in the figure, the second negative transition edge of the frequency deviation curve is detected. Before the second negative transition edge is detected, the value of Marker remains at 1. If the above conditions (3) and (4) are met, the second negative transition edge is considered to be detected (for example, at Fig.14 The second negative transition edge is detected at t5 in FIG.

[0155] Through the above process, we can know that the marker is 1 between two adjacent positive pulses and 1 between two adjacent negative pulses, and is 0 between adjacent positive pulses and negative pulses. Therefore, we can get the following Fig.15 That is, after the positive jump edge and the negative jump edge of the frequency deviation are determined, the frequency deviation between the adjacent positive jump edge and the negative jump edge is determined as the frequency deviation interval that needs to be adjusted.

[0156] In an exemplary embodiment of the present application, the frequency deviation curve is divided into three intervals according to the value of the Marker and whether the frequency deviation is positive or negative, and then the frequency deviation interval that needs to be adjusted is determined. Specifically, the frequency deviation curve can be divided into the following three intervals:

[0157] Interval 1: Marker = 0 and frequency deviation is negative;

[0158] Interval 2: Marker = 1;

[0159] Interval 3: Marker = 0 and the frequency deviation is positive.

[0160] That is to say, interval 1 and interval 3 are intervals between adjacent positive transition edges and negative transition edges, and interval 2 is an interval between two adjacent positive transition edges or an interval between two adjacent negative transition edges. Fig.16 As shown in Fig.16 The upper figure is the estimated frequency deviation curve within 10,000 transmission time intervals (Time to Interactive, TTI), and the lower figure is an enlarged view near the first jump edge. Fig.16It can be seen that the frequency offsets in intervals 1 and 3 are the frequency offset intervals that need to be adjusted (hereinafter referred to as "frequency offset intervals to be adjusted"), and the frequency offset in interval 2 does not need to be adjusted.

[0161] In step S1130, the frequency offset is adjusted for the frequency offset interval determined in step S1120.

[0162] like Fig. 11B As shown in , after the frequency offset interval to be adjusted is determined by performing the edge detection operation, the frequency offset adjustment operation is performed. Specifically, the step of adjusting the frequency offset of the frequency offset interval to be adjusted includes: reducing the positive frequency offset in the frequency offset interval to be adjusted by a first predetermined value; and increasing the negative frequency offset in the frequency offset interval to be adjusted by a first predetermined value.

[0163] For example, Fig.17 As shown in , for interval 1 in the frequency offset interval to be adjusted (i.e., a negative frequency offset interval in the frequency offset interval to be adjusted), for example, Fig.17 The marker curve in (b) indicates Fig.17 The blue area in (a) of FIG. 1 increases the frequency deviation in interval 1 by a first predetermined value (e.g., 2000 Hz), such as Fig.17 As shown in the blue area of ​​(c) in FIG. 1 ; for interval 3 in the interval to be adjusted (ie, the frequency offset interval that is a positive number in the frequency offset interval to be adjusted, for example, Fig.17 The marker curve in (b) indicates Fig.17 The orange area in (a) of FIG. 3 ) reduces the frequency deviation in the interval 3 by a first predetermined value (eg, 2000 Hz), such as Fig.17 As shown in the orange interval of (c) in FIG. 2 ; for interval 2, there is no need to adjust the frequency deviation therein, such as Fig.17 As shown in the red area of ​​(c). Fig.17 As shown in (c) of FIG. 1 , the adjusted frequency deviation curve can accurately reflect the following Fig.17 The ideal frequency deviation is shown in (d). Fig. 11B As shown in , the adjusted frequency offset is used to perform frequency offset compensation, which is described below.

[0164] Return to reference Fig. 6A After executing step S640, execute step S650 to obtain a second channel estimation value based on the frequency offset and the result of the frequency domain channel estimation.

[0165] like Figure 6B As shown in , frequency offset compensation is performed after frequency offset adaptive adjustment is performed. Specifically, a second channel estimation value is obtained based on the adjusted frequency offset and the first channel estimation value.

[0166] For example, the first channel estimation value H of the frequency domain channel estimation obtained in step S610 above can be calculated according to the following equation (20): FD Frequency offset compensation is performed to obtain a second channel estimation value h_foc.

[0167] h_foc=h*e -j*2π*fo*t (20)

[0168] Wherein, fo represents the frequency offset used for frequency offset compensation, and t represents the time interval of each OFDM symbol relative to the frame header.

[0169] In addition, in step S630, when it is determined that the channel type of the channel is not a high-speed mobile channel (ie, a normal channel), the process may proceed directly to step S650, for example, Fig. 11B As shown in , when the channel type is a normal channel, the second channel estimation value can be obtained based on the frequency offset obtained by frequency offset estimation and the first channel estimation value, that is, the first channel estimation value is frequency offset compensated using the frequency offset obtained by frequency offset estimation, that is, without adjusting the frequency offset determined according to the above equation (14), the first channel estimation value is frequency offset compensated based on the frequency offset using equation (20) to obtain the second channel estimation value.

[0170] Fig.18 is a diagram of a system (eg, a Physical Uplink Shared Channel (PUSCH) receiver) to which a method performed by a network node according to an exemplary embodiment of the present application is applied.

[0171] like Fig.18 As shown in , the base station receives a signal from a user equipment, and then the base station performs channel estimation based on the received signal. Specifically, the DMRS decorrelator decorrelates the DMRS symbols in the received signal to obtain a decorrelated signal, and then the frequency domain channel estimator performs frequency domain channel estimation based on the decorrelated signal to obtain a first channel estimation value of the channel used by the user equipment. After that, the channel type determiner determines the channel type of the channel based on the decorrelated signal received from the DMRS decorrelator and the time offset and signal-to-noise ratio of the channel received from the frequency domain channel estimator. Then, according to the channel type of the channel, the time domain channel estimator performs time domain channel estimation. Specifically, if the channel type of the channel is a normal channel, the time domain channel estimator directly uses the frequency offset of the channel to perform frequency offset compensation on the first channel estimation value to obtain a second channel estimation value as shown in the above step S650. If the channel type of the channel is a high-speed mobile channel, the time domain channel estimator first adjusts the frequency offset of the channel as shown in the above step S640, and then uses the adjusted frequency offset to perform frequency offset compensation on the first channel estimation value to obtain a second channel estimation value. Since the above reference Fig. 6AThe channel estimation process has been described in detail, so it will not be repeated here.

[0172] After the base station obtains the second channel estimation value, the base station performs equalization and demodulation, log likelihood ratio buffer and channel decoding according to the existing method.

[0173] The method proposed by the present application and executed by the network node can effectively improve the accuracy of Doppler compensation, thereby improving the accuracy of channel estimation and improving the system throughput; and the present application can effectively extract the propagation difference between non-high-speed mobile channels and high-speed mobile channels by extracting the path quantity feature as the main feature and inputting it into the neural network, and further using the phase feature, amplitude feature and signal-to-noise ratio feature of the signal as the auxiliary feature input into the neural network, which can help the neural network learn the characteristics of the data faster and better, improve the robustness of the neural network to different environments and different noise scenes, and further improve the system capacity. In addition, the method executed by the network node of the present application adopts the process of first determining the channel type (also known as classifying the channel) and then processing the signal according to the channel type, which reduces the number of retransmissions of the system and reduces the complexity of the system. In addition, the edge detection operation in the channel estimation method of the present application can accurately identify the changing trend of the frequency deviation under the high-speed mobile channel, distinguish different frequency deviation intervals and make corresponding frequency deviation adjustments, so as to obtain accurate frequency deviation for frequency deviation compensation, and finally obtain accurate channel estimation value.

[0174] Fig.19 is a block diagram illustrating a network node according to an exemplary embodiment of the present application.

[0175] like Fig.19 As shown in FIG. 1 , the network node 1900 includes a transceiver 1911 and a processor 1912, wherein the processor 1912 is coupled to the transceiver 1911 and is configured to execute the above reference FIG. 6A to FIG. 17 For details of the operation of the method performed by the network node described above, see FIG. 6A to FIG. 17 The description will not be repeated here.

[0176] An electronic device is also provided in an embodiment of the present disclosure, which includes at least one processor and, optionally, may also include at least one transceiver and / or at least one memory coupled to the at least one processor, and the at least one processor is configured to execute the steps of the method provided in any optional embodiment of the present disclosure.

[0177] Fig. 20 A schematic diagram of the structure of an electronic device applicable to an embodiment of the present invention is shown in FIG. Fig. 20 As shown, Fig. 20The electronic device 4000 shown includes: a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, such as through a bus 4002. Optionally, the electronic device 4000 may also include a transceiver 4004, which can be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, each of the processor 4001, the memory 4003 and the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present disclosure. Optionally, the electronic device may be a first network node, a second network node or a third network node.

[0178] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) 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 conjunction with the disclosure of the present invention. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0179] The bus 4002 may include a path to transmit information between the above components. The bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig. 20 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0180] The memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium that can be used to carry or store computer programs and can be read by a computer, without limitation herein.

[0181] The memory 4003 is used to store computer programs or executable instructions for executing the embodiments of the present disclosure, and the execution is controlled by the processor 4001. The processor 4001 is used to execute the computer programs or executable instructions stored in the memory 4003 to implement the steps shown in the above method embodiments.

[0182] An embodiment of the present disclosure provides a computer-readable storage medium having a computer program or instructions stored thereon. When the computer program or instructions are executed by at least one processor, the steps and corresponding contents of the aforementioned method embodiment can be executed or implemented.

[0183] The embodiments of the present disclosure also provide a computer program product, including a computer program, which can implement the steps and corresponding contents of the aforementioned method embodiments when executed by a processor.

[0184] The terms "first", "second", "third", "fourth", "1", "2", etc. (if any) in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than that shown or described in the text.

[0185] It should be understood that, although the flowchart of the embodiment of the present disclosure indicates each operation step by arrows, the implementation order of these steps is not limited to the order indicated by the arrows. Unless clearly stated herein, in some implementation scenarios of the embodiment of the present disclosure, the implementation steps in each flowchart can be executed in other orders according to demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages based on the actual implementation scenario. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage in these sub-steps or stages can also be executed at different times. In scenarios with different execution times, the execution order of these sub-steps or stages can be flexibly configured according to demand, and the embodiment of the present disclosure does not limit this.

[0186] The above text and drawings are provided only as examples to help readers understand the present disclosure. They are not intended and should not be interpreted as limiting the scope of the present disclosure in any way. Although certain embodiments and examples have been provided, based on the contents disclosed herein, it is obvious to those skilled in the art that the embodiments and examples shown can be changed without departing from the scope of the present disclosure, and other similar implementation means based on the technical ideas of the present disclosure are adopted, which also fall within the protection scope of the embodiments of the present disclosure.

Claims

1. A method performed by a network node, include: Perform frequency domain channel estimation based on a decorrelated signal of a demodulation reference signal DMRS symbol received from a user equipment to obtain a first channel estimation value of a channel used by the user equipment; determining a channel type of the channel; as well as When the channel type of the channel is a high-speed mobile channel, the frequency offset of the channel is adjusted, and a second channel estimation value is obtained based on the adjusted frequency offset and the first channel estimation value.

2. The method according to claim 1, in, The step of performing frequency domain channel estimation based on a decorrelated signal of a demodulation reference signal DMRS symbol received from a user equipment and obtaining a first channel estimation value of a channel used by the user equipment comprises: Performing time offset estimation based on the decorrelated signal to obtain a time offset on the channel; and Time offset compensation is performed on the decorrelated signal based on the time offset to obtain the first channel estimation value.

3. The method according to claim 1, in, The step of determining the channel type of the channel comprises: Based on the decorrelated signal, a channel type of the channel is determined by a neural network.

4. The method according to claim 3, in, The neural network is one of a support vector machine, a convolutional neural network, a recurrent neural network, a long short-term memory network, a Transformer network, and a multi-layer perceptron mixer.

5. The method according to claim 3 or 4, in, The step of determining the channel type of the channel by a neural network based on the decorrelated signal comprises: According to the decorrelated signal, at least one of a phase characteristic, an amplitude characteristic, a radial quantity characteristic of the channel, and a signal-to-noise ratio of the channel is obtained; Based on the at least one feature, a channel type of the channel is determined by a neural network.

6. The method according to claim 5, in, The step of obtaining at least one of a phase characteristic, an amplitude characteristic, a channel diameter quantity characteristic, and a signal-to-noise ratio of the channel according to the decorrelated signal comprises: determining a signal-to-noise ratio of the channel according to the decorrelated signal; and / or By matching the dimension of the decorrelated signal with the frequency domain resource dimension of the neural network, a first signal matching the frequency domain resource dimension of the neural network is obtained, and based on the first signal, the phase characteristic and amplitude characteristic of the first signal on each resource element and at least one of the path quantity characteristics of the channel are extracted.

7. The method according to claim 6, in, According to at least one of the features, the step of determining the channel type of the channel by a neural network comprises: According to the at least one feature and the first signal, a channel type of the channel is determined by a neural network.

8. The method according to claim 6, in, The step of obtaining a first signal matching the frequency domain resource dimension of the neural network by matching the dimension of the decorrelated signal with the frequency domain resource dimension of the neural network comprises: If the dimension of the decorrelation signal is equal to the frequency domain resource dimension of the neural network, determining the decorrelation signal as the first signal; If the dimension of the decorrelation signal is smaller than the frequency domain resource dimension of the neural network, expanding the dimension of the decorrelation signal to the frequency domain resource dimension of the neural network based on the time offset of the channel to obtain the first signal; If the dimension of the decorrelation signal is greater than the frequency domain resource dimension of the neural network, a portion of the signal is intercepted from the decorrelation signal based on the frequency domain resource dimension of the neural network as the first signal.

9. The method according to claim 6, in, The step of extracting the path quantity characteristics of the channel according to the first signal comprises: Determining a magnitude of a time domain impulse response of the channel according to the first signal; According to the number of times that the amplitude of the time domain impulse response exceeds a predetermined threshold, the path quantity feature of the channel is extracted.

10. The method according to claim 1, in, The step of adjusting the frequency offset of the channel includes: Determining a frequency deviation interval that needs to be adjusted in the frequency deviation; A frequency offset adjustment is performed on the frequency offset interval.

11. The method according to claim 10, in, The step of determining the frequency deviation interval that needs to be adjusted in the frequency deviation includes: By performing edge detection on the frequency deviation, a frequency deviation interval that needs to be adjusted in the frequency deviation is determined.

12. The method according to claim 11, in, The step of determining the frequency deviation interval that needs to be adjusted in the frequency deviation also includes: Before the edge detection, filtering and / or sliding averaging are performed on the frequency deviation.

13. The method according to claim 11 or 12, in, The step of determining a frequency deviation interval that needs to be adjusted in the frequency deviation by performing edge detection on the frequency deviation comprises: Calculating the slope of the frequency deviation; Determining a positive transition edge and a negative transition edge of the frequency deviation according to the slope of the frequency deviation; The frequency deviation between adjacent positive transition edges and negative transition edges is determined as the frequency deviation interval that needs to be adjusted.

14. The method according to claim 10, in, The step of adjusting the frequency offset in the frequency offset interval includes: reducing the positive frequency deviation in the frequency deviation interval by a first predetermined value; and / or The frequency deviation that is a negative number in the frequency deviation interval is increased by a first predetermined value.

15. A network node, include: transceiver, used to send and receive signals; as well as A processor is coupled to the transceiver and configured to execute the method according to any one of claims 1 to 14.

16. A computer-readable storage medium storing instructions, in, When the instructions are executed by at least one processor, the at least one processor is prompted to perform the method according to any one of claims 1 to 14.

Images