Pilot signal transmission method, channel estimation method, device and equipment

Abstract

The present application discloses a pilot signal sending method, a channel estimation method, an apparatus and a device, the method comprising: a first communication device sends a first pilot signal in a first matrix; wherein the first matrix is ​​an N×M matrix, the elements of the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and specifically relates to a pilot signal sending method, a channel estimation method, a device and equipment. Background Art

[0002] Using massive multiple-input multiple-output (MIMO) technology to form a large-scale antenna array can support more users to send and receive signals at the same time, thereby increasing the channel capacity and data traffic of the mobile network by dozens of times or more, and at the same time can achieve a sharp reduction in interference between multiple users. Therefore, it has been receiving continuous and high attention from researchers since it was proposed. In order to support broadband wireless communications, orthogonal frequency division multiplexing (OFDM) has become the underlying technology of mobile communications since the fourth generation of mobile communication technology (4G). It can effectively combat multipath interference and divide the frequency-domain frequency selective channel into multiple flat-fading sub-channels to support wireless transmission. OFDM combined with massive MIMO is already the basic framework for current and future wireless communications.

[0003] However, in a massive MIMO system, due to the increase in the number of antennas, the pilot overhead and the complexity of channel estimation have increased by orders of magnitude. How to reduce the pilot overhead is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The embodiments of the present application provide a pilot signal sending method, a channel estimation method, an apparatus and a device, which can solve the problem of how to reduce pilot overhead.

[0005] In a first aspect, a pilot signal sending method is provided, comprising:

[0006] The first communication device sends a first pilot signal in a first matrix;

[0007] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and N and M are both integers greater than 1.

[0008] In a second aspect, a channel estimation method is provided, characterized by comprising:

[0009] The second communication device receives the first pilot signal in the first matrix sent by the first communication device;

[0010] The second communication device performs channel estimation according to the first pilot signal;

[0011] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and N and M are both integers greater than 1.

[0012] According to a third aspect, a pilot signal sending device is provided, including:

[0013] A first sending module, used for sending a first pilot signal in a first matrix;

[0014] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and N and M are both integers greater than 1.

[0015] According to a fourth aspect, a channel estimation device is provided, comprising:

[0016] A first receiving module, configured to receive a first pilot signal in a first matrix sent by a first communication device;

[0017] A channel estimation module, configured for the second communication device to perform channel estimation according to the first pilot signal;

[0018] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and N and M are both integers greater than 1.

[0019] In a fifth aspect, a communication device is provided, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the method described in the first aspect or the second aspect.

[0020] In a sixth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect or the second aspect are implemented.

[0021] In the seventh aspect, a computer program / program product is provided, wherein the computer program / program product is stored in a non-volatile storage medium, and the computer program / program product is executed by at least one processor to implement the steps of the processing method as described in the first aspect or the second aspect.

[0022] In an eighth aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the processing method described in the first aspect or the second aspect.

[0023] In an embodiment of the present application, a first communication device sends a first pilot signal in a first matrix. Since the first matrix is ​​an N×M matrix, where N is the number of antennas of the first communication device and M is the number of REs that send pilot signals on each RB, the purpose of pilot overhead compression is achieved by setting M to be less than N, thereby reducing the pilot overhead while maintaining the channel estimation accuracy of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a schematic diagram of a wireless communication system to which the embodiments of the present application can be applied;

[0025] Figure 2 is a flowchart of a method for sending a pilot signal in an embodiment of the present application;

[0026] Figure 3 is a flow chart of a channel estimation method in an embodiment of the present application;

[0027] Figure 4 is a schematic diagram of a massive MIMO antenna system based on OFDM in an embodiment of the present application;

[0028] Figure 5 is a schematic diagram of the training of the perception matrix in an embodiment of the present application;

[0029] Figure 6 This is a schematic diagram of the LAMP network structure in an embodiment of the present application;

[0030] Figure 7It is a schematic diagram of the t-th layer structure of the LAMP network in an embodiment of the present application;.

[0031] Figure 8 is a schematic diagram of a pilot signal sending device in an embodiment of the present application;

[0032] Fig. 9 is a schematic diagram of a channel estimation device in an embodiment of the present application;

[0033] Fig.10 is a schematic diagram of a terminal in an embodiment of the present application;

[0034] Fig.11 is a schematic diagram of a network side device in an embodiment of the present application;

[0035] Fig.12 It is a schematic diagram of a communication device in an embodiment of the present application. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

[0037] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specified order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "and" in the specification and claims represents at least one of the connected objects, and the character " / " generally represents that the objects associated with each other are in an "or" relationship.

[0038] It is worth noting that the technology described in the embodiments of the present application is not limited to the New Radio (NR) system, Long Term Evolution (LTE) / LTE's evolution (LTE-Advanced, LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA) and other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used for the systems and radio technologies mentioned above as well as for other systems and radio technologies. The following description describes the NR system for illustrative purposes, and NR terminology is used in most of the following description, but these technologies can also be applied to applications other than NR system applications, such as the 6th generation (6 th Generation, 6G) communication system.

[0039] In massive MIMO systems, due to the huge number of antennas, the pilot overhead and feedback overhead required for channel estimation and feedback are huge. In order to reduce the pilot overhead and feedback overhead, a channel estimation scheme based on compressed sensing theory is proposed by utilizing the sparse characteristics of massive MIMO channels in the spatial angle domain.

[0040] This scheme uses the limited scattering characteristics of large-scale MIMO channels and the characteristics of closely arranged base station antenna arrays to analyze the structured sparse characteristics between channels in the spatial angle domain of different antennas. Using this characteristic, the large-scale MIMO channel estimation is transformed into a sparse signal recovery problem of structured compressed sensing. Specifically, a small number of non-orthogonal pilots sent by the base station are used to reduce the pilot overhead. The number of pilots is much smaller than the number of antennas. After the mobile user receives the pilot signal, it uses compressed sensing sparse signal recovery algorithms such as orthogonal matching pursuit (OMP- Orthogonal Matching Pursuit) or approximate message passing (Approximate Message Passing, AMP) to solve this problem. Both theoretical analysis and simulation results show that in the case of obvious spatial sparsity, the channel estimation and feedback method based on structured compressed sensing can accurately obtain large-scale MIMO channel state information with low pilot overhead.

[0041] Although compressed sensing (CS) is considered to be a promising method for reducing channel state information (CSI) overhead, this type of solution has some inherent problems.

[0042] First, CS solutions strongly rely on the assumption of channel sparsity in the spatial angle domain, but many real channels in actual scenarios are not absolutely sparse. Therefore, CS-based channel estimation and feedback solutions rely on the prior assumption that the CSI matrix is ​​perfectly sparse, and perform poorly on large-scale MIMO channels that only meet approximate sparsity conditions. The performance is difficult to meet the requirements of actual systems.

[0043] Second, when compressed sensing is used in a massive MIMO system based on OFDM, it is not possible to effectively integrate frequency domain and spatial domain resources. Compressed sensing is difficult to apply in actual systems. In other words, in a massive MIMO system based on OFDM, due to the frequency selectivity of the channel, the channels on different sub-bands are different. Therefore, it is necessary to perform channel estimation for massive MIMO on each sub-band at the same time. Therefore, even if compressed sensing is used at this time, there is still a huge pilot overhead.

[0044] See also Figure 1, the figure shows a block diagram of a wireless communication system applicable to the embodiment of the present application. The wireless communication system includes a terminal 11, a terminal 12 and a network side device 13. Among them, the terminal can also be called a terminal device or UE, and the terminal can be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer) or a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), augmented reality (augmented reality, AR) / virtual reality (virtual reality, VR) equipment, a robot, a wearable device (Wearable Device), a vehicle-mounted device (VUE), a pedestrian terminal (PUE), a smart home (home equipment with wireless communication function, such as a refrigerator, a TV, a washing machine or furniture, etc.) and other terminal side devices, and the wearable device includes: smart watches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, game consoles, etc. It should be noted that the embodiments of the present application do not limit the specific types of the terminals 11 and 12.

[0045] The network side device 13 can be a base station or a core network, wherein the base station can be called a node B, an evolved node B, an access point, a base transceiver station (Base Transceiver Station, BTS), a radio base station, a radio transceiver, a basic service set (Basic Service Set, BSS), an extended service set (Extended Service Set, ESS), a B node, an evolved B node (gNB), a home B node, a home evolved B node, a WLAN access point, a WiFi node, a transmitting and receiving point (Transmitting Receiving Point, TRP), a wireless access network node or some other suitable term in the field. As long as the same technical effect is achieved, the base station is not limited to the specified technical vocabulary. It should be noted that in the embodiment of the present application, only the base station in the NR system is taken as an example, but the specific type of the base station is not limited.

[0046] See also Figure 2 , the embodiment of the present application provides a method for sending a pilot signal, and the specific steps are as follows:

[0047] Step 201: A first communication device sends a first pilot signal in a first matrix;

[0048] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent by the i-th resource element (RE) and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N (that is, i=1~N), j is a positive integer greater than or equal to 1 and less than or equal to M (that is, j=1~M), N is the number of antennas of the first communication device, and M is the number of REs that send pilot signals on each resource block (RB), M is less than N, and N and M are both integers greater than 1.

[0049] Taking the first communication device as a base station as an example, the base station sends a pilot channel state information reference signal (CSI-RS). In pilot design and resource allocation, resource blocks (RBs) are used as the basic unit. Each RB contains 12 subcarriers in the frequency domain and 6-7 OFDM symbols in the time domain. For each RB, the compressed sensing method is used to perform pilot compression.

[0050] In one embodiment of the present application, the first matrix is ​​determined based on the second matrix and the third matrix, for example, the first matrix = the second matrix (W1) × the third matrix (W2), the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain. In this way, channels with frequency selectivity in the frequency domain of multiple antennas can be converted into multipath channels with a limited number of paths in the beam domain and delay domain for joint estimation, thereby greatly reducing the pilot overhead.

[0051] In one embodiment of the present application, the third matrix is ​​an N-dimensional Discrete Fourier Transform (DFT) matrix;

[0052] or,

[0053] The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

[0054] In an embodiment of the present application, the first communication device uses a DFT matrix in the spatial domain as a pilot transmission symbol on each RB, and uses a perception matrix operation to compress the transmitted first pilot signal.

[0055] In one embodiment of the present application, the method further includes:

[0056] The second matrix is ​​trained based on the first pilot signal.

[0057] In one implementation manner of the present application, the training the second matrix based on the first pilot signal includes:

[0058] The first communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0059] The first communication device estimates the pilot signal of the delay path according to the AMP algorithm to obtain a first estimation error;

[0060] The first communication device trains the second matrix according to the first estimation error.

[0061] In one embodiment of the present application, the method further includes:

[0062] The first communication device sends the trained second matrix to the second communication device.

[0063] In one embodiment of the present application, the method further includes:

[0064] The parameters of LAMP are trained based on the first pilot signal.

[0065] In one implementation manner of the present application, the step of training the parameters of the LAMP based on the first pilot signal includes:

[0066] The first communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0067] The first communication device performs overall channel estimation using a learned approximate message passing (LAMP) method according to the pilot signal of the delay path to obtain a second estimation error;

[0068] The first communication device trains parameters of LAMP according to the second estimation error.

[0069] In one embodiment of the present application, the method further includes:

[0070] The first communication device sends the trained parameters of the LAMP to the second communication device.

[0071] In the embodiment of the present application, the pilot overhead is reduced while maintaining the channel estimation accuracy of the system.

[0072] See also Figure 3, an embodiment of the present application provides a channel estimation method, and the specific steps include: step 301 and step 302.

[0073] Step 301: A second communication device receives a first pilot signal in a first matrix sent by a first communication device;

[0074] Step 302: The second communication device performs channel estimation according to the first pilot signal;

[0075] Among them, the first matrix is ​​an N×M matrix, the elements in the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1.

[0076] In an embodiment of the present application, the second communication device jointly processes the pilot signals of all RBs in the frequency domain, first performing an inverse fast Fourier transform (IFFT) on the pilot received signals of each RB of each port and converting them into the delay domain. Then, the signals of each port and each delay are combined to perform overall channel estimation of learned approximate message passing (LAMP). Finally, the estimated signal is transferred to the frequency domain again to obtain the channel of each antenna in each RB, and the final channel estimation is completed.

[0077] In one embodiment of the present application, the first matrix is ​​determined based on the second matrix and the third matrix, for example, the first matrix = the second matrix × the third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain. In this way, channels showing frequency selectivity in multiple antenna frequency domains can be converted into multipath channels with a limited number of paths in the beam domain and delay domain for joint estimation, thereby greatly reducing the pilot overhead.

[0078] In one implementation of the present application, the third matrix is ​​an N-dimensional DFT matrix;

[0079] or,

[0080] The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

[0081] In one implementation manner of the present application, the second communication device performs channel estimation according to the first pilot signal, including:

[0082] The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0083] The second communication device performs overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a first channel estimation result.

[0084] Optionally, the second communication device performs channel estimation according to the first pilot signal, further comprising:

[0085] The second communication device obtains a second channel estimation result of a channel of each antenna in a corresponding RB according to the first channel estimation result.

[0086] In one implementation of the present application, the second communication device obtains, according to the first channel estimation result, a second channel estimation result of each antenna in a channel corresponding to the RB, including:

[0087] The second communication device converts the first channel estimation result into an M×N channel matrix, where the first channel estimation result is a channel vector of a spatial domain beam;

[0088] The second communication device converts the M×N channel matrix into the frequency domain and the spatial domain to obtain a second channel estimation result of each antenna in the channel corresponding to the RB.

[0089] In one embodiment of the present application, the method further includes:

[0090] The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0091] The second communication device estimates the pilot signal of each delay path according to the AMP algorithm to obtain a first estimation error;

[0092] The second communication device trains the second matrix according to the first estimation error.

[0093] In one embodiment of the present application, the method further includes:

[0094] The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0095] The second communication device performs overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a second estimation error;

[0096] The second communication device trains parameters of LAMP according to the second estimated error.

[0097] In one implementation of the present application, the second communication device performs LAMP overall channel estimation according to the pilot signal of the delay path to obtain a second estimation error, including:

[0098] The second communication device splices the pilot signals of each delay path together, and the spliced ​​signal matrix is ​​a fourth matrix, and the fourth matrix is ​​a Kronnick product of a unit matrix with a dimension of the number of RBs and a perception matrix on each RB;

[0099] The second communication device performs LAMP channel estimation according to the fourth matrix to obtain a second estimation error.

[0100] In the embodiment of the present application, the pilot overhead is reduced while maintaining the channel estimation accuracy of the system.

[0101] The following takes the first communication device as a transmitting end and the second communication device as a receiving end as an example.

[0102] In the present application embodiment, using Figure 4 The OFDM-based massive MIMO antenna system shown combines the sparsity of the channel in the spatial domain (or described as the space domain) and the sparsity in the delay domain. Through the pilot design and channel estimation that are jointly compressed and optimized in the spatial domain and the delay domain, the final channel estimation of all antennas in the OFDM-based massive MIMO antenna system on all RBs can be completed at one time.

[0103] In a massive MIMO antenna system based on OFDM, the transmitting end has N antennas. Since each receiving antenna at the receiving end performs channel estimation independently, only one receiving antenna can be considered for design. That is, an N×1 massive antenna system is considered. At the transmitting end, pilot signals are sent in units of RB, so channel estimation can be performed separately on each RB. The number of REs that send pilot signals on each RB is M, which is much smaller than the number of transmitting antennas N, thereby achieving the purpose of compressing pilot overhead.

[0104] If on the i-th RE, the transmitted symbol of the j-th antenna is represented by c i,j , then the transmitted symbols on all M REs on N antennas can be represented as an N×M matrix C.

[0105]

[0106] C is obtained by multiplying two matrices, that is, C=W2W1, where W1 is an N×M sensing matrix used to compress pilot overhead.

[0107] In the traditional compressed sensing method, W1 generally adopts a random matrix. In the embodiment of the present application, W1 is trained by a large amount of channel data using a deep learning method. W2 is the conversion matrix from the spatial domain to the beam domain. For a linear array, W2 is an N-dimensional DFT matrix. For an N1×N2 array, W2 is the Kronnick product of the N1-dimensional DFT matrix and the N2-dimensional DFT matrix (W2=kron(N1-dimensional DFT matrix, N2-dimensional DFT matrix).

[0108] In each RB, based on C, pilot symbols are sent on M REs, and after IDFT (OFDM modulation), they are sent through N antennas.

[0109] At the receiving end, only one receiving antenna is considered. The received signal on this antenna is subjected to FFT (after OFDM demodulation) to obtain the signal on each RB. The pilot signal is received on the M REs on each RB, which is a 1×M vector. The pilot signal r is received on the i-th RB. i =[r i,1 r i,2 … r i,M ] can be expressed as r i =h i C+n i . where h i is the channel on the i-th RB, which is a 1×N channel vector, and the mathematical expression is:

[0110] h i =[h i,1 h i,2 … h i,N ]

[0111] n i is the noise received on M REs on the i-th RB, n i is a 1×M vector.

[0112] In the embodiment of the present application, the received signals on K RBs are processed jointly. Define R = [r1 r2 … r K ] T , R is the matrix of pilot signals received on all RBs, r k The pilot signal received on the kth RB, T is the transposed symbol. From this we can get:

[0113]

[0114] definition is the channel of N transmitting antennas on K RBs, and is the value to be estimated.

[0115] definition is a beam-domain sparse channel with N transmit antennas on K RBs.

[0116] Available,

[0117] In the embodiment of the present application, in order to perform channel estimation in the delay domain, the received pilot signal r is first subjected to a K-dimensional IDFT in the RB dimension to obtain

[0118] Definition H d =W IDFT H and We can get:

[0119] R d is a K×M matrix.

[0120] in is the channel vector in the beam domain on the lth delay path, is a 1×N vector.

[0121] Here we first estimate That is, the channel gain in the beam domain and delay domain.

[0122] because

[0123] (R d ) T Written as a KM×1 vector Written as a KM×1 vector Defining the Perception Matrix is a KM×KN matrix, A is the equivalent sensing matrix for overall channel estimation in beam domain and delay domain, I k is the k-order identity matrix.

[0124] in and A is known, Since it has sparsity in both the beam domain and the delay domain, it can be solved using compressed sensing methods.

[0125] In the embodiment of the present application, a two-step deep learning training method is used to achieve optimal performance. The first step is to train the optimal perception matrix A. This allows only W1 to be trained, see Figure 5 .

[0126] Since W1 operates on each RB and W1 on each RB is the same, W1 is trained separately on each RB and does not need to be operated on all RBs as a whole. Since the final channel estimation is performed in the delay domain, W1 is trained separately on each delay path in the delay domain. Perform IFFT on the pilot received signal of each RB of each port, convert it to the delay domain for IDFT, and then use the AMP algorithm to estimate the pilot signal of each delay path, and use the estimation error to train W1.

[0127] This breaks down into:

[0128]

[0129] Without loss of generality, for the i-th delay path, the pilot signal expression is where r di is a 1×M vector, is a 1×N vector. At this time, using a large number of r di and As training data, W1 is trained to obtain the optimal perception matrix.

[0130] In order to find the optimal perception matrix through the neural network method, the perception matrix is ​​regarded as a linear neural network. Compared with the traditional neural network, it has no offset and activation function. The optimal perception matrix is ​​found through training.

[0131] During training, the sensing matrix at the transmitter, the AMP recovery algorithm at the receiver, the massive MIMO channel, and the introduced noise are considered as a neural network for training.

[0132] The training data comes from a large number of beam domain and delay domain channels, and the training optimization target (cost function) is the output of the output neural network. and the actual channel The mean square error between them is the smallest.

[0133] Right now

[0134] After the training is completed, the optimized perception matrix W1 obtained through training can be used for pilot transmission and channel estimation. The base station uses the transmission signal (symbol) of each transmitting antenna on each pilot RE of each RB according to the optimized perception matrix. At the mobile user end, the mobile user first performs IFFT on the pilot of each RB of each port and converts it into the delay domain. Then, the signals of each port and each delay are combined to perform the overall channel estimation of LAMP. The LAMP estimation result is the channel of each beam on each delay path. LAMP training is based on During training, a large amount of and Data, get the B of each LAMP iteration t and Complete the parameter training of LAMP, B t and They correspond to the linear weight matrix and nonlinear activation function in the neural network respectively. The specific training process is as follows:

[0135] The LAMP network is built based on the traditional compressed sensing algorithm, AMP. The iterative solution process of the AMP algorithm is expanded into a neural network, and its linear operation coefficients and nonlinear shrinkage parameters are jointly optimized. The values ​​of these distribution parameters can be obtained through the training process of the deep neural network. It is worth noting that during training, the transmitter uses the optimized perception matrix that has been trained in the first stage.

[0136] Different from the general neural network structure, the LAMP network structure is built based on the AMP algorithm. In fact, the LAMP algorithm is to expand the iterative operation process of the AMP algorithm into a deep neural network. The specific network structure is as follows: Figure 6 shown.

[0137] For the y=Ax+n scenario commonly used in compressed sensing, the sparse signal of x is estimated from the known (y, A) and recovered using represents an estimated vector of the original sparse signal, y is the observation vector, and v is the residual vector. The observation vector y is used as the common input of all layers, and the previous layer estimates the vector The output of the residual vector v is used as the next layer estimation vector And the input of the residual vector v. After passing through the T-layer neural network, the final estimate of the original sparse signal is output The network structure of the tth layer is as follows Figure 7 This structure completely corresponds to the iterative process in the AMP algorithm, which gradually realizes the estimation of vector And the update of the residual vector v. Where B t and They correspond to the linear weighting matrix and nonlinear activation function in the neural network respectively.

[0138] The LAMP network structure has some unique designs, as follows: (1) There is an extra branch in the LAMP network, such as Figure 7 As shown in ①, it corresponds to the Onsager correction term in the AMP algorithm to accelerate convergence; (2) The nonlinear function in the LAMP algorithm is a shrinkage function derived from a specific signal estimation problem, rather than an activation function in a general neural network that has no clear physical meaning in order to introduce nonlinear functions; (3) The noise parameter in the shrinkage function It is related to the residual and can be updated layer by layer. It is precisely because of the above characteristics that the LAMP network is more suitable for sparse signal recovery problems than general neural networks.

[0139] The LAMP algorithm combines deep learning and the AMP algorithm, taking the advantages of both. It not only utilizes the powerful learning ability of deep neural networks but also retains the function of the AMP algorithm to achieve sparse signal recovery.

[0140] In this embodiment, supervised learning is adopted by inputting a data set. Perform network parameter training, where y d represents the low-dimensional observation signal, x d In order to further improve the performance of the algorithm and make full use of the powerful learning ability of the neural network, in each layer, in addition to the nonlinear parameter θ t In addition to updating layer by layer, the linear operator B can also be updated layer by layer (in the AMP algorithm, B = A T ). Therefore, in the LAMP network, the parameter set that needs to be trained is

[0141] Since the LAMP network is built on the basis of the AMP iterative algorithm, the network can be trained layer by layer during the training process to achieve joint optimization of linear operation coefficients and nonlinear shrinkage parameters. Different from the general neural network that only defines one loss function, in the LAMP network, in order to achieve layer-by-layer training, each layer has a loss function defined separately, which is specifically defined as follows:

[0142]

[0143] y d represents the low-dimensional observation signal, x d Representing high-dimensional sparse signals

[0144] L t (Φ) is the loss function of the tth layer, is the t-th layer pair x d The estimated results.

[0145] In addition, in order to avoid network overfitting, the linear operation coefficient matrix B t and the nonlinear shrinkage parameter θ t In each layer, the training is carried out by first optimizing separately and then jointly optimizing.

[0146] After the parameter training of the perception matrix W1 and LAMP is completed, pilot transmission and channel estimation can be performed. At the transmitting end, the perception matrix W1 operation can also be performed independently on each RB. However, at the receiving end, the pilot signals of all RBs in the frequency domain are jointly processed to perform channel estimation. First, the pilot signals received by each port on each RB are collected together for IFFT and converted into the delay domain. Then, the signals of each port and each delay are combined to perform the overall channel estimation of LAMP. At this time, the channel vector of the spatial domain beam is estimated. First, the channel vector Convert to M×N channel matrix Then perform DFT operation to convert it into frequency domain and beam domain, and finally use W2 to transfer the result to frequency domain and spatial domain. Mathematical expression is Thus, the channel of each antenna in each RB is obtained, and the final channel estimation is completed.

[0147] See also Figure 8 The embodiment of the present application provides a pilot signal sending device, which is applied to a first communication device. The device 800 includes:

[0148] A first sending module 801, configured to send a first pilot signal in a first matrix;

[0149] Among them, the first matrix is ​​an N×M matrix, and the elements in the i-th row and j-th column of the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and N and M are both integers greater than 1.

[0150] In one embodiment of the present application, the first matrix is ​​determined based on a second matrix and a third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

[0151] In one embodiment of the present application, the third matrix is ​​an N-dimensional DFT matrix;

[0152] or,

[0153] The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

[0154] In one embodiment of the present application, the device further comprises:

[0155] In one embodiment of the present application, the device further comprises:

[0156] The first training module is used to train the second matrix based on the first pilot signal.

[0157] In one embodiment of the present application, the first training module includes:

[0158] A first processing unit, configured to perform a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0159] A second processing unit, configured to estimate the pilot signal of the delay path according to an AMP algorithm to obtain a first estimation error;

[0160] A third processing unit is used to train the second matrix according to the first estimation error.

[0161] In one embodiment of the present application, the device further comprises:

[0162] A second sending module is configured to send the trained second matrix to the second communication device.

[0163] In one embodiment of the present application, the device further comprises:

[0164] The second training module is used to train the parameters of LAMP based on the first pilot signal.

[0165] In one embodiment of the present application, the second training module includes:

[0166] a fourth processing unit, configured to perform a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0167] a fifth processing unit, configured to perform overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a second estimation error;

[0168] The sixth processing unit is used to train the parameters of LAMP according to the second estimation error.

[0169] In one embodiment of the present application, the device further comprises:

[0170] The third sending module is used to send the trained parameters of the LAMP to the second communication device.

[0171] The device provided in the embodiment of the present application can achieve Figure 2 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0172] See also Fig. 9, an embodiment of the present application provides a channel estimation device, applied to a second communication device, the device 900 comprising:

[0173] A first receiving module 901 is configured to receive a first pilot signal in a first matrix sent by a first communication device;

[0174] A channel estimation module 902, configured for the second communication device to perform channel estimation according to the first pilot signal;

[0175] Among them, the first matrix is ​​an N×M matrix, the elements in the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1.

[0176] In one embodiment of the present application, the first matrix is ​​determined based on a second matrix and a third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

[0177] In one embodiment of the present application, the third matrix is ​​an N-dimensional DFT matrix;

[0178] or,

[0179] The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

[0180] In one implementation of the present application, the channel estimation module 902 is further configured to:

[0181] Performing a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0182] According to the pilot signal of the delay path, the overall channel estimation of LAMP is performed to obtain a first channel estimation result.

[0183] In one implementation of the present application, the channel estimation module 902 is further configured to:

[0184] According to the first channel estimation result, a second channel estimation result of each antenna in the channel corresponding to the RB is obtained.

[0185] In one implementation of the present application, the channel estimation module 902 is further configured to:

[0186] The first channel estimation result is converted into an M×N channel matrix, where the first channel estimation result is a channel vector of a spatial domain beam; the M×N channel matrix is ​​converted into a frequency domain and a spatial domain to obtain a second channel estimation result of each antenna in a channel corresponding to a RB.

[0187] In one embodiment of the present application, the device further comprises:

[0188] a seventh processing module, configured to perform a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0189] an eighth processing module, configured to estimate the pilot signal of each delay path according to an AMP algorithm to obtain a first estimation error;

[0190] A ninth processing module is used to train the second matrix according to the first estimation error.

[0191] In one embodiment of the present application, the device further comprises:

[0192] a tenth processing module, configured to perform a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path;

[0193] an eleventh processing module, configured to perform overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a second estimation error;

[0194] A twelfth processing module is used to train the parameters of LAMP according to the second estimation error.

[0195] Optionally, the eleventh processing module is further used to: splice the pilot signals of each delay path together, the spliced ​​signal matrix is ​​a fourth matrix, and the fourth matrix is ​​the Kronnick product of a unit matrix with a dimension of the number of RBs and a perception matrix on each RB; according to the fourth matrix, perform LAMP channel estimation to obtain a second estimation error.

[0196] The device provided in the embodiment of the present application can achieve Figure 3 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0197] Fig.10 To implement a hardware structure diagram of a terminal in an embodiment of the present application, the terminal 1000 includes but is not limited to: a radio frequency unit 1001, a network module 1002, an audio output unit 1003, an input unit 1004, a sensor 1005, a display unit 1006, a user input unit 1007, an interface unit 1008, a memory 1009, and at least some of the components of the processor 1100.

[0198] Those skilled in the art will appreciate that the terminal 1000 may also include a power source (such as a battery) for supplying power to various components, and the power source may be logically connected to the processor 1100 through a power management system, thereby implementing functions such as managing charging, discharging, and power consumption management through the power management system. Fig.10 The terminal structure shown in the figure does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which will not be described in detail here.

[0199] It should be understood that in the embodiment of the present application, the input unit 1004 may include a graphics processor (Graphics Processing Unit, GPU) 10041 and a microphone 10042, and the graphics processor 10041 processes the image data of the static picture or video obtained by the image capture device (such as a camera) in the video capture mode or the image capture mode. The display unit 1006 may include a display panel 10061, and the display panel 10061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 1007 includes a touch panel 10061 and other input devices 10072. The touch panel 10061 is also called a touch screen. The touch panel 10061 may include two parts: a touch detection device and a touch controller. Other input devices 10072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0200] In the embodiment of the present application, the radio frequency unit 1001 receives downlink data from the network side device and sends it to the processor 1100 for processing; in addition, the uplink data is sent to the network side device. Generally, the radio frequency unit 1001 includes but is not limited to an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

[0201] The memory 1009 can be used to store software programs or instructions and various data. The memory 1009 can mainly include a program or instruction storage area and a data storage area, wherein the program or instruction storage area can store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 1009 can include a high-speed random access memory, and can also include a non-volatile memory, wherein the non-volatile memory can be a read-only memory (Read-Only Memory, ROM), a programmable read-only memory (Programmable ROM, PROM), an erasable programmable read-only memory (Erasable PROM, EPROM), an electrically erasable programmable read-only memory (Electrically EPROM, EEPROM) or a flash memory. For example, at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device.

[0202] The processor 1100 may include one or more processing units; optionally, the processor 1100 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs or instructions, etc., and the modem processor mainly processes wireless communications, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 1100.

[0203] The terminal provided in the embodiment of the present application can achieve Figure 2 or Figure 3 The various processes implemented by the method embodiment shown achieve the same technical effect and will not be described again here to avoid repetition.

[0204] See also Fig.11 , the embodiment of the present application also provides a network side device. Fig.11 As shown, the network side device 1100 includes: an antenna 1101, a radio frequency device 1102, and a baseband device 1103. The antenna 1101 is connected to the radio frequency device 1102. In the uplink direction, the radio frequency device 1102 receives information through the antenna 1101 and sends the received information to the baseband device 1103 for processing. In the downlink direction, the baseband device 1103 processes the information to be sent and sends it to the radio frequency device 1102. The radio frequency device 1102 processes the received information and sends it out through the antenna 1101.

[0205] The frequency band processing device may be located in the baseband device 1103 . The method performed by the network-side device in the above embodiment may be implemented in the baseband device 1103 . The baseband device 1103 includes a processor 1104 and a memory 1105 .

[0206] The baseband device 1103 may include, for example, at least one baseband board on which a plurality of chips are arranged. Fig.11 As shown, one of the chips is, for example, a processor 1104, which is connected to a memory 1105 to call a program in the memory 1105 to execute the network device operations shown in the above method embodiment.

[0207] The baseband device 1103 may further include a network interface 1106 for exchanging information with the radio frequency device 1102 . The interface may be, for example, a common public radio interface (CPRI).

[0208] Specifically, the network side device in the embodiment of the present application also includes: instructions or programs stored in the memory 1105 and executable on the processor 1104 .

[0209] It is understood that the processor 1104 calls the instructions or programs in the memory 1105 to execute Figure 8 or Fig. 9 The methods executed by the modules shown achieve the same technical effects, and therefore will not be described here in detail to avoid repetition.

[0210] Optional, such as Fig.12 As shown, the embodiment of the present application also provides a communication device 1200, including a processor 1201, a memory 1202, and a program or instruction stored in the memory 1202 and executable on the processor 1201. For example, when the communication device 1200 is a terminal, the program or instruction is executed by the processor 1201 to implement the above Figure 2 or Figure 3 When the communication device 1200 is a network side device, the program or instruction is executed by the processor 1201 to implement the above Figure 2 or Figure 3 The various processes of the method embodiments can achieve the same technical effect, and will not be described again here to avoid repetition.

[0211] The embodiment of the present application also provides a computer program / program product, wherein the computer program / program product is stored in a non-volatile storage medium, and the computer program / program product is executed by at least one processor to implement the following Figure 2 or Figure 3 The steps of the processing method.

[0212] The embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the above Figure 2 or Figure 3The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0213] The processor is a processor in the terminal or network side device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0214] The embodiment of the present application also provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, the processor is used to run a program or instruction to implement the above Figure 4 or Figure 5 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0215] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0216] It should be noted that, in this article, the term "comprises", "includes" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0217] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0218] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A method for sending a pilot signal, characterized in that: include: The first communication device sends a first pilot signal in a first matrix; The first matrix is ​​an N×M matrix, the elements of the i-th row and the j-th column in the first matrix represent the pilot signal sent by the i-th resource unit RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each resource block RB, M is less than N, and both N and M are integers greater than 1; The first matrix is ​​determined based on the second matrix and the third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

2. The method according to claim 1, characterized in that The third matrix is ​​an N-dimensional discrete Fourier transform DFT matrix; or, The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

3. The method according to claim 1, characterized in that The method further comprises: The second matrix is ​​trained based on the first pilot signal.

4. The method according to claim 3, characterized in that The step of training the second matrix based on the first pilot signal comprises: The first communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path; The first communication device estimates the pilot signal of the delay path according to an approximate message passing AMP algorithm to obtain a first estimation error; The first communications device trains the second matrix according to the first estimation error.

5. The method according to claim 3 or 4, characterized in that: The method further comprises: The first communication device sends the trained second matrix to the second communication device.

6. The method according to claim 1, characterized in that The method further comprises: The parameters of LAMP are trained based on the first pilot signal.

7. The method according to claim 6, characterized in that The step of training the parameters of LAMP based on the first pilot signal comprises: The first communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path; The first communication device performs overall channel estimation using a learning approximate message passing (LAMP) method according to the pilot signal of the delay path to obtain a second estimation error; The first communication device trains parameters of LAMP according to the second estimation error.

8. The method according to claim 6 or 7, characterized in that: The method further comprises: The first communication device sends the trained parameters of the LAMP to the second communication device.

9. A channel estimation method, characterized in that: include: The second communication device receives the first pilot signal in the first matrix sent by the first communication device; The second communication device performs channel estimation according to the first pilot signal; The first matrix is ​​an N×M matrix, the elements of the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1; The first matrix is ​​determined based on the second matrix and the third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

10. The method according to claim 9, characterized in that The third matrix is ​​an N-dimensional DFT matrix; or, The third matrix is ​​the Kronecker product of an N1-dimensional DFT matrix and an N2-dimensional DFT matrix, wherein N1 is the number of rows in the first antenna array, and N2 is the number of columns in the first antenna array, and the first antenna array includes N antennas.

11. The method according to claim 9, characterized in that The second communication device performs channel estimation according to the first pilot signal, including: The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path; The second communication device performs overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a first channel estimation result.

12. The method according to claim 11, characterized in that The second communication device performs channel estimation according to the first pilot signal, further comprising The second communication device obtains a second channel estimation result of a channel of each antenna in a corresponding RB according to the first channel estimation result.

13. The method according to claim 12, characterized in that The second communication device obtains, according to the first channel estimation result, a second channel estimation result of each antenna in a channel corresponding to the RB, including: The second communication device converts the first channel estimation result into an M×N channel matrix, where the first channel estimation result is a channel vector of a spatial domain beam; The second communication device converts the M×N channel matrix into the frequency domain and the spatial domain to obtain a second channel estimation result of each antenna in the channel corresponding to the RB.

14. The method according to claim 9, characterized in that The method further comprises: The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path; The second communication device estimates the pilot signal of each delay path according to the AMP algorithm to obtain a first estimation error; The second communication device trains the second matrix according to the first estimation error.

15. The method according to claim 11, characterized in that The method further comprises: The second communication device performs a fast Fourier transform on the first pilot signal to obtain a pilot signal of a delay path; The second communication device performs overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a second estimation error; The second communication device trains parameters of LAMP according to the second estimated error.

16. The method according to claim 15, characterized in that The second communication device performs overall channel estimation of LAMP according to the pilot signal of the delay path to obtain a second estimation error, including: The second communication device splices the pilot signals of each delay path together, and the spliced ​​signal matrix is ​​a fourth matrix, and the fourth matrix is ​​a Kronnick product of a unit matrix with a dimension of the number of RBs and a perception matrix on each RB; The second communication device performs LAMP channel estimation according to the fourth matrix to obtain a second estimation error.

17. A pilot signal sending device, characterized in that: include: A first sending module, used for sending a first pilot signal in a first matrix; The first matrix is ​​an N×M matrix, the elements of the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1; The first matrix is ​​determined based on the second matrix and the third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

18. A channel estimation device, characterized in that: include: A first receiving module, configured to receive a first pilot signal in a first matrix sent by a first communication device; A channel estimation module, configured for the second communication device to perform channel estimation according to the first pilot signal; The first matrix is ​​an N×M matrix, the elements of the i-th row and the j-th column in the first matrix represent the pilot signal sent through the i-th RE and the j-th antenna, i is a positive integer greater than or equal to 1 and less than or equal to N, j is a positive integer greater than or equal to 1 and less than or equal to M, N is the number of antennas of the first communication device, M is the number of REs that send pilot signals on each RB, M is less than N, and both N and M are integers greater than 1; The first matrix is ​​determined based on the second matrix and the third matrix, the second matrix is ​​an N×M perception matrix, and the third matrix is ​​a conversion matrix from the spatial domain to the beam domain.

19. A communication device, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of the method according to any one of claims 1 to 16 are implemented.

20. A readable storage medium, characterized in that: The readable storage medium stores a program or an instruction, and when the program or the instruction is executed by a processor, the steps of the method according to any one of claims 1 to 16 are implemented.

Images