Channel estimation method, device, equipment, storage medium and program product

Abstract

The embodiment of the invention discloses a channel estimation method and device, equipment, a storage medium and a program product. Determining scatterer space distribution information based on the channel map; obtaining non-line of sight observation data and a non-line of sight correlation matrix; wherein the non-line-of-sight observation data is obtained through pilot frequency training; determining a feature subspace based on the non-line-of-sight correlation matrix or the scatterer space distribution information; determining a non-line-of-sight channel component based on the feature subspace and the non-line-of-sight observation data; and determining a target channel estimation result based on the non-line-of-sight channel component. According to the channel estimation method provided by the embodiment of the invention, the feature subspace is determined based on the non-line-of-sight correlation matrix, and the high-dimensional correlation matrix is randomly projected to the low-dimensional correlation matrix, so that the calculation complexity is remarkably reduced; a feature subspace is determined based on scatterer space distribution information, the calculation complexity is further reduced, and fine estimation close to the optimal performance is obtained; the estimation precision and the calculation efficiency can be improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of communication technology, and in particular to a channel estimation method, apparatus, device, storage medium and program product. Background Technology

[0002] As the information society continues to advance towards the stage of "Internet of Everything and ubiquitous intelligence," future 6G wireless communication systems are expected to further break through the limitations of capacity, coverage, and connection density based on 5G technology to meet the communication needs of all scenarios, multiple terminals, and high reliability. In this evolutionary trend, massive MIMO (Multiple-Input Multiple-Output) technology continues to be promoted as a key technology due to its resource reuse capabilities and spatial diversity advantages. Especially in the face of the challenges of channel state awareness brought about by the surge in the number of users and the increasing complexity of application scenarios, massive MIMO under traditional cellular architectures has gradually exposed performance bottlenecks such as pilot collisions and difficulties in interference suppression. To address this, the industry has proposed ultra-large-scale MIMO technology, which, by introducing antenna arrays of orders of magnitude larger, supports more refined beamforming and stronger spatial resolution, becoming an important pillar technology of 6G networks.

[0003] In ultra-large-scale MIMO systems, the aperture of the antenna array is much larger than the wavelength, causing the system to operate in the near-field propagation region. The channel no longer satisfies the far-field plane wave assumption, but exhibits distinct spherical wave characteristics. This near-field characteristic brings new opportunities to communication systems, such as beam focusing to enhance signal power and distance-aware support for sensing communication fusion, but it also presents significant challenges to channel modeling and estimation. Compared to far-field channel models, near-field channels exhibit more complex structures in the angle, distance, and amplitude directions, making traditional angle-domain-based estimation methods difficult to apply. Furthermore, since the number of antennas in ultra-large-scale MIMO systems can reach thousands or even more, the channel dimension grows exponentially, leading to problems such as limited pilot resources, extremely high training overhead, and unacceptable computational complexity for traditional pilot-based channel estimation methods. Therefore, how to reduce the channel estimation overhead without sacrificing estimation accuracy has become one of the key bottlenecks in realizing the practical application of ultra-large-scale MIMO systems. Traditional estimation methods such as Least Squares (LS) and Minimum Mean-Squared Error (MMSE) either fail to utilize channel statistical properties or rely on the difficult-to-obtain covariance matrix, making it difficult to achieve both estimation performance and feasibility. Summary of the Invention

[0004] This invention provides a channel estimation method, apparatus, device, storage medium, and program product, which can improve estimation accuracy and computational efficiency.

[0005] In a first aspect, embodiments of the present invention provide a channel estimation method, including: Determine the spatial distribution information of scatterers based on channel maps; Obtain non-line-of-sight observation data and non-line-of-sight correlation matrix; wherein, the non-line-of-sight observation data is obtained through pilot training; The feature subspace is determined based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterers; The non-line-of-sight channel components are determined based on the feature subspace and the non-line-of-sight observation data; The target channel estimation result is determined based on the non-line-of-sight channel components.

[0006] Secondly, embodiments of the present invention also provide a channel estimation apparatus, comprising: The scatterer spatial distribution information determination module is used to determine the scatterer spatial distribution information based on the channel map; The data acquisition module is used to acquire non-line-of-sight observation data and non-line-of-sight correlation matrix; wherein, the non-line-of-sight observation data is obtained through pilot training; A feature subspace determination module is used to determine the feature subspace based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterers; A non-line-of-sight channel component determination module is used to determine non-line-of-sight channel components based on the feature subspace and the non-line-of-sight observation data; The target channel estimation result determination module is used to determine the target channel estimation result based on the non-line-of-sight channel components.

[0007] Thirdly, embodiments of the present invention also provide an electronic device, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the channel estimation method described in the embodiments of the present invention.

[0008] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions that are used to cause a processor to execute the channel estimation method described in the embodiments of the present invention.

[0009] Fifthly, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the channel estimation method as described in the embodiments of the present invention.

[0010] This invention discloses a channel estimation method, apparatus, device, storage medium, and program product. The method involves: determining the spatial distribution information of scatterers based on a channel map; acquiring non-line-of-sight (NLS) observation data and a NLS correlation matrix; wherein the NLS observation data is obtained through pilot training; determining a feature subspace based on the NLS correlation matrix or the spatial distribution information of scatterers; determining NLS channel components based on the feature subspace and the NLS observation data; and determining the target channel estimation result based on the NLS channel components. The channel estimation method provided by this invention determines the feature subspace based on the NLS correlation matrix by randomly projecting the high-dimensional correlation matrix to a low-dimensional dimension, significantly reducing computational complexity and enabling rapid extraction of the main channel subspace; determining the feature subspace based on the spatial distribution information of scatterers further reduces computational complexity and obtains a refined estimate with near-optimal performance; thus improving estimation accuracy and computational efficiency. Attached Figure Description

[0011] Figure 1 This is a flowchart of a channel estimation method according to Embodiment 1 of the present invention; Figure 2 This is a scenario diagram of channel estimation using a channel map in an embodiment of the present invention; Figure 3 This is a flowchart of a channel estimation method according to Embodiment 1 of the present invention; Figure 4 This is a schematic diagram illustrating the NMSE performance of the two channel estimation methods in this invention under different antenna sizes. Figure 5 This is a comparison chart of the running time required to construct the subspace matrix under different antenna sizes in this embodiment of the invention; Figure 6 This is a schematic diagram of the structure of a channel estimation device provided in Embodiment 2 of the present invention; Figure 7 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0012] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0013] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0014] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0015] A channel map is a channel state information (CSI) data platform covering a specific area. Built using an environmental database or sensing platform, it outputs prior information including path loss, direction of arrival, scattering structure, and power distribution based on the spatial location of terminals or scatterers. Modeling the channel characteristics of a covered area using channel maps and making them available to communication systems holds promise as a breakthrough for revolutionizing UMIMO channel estimation paradigms. Therefore, it is necessary to research a channel map-enabled UMIMO channel estimation method that utilizes the high-dimensional channel structure information provided by the channel map to constrain and optimize the algorithm flow, thereby improving estimation accuracy and computational efficiency.

[0016] Currently, the commonly used channel estimation methods in ultra-large-scale MIMO systems mainly fall into the following categories: The first category is traditional estimation methods based on pilot received signals, with LS estimation being a typical example. These methods typically assume an unknown channel at the receiver and directly deduce the channel coefficients from the receiver's response to a known pilot sequence. This method is simple in structure, low in implementation cost, and has some adaptability in scenarios where channel changes are not drastic. However, this method does not consider the structural characteristics of the channel and environmental statistics, therefore its estimation accuracy is limited in noisy environments or complex channel paths, and it is prone to bias accumulation and performance bottlenecks.

[0017] The second type of method introduces the statistical properties of the channel as prior information to help improve estimation accuracy, with linear MMSE estimation being a representative example. Its core idea is to use the spatial correlation of the channel to weight and optimize the estimation process. Although this method can theoretically achieve superior performance, i.e., obtaining the lowest normalized mean square error (NMSE), it places high demands on the accurate acquisition of statistical properties such as the channel covariance matrix. Furthermore, when the number of antennas is very large, the storage, updating, and matrix operations of this statistical information can impose a significant system burden.

[0018] The third category comprises estimation methods based on reduced-dimensional channel subspaces, which have emerged in recent years, with Reduced-Subspace LS (RSLS) estimation being a prime example. These methods leverage the characteristics of high-frequency ultra-large-scale MIMO sparse channels, such as limited main directions and concentrated energy, to perform estimation within a subspace with a significantly lower dimension than the original channel space, thereby reducing computational overhead while maintaining accuracy. Although this method has strong practical value in high-dimensional systems, its performance still depends on the effective construction of the main channel subspace, a process that often requires eigenvalue decomposition of high-dimensional matrices, making it difficult to meet the requirements of low latency and low complexity in practical systems.

[0019] In summary, existing technologies have limitations in terms of channel estimation accuracy, utilization of prior information, and system overhead control. In particular, they cannot fully exploit structural priors in real propagation environments, leading to practical problems such as insufficient generalization and computational resource constraints. Therefore, there is an urgent need for an efficient channel estimation method that can incorporate environmental structural information and is applicable to ultra-large-scale MIMO systems.

[0020] Existing ultra-large-scale MIMO channel estimation schemes mainly suffer from the following technical shortcomings: 1) Lack of environmental awareness leads to high estimation errors. Traditional LS estimation does not utilize any prior information, resulting in large bit errors under low signal-to-noise ratio and severe multipath interference. Models based on statistical averaging, such as the classic clustered scattering channel model, ignore the specific location of scatterers and occlusion effects in the actual scene, and cannot provide accurate prior information, causing the estimation results to deviate from the real channel.

[0021] 2) Obtaining high-dimensional statistics is difficult. Although methods such as MMSE are theoretically optimal, they require high-dimensional correlation matrices to be obtained in advance. In rapidly changing or newly introduced scenarios, collecting sufficient training data for estimation is extremely costly, especially when the number of antennas reaches hundreds or thousands, the number of correlation matrix elements is huge, and the training overhead and storage cost are unacceptable.

[0022] 3) Algorithm complexity increases dramatically with the number of antennas. Some improved methods, such as RSLS estimation, reduce dimensionality through eigenvalue decomposition, but their complexity is still the cube of the number of antennas, which cannot meet the actual processing requirements of ultra-large-scale MIMO and exceeds the computing power of existing baseband processors. Furthermore, dimensionality reduction approximation often comes at the cost of performance, making it difficult to meet the high-quality CSI requirements of coherent transmission.

[0023] 4) Lack of utilization of channel map information. Current schemes rarely use scatterer location information and visibility data provided by environmental databases for channel estimation. Due to the lack of full utilization of these structured priors, the estimator cannot know the intrinsic rank structure and energy distribution of the channel matrix, resulting in suboptimal performance and wasted resources.

[0024] Example 1 Figure 1 This is a flowchart of a channel estimation method provided in Embodiment 1 of the present invention. This embodiment is applicable to the estimation of ultra-large-scale MIMO channels. The method can be executed by a channel estimation device, which can be implemented in software and / or hardware, optionally through an electronic device, such as a mobile terminal, PC, or server. Specifically, it includes the following steps: S110, determine the spatial distribution information of scatterers based on the channel map.

[0025] The channel map stores a mapping relationship between "location and channel characteristics." By analyzing and inverting the large-scale or small-scale parameters of the channel it carries, the spatial distribution information of users (User Equipment, UE), base stations, and major scatterers can be accurately obtained; the number of major scatterers is one or more. The user spatial distribution information can be represented as: The spatial distribution information of base stations can be represented as: The spatial distribution information of the scatterer can be represented as: Where L > 0.

[0026] In this embodiment, the spatial distribution information of scatterers determined based on the channel map has a certain error compared to the actual spatial distribution information of scatterers, namely... ,in, For true spatial distribution information of scatterers, The channel map error is used to determine the channel estimation method. If the channel map error is greater than or equal to a set value, SketchAlgorithm-based Receiver Sidelobe Suppression (SA-RSLS) estimation is used to estimate non-line-of-sight components. If the channel map error is less than the set value, Channel Map-based RSLS (CM-RSLS) estimation is used to estimate non-line-of-sight components.

[0027] In this embodiment, after determining the spatial distribution information of users, base stations, and scatterers, the distance between users and base stations can be obtained based on the user spatial distribution information and the base station spatial distribution information. azimuth of arrival Reaching the angle of elevation In addition to occlusion conditions, the line-of-sight (LoS) component of the channel (also known as the line-of-sight channel component) can be constructed. On the other hand, the distance between each scatterer and the base station can be obtained based on the spatial distribution information of the base station and the spatial distribution information of the scatterers. azimuth of arrival Reaching the angle of elevation And the situation of obstruction, This is used to estimate the non-line-of-sight (NLoS) components of the channel (also known as non-line-of-sight channel components). For example, Figure 2 This is a scenario diagram of channel estimation using a channel map in an embodiment of the present invention, such as... Figure 2 As shown, this scenario includes one base station, one user, and multiple main scatterers.

[0028] S120: Obtain non-line-of-sight observation data and non-line-of-sight correlation matrix.

[0029] Non-line-of-sight (NLS) observation data is obtained through pilot training. The principle of pilot training is that the transmitter sends a known pilot sequence to the receiver, which then uses this known sequence to deduce the NLS observation data. In this application scenario, consider the following Ricean channel, where the channel between any user and the base station is a superposition of line-of-sight and non-line-of-sight components, denoted as: ,in, For the Loss component, For large-scale fading components of the LoS component, the following condition must be satisfied: N > 0 represents antenna data. This is the array response vector from the user to the base station antenna array. (n=1, 2, ..., N) represents the distance from the user to the nth antenna of the base station antenna array. The calculation formula can be expressed as: ,in, Let represent the y and z coordinates of the nth antenna, respectively; The carrier wavelength. For NLoS components, For from the scatterer l The gain of multipath propagation, For large-scale fading components of NLoS components, This is the non-line-of-sight (NLoS) correlation matrix.

[0030] If the channel map indicates that a LosS path exists, the LosS component is calculated according to the method described above. Therefore, the Loss component can be removed from the pilot signal observed by the base station first, and then the NLoS component can be estimated. Note: , is the residual pilot signal after removing the Loss component, where The signal-to-noise ratio of the pilot signal transmitted by the user. This refers to the observation noise at the base station. Subsequent steps will focus on... The NLoS components of the channel contained in it are estimated.

[0031] S130, determine the feature subspace based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterer.

[0032] One approach to determining the feature subspace based on the non-line-of-sight correlation matrix is ​​to approximate the high-dimensional non-line-of-sight correlation matrix by projecting it onto a low-dimensional subspace using SA-RSLS estimation, thereby obtaining the main signal subspace with low complexity. Another approach to determining the feature subspace based on the spatial distribution information of the scatterers is to construct a set of array response vectors for the scattering paths based on the spatial distribution information of the scatterers, and then determine the feature subspace based on this set of array response vectors.

[0033] Optionally, the method for determining the feature subspace based on the non-line-of-sight correlation matrix can be as follows: performing random sketch projection on the non-line-of-sight correlation matrix to obtain a sketch matrix; extracting orthogonal bases from the sketch matrix to obtain an orthogonal basis matrix; performing low-dimensional eigenvalue decomposition on the non-line-of-sight correlation matrix based on the orthogonal basis matrix to obtain an eigenvector matrix; and performing subspace recovery on the eigenvector matrix based on the orthogonal basis matrix to obtain the feature subspace.

[0034] In this embodiment, the sketch matrix can be obtained by performing random sketch projection on the non-line-of-sight correlation matrix. This can be achieved by: generating a random matrix; and then applying the random matrix to the non-line-of-sight correlation matrix. Perform random sketch projection to obtain the sketch matrix.

[0035] Among them, random matrix The elements are independent and follow a complex Gaussian distribution with zero mean and unit variance. .in, , The sketch dimension can be determined by the rank of the non-line-of-sight correlation matrix. This is an oversampling margin used to ensure approximate accuracy. It is typically expressed as a percentage of the number of antennas N. The formula for calculating the sketch matrix can be expressed as: Its column space is approximately equivalent to The column space, that is, the reserved The subspace containing the main rank components.

[0036] In this embodiment, orthogonal basis extraction is performed on the sketch matrix. The orthogonal basis matrix can be obtained by performing a compact QR decomposition on the sketch matrix. ,in The column vectors form an orthogonal basis. It is an upper triangular matrix. Because... column space approximation The signal subspace, therefore Can be regarded as An orthogonal approximate basis for the subspace. Compared to direct pairs Eigenvalue decomposition and QR decomposition have lower overhead, and The dimension is much smaller than the number of antennas N.

[0037] In this embodiment, the method of obtaining the eigenvector matrix by performing low-dimensional eigenvalue decomposition on the non-line-of-sight correlation matrix based on the orthogonal basis matrix can be as follows: project the non-line-of-sight correlation matrix into the orthogonal basis matrix space to obtain a low-dimensional matrix; perform eigenvalue decomposition on the low-dimensional matrix to obtain the eigenvector matrix.

[0038] Projecting the non-line-of-sight correlation matrix onto the orthogonal basis matrix space yields a low-dimensional matrix that can be represented as: The eigenvalue decomposition of a low-dimensional matrix can be expressed as: Obtain the eigenvector matrix and only keep the largest The eigenvalue diagonal matrix of eigenvalues. .because Dimensions are much smaller Compared to directly targeting By performing compact eigenvalue decomposition, the computational complexity of this step is significantly reduced.

[0039] In this embodiment, the eigenvector matrix is ​​subspaced based on orthogonal basis matrices to obtain the eigenspace. One possible method is to map the low-dimensional eigenvectors back to the original space. In order to achieve An approximation of the signal subspace. Matrix The column vectors constitute A set of bases for the principal feature subspace. When the number of antennas... hour, It will converge to the spatial correlation matrix. The characteristic subspace.

[0040] The main computational overhead of the aforementioned SA-RSLS scheme comes from matrix multiplication. and The complexity is approximately Compared to the original RSLS method The complexity, in , , At that time, SA-RSLS reduced the complexity from Down to The magnitude is reduced by approximately three orders of magnitude. Therefore, SA-RSLS achieves a leap from non-real-time to real-time estimation of ultra-large-scale MIMO channels with almost no loss of estimation accuracy. This is especially true for spatial correlation matrices. When the data can be approximated by a model or finite sampling, SA-RSLS provides a practical and efficient method for estimating NLoS channels.

[0041] Optionally, the method for determining the feature subspace based on the spatial distribution information of the scatterers can be: determining the array response vector of each scatterer to the base station antenna array based on the spatial distribution information of the scatterers and the spatial distribution information of the base station; and performing compact QR decomposition on the array response matrix composed of the array response vectors of each scatterer to the base station antenna array to obtain the feature subspace.

[0042] In this embodiment, the prior scatterer position is provided by the channel map. Calculate the array response vector of the base station antenna array incident on each scatterer. , The array response vectors of the L main scatterers are arranged sequentially to form a matrix. .right Perform compact QR decomposition ,in The columns are full rank and the column vectors are pairwise orthogonal. It is an upper triangular matrix. Where, the matrix... The column space is spanned by the direction vectors of L scattering paths, and can be considered as described in the above embodiment. An approximation signal subspace, i.e., a characteristic subspace. When the number of antennas... hour, Will converge to the truth ,in To The eigenvector matrix obtained by performing compact eigenvalue decomposition. for Rank.

[0043] The CM-RSLS scheme described above requires only L principal scatterer coordinates and a total of 3L real parameters, relative to the complete spatial correlation matrix. of In terms of the number of complex parameters, the prior cost is extremely low. Furthermore, the main computational cost comes from the computational complexity of the complex numbers. The compact QR decomposition has a complexity of approximately In real-world scenarios, the number of effective scatterers L in the environment is usually much smaller than the number of antennas N, therefore, compared to the original RSLS method... The CM-RSLS scheme can significantly reduce computational complexity compared to the previous method. While the estimation performance of CM-RSLS may slightly decrease when there are small errors in the scatterer position, this is because... It still captures the main channel energy direction and is generally superior to traditional methods that do not utilize scattering information. In summary, CM-RSLS makes full use of the environmental prior of the channel map, and is superior to existing estimation schemes without prior knowledge in both complexity and accuracy.

[0044] S140, determine the non-line-of-sight channel components based on the feature subspace and non-line-of-sight observation data.

[0045] Optionally, the method for determining the non-line-of-sight channel components based on the feature subspace and the non-line-of-sight observation data can be as follows: obtaining the pilot signal-to-noise ratio; performing dimensionality reduction processing on the non-line-of-sight observation data based on the feature subspace to obtain dimensionality-reduced non-line-of-sight observation data; performing channel estimation based on the dimensionality-reduced non-line-of-sight observation data and the pilot signal-to-noise ratio to obtain the non-line-of-sight channel components within the subspace; and performing dimensionality recovery on the non-line-of-sight channel components within the subspace based on the feature subspace to obtain the non-line-of-sight channel components.

[0046] In this embodiment, for the feature subspace determined based on SA-RSLS estimation NLoS observation data Projected onto a lower-dimensional subspace The channel coefficients are estimated using LS and then restored to the original space. Specifically, the channel coefficients are first calculated... This yields effective observations after dimensionality reduction; LS estimation is performed in the N-dimensional subspace, and then projected back to the N-dimensional subspace to obtain the NLoS channel estimate, denoted as: , The signal-to-noise ratio of the pilot signal transmitted by the user.

[0047] In this embodiment, for the feature subspace determined based on CM-RSLS estimation NLoS observation signal Projected onto a lower-dimensional subspace The channel coefficients are estimated using LS and then restored to the original space. Specifically, the channel coefficients are first calculated... We obtain the effective observations after dimensionality reduction; we perform LS estimation in the L-dimensional subspace, and then project it back to N dimensions to obtain the NLoS channel estimate, denoted as: .

[0048] S150, determine the target channel estimation result based on the non-line-of-sight channel components.

[0049] Optionally, the method for determining the target channel estimation result based on the non-line-of-sight channel component can be as follows: determine the user spatial distribution information and base station spatial distribution information based on the channel map; determine the line-of-sight channel component based on the user spatial distribution information and base station spatial distribution information; and accumulate the non-line-of-sight channel component and the line-of-sight channel component to obtain the target channel estimation result.

[0050] The determination of the line-of-sight channel components based on user spatial distribution information and base station spatial distribution information can be found in the above embodiments and will not be repeated here. Specifically, after obtaining the estimate of the NLoS channel components... Then the Los component Adding them together gives the complete channel estimate. .

[0051] The technical solution of this embodiment determines the spatial distribution information of scatterers based on a channel map; acquires non-line-of-sight (NLS) observation data and a NLS correlation matrix; wherein the NLS observation data is obtained through pilot training; determines a feature subspace based on the NLS correlation matrix or the spatial distribution information of scatterers; determines the NLS channel components based on the feature subspace and the NLS observation data; and determines the target channel estimation result based on the NLS channel components. The channel estimation method provided by this embodiment determines the feature subspace based on the NLS correlation matrix, randomly projects the high-dimensional correlation matrix to a low dimension, significantly reducing computational complexity and achieving rapid extraction of the main channel subspace; determining the feature subspace based on the spatial distribution information of scatterers further reduces computational complexity and obtains a refined estimate with near-optimal performance; it can improve estimation accuracy and computational efficiency.

[0052] Based on the above embodiments, Figure 3 This is a flowchart of a channel estimation method according to an embodiment of the present invention, such as... Figure 3 As shown, the method includes the following steps: S310 obtains spatial distribution information of users, base stations, and major scatterers through channel maps.

[0053] S320 obtains the NLoS part of the observation data through pilot training.

[0054] S330: Determine if the channel map is accurate enough (or determine if the channel map error is less than the set value). If yes, proceed to S340; otherwise, proceed to S360.

[0055] S340, obtains the dimension-reduced subspace matrix based on the channel map.

[0056] S350, LS channel estimation in a reduced-dimensional subspace.

[0057] S360 obtains the dimension-reduced subspace matrix based on the sketch algorithm.

[0058] S370, LS channel estimation in a reduced-dimensional subspace.

[0059] S380 outputs the complete channel estimation results.

[0060] Figure 4 The figure illustrates the NMSE performance of two channel estimation methods included in this invention under different antenna sizes. The comparison schemes include MMSE channel estimation, LS channel estimation, and GA-RSLS channel estimation using a real spatial correlation matrix. As shown in the figure, the estimation error of all schemes decreases with increasing antenna number. However, the CM-RSLS and SA-RSLS schemes proposed in this invention significantly outperform traditional LS estimation and RSLS methods based on real or approximate spatial correlation matrices. SA-RSLS, through a low-dimensional subspace constructed using a sketch matrix, stably approximates the geometrically approximated GA-RSLS method, exhibiting highly consistent performance across the entire antenna size range. CM-RSLS, relying on the scatterer structure prior provided by the channel map, gradually converges to the optimal result of GA-RSLS as the antenna size increases. In ultra-large-scale MIMO scenarios, both schemes of this invention can approach the upper limit of the ideal MMSE performance.

[0061] Figure 5The comparison results of the running time required to construct the subspace matrix under different antenna scales are presented. It can be seen that GA-RSLS is the most time-consuming due to the need for eigenvalue decomposition of the complete high-dimensional covariance matrix; SA-RSLS, with the help of sketch dimensionality reduction, only needs to perform QR decomposition and eigenvalue decomposition on the low-dimensional matrix, significantly reducing its running time compared to GA-RSLS; while CM-RSLS directly generates the subspace basis using the scatterer positions in the channel map, without performing eigenvalue decomposition, thus achieving the highest overall running efficiency. Specifically, under a typical 1024 antenna scale, the SA-RSLS and CM-RSLS methods of this invention reduce the running time by approximately two and three orders of magnitude, respectively, compared to GA-RSLS. This demonstrates that this invention not only approaches optimal performance in estimation accuracy but also significantly outperforms existing schemes in running efficiency, making it particularly suitable for efficient, real-time channel estimation scenarios in 6G ultra-large-scale MIMO environments.

[0062] Example 2 Figure 6 This is a schematic diagram of the structure of a channel estimation device provided in Embodiment 2 of the present invention, as shown below. Figure 6 As shown, the device includes: The scatterer spatial distribution information determination module 610 is used to determine the scatterer spatial distribution information based on the channel map; The data acquisition module 620 is used to acquire non-line-of-sight observation data and a non-line-of-sight correlation matrix; wherein, the non-line-of-sight observation data is obtained through pilot training; The feature subspace determination module 630 is used to determine the feature subspace based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterer. The non-line-of-sight channel component determination module 640 is used to determine the non-line-of-sight channel components based on the feature subspace and the non-line-of-sight observation data; The target channel estimation result determination module 650 is used to determine the target channel estimation result based on the non-line-of-sight channel components.

[0063] Optionally, the target channel estimation result determination module 650 is also used for: Based on the channel map, determine the user spatial distribution information and the base station spatial distribution information; The line-of-sight channel component of the channel is determined based on the user spatial distribution information and the base station spatial distribution information; The non-line-of-sight channel component is summed with the line-of-sight channel component to obtain the target channel estimation result.

[0064] Optionally, the feature subspace determination module 630 is also used for: Random sketch projection is performed on the non-line-of-sight correlation matrix to obtain a sketch matrix; Orthogonal basis extraction is performed on the sketch matrix to obtain the orthogonal basis matrix; Based on the orthogonal basis matrix, the non-line-of-sight correlation matrix is ​​decomposed into a low-dimensional feature vector matrix. Based on the orthogonal basis matrix, the eigenvector matrix is ​​subspaced to obtain the eigenspace.

[0065] Optionally, the feature subspace determination module 630 is also used for: Generate a random matrix; wherein the elements of the random matrix independently follow a complex Gaussian distribution with zero mean and unit variance; Based on the random matrix, a random sketch projection is performed on the non-line-of-sight correlation matrix to obtain a sketch matrix.

[0066] Optionally, the feature subspace determination module 630 is also used for: The non-line-of-sight correlation matrix is ​​projected onto the orthogonal basis matrix space to obtain a low-dimensional matrix; The low-dimensional matrix is ​​subjected to eigenvalue decomposition to obtain the eigenvector matrix.

[0067] Optionally, the feature subspace determination module 630 is also used for: The array response vector of each scatterer to the base station antenna array is determined based on the spatial distribution information of the scatterers and the spatial distribution information of the base station. Compact QR decomposition is performed on the array response matrix, which is composed of the array response vectors from each scatterer to the base station antenna array, to obtain the characteristic subspace.

[0068] Optionally, the non-line-of-sight channel component determination module 640 is also used for: Obtain the signal-to-noise ratio of the pilot signal; The non-line-of-sight observation data is dimensionality reduced based on the feature subspace to obtain dimensionality-reduced non-line-of-sight observation data. Based on the dimensionality-reduced non-line-of-sight observation data and the signal-to-noise ratio of the pilot signal, channel estimation is performed to obtain the non-line-of-sight channel components in the subspace. Based on the feature subspace, the dimension of the non-line-of-sight channel components within the subspace is restored to obtain the non-line-of-sight channel components.

[0069] The above-described apparatus can execute the methods provided in all the foregoing embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the above methods. Technical details not described in detail in this embodiment can be found in the methods provided in all the foregoing embodiments of the present invention.

[0070] Example 3 Figure 7A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components, connections and relationships between components, and their functions shown herein are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0071] like Figure 7 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0072] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0073] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as channel estimation methods.

[0074] In some embodiments, the channel estimation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded into and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the channel estimation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the channel estimation method by any other suitable means (e.g., by means of firmware).

[0075] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0076] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0077] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0078] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0079] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0080] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0081] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the channel estimation method provided in any embodiment of this application.

[0082] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0083] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and no limitation is imposed herein.

[0084] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A channel estimation method, characterized in that, include: Determine the spatial distribution information of scatterers based on channel maps; Obtain non-line-of-sight observation data and non-line-of-sight correlation matrix; wherein, the non-line-of-sight observation data is obtained through pilot training; The feature subspace is determined based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterers; The non-line-of-sight channel components are determined based on the feature subspace and the non-line-of-sight observation data; The target channel estimation result is determined based on the non-line-of-sight channel components.

2. The method according to claim 1, characterized in that, The target channel estimation result is determined based on the non-line-of-sight channel components, including: Based on the channel map, determine the user spatial distribution information and the base station spatial distribution information; The line-of-sight channel component of the channel is determined based on the user spatial distribution information and the base station spatial distribution information; The non-line-of-sight channel component is summed with the line-of-sight channel component to obtain the target channel estimation result.

3. The method according to claim 1, characterized in that, Determining the feature subspace based on the non-line-of-sight correlation matrix includes: Random sketch projection is performed on the non-line-of-sight correlation matrix to obtain a sketch matrix; Orthogonal basis extraction is performed on the sketch matrix to obtain the orthogonal basis matrix; Based on the orthogonal basis matrix, the non-line-of-sight correlation matrix is ​​decomposed into a low-dimensional feature vector matrix. Based on the orthogonal basis matrix, the eigenvector matrix is ​​subspaced to obtain the eigenspace.

4. The method according to claim 3, characterized in that, Perform random sketch projection on the non-line-of-sight correlation matrix to obtain a sketch matrix, including: Generate a random matrix; wherein the elements of the random matrix are independent and follow a complex Gaussian distribution with zero mean and unit variance; Based on the random matrix, a random sketch projection is performed on the non-line-of-sight correlation matrix to obtain a sketch matrix.

5. The method according to claim 3, characterized in that, Based on the orthogonal basis matrix, a low-dimensional eigenvalue decomposition is performed on the non-line-of-sight correlation matrix to obtain an eigenvector matrix, including: The non-line-of-sight correlation matrix is ​​projected onto the orthogonal basis matrix space to obtain a low-dimensional matrix; The low-dimensional matrix is ​​subjected to eigenvalue decomposition to obtain the eigenvector matrix.

6. The method according to claim 2, characterized in that, Determining the feature subspace based on the spatial distribution information of the scatterers includes: The array response vector of each scatterer to the base station antenna array is determined based on the spatial distribution information of the scatterers and the spatial distribution information of the base station. Compact QR decomposition is performed on the array response matrix, which is composed of the array response vectors from each scatterer to the base station antenna array, to obtain the characteristic subspace.

7. The method according to claim 1, characterized in that, Determining the non-line-of-sight channel components based on the feature subspace and the non-line-of-sight observation data includes: Obtain the signal-to-noise ratio of the pilot signal; The non-line-of-sight observation data is dimensionality reduced based on the feature subspace to obtain dimensionality-reduced non-line-of-sight observation data. Based on the dimensionality-reduced non-line-of-sight observation data and the signal-to-noise ratio of the pilot signal, channel estimation is performed to obtain the non-line-of-sight channel components in the subspace. Based on the feature subspace, the dimension of the non-line-of-sight channel components within the subspace is restored to obtain the non-line-of-sight channel components.

8. A channel estimation device, characterized in that, include: The scatterer spatial distribution information determination module is used to determine the scatterer spatial distribution information based on the channel map; The data acquisition module is used to acquire non-line-of-sight observation data and non-line-of-sight correlation matrix; wherein, the non-line-of-sight observation data is obtained through pilot training; A feature subspace determination module is used to determine the feature subspace based on the non-line-of-sight correlation matrix or the spatial distribution information of the scatterers; A non-line-of-sight channel component determination module is used to determine non-line-of-sight channel components based on the feature subspace and the non-line-of-sight observation data; The target channel estimation result determination module is used to determine the target channel estimation result based on the non-line-of-sight channel components.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the channel estimation method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the channel estimation method according to any one of claims 1-7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the channel estimation method as described in any one of claims 1-7.

Images