A near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method and system

Abstract

The present invention discloses a method and system for designing ultra-large-scale MIMO non-orthogonal pilots assisted by near-field channel maps. Considering the near-field communication scenario, the base station side obtains the statistical channel information of each user based on the uplink detection signal sent by each user on different time-frequency resources, extracts the spatial characteristics of the polar channel power distribution, and then draws the near-field channel map; using the second-order covariance matrix of the statistical channel information of each user, an integer programming problem is constructed to minimize the channel correlation of users who reuse the same pilot, and low-complexity non-orthogonal pilot scheduling is achieved with the assistance of the near-field channel map. The present invention takes into account the near-field polar domain sparse characteristics of the channel between the user and the ultra-large-scale antenna array of the base station, extracts the polar channel spatial characteristics from the statistical channel information to draw the near-field channel map, and allocates non-orthogonal pilot signals to each scheduled user accordingly. The present invention can further improve the accuracy of system channel estimation while reducing the pilot overhead of the system.

Description

Technical Field

[0001] The present invention relates to the field of multiple-input multiple-output (MIMO) wireless communications using ultra-large-scale antenna arrays, and in particular to a near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method and system. Background Art

[0002] With the continued growth in demand for mobile internet and Internet of Things applications, wireless mobile communications are developing rapidly. To meet the demands of future mobile communication applications, it is necessary to fully exploit spatial wireless resources and significantly improve the spectrum and power efficiency of wireless communications. MIMO wireless transmission technology, which employs multiple antennas for transmission and reception, is a fundamental technology for improving the spectrum and power efficiency of wireless communications. However, compared to massive MIMO, which deploys a massive antenna array (more than dozens) at the base station, ultra-massive MIMO further increases the number of antenna elements and the array aperture. This introduces near-field propagation characteristics into the wireless channel, leading to an order of magnitude increase in the spatial degrees of freedom and the number of channel parameters to be estimated. Traditional methods for acquiring channel state information based on orthogonal pilots suffer from the bottleneck problem of pilot overhead, which increases linearly with factors such as the number of transmit antennas or space-time-frequency. To overcome this limitation, non-orthogonal pilot technology has been introduced, where the pilot signals used by different users can be non-completely orthogonal. By reducing the system's pilot overhead, non-orthogonal pilot design and ultra-massive MIMO can be combined for more efficient information transmission.

[0003] Existing non-orthogonal pilot design inventions are all based on the far-field plane wave assumption. However, due to the increase in the number of VLSI antenna elements and array aperture, VLSI wireless channels exhibit near-field propagation characteristics. User channels exhibit polar sparsity. Based on the scattering environment information implicit in statistical channel state information (s-CSI) or instantaneous channel state information (i-CSI), channel charting is a technology that uses the steady-state characteristics of the wireless channel to characterize the wireless transmission environment. Channel charting aims to map high-dimensional s-CSI or i-CSI into low-dimensional virtual coordinates that match the channel characteristics, so that users with similar low-dimensional virtual coordinates have similar channel characteristics. Generally speaking, in real environments, users with close physical locations often have strong correlations with steady-state characteristics in the channel that are strongly correlated with the environment. Conversely, if two users have highly correlated steady-state characteristics, they are likely to be physically close. Channel charting technology helps fully exploit the spatial characteristics of user channels and reconstruct a polar-domain neighbor relationship map of users at the base station, thereby achieving low-complexity pilot scheduling. Taking advantage of the fact that ultra-large-scale MIMO near-field channels usually exhibit strong polar sparsity and the technical tools of channel maps, it is necessary to study a ultra-large-scale MIMO non-orthogonal pilot design scheme that takes into account the near-field characteristics of the channel. Summary of the Invention

[0004] Purpose of the invention: In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a method and system for designing ultra-large-scale MIMO non-orthogonal pilots that utilize the sparse characteristics of the near-field polar domain. For the scenario where an ultra-large-scale antenna array is deployed on the base station side, it can reduce the system's pilot overhead and further reduce the complexity of implementing this non-orthogonal pilot scheduling process, making it easier to implement.

[0005] Technical solution: To achieve the above-mentioned purpose, the present invention adopts the following technical solution:

[0006] A near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method comprises the following steps:

[0007] Based on the uplink sounding signals received from each user on different time-frequency resources, the channel covariance matrix of each user is obtained. The user channel covariance matrix is ​​decomposed using the near-field channel codebook, and the spatial characteristics of the polar domain (angular distance domain) channel power distribution are extracted to draw the near-field channel spectrum.

[0008] The near-field channel map that reflects the user channel correlation is used to assist in constructing an integer programming problem to minimize the channel correlation of users reusing the same pilot. The near-field channel map is used to assist in implementing a non-orthogonal pilot scheduling scheme.

[0009] Furthermore, a very large-scale antenna array is deployed on the base station side. The wireless channel between the user and the base station antenna array is a spatial non-stationary channel. The channel vector between user k and the base station antenna array is The expression is: Where b(θ,ρ) is the base station antenna array response vector that obeys the spherical wave assumption corresponding to the incident angle θ and the distance ρ between the user terminal or scatterer and the base station, and h k (θ,ρ) is the path gain of the wireless channel of user k corresponding to the incident angle θ and spacing ρ, is the near-field area of ​​the base station antenna array, and M is the number of base station antennas.

[0010] Furthermore, the sum of the user channel correlations that reuse the same pilot is expressed as in, is the orthogonal pilot set available at the base station side, is the second-order statistical characteristic of the statistical channel information of user k, that is, the channel covariance matrix, is the set of users using the pilot sequence t, ||·||2 represents the two-norm of the matrix, represents expectation, and ∑{·} represents summation.

[0011] Furthermore, the ultra-large-scale array antenna on the base station side of the ultra-large-scale MIMO wireless communication system includes more than hundreds of antenna units, and the spacing between each antenna unit is less than the wavelength of the carrier. When each antenna adopts an omnidirectional antenna, a 120-degree sector antenna, or a 60-degree sector antenna, the spacing between each antenna is 1 / 2 wavelength, 1 / 3 wavelength, or 1 wavelength, respectively; each antenna unit adopts a single-polarization or multi-polarization antenna; and the uplink and downlink communications adopt time division duplexing TDD or frequency division duplexing FDD.

[0012] Furthermore, the uplink detection signals sent by each user on different time-frequency resources are orthogonal to each other. The base station estimates the second-order statistical characteristics of the statistical channel information of each user by using the sample enhanced averaging method based on the received uplink detection signals, that is, the channel covariance matrix Φ of each user channel on the base station antenna array k .

[0013] Furthermore, according to the formula Use the near-field channel codebook W to calculate the user channel covariance matrix Φ k Decompose and extract the spatial characteristics of the polar channel power distribution k=1,…K, where K is the number of users.

[0014] Furthermore, the spatial characteristics of user polar channel power distribution are used k=1,…K, as the input features of the near-field channel map, the dissimilarity measure matrix is ​​calculated using the cosine similarity criterion; the calculated dissimilarity measure matrix is ​​reduced in dimension to obtain the two-dimensional virtual coordinates c of each user node on the map k , generate the near-field channel spectrum C=[c1,…,c K ]; The distance between user nodes in the obtained near-field channel map not only reflects their actual physical distance but also indicates the size of their channel correlation.

[0015] Furthermore, based on the criterion of minimum channel correlation for users using the same pilot, the Euclidean distance between the two-dimensional virtual coordinates of each user node in the near-field channel map is used to characterize the size of the channel correlation, and the original integer programming problem is converted into a node grouping problem. The sum of the channel correlations between users using the same pilot is minimized by maximizing the distance between nodes in the same group. Based on the generated pilot codebook and near-field channel map, the users and available pilot resources in the cell are scheduled, and the non-orthogonal pilot scheduling scheme, that is, the pilot signal used by each user, is determined. The non-orthogonal pilot scheduling is completed by an exhaustive search algorithm or a nearest neighbor search algorithm (greedy algorithm).

[0016] The present invention also provides a near-field channel spectrum-assisted ultra-large-scale MIMO wireless communication method, comprising:

[0017] The base station is equipped with a very large-scale array antenna. Based on its available pilot resources, the base station generates a pilot codebook consisting of several orthogonal pilot sequences.

[0018] During the uplink channel detection and pilot scheduling phase, a non-orthogonal pilot scheduling scheme is obtained based on the near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method; the base station performs non-orthogonal pilot scheduling based on the generated pilot codebook to determine the pilot signal used by each user;

[0019] During the uplink channel training phase, each user periodically transmits its allocated uplink pilot signal on the same time-frequency resource. The base station uses the received uplink pilot signal and the statistical information of each user's channel to estimate the channel of each user.

[0020] The present invention also provides a computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the computer system implements the steps of the near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method or the steps of the near-field channel spectrum-assisted ultra-large-scale MIMO wireless communication method.

[0021] Beneficial effects: The present invention models the uplink training phase of ultra-large-scale MIMO based on statistical channel state information that is relatively easy to obtain, and constructs a near-field channel spectrum that reflects channel correlation by extracting the spatial characteristics of the polar channel power distribution based on the channel covariance matrix; compared to the channel estimation method that relies on orthogonal pilot training, the non-orthogonal pilot design method adaptively schedules the available pilot resources of the base station based on the statistical channel information and second-order statistical characteristics of each user, which can reduce the training overhead of user channel information acquisition in the ultra-large-scale MIMO system. The present invention further uses the near-field channel spectrum to assist in the non-orthogonal pilot scheduling of the ultra-large-scale MIMO system, which can achieve the goal of reducing pilot overhead while ensuring the accuracy of channel estimation, and reduce the complexity of solving the optimization problem and the physical layer implementation, thereby speeding up the operation and improving the flexibility of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 Schematic diagram of the overall method flow of an embodiment of the present invention;

[0023] Figure 2 Schematic diagram of the far-field and near-field propagation characteristics of a wireless channel in an embodiment of the present invention;

[0024] Figure 3 Schematic diagram of the ultra-large-scale MIMO channel estimation process assisted by near-field channel spectrum in an embodiment of the present invention;

[0025] Figure 4Schematic diagram of the steps for generating a near-field channel map in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0027] An embodiment of the present invention discloses a near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method. The base station obtains the channel covariance matrix of each user based on the uplink detection signal sent by each user on different time-frequency resources, and decomposes the user channel covariance matrix using a near-field channel codebook to extract the spatial characteristics of the polar channel power distribution and draw a near-field channel spectrum. The near-field channel spectrum that reflects the user channel correlation is used to assist in constructing an integer programming problem that minimizes the user channel correlation of multiplexing the same pilot, and the non-orthogonal pilot scheduling scheme is implemented with the assistance of the near-field channel spectrum.

[0028] Figure 1 In the exemplary scenario shown, an ultra-large-scale antenna array is deployed on the base station side. The figure considers the case of a single-cell base station, and the base station side is configured with an antenna array containing more than hundreds of antenna units; then a corresponding near-field spherical wave channel model is established, and the base station side generates a pilot codebook consisting of several orthogonal pilot sequences based on its available pilot resources and obtains statistical channel state information of each user; then, based on the statistical channel information, the user polar channel power distribution is extracted as the spatial feature to calculate the dissimilarity measure matrix, construct a near-field channel spectrum, and the non-orthogonal pilot scheduling problem in the multi-user scenario is established as an integer programming problem to minimize the channel correlation of users reusing the same pilot; the base station side completes non-orthogonal pilot scheduling based on the generated near-field channel spectrum and allocates pilot signals to each user; the base station side performs channel estimation based on the pilot signals sent by the users.

[0029] In this embodiment, a nearest neighbor search algorithm assisted by a near-field channel spectrum is used to solve the optimization problem, specifically including:

[0030] Based on user statistical channel information, the user polar channel power distribution is extracted as a spatial feature. Cosine similarity is introduced as a dissimilarity metric to calculate the dissimilarity matrix and construct a near-field channel map. This ensures that the user spacing on the near-field channel map reflects both the actual physical location spacing and the channel correlation between them. Based on this map, a nearest neighbor search algorithm is used to assign different pilot sequences to adjacent user nodes, thereby completing non-orthogonal pilot scheduling.

[0031] The integer programming problem is converted into a node search problem, and non-orthogonal pilot scheduling is achieved by grouping users.

[0032] The method of this embodiment is described in more detail below with reference to a specific scenario.

[0033] Part 1: Building a Very Large Scale MIMO System Model Deploying Very Large Scale Antenna Arrays

[0034] Specifically, consider a single-cell base station ultra-large-scale MIMO system. The base station is configured with an antenna array containing hundreds or more antenna elements. The ultra-large-scale antenna array can adopt linear arrays, circular arrays, plate arrays, or other array structures. Each antenna element can adopt an omnidirectional antenna or a sector antenna. When each antenna element adopts an omnidirectional antenna, a 120-degree sector antenna, or a 60-degree sector antenna, the spacing between antennas can be configured to be 1 / 2 wavelength, 1 / 3 wavelength, and 1 wavelength. Each antenna element can adopt a single-polarization or multi-polarization antenna. A corresponding near-field spherical wave channel model is then established. The base station generates a pilot codebook consisting of τ orthogonal pilot sequences based on its available pilot resources and obtains statistical channel state information for each user. Based on this statistical channel information, the user polar domain channel power distribution is extracted as a spatial feature, and a dissimilarity metric matrix is ​​calculated to construct a near-field channel map.

[0035] In this embodiment, only narrowband channels are considered. In this narrowband channel, there is only a single composite path. This narrowband channel can be considered a single subcarrier channel in a conventional wideband OFDM system. Time division duplex (TDD) or frequency division duplex (FDD) transmission schemes are considered. Assume that the number of antennas deployed at the base station is M, the number of users is K, and each user is equipped with a single antenna.

[0036] Figure 2 Schematic diagram of the far-field and near-field propagation characteristics of ultra-large-scale MIMO wireless channels, where the electromagnetic wavefront propagates as a plane wave in the far field (the array steering vector is only related to the angle parameter) and as a spherical wave in the near field (the array steering vector is related to both the angle and distance parameters).

[0037] Figure 3 This is a schematic diagram of the ultra-large-scale MIMO channel estimation process assisted by the near-field channel map in an embodiment of the present invention, in which the base station side is equipped with an ultra-large-scale antenna array, estimates the high-dimensional quasi-static channel based on the uplink detection signal sent by the user, obtains statistical channel information to construct the near-field channel map, and then allocates the pilot sequence to estimate the low-dimensional time-varying channel.

[0038] Figure 4 Schematic diagram of the steps for generating a near-field channel map in an embodiment of the present invention.

[0039] To reduce pilot overhead in ultra-large-scale MIMO wireless communication systems, different users within a cell can simultaneously use the same pilot from a base station-generated codebook to estimate uplink pilot channel parameters by leveraging the extreme-domain sparseness of each user's channel. During the uplink channel training phase, each scheduled user transmits its assigned pilot signal. The pilot signals used by different users are not required to be completely orthogonal, and different users can simultaneously use the same pilot from the codebook. The base station processes the received pilot signals to achieve minimum mean square error channel estimation for each scheduled user.

[0040] Assume that there are K scheduled single-antenna users in the cell, the number of pilots is τ, and Represents the set of scheduled users, represents the set of available orthogonal pilot sequence numbers, and k represents the user number. is the set of users using pilot sequence t. In the non-orthogonal pilot scheduling scheme In this case, the number of pilots is less than the number of scheduled users in the cell, that is, τ is less than K, and the pilot sequence length is equal to τ. When the number of pilots is not less than 80% of the number of scheduled users in the cell, this non-orthogonal pilot design scheme can still guarantee the system base station side channel estimation performance.

[0041] Assume that τ pilot sequences are orthogonal. In the uplink training phase, the pilot sequence assigned by the base station is sent to the kth user, that is, the pilot signal vector is sent, with [h k ] m represents the channel parameters between the kth user and the mth antenna on the base station side in the current training cycle, is the channel vector between user k and the base station antenna array, in, corresponds to the incident angle The base station antenna array response vector obeys the spherical wave assumption with the user terminal (or scatterer) and the base station spacing ρ, and the carrier frequency is f c , the speed of light is c, is the incident angle in radians, d is the distance between antenna elements in the base station antenna array, h k (θ,ρ) is the path gain of the wireless channel of user k corresponding to the incident angle θ and spacing ρ, is the near-field area of ​​the base station antenna array.

[0042] is the second-order statistical characteristic of the statistical channel information of user k, that is, the channel covariance matrix, and the specific expression is:

[0043]

[0044] Among them, ξ k is the large-scale fading coefficient of user k, fk (θ,ρ) is the channel polar domain power spectrum of user k.

[0045] The base station performs channel estimation based on the received pilot signal, obtaining the estimated value and mean square error of each user channel. Taking the minimum mean square error (MMSE) channel estimation performed on the base station side as an example, the total error of the channel estimation for all users in the system is calculated as follows:

[0046]

[0047] Among them, ε k is the channel estimation error of user k, ζ is the uplink training signal-to-noise ratio, and I is the identity matrix.

[0048] Part II: Design and Scheduling of Non-Orthogonal Pilots in Very Large-Scale MIMO Systems

[0049] A very large-scale MIMO non-orthogonal pilot design method utilizes the sparse characteristics of the channel extreme region. A very large-scale antenna array is deployed on the base station side. By establishing a near-field channel model and constructing an integer programming problem to minimize the spatial correlation of users reusing the same pilot based on user statistical channel information, a near-field channel map is used to assist in the nearest neighbor search method to optimize the non-orthogonal pilot scheduling scheme. The optimization problem is expressed as:

[0050]

[0051] in, is the set of users communicating on the same time-frequency resource, ||·||2 represents the two-norm of the matrix, represents expectation, and ∑{·} represents summation. The optimization problem is solved using a nearest neighbor search method assisted by near-field channel graph.

[0052] The uplink detection signals sent by each user are mutually orthogonal. The base station estimates the statistical channel information and second-order statistical characteristics of each user, that is, the channel covariance matrix of each user channel on the base station antenna array, based on the received uplink detection signals using the sample enhanced averaging method.

[0053] By introducing the spatial characteristic matrix of the extreme channel power distribution of user k and matrix Establish the channel covariance matrix of user k and its polar channel power distribution spatial characteristics The specific mathematical expression is as follows:

[0054]

[0055] in, represents the near-field polar channel codebook consisting of MS array rudder vectors as column vectors, represents a unitary matrix consisting of M array rudder vectors as column vectors, which satisfies The angle domain shown is uniformly sampled, The distance domain is non-uniformly sampled, S is the number of sampling points in the distance domain, d is the distance between base station array antennas, and λ c is the frequency f c The carrier wavelength, β is a positive adjustment variable (1.8 in this example), and They are respectively the matrix formed by merging the column vectors of the equidistant sampling values ​​of the angular power spectrum of user k at S different sampling distances on the radian [0,2π] and the diagonal matrix formed as the main element after vectorization, that is, the polar domain (angular distance domain) distribution information of the channel power of user k, and the values ​​of each element satisfy Since there is no widely used expression for the near-field user's polar channel power distribution, this example uses the simultaneous orthogonal matching pursuit algorithm to estimate the near-field user's polar channel h k p =W H h k , using 500 Monte Carlo calculation results h k p (h k p ) H The average value of the polar channel power distribution spatial characteristics is used to approximate

[0056] If cosine similarity is selected as the user spatial feature dissimilarity measurement indicator, the dissimilarity index can reflect the correlation of user channel features. The specific formula is as follows:

[0057]

[0058] Among them, e k is a column vector.

[0059] Based on the spatial feature of the user polar area channel power distribution extracted from the channel covariance matrix at the base station side, the global inter-user dissimilarity measure matrix Ξ is calculated. The specific calculation method of each element in Ξ is as follows:

[0060]

[0061] The framework of manifold learning technology assumes that the user channel characteristics are functions of the user's actual physical location on a certain manifold. Using manifold learning, the low-dimensional manifold structure can be restored from the high-dimensional sampled data, so that the geodesic distance between the user's virtual coordinates (the distance between the user points on the manifold) can fully reflect the distance between the user's real locations. Manifold learning technology mainly includes linear dimensionality reduction techniques, such as principal component analysis, and nonlinear dimensionality reduction techniques, such as isometric mapping. This example uses principal component analysis or isometric mapping method as the data dimensionality reduction function f, generates a near-field channel map based on Ξ, and obtains the two-dimensional virtual coordinates c of each user node on the map. k (k=1,…,K), that is The Euclidean distance between the two-dimensional virtual coordinates of each user node in the obtained near-field channel map not only reflects its actual physical distance but also indicates the size of its channel correlation.

[0062] The base station schedules the users and available pilot resources in the cell based on the generated pilot codebook according to the minimum channel correlation criterion of users using the same pilot, and determines the non-orthogonal pilot scheduling scheme. That is, the pilot sequence used by each user, non-orthogonal pilot scheduling is completed by exhaustive or nearest neighbor search algorithm (greedy algorithm), and the node spacing on the near-field channel spectrum is calculated according to the Euclidean distance, that is, Where a and b are n-dimensional vectors.

[0063] The τ nearest neighbor node search algorithm assisted by the near-field channel spectrum is used as an example for explanation.

[0064] The implementation process of the τ nearest neighbor node search algorithm assisted by the near-field channel spectrum includes the following steps:

[0065] Step S1: Initialize user set and pilot set: User set Pilot Set Remaining user set Each pilot sequence multiplexes the user set

[0066] Step S2, initialize pilot allocation: user 1 is used as the k* node, and the k* node uses pilot number 1. The remaining pilot sequences are multiplexed into user sets Update the remaining user collection

[0067] Step S3: For each unassigned pilot signal t=2:τ, select the remaining user sets for it in turn. The user node k with the smallest distance from the k* node in the near-field channel spectrum. For pilot t = 2:τ, the user selection formula is:

[0068]

[0069] Step S4: assign pilot t to user k and update the multiplexed user set of pilot t Remaining user set t←t+1.

[0070] Step S5: If t < τ, return to step S3 to loop; otherwise, go to step S6.

[0071] Step S6: For pilot t=1, verify the distances between all user nodes in the remaining user set and the current k* node in the near-field channel spectrum in turn, and select the node that meets the requirement. If there is user node k, user k is regarded as k* node, and k* node uses pilot number 1.

[0072] Step S7: Update the multiplexed user set of pilot 1 Remaining user set

[0073] Step S8: If Return to step S3 to loop; otherwise, terminate the scheduling.

[0074] An embodiment of the present invention discloses a near-field channel map-assisted ultra-large-scale MIMO wireless communication method, comprising:

[0075] The base station is equipped with ultra-large-scale array antennas (hundreds or more). The base station conducts wireless communications with multiple users on the same time-frequency resources. The uplink and downlink communications use time division duplex (TDD) or frequency division duplex (FDD). The base station generates a pilot codebook consisting of several orthogonal pilot sequences based on its available pilot resources.

[0076] During the uplink channel sounding and pilot scheduling phase, each user sends an uplink sounding signal on different time-frequency resources. The base station obtains a non-orthogonal pilot scheduling scheme based on the near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method. The base station performs non-orthogonal pilot scheduling based on the generated pilot codebook to determine the pilot signal used by each user.

[0077] During the uplink channel training phase, each user periodically transmits its allocated uplink pilot signal on the same time-frequency resource, and the base station performs channel estimation for each user based on the received pilot signal.

[0078] In the communication method of this embodiment, based on the pilot codebook generated by the base station, the pilot signals used by different users communicating on the same resources within the cell are not required to be completely orthogonal. Different users can use the same pilot in the codebook at the same time. The length of the pilot signal and the number of orthogonal pilots can be less than the number of users communicating on the same time-frequency resources within the cell.

[0079] A computer system disclosed in an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the computer system implements the steps of the near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method or the near-field channel spectrum-assisted ultra-large-scale MIMO wireless communication method.

[0080] Anything not described in detail in the present invention is well known to those skilled in the art.

[0081] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method, characterized in that: The steps include: According to the uplink detection signal sent by each user on different time-frequency resources, the channel covariance matrix of each user is obtained, and the user channel covariance matrix is ​​decomposed using the near-field channel codebook to extract the spatial characteristics of the polar channel power distribution and draw the near-field channel map; wherein, according to the formula The channel covariance matrix Φ for user k using the near-field channel codebook W is k Decompose and extract the spatial characteristics of the polar channel power distribution K is the number of users; the spatial characteristics of the user's polar channel power distribution are used As the input feature of the near-field channel map, the cosine similarity criterion is used to calculate the dissimilarity measure matrix; the calculated dissimilarity measure matrix is ​​reduced in dimension to obtain the two-dimensional virtual coordinates c of each user node on the map k , generate the near-field channel spectrum C=[c1,…,c K ]; The near-field channel map that reflects the user channel correlation is used to assist in constructing an integer programming problem to minimize the channel correlation of users reusing the same pilot. The near-field channel map is used to assist in implementing a non-orthogonal pilot scheduling scheme.

2. The near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to claim 1, characterized in that: A very large-scale antenna array is deployed on the base station side. The wireless channel between the user and the base station antenna array is a near-field channel. The channel vector between user k and the base station antenna array is The expression is: Where b(θ,ρ) is the base station antenna array response vector that obeys the spherical wave assumption corresponding to the incident angle θ and the distance ρ between the user terminal or scatterer and the base station, and h k (θ,ρ) is the path gain of the wireless channel of user k corresponding to the incident angle θ and spacing ρ, is the near-field area of ​​the base station antenna array, and M is the number of base station antennas.

3. The near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to claim 1, characterized in that: The sum of the channel correlations of users that reuse the same pilot is expressed as in, is the orthogonal pilot set available at the base station side, is the channel covariance matrix of user k, is the channel vector between user k and the base station antenna array, M is the number of base station antennas, is the set of users using the pilot sequence t, ||·||2 represents the two-norm of the matrix, represents expectation, and ∑{·} represents summation.

4. The near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to claim 1, characterized in that: In the ultra-large-scale MIMO wireless communication system, the ultra-large-scale array antenna on the base station side contains more than hundreds of antenna units. The spacing between each antenna unit is less than the wavelength of the carrier. When each antenna adopts an omnidirectional antenna, a 120-degree sector antenna, or a 60-degree sector antenna, the spacing between each antenna is 1 / 2 wavelength, 1 / 3 wavelength, or 1 wavelength respectively; each antenna unit adopts a single-polarization or multi-polarization antenna; the uplink and downlink communications adopt time division duplex TDD or frequency division duplex FDD.

5. The near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to claim 1, characterized in that: The uplink detection signals sent by each user on different time-frequency resources are mutually orthogonal. The base station uses the sample enhanced averaging method based on the received uplink detection signals to estimate the second-order statistical characteristics of each user's statistical channel information, that is, the channel covariance matrix of each user channel on the base station antenna array.

6. The near-field channel profile-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to claim 1, characterized in that: According to the criterion of minimizing the channel correlation of users using the same pilot, the channel correlation size is characterized by the Euclidean distance between the two-dimensional virtual coordinates of each user node in the near-field channel map, and the integer programming problem is converted into a node grouping problem. The sum of the channel correlations between users using the same pilot is minimized by maximizing the distance between nodes in the same group. Based on the generated pilot codebook and near-field channel map, users and available pilot resources in the cell are scheduled to determine a non-orthogonal pilot scheduling scheme, that is, the pilot signal used by each user. Non-orthogonal pilot scheduling is completed by an exhaustive search or nearest neighbor search algorithm.

7. A near-field channel map-assisted ultra-large-scale MIMO wireless communication method, characterized in that: include: The base station is equipped with a very large-scale array antenna. Based on its available pilot resources, the base station generates a pilot codebook consisting of several orthogonal pilot sequences. In the uplink channel detection and pilot scheduling phase, a non-orthogonal pilot scheduling scheme is obtained according to the near-field channel spectrum-assisted ultra-large-scale MIMO non-orthogonal pilot design method according to any one of claims 1 to 6; The base station performs non-orthogonal pilot scheduling based on the generated pilot codebook and near-field channel spectrum to determine the pilot signal used by each user; During the uplink channel training phase, each user periodically transmits its allocated uplink pilot signal on the same time-frequency resource. The base station uses the received uplink pilot signal and the statistical information of each user's channel to estimate the channel of each user.

8. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is loaded into the processor, it implements the steps of the near-field channel spectrum assisted ultra-large-scale MIMO non-orthogonal pilot design method according to any one of claims 1 to 6, or implements the steps of the near-field channel spectrum assisted ultra-large-scale MIMO wireless communication method according to claim 7.

Images