A low-complexity channel estimation method based on user frequency band and simplified basis expansion

Abstract

The present invention discloses a low-complexity channel estimation method based on user frequency band and simplified basis expansion, which relates to the field of digital communication technology. The method comprises the following steps: obtaining a local time domain signal and extracting a local user frequency band from the local time domain signal; obtaining a frequency domain signal received by a user and extracting a user reception frequency band from the frequency domain signal received by the user; generating basis vectors according to a basis expansion model and performing frequency domain transformation on the basis vectors to obtain a frequency domain transformation matrix; transforming the local user frequency band to obtain a transformation domain symbol A; merging the frequency domain transformation matrices corresponding to all basis vectors to obtain a vector B; multiplying the transformation domain symbol A and the vector B to obtain a solution matrix Q; obtaining a basis coefficient matrix according to the user reception frequency band and the solution matrix; obtaining a time domain channel response matrix according to the basis coefficient matrix, and performing frequency domain transformation on the time domain channel response matrix to obtain a frequency domain channel response matrix. The present invention can provide a channel estimation algorithm with low computational complexity.

Description

Technical Field

[0001] The present invention belongs to the technical field of digital communications, and in particular relates to a low-complexity channel estimation method based on user frequency band and simplified basis expansion. Background Art

[0002] With the rise of low-orbit broadband satellite internet networks such as Starlink and Globalstar, satellite mobile communication systems are developing rapidly. Typical examples include the SpaceX Starlink system, the Iridium satellite system, the Oneweb satellite system, the Teledesic satellite system, and other constellations such as the Hongyan constellation. These constellations often utilize low-orbit satellite communications. The high speeds of low-orbit satellites relative to ground platforms or user equipment can generate significant Doppler offsets, which severely impact communication transmission performance and pose a significant challenge to high-quality transmission. Although terminals perform Doppler pre-compensation before accessing the system receiver, residual Doppler offsets can still be as high as 5kHz, far exceeding the Doppler offsets experienced in terrestrial 5G communications, causing rapid deterioration in system performance.

[0003] Therefore, there is an urgent need to improve the above-mentioned defects in the existing technology and improve the transmission performance of communications. Summary of the Invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a low-complexity channel estimation method based on user frequency band and simplified basis extension. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0005] In a first aspect, the present invention provides a low-complexity channel estimation method based on user frequency band and simplified basis extension, comprising:

[0006] Acquire a local time domain signal, and extract a local user frequency band from the local time domain signal;

[0007] Acquire a frequency domain signal received by the user, and extract a user receiving frequency band from the frequency domain signal received by the user;

[0008] According to the basis expansion model, the basis vectors are generated, and the basis vectors are transformed in the frequency domain to obtain the frequency domain transformation matrix;

[0009] Transform the local user frequency band to obtain the transform domain symbol A; combine the frequency domain transform matrices corresponding to all basis vectors to obtain vector B; multiply the transform domain symbol A and vector B to obtain the solution matrix Q;

[0010] According to the user receiving frequency band and the solution matrix, the basis coefficient matrix is ​​obtained;

[0011] According to the basis coefficient matrix, a time domain channel response matrix is ​​obtained, and the time domain channel response matrix is ​​transformed into the frequency domain to obtain a frequency domain channel response matrix.

[0012] Beneficial effects of the present invention:

[0013] The present invention provides a low-complexity channel estimation method based on user frequency bands and simplified basis extension. This method implements the BEM channel estimation algorithm with low complexity, without compromising the performance of the original BEM channel estimation algorithm. Based on the traditional BEM algorithm, the method first performs BEM channel estimation based on the user frequency band to reduce the matrix computational dimension of the basis extension. Then, by deriving a direct relationship between the basis coefficients and the frequency-domain channel matrix, matrix multiplication is converted into vector multiplication, reducing the complexity of the matrix multiplication. Combining these two points, the BEM channel estimation algorithm is implemented with low computational complexity to obtain channel estimates, enabling its application in practical systems.

[0014] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flow chart of a low-complexity channel estimation method based on user frequency band and simplified basis expansion provided by an embodiment of the present invention;

[0016] Figure 2 is another schematic diagram of a low-complexity channel estimation method based on user frequency band and simplified basis extension provided by an embodiment of the present invention;

[0017] Figure 3 This is a system framework diagram used in the simulation experiment provided by the embodiment of the present invention;

[0018] Figure 4 1 is a schematic diagram of the structure of the pilot sub-band provided in an embodiment of the present invention;

[0019] Figure 5 This is a simulation diagram of the system bit error rate performance when the residual Doppler frequency offset is 5 kHz and 10 kHz, provided by an embodiment of the present invention;

[0020] Figure 6 2. FIG. 1 is a schematic diagram comparing the number of complex multiplications between the UB-F-BEM algorithm provided by an embodiment of the present invention and the traditional BEM algorithm;

[0021] Figure 7 2. FIG. 1 is a schematic diagram comparing the number of complex number additions of the UB-F-BEM algorithm provided by an embodiment of the present invention and the traditional BEM algorithm;

[0022] Figure 8 1. This is a schematic diagram comparing the number of complex multiplications of the UB-F-BEM algorithm provided by an embodiment of the present invention and the patent [CN107018101A];

[0023] Figure 9 2 is a schematic diagram comparing the number of complex number additions of the UB-F-BEM algorithm provided by an embodiment of the present invention and the patent [CN107018101A]. DETAILED DESCRIPTION

[0024] The present invention will be further described in detail below with reference to specific examples, but the embodiments of the present invention are not limited thereto.

[0025] In the prior art, the commonly used basis expansion model channel estimation algorithm (BEM algorithm) uses a limited number of basis functions to fit the channel response of dynamic multipath channels, improving the estimation performance of dynamic multipath channels and being well-suited for dynamic multipath channel systems. The complex exponential basis expansion model (CE-BEM) was first proposed to fit time-varying channels. However, this model suffers from spectral leakage when the Doppler shift is low, resulting in the Gibbs effect. To eliminate the Gibbs effect of CE-BEM, the oversampled basis expansion model (GCE-BEM) was proposed. This model increases the frequency resolution by increasing the sampling frequency, thereby effectively improving modeling accuracy. Subsequently, the discrete Karlo basis expansion model (DKL-BEM) was proposed to fit time-varying channels. This model uses the eigenvectors of the channel autocorrelation function as basis function vectors and exhibits good fitting performance. However, these traditional BEM algorithms involve frequency and time domain data processing, involve a large number of matrix operations, and have high computational complexity, making them difficult to implement in practical hardware. Therefore, simplified algorithms are needed. In the "Fast Time-Varying Channel Estimation Method Based on Simplified Basis Expansion Model", the channel characteristics and the transformation relationship between the time domain and the frequency domain are used to simplify the transformation formula of the channel response from the time domain to the frequency domain, and the frequency domain channel response of the received signal is estimated with lower complexity, which greatly reduces the computational complexity of the BEM algorithm. However, in the process of simplifying the time domain-frequency domain transformation formula, the algorithm approximates the time domain channel matrix as a Toeplitz circulant matrix, resulting in a slight loss in performance. In addition, high-dimensional matrix operations are used in the transformation formula, and the computational complexity is still relatively high.

[0026] In summary, the problems with existing technologies are: although the BEM algorithm shows good performance in dynamic multipath channels, its computational complexity is too high. A large number of high-dimensional matrix operations will result in huge resource utilization and long delays, making it difficult to apply to practical systems.

[0027] The difficulty and significance of solving the above technical problems: The channel estimation algorithm aims to accurately estimate the channel state, thereby eliminating the interference of the channel on the signal, improving the performance of the receiver, and ensuring the correct transmission of the signal. With the development of transportation, the channel environment at the receiving end has become increasingly harsh. High-speed movement produces the Doppler effect, which causes the channel to change dramatically. Traditional channel estimation algorithms are no longer applicable to dynamic multipath channels at this time. The emergence of the BEM algorithm provides a new approach. It maps the channel to the basis vector subspace and uses a small number of basis coefficients to fit the channel gain. It transforms the problem of estimating the rapidly changing channel response into the problem of solving the slowly changing basis function coefficients. Therefore, the BEM algorithm has become one of the main algorithms for accurately estimating fast time-varying channels. However, the complexity of the BEM algorithm and its improved algorithms is still relatively high.

[0028] In view of this, the present invention provides a low-complexity channel estimation method based on user frequency band and simplified basis extension, which completes the BEM channel estimation algorithm with lower complexity without affecting the performance.

[0029] See Figures 1 and 2 As shown, Figure 1 1 is a flow chart of a low-complexity channel estimation method based on user frequency band and simplified basis expansion provided by an embodiment of the present invention. Figure 2 is another schematic diagram of a low-complexity channel estimation method based on user frequency band and simplified basis extension provided by an embodiment of the present invention. The low-complexity channel estimation method based on user frequency band and simplified basis extension provided by the present invention includes:

[0030] Acquire a local time domain signal, and extract a local user frequency band from the local time domain signal;

[0031] Acquire a frequency domain signal received by a user, and extract a user reception frequency band from the frequency domain signal received by the user;

[0032] Generate basis vectors according to the basis expansion model, and perform frequency domain transformation on the basis vectors to obtain a frequency domain transformation matrix;

[0033] Transforming the local user frequency band to obtain a transform domain symbol; merging the frequency domain transform matrices corresponding to all the basis vectors to obtain a vector; and multiplying the transform domain symbol and the vector to obtain a solution matrix;

[0034] Obtaining a basis coefficient matrix according to the user receiving frequency band and the solution matrix;

[0035] A time domain channel response matrix is ​​obtained according to the base coefficient matrix, and the time domain channel response matrix is ​​transformed into a frequency domain to obtain a frequency domain channel response matrix.

[0036] Specifically, a low-complexity channel estimation method based on user band and simplified basis expansion is provided in this embodiment. First, the time domain received signal is synchronized, the cyclic prefix is ​​removed, and the FFT transformation is performed to obtain the frequency domain signal. The received user band signal is extracted and the local user band signal is generated to obtain the frequency domain local user band signal. and user receiving band signals Secondly, basis vectors are generated according to the basis expansion model, and the basis vectors are transformed in the frequency domain to obtain the frequency domain transformation matrix; thirdly, the basis coefficient matrix is ​​obtained according to the local user frequency band, the user receiving frequency band and the frequency domain transformation matrix; finally, the time domain channel response matrix is ​​obtained according to the basis coefficient matrix, and the frequency domain channel response matrix is ​​further obtained; in this way, the BEM channel estimation algorithm is completed with lower complexity without affecting the performance of the original BEM channel estimation algorithm; based on the traditional BEM algorithm, the present invention first performs BEM channel estimation based on the user frequency band to reduce the matrix calculation dimension of the basis expansion; then, by deriving the direct relationship between the basis coefficient matrix and the frequency domain channel matrix, the matrix multiplication is converted into vector multiplication, thereby reducing the complexity of the matrix multiplication; combining the above two points, the BEM channel estimation algorithm is implemented with lower computational complexity to obtain the channel estimation value, so that it can be better applied to actual systems.

[0037] In an optional embodiment of the present invention, the CE-BEM base extension model is adopted, and the base extension model is generated by base vectors, and the base vector b m The expression is:

[0038]

[0039] Wherein, G is the length of the basis vector, that is, the user frequency band performs the FFT operation of point G, M is the number of basis vectors, m is the mth basis vector, and n is the nth element of the mth basis vector.

[0040] In an optional embodiment of the present invention, the expression of the transform domain symbol A is:

[0041]

[0042] Where L is the number of separation paths of the channel, L is the first L columns of the Fourier transform matrix F, whose dimension is G×L, I M is the M-dimensional identity matrix, is the Knone column product, diag{} is the vector converted into a diagonal matrix, is the local user frequency band.

[0043] It should be noted that the expression of the Fourier transform matrix F is:

[0044]

[0045] Here, G is the length of the basis vector, that is, the user frequency band performs an FFT operation of point G.

[0046] In an optional embodiment of the present invention, the frequency domain transformation matrix B m The expression is:

[0047]

[0048] Among them, d m =FFT{b m ,G} / G is the basis vector b m Find FFT, a m =IFFT{b m ,G} is the Fourier transform result of IFFT, d m and a m The vector dimension is G×1, d m The gth vector element of for a m The gth vector element of F is the Fourier transform matrix, whose dimension is G×G, (·) H is the conjugate transpose operation of the matrix.

[0049] In an optional embodiment of the present invention, the expression of vector B is:

[0050] B=[Fdiag{b0}F H ,Fdiag{b1}F H ,...,Fdiag{b M-1}F H ];

[0051] It should be noted that the transformation domain symbol A and the vector B are multiplied to obtain the solution matrix Q, where the expression of Q is:

[0052] Q=BA.

[0053] In an optional embodiment of the present invention, the user receiving frequency band and the solution matrix are subjected to the least square method to obtain the base coefficient matrix Its expression is:

[0054]

[0055] in,(.) + is the pseudo-inverse of the matrix.

[0056] In an optional embodiment of the present invention, the expression of the time domain channel response matrix is:

[0057]

[0058] Among them, C m is the basis coefficient matrix corresponding to the mth basis vector.

[0059] It should be noted that the basis coefficient matrix C corresponding to the mth basis vector m The expression is:

[0060]

[0061] The final time domain channel matrix is ​​sparse, so only the time domain tap calculation of the tap position can be performed. The time domain channel matrix tap vector h of the lth branch is l , the vector dimension is G×1, which can be obtained by the following formula:

[0062]

[0063] in, is the coefficient of the lth branch and the mth basis vector, b m is the mth basis vector, and the vector dimension is G × 1. By solving the RL time domain matrix tap vectors, the entire time domain channel matrix can be obtained.

[0064] In an optional embodiment of the present invention, the frequency domain channel response matrix is ​​expressed as:

[0065]

[0066]

[0067] Among them, T l is the radial vector of the lth time domain channel, the vector dimension is G×1, is the radial dimension of the lth time domain channel, the nth channel tap response, h is the time domain channel response matrix, and its dimension is G×G.

[0068] In an optional embodiment of the present application, the method provided by the present invention is verified through experiments.

[0069] 1. Simulation conditions

[0070] See Figure 3 , Figure 3 This is a system framework diagram used in the simulation experiment provided by the embodiment of the present invention. The simulation system used is the multi-carrier frequency division multiplexing 5G NR PUSCH system specified in the 3GPP standard, using a 50MHz system bandwidth, a 30GHz carrier frequency, N=1024 subcarriers, a 120kHz subcarrier spacing, and a QPSK modulation scheme. The system uses a four-pilot symbol structure, see Figure 4 , Figure 4There are 12 single-carrier frequency division multiplexing symbols in a time slot, of which four block pilot symbols are located at the 3rd, 6th, 9th and 12th symbols.

[0071] The simulated channel is the non-terrestrial network NTN-TDL-D channel in the 3GPP standard. The multipath channel delay is [0, 0, 11.19, 146.68] ns, and the power attenuation corresponding to each path is [-0.284, -11.991, -9.887, -16.771] dB.

[0072] 2. Simulation Results

[0073] The performance of the method provided by the present invention (UB-F-BEM) is further illustrated by comparing the traditional BEM algorithm with the method provided by the present invention.

[0074] See Figure 5 As shown, Figure 5 This is a simulation diagram of the system bit error rate performance when the residual Doppler frequency deviation is 5kHz and 10kHz provided by an embodiment of the present invention. The channel path size is set to L=20, the basis vector dimension M=2 at 5kHz, and the basis vector dimension M=3 at 10kHz.

[0075] Depend on Figure 5 The performance curve shows that as the Doppler frequency offset increases, the system performance decreases. The UB-F-BEM algorithm provided by the present invention has basically the same performance as the traditional BEM algorithm.

[0076] 3. Complexity Analysis

[0077] By comparing the complexity of the traditional BEM algorithm, the patented "Fast Time-Varying Channel Estimation Method Based on Simplified Basis Extension Model" and the UB-F-BEM algorithm provided by the present invention, the performance of the method provided by the present invention is further illustrated.

[0078] The computational complexity is measured by the number of complex multiplications and complex additions. When the system parameters are relatively random, the solution matrix cannot be saved first. In this case, the total complexity of the traditional BEM algorithm and the UB-F-BEM algorithm provided by the present invention is compared as shown in Table 1 below.

[0079] Table 1 Complexity quantification table When the system parameters are completely fixed, the solution matrix can be saved first. At this time, the complexity of the subsequent modules of the patent "Fast Time-Varying Channel Estimation Method Based on Simplified Basis Extension Model" and the UB-F-BEM algorithm provided by the present invention is compared as shown in Table 2 below.

[0080] Table 2 Complexity quantification table

[0081]

[0082] In the formula describing complexity, there are four variables: L, M, N, and G. Variables N and G are both FFT points, and their values ​​are much larger than the other three variables. Therefore, the algorithm computational complexity depends on variables N and G. Assume L = 10, M = 3, N = 1024, and plot R as the variable. See Figure 6 and Figure 7 As shown, Figure 6 : is a schematic diagram comparing the number of complex multiplications between the UB-F-BEM algorithm provided by an embodiment of the present invention and the traditional BEM algorithm. Figure 7 This is a schematic diagram comparing the number of complex number additions of the UB-F-BEM algorithm provided by the embodiment of the present invention and the traditional BEM algorithm. Assume L = 10, M = 3, N = 1024, and draw a curve with R as the variable. Figure 8 As shown, Figure 8 This is a schematic diagram comparing the number of complex multiplications of the UB-F-BEM algorithm provided by the embodiment of the present invention and the patent [CN107018101A], Figure 9 2 is a schematic diagram comparing the number of complex number additions of the UB-F-BEM algorithm provided by an embodiment of the present invention and the patent [CN107018101A].

[0083] It should be noted that, in this document, relational terms such as first and second are used solely to distinguish one entity or operation from another, and do not necessarily require or imply any actual relationship or order between these entities or operations. Furthermore, the terms "comprise," "include," or any other variations thereof are intended to encompass non-exclusive inclusion, such that an article or device comprising a list of elements includes not only those elements but also other elements not explicitly listed. Without further limitation, an element defined by the phrase "comprising a..." does not preclude the presence of additional identical elements in the article or device comprising the element. Terms such as "connected" or "connected" are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. References to orientations or positional relationships, such as "upper," "lower," "left," and "right," are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate description and simplify the description of the present invention. They do not indicate or imply that the device or element referred to must have, be constructed, or operate in a specific orientation, and are therefore not to be construed as limiting the present invention.

[0084] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described can be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification.

[0085] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. A low-complexity channel estimation method based on user band and simplified basis expansion, characterized in that: include: Acquire a local time domain signal, and extract a local user frequency band from the local time domain signal; Acquire a frequency domain signal received by a user, and extract a user reception frequency band from the frequency domain signal received by the user; Generate basis vectors according to the basis expansion model, and perform frequency domain transformation on the basis vectors to obtain a frequency domain transformation matrix; Transform the local user frequency band to obtain a transform domain symbol ; Combine the frequency domain transformation matrices corresponding to all the basis vectors to obtain the vector ; The transform domain symbol and the vector Multiply to get the solution matrix ; Obtaining a basis coefficient matrix according to the user receiving frequency band and the solution matrix; A time domain channel response matrix is ​​obtained according to the base coefficient matrix, and the time domain channel response matrix is ​​transformed into a frequency domain to obtain a frequency domain channel response matrix.

2. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 1, characterized in that: The basis vectors The expression is: ； in, G is the length of the basis vector, i.e. the user band is G Point FFT operation, is the number of basis vectors, m For the m basis vectors, n For the m The basis vector n elements.

3. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 2, characterized in that: The transform domain symbol The expression is: ； in, is the number of separation paths of the channel, is the front of the Fourier transform matrix F Columns, whose dimensions are , for dimensional identity matrix, is the Knone product, diag {} is to convert the vector into a diagonal matrix, is the local user frequency band.

4. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 2, characterized in that: The frequency domain transformation matrix The expression is: ； in, is the basis vector Find the FFT, is the Fourier transform result of IFFT, and The vector dimension is , for No. vector elements, for No. vector elements, , F is the Fourier transform matrix, its dimension is G × G , is the conjugate transpose operation of the matrix.

5. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 2, characterized in that: The vector The expression is: ； Where F is the Fourier transform matrix, is the conjugate transpose operation of the matrix.

6. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 1, characterized in that: The user receiving frequency band and the solution matrix are subjected to the least square method to obtain the base coefficient matrix , whose expression is: ； in,(.) + is the pseudo-inverse of the matrix.

7. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 2, characterized in that: The expression of the time domain channel response matrix is: ； in, For the m The basis coefficient matrix corresponding to the basis vectors.

8. The low-complexity channel estimation method based on user frequency band and simplified basis extension according to claim 7, characterized in that: The expression of the frequency domain channel response matrix is: ； ； in, For the The time domain channel is divided into radial vectors, and the vector dimension is G ×1, For the The time domain channel is divided into radial quantities, channel tap responses, is the time domain channel response matrix, whose dimension is G × G .

Images