Improved estimation method, system and equipment of MIMO-OTFS system based on block sparse Bayesian, and medium
By introducing virtual angle dimensions and edge optimization algorithms in the MIMO-OTFS system, using block sparse Bayesian hierarchical model for channel vector modeling, the problem of channel estimation complexity in high-speed mobile scenarios is solved, and higher accuracy is achieved and efficiency.
Patent Information
- Application Number
- CN202410978888.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-22
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-07-22
AI Technical Summary
In high-speed mobile scenarios, the channel estimation complexity of the MIMO-OTFS system is high, and the existing sparse Bayesian learning algorithm does not fully utilize the block sparse characteristics, resulting in low channel estimation accuracy and computational efficiency.
Introduce virtual angle dimensions, convert burst block sparse vectors into high-dimensional block sparse vectors by improving the transformation matrix, and combine with edge optimization algorithms to model channel vectors using block sparse Bayesian hierarchical model to accurately solve burst block sparse Channel vector.
It improves the accuracy and computing efficiency of channel estimation, reduces the complexity of the algorithm, is more adaptable, and is suitable for high-complexity channel estimation problems.
Smart Images

Figure CN118972206A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wireless communication technology, and in particular to an improved estimation method, system, device and medium for a MIMO-OTFS system based on block sparse Bayes. Background Art
[0002] At present, my country has the longest and fastest high-speed railway network in the world, and high-speed rail has become an indispensable means of transportation for many people in their daily travel. Therefore, how to provide diversified and high-quality wireless communication services in high-speed mobile scenarios has become one of the main challenges facing future wireless communication systems. However, the channel in high-speed mobile scenarios has the characteristics of dual selection of time and frequency, which brings more severe challenges to the accuracy and effectiveness of channel estimation.
[0003] In high-speed mobile scenarios, the dual-frequency characteristics of the channel will greatly affect the orthogonality of the subcarriers of the Orthogonal Frequency Division Multiplexing (OFDM) system, resulting in a sharp decline in its performance. Therefore, the Orthogonal Time Frequency Space (OTFS) system is used in high-speed mobile scenarios as an alternative modulation scheme for the OFDM system. Unlike the OFDM system, which modulates data symbols in the time-frequency domain, OTFS modulates data symbol signals in a two-dimensional delay-Doppler channel that is independent of time. Since the delay-Doppler channel intuitively reflects the position of each scatterer that constitutes the wireless communication link at this time, it significantly reduces the impact of the dual-frequency channel on the signal, so that the diversified and high-quality communications in the high-speed mobile environment are met.
[0004] In addition, in the development of wireless communication systems, the demand for capacity is growing rapidly and the problem of excessive spectrum resource consumption is becoming increasingly serious. In order to further improve the spectrum utilization of the OTFS system, the Multiple Input Multiple Output (MIMO) technology is combined with the OTFS technology, and spatial multiplexing is used to improve the channel capacity and channel transmission rate in high-speed mobile scenarios. In the MIMO-OTFS system, the addition of multiple antennas introduces a spatial dimension to the delay-Doppler domain, which not only increases the complexity of the channel, but also greatly increases the difficulty of channel estimation.
[0005] Since the MIMO-OTFS system has good common sparsity characteristics in the delay-Doppler-angle domain and burst block sparsity characteristics in the angle domain, the channel estimation matrix can be converted into a compressed sensing sparse vector solution problem. The reconstruction algorithm based on the Bayesian framework occupies an important position in the compressed sensing reconstruction algorithm, so it can usually obtain a more accurate sparse solution than the convex optimization algorithm and the greedy algorithm. The commonly used algorithm is the sparse Bayesian learning algorithm (SparseBayesian Learning, SBL), but the SBL estimation algorithm does not fully utilize the block sparsity characteristics.
[0006] The patent application with publication number CN116471148A proposes a channel estimation method based on sparse Bayesian learning. The method adopts two layers of sparse prior information, including sparse channel parameters and phase shift matrix in the angle domain. The description of channel characteristics is relatively simple, and the sparse Bayesian learning (SBL) algorithm framework is adopted. The block sparsity characteristics are not fully utilized, and the algorithm performance is relatively low. Summary of the invention
[0007] In order to overcome the defects of the above-mentioned prior art, the purpose of the present invention is to provide an improved estimation method, system, device and medium for a MIMO-OTFS system based on block sparse Bayesian. A virtual angle dimension is introduced on the basis of signal reconstruction using a block sparse Bayesian (BSBL) algorithm. The angle dimension related vector is calculated and mapped into a high-dimensional block sparse vector using a lifting transformation matrix, and then inversely quantized into a traditional block sparse vector, so as to effectively deal with the sparsity problem of the angle dimension; a marginal optimization (BO) algorithm with low time cost for estimating super-prior parameters is combined to obtain a faster convergence speed, reduce the time required for the algorithm to run, and thus reduce the complexity of the algorithm; the application of the present invention in a MIMO-OTFS system is not only superior to traditional algorithms in channel estimation accuracy, but also shows significant advantages in computational efficiency and adaptability, and is an effective method for solving high-complexity channel estimation problems.
[0008] In order to achieve the above object, the present invention adopts the following technical solutions:
[0009] An improved estimation method for a MIMO-OTFS system based on block sparse Bayes is characterized by comprising the following steps:
[0010] Step S1. Using the lifting transformation matrix Convert the burst block sparse vector into the traditional block sparse vector D θi ;
[0011] Step S2. Based on the traditional block sparse vector D θi Calculate the start and end positions of the burst block sparse vector, and find the distribution of non-zero blocks in the angle dimension Λ in combination with the angle burst block length Dθi , block length d for the angle dimension i , the distribution of non-zero blocks in the angular dimension Λ θi And the corresponding columns of the block dictionary matrix Φ are updated;
[0012] Step S3. According to step S1 and step S2, a block sparse Bayesian hierarchical model is used to model the channel vector, calculate the posterior probability density and likelihood function, accurately solve the burst block sparse channel vector, and realize block sparse channel vector estimation in a noisy environment.
[0013] The specific process of converting the burst block sparse vector into the traditional block sparse vector by using the lifting transformation matrix in step S1 is as follows:
[0014] Step 1.1) Calculate the angle dimension related vector in, and y are the block dictionary matrix and the received signal respectively;
[0015] Step 1.2) The angle dimension correlation vector x obtained in step 1.1) is i , using the lifting transformation matrix Map the burst block sparse vector to a high-dimensional block sparse vector, and increase it to N t A high-dimensional block sparse vector d of D×1 θi =Ω H μ i , which is then inversely quantized to obtain the traditional block sparse vector
[0016] In step S2, the start and end positions of the burst block sparse vector are calculated according to the traditional block sparse vector, and the distribution of the non-zero blocks in the angle dimension is found in combination with the angle burst block length D. The specific process of updating the angle dimension block length, the distribution of the non-zero blocks in the angle dimension and the block dictionary matrix is as follows:
[0017] Step 2.1) The traditional block sparse vector D obtained by step 1.2) θi , calculate the traditional block sparse vector D θi The l2 norm of all row vectors is used to obtain the correlation vector g θi (r)=||D θi (r,:)||2, and then through the relevant vector g θi The maximum element of gets the start and end position of the burst block sparse vector t θi = arg max(g θi (r));
[0018] Step 2.2) Update the angle dimension correlation vector x according to the angle burst block length D i The corresponding angle dimension block length d i, and then the starting and ending positions t of the burst block sparse vector obtained in step 2.1) θi To update the distribution of non-zero blocks in the angle dimension Λ θi , and finally update the block dictionary matrix Φ. The update formula is as follows:
[0019]
[0020] The specific process in step S3 is as follows:
[0021] Step 3.1) Based on the block dictionary matrix Φ obtained in step 2.2), the channel vector is modeled using a block sparse Bayesian hierarchical model. The sparse relation is:
[0022] y=Φh+ε
[0023] Among them, y is the received signal; h is the block sparse signal, which is determined by the intra-block correlation parameter γ i and the intra-block correlation matrix B i Describe together; ε is a noise signal with a mean of zero and a covariance of λ -1 Complex Gaussian distribution of I, λ = δ -2 is the noise accuracy parameter; γ i , B i and λ is called the hyper-prior parameter;
[0024] Step 3.2) By calculating the posterior probability density P(x|y,γ) of the block sparse signal h i ,B i ,λ) and the likelihood function P(y|γ i ,B i ,λ), and obtain the calculation formula of mean μ, covariance Σ and variance C of likelihood function:
[0025]
[0026] P(y|γ i ,B i ,λ)=CN(0,C)
[0027] Among them, P(x|y,γ i ,B i ,λ) is a complex Gaussian distribution with mean μ and covariance Σ, P(y|γ i ,B i ,λ) is a complex Gaussian distribution with mean 0 and covariance C;
[0028] By calculation, we get:
[0029] μ=λΣΦ H y,Σ=(Γ -1 +λΦ HΦ) -1 , C = λ -1 I+ΦΓΦ H
[0030] In the above formula, the intra-block correlation diagonal matrix Γ = diag(γ1B1,γ2B2,…,γ g B g );
[0031] Step 3.3) For the hyper-prior parameters λ and B i Update; According to the likelihood function, the cost function can be obtained, and using the Woodbury matrix identity, we have:
[0032]
[0033] Find λ and B from the above formula i The partial derivatives of λ and B are obtained i The update formula is:
[0034]
[0035] Step 3.4) For the hyper-prior parameter γ i Update: Use the marginal optimization (BO) algorithm to calculate the hyper-prior parameters, find a substitute function as the lower bound of the cost function, and then find the extreme point of the substitute function to obtain the updated value of the hyper-prior parameters. The formula of the derived cost function is as follows:
[0036]
[0037] Among them, const is an irrelevant term; γ i Take the partial derivative and get the super prior parameter γ i The update formula is:
[0038]
[0039] Step 3.5) For the intra-block correlation matrix B i Add constraints: Take a single relevant parameter r i The symmetric positive definite matrix To the intra-block correlation matrix B i Replace; when the symmetric positive definite matrix is a Toeplitz matrix, it is considered a symmetric positive definite matrix It can well reflect the intra-block correlation of the block sparse signal h;
[0040]
[0041] In the above formula, the correlation parameter r of the block sparse signal h is iAccording to step 3.3) i By calculation:
[0042]
[0043] in, is the intra-block correlation matrix B i The average of the subdiagonal elements of For B i The average correlation parameter of the block sparse signal h is the correlation parameter r of each block sparse signal h. i The average is:
[0044] Finally, the replaced intra-block correlation matrix of the block sparse signal h is obtained, that is, the symmetric positive definite matrix
[0045] Step 3.6) Repeat the iteration to solve: Through steps 3.1) to 3.5), the burst block sparse channel vector h new To solve, if The algorithm ends and the result is output as burst block sparse channel vector h new (Λ θi )=μ, otherwise repeat the above steps 3.2)-3.5).
[0046] The improved estimation method of the MIMO-OTFS system based on block sparse Bayes is applied to various complex application scenarios such as wireless communication systems under high-speed mobility in the 5G standard, in vehicle networking systems to provide a reliable data transmission basis for intelligent driving, and in underwater acoustic communication technology.
[0047] A system for an improved estimation method of a MIMO-OTFS system based on the block sparse Bayesian method, comprising:
[0048] The block sparse vector conversion module, used in step S1, uses the lifting transformation matrix Convert the burst block sparse vector into the traditional block sparse vector D θi ;
[0049] The traditional block sparse vector calculation module is used in step S2 to calculate the traditional block sparse vector D θi Calculate the start and end positions of the burst block sparse vector, and find the distribution of non-zero blocks in the angle dimension Λ in combination with the angle burst block length D θi , block length d for the angle dimension i , the distribution of non-zero blocks in the angular dimension Λ θi And the corresponding columns of the block dictionary matrix Φ are updated;
[0050] The block sparse Bayes solution module is used in step S3. According to step S1 and step S2, the block sparse Bayes hierarchical model is used to model the channel vector, calculate the posterior probability density and the likelihood function, accurately solve the burst block sparse channel vector, and realize the block sparse channel vector estimation in a noisy environment.
[0051] A device for an improved estimation method of a MIMO-OTFS system based on the block sparse Bayesian method, comprising:
[0052] Memory for storing computer programs;
[0053] A processor is used to implement the improved estimation method for the MIMO-OTFS system based on block sparse Bayesian when executing the computer program.
[0054] A computer-readable storage medium is used to store a computer program. When the computer program is executed by a processor, it can solve the high-complexity channel estimation problem based on the improved estimation method of the MIMO-OTFS system based on block sparse Bayes described in steps S1 to S3.
[0055] Compared with the prior art, the advantages and positive effects of the present invention are:
[0056] 1. The present invention introduces a virtual angle dimension on the basis of signal reconstruction using a block sparse Bayesian (BSBL) algorithm. By calculating the angle dimension related vector and mapping it into a high-dimensional block sparse vector using a lifting transformation matrix, and then inversely quantizing it into a traditional block sparse vector, the sparsity problem of the angle dimension can be effectively handled. Not only is the ability to capture the sparsity of the angle dimension enhanced, but the start and end positions of the burst block sparse vector can also be calculated more accurately, further improving the accuracy of channel estimation, so that the probability of signal reconstruction is higher.
[0057] 2. The present invention combines the edge optimization (BO) algorithm with low time cost for estimating super-prior parameters to obtain faster convergence speed, reduce the time required for the algorithm to run, and thus reduce the complexity of the algorithm. By finding a substitute function as the lower bound of the cost function, and then finding the extreme point of the substitute function to obtain the updated value of the super-prior parameter. This process significantly reduces the computational complexity of the super-prior parameter and can converge quickly, making the entire algorithm more efficient when dealing with large-scale channel estimation problems.
[0058] 3. Compared with other traditional algorithms, the algorithm of the present invention is more suitable for high-complexity channel estimation problems in MIMO-OTFS scenarios. Traditional algorithms often face the problems of high computational complexity and slow convergence when dealing with multi-dimensional and large-scale sparse channel estimation. However, this algorithm not only improves the probability and accuracy of signal reconstruction, but also greatly reduces the computational complexity and convergence time by introducing virtual angle dimensions and edge optimization algorithms, thereby showing superior performance in practical applications.
[0059] In summary, the application of the present invention in the MIMO-OTFS system is not only superior to the traditional algorithm in terms of channel estimation accuracy, but also shows significant advantages in terms of computational efficiency and adaptability. It is an effective method for solving high-complexity channel estimation problems. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0061] Figure 1 It is a flow chart of a method according to an embodiment of the present invention.
[0062] Figure 2 It is a hierarchical model diagram of block sparse Bayes according to an embodiment of the present invention.
[0063] Figure 3 4 is a diagram showing the relationship between NMSE performance and SNR of different estimation algorithms according to an embodiment of the present invention.
[0064] Figure 4 It is a relationship diagram between BER and SNR of different estimation algorithms under QPSK modulation in an embodiment of the present invention. DETAILED DESCRIPTION
[0065] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0066] In view of the problems existing in the prior art, the present invention provides an improved estimation method for a MIMO-OTFS system based on block sparse Bayesian. The present invention is described in detail below with reference to the accompanying drawings.
[0067] In order to enable those skilled in the art to fully understand how to implement the present invention in detail, this section is an explanatory embodiment that expands and describes the technical solution of the claims.
[0068] like Figure 1 As shown, the improved estimation algorithm based on the MIMO-OTFS system provided by the embodiment of the present invention includes the following steps:
[0069] Step S1. Using the lifting transformation matrix Convert the burst block sparse vector into the traditional block sparse vector D θi ; The specific process is as follows:
[0070] 1.1) Calculate the angle dimension related vector in and y are the dictionary matrix and the received signal respectively;
[0071] 1.2) The angle dimension correlation vector χ obtained in 1.1) i , by using the lifting transformation matrix Map the burst block sparse vector to a high-dimensional block sparse vector, and increase it to N t A high-dimensional block sparse vector d of D×1 θi =Ω H μ i , which is then inversely quantized to obtain the traditional block sparse vector
[0072] Step S2. Based on the traditional block sparse vector D θi Calculate the start and end positions of the burst block sparse vector, and find the distribution of non-zero blocks in the angle dimension Λ in combination with the angle burst block length D θi , block length d for the angle dimension i , the distribution of non-zero blocks in the angular dimension Λ θi And the corresponding columns of the block dictionary matrix Φ are updated; the specific process is as follows:
[0073] 2.1) The traditional block sparse vector D obtained by 1.2) θi , calculate D θi The l2 norm of all row vectors is used to obtain the correlation vector g θi (r)=||D θi (r,:)||2, then through g θi The maximum element of gets the start and end position of the burst block sparse vector t θi = argmax(g θi (r));
[0074] 2.2) Update the angle dimension correlation vector χ according to the angle burst block length D i The corresponding angle dimension block length d i , and then the starting and ending positions t of the burst block sparse vector obtained in step 2.1) θi To update the distribution of non-zero blocks in the angle dimension Λ θi, and finally update the block dictionary matrix Φ. The update formula is as follows:
[0075]
[0076] Step S3. According to step S1 and step S2, a block sparse Bayesian hierarchical model is used to model the channel vector, calculate the posterior probability density and likelihood function, accurately solve the burst block sparse channel vector, and realize the block sparse channel vector estimation in a noisy environment; the specific process is as follows:
[0077] 3.1) Based on the dictionary matrix Φ obtained in 2.2), the channel vector is modeled using a block sparse Bayesian hierarchical model, such as Figure 2 As shown, the sparse relations are:
[0078] y=Φh+ε
[0079] Among them, y is the received signal; h is the block sparse signal, which is determined by the intra-block correlation parameter γ i and the intra-block correlation matrix B i Describe together; ε is a noise signal with a mean of zero and a covariance of λ -1 Complex Gaussian distribution of I, λ = δ -2 is the noise accuracy parameter; γ i , B i and λ is called the hyper-prior parameter;
[0080] 3.2) By calculating the posterior probability density P(x|y,γ) of the block sparse signal h i ,B i ,λ) and the likelihood function P(y|γ i ,B i ,λ), and obtain the calculation formula of mean μ, covariance Σ and variance C of likelihood function.
[0081]
[0082] P(y|γ i ,B i ,λ)=CN(0,C)
[0083] Among them, P(x|y,γ i ,B i ,λ) is a complex Gaussian distribution with mean μ and covariance Σ, P(y|γ i ,B i ,λ) is a complex Gaussian distribution with mean 0 and covariance C.
[0084] By calculation, we can get:
[0085] μ=λΣΦ H y,Σ=(Γ-1 +λΦ H Φ) -1 , C = λ -1 I+ΦΓΦ H
[0086] In the above formula, the intra-block correlation diagonal matrix Γ = diag(γ1B1,γ2B2,…,γ g B g ).
[0087] 3.3) For the hyper-prior parameters λ and B i Update; According to the likelihood function, the cost function can be obtained, and using the Woodbury matrix identity, we have:
[0088]
[0089] Find λ and B from the above formula i The partial derivatives of can get their update formula:
[0090]
[0091] 3.4) For the hyper-prior parameter γ i Update; use the edge optimization (BO) algorithm to calculate the super prior parameters, find a substitute function as the lower bound of the cost function, and then find the extreme point of the substitute function to obtain the updated value of the super prior parameters. The formula of the derived cost function is as follows:
[0092]
[0093] Among them, const is an irrelevant term; γ i Taking partial derivatives, we can get the super prior parameter γ i The update formula is:
[0094]
[0095] 3.5) For the intra-block correlation matrix B i Add constraints; use a single relevant parameter r i The symmetric positive definite matrix To the intra-block correlation matrix B i Replace; when the symmetric positive definite matrix is a Toeplitz matrix, it is considered a symmetric positive definite matrix It can well reflect the intra-block correlation of the block sparse signal h;
[0096]
[0097] In the above formula, the correlation parameter r of the block sparse signal h isi According to 3.3) i By calculation:
[0098]
[0099] in, is the intra-block correlation matrix B i The average of the subdiagonal elements of For B i The average correlation parameter of the block sparse signal h is the correlation parameter r of each block sparse signal h. i The average is:
[0100] Finally, the replaced intra-block correlation matrix of the block sparse signal h is obtained, that is, the symmetric positive definite matrix
[0101] 3.6) Repeat the iteration to solve: Through steps 3.1) to 3.5), the burst block sparse channel vector h new To solve, if The algorithm ends and the result is output as burst block sparse channel vector h new (Λ θi )=μ, otherwise repeat the above steps 3.2)-3.5).
[0102] In order to prove the creativity and technical value of the technical solution of the present invention, the following are application examples and verifications of the claimed technical solution on specific products or related technologies.
[0103] The present invention can be used in wireless communication systems under high-speed mobility in the 5G standard, for example, to ensure reliable wireless communication services in the high-speed mobility environment of high-speed railways, to be used in vehicle networking systems to provide a reliable data transmission basis for intelligent driving, and to be used in various complex application scenarios such as underwater acoustic communication technology.
[0104] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.
[0105] The embodiments of the present invention have achieved some positive effects during the development or use process, and indeed have great advantages over the prior art. The following content is described in conjunction with data, charts, etc. of the test process.
[0106] 1. Simulation conditions
[0107] A simulation platform was built in MATLAB software for simulation testing. The channel model adopted the 3GPP standard spatial channel model, namely the urban macro unit environment of the SCM channel model. The specific parameter configuration is shown in the following table:
[0108] parameter Parameter Value Carrier frequency (GHz) 2.15 Subcarrier spacing (kHz) 15 OTFS frame size M×N 256×64 Channel Model 3GPPSCM Array Antenna ULA User speed (m / s) 100 Number of antennas at the base station 8 Number of user antennas 1 Data modulation method QPSK
[0109] 2. Simulation content and results
[0110] The performance simulation mainly compares the performance of three algorithms:
[0111] 1) Existing BOMP-based estimation algorithms.
[0112] 2) Existing SBL-based estimation algorithms.
[0113] 3) The estimation algorithm of the present invention (an improved estimation algorithm for MIMO-OTFS system based on block sparse Bayes, MABSBL).
[0114] Figure 3 The normalized mean square error (NMSE) performance comparison of the BOMP estimation algorithm, the SBL estimation algorithm, and the proposed MABSBL estimation algorithm under different signal-to-noise ratio conditions is shown. Figure 3It can be seen that the MABSBL estimation algorithm shows excellent performance under all signal-to-noise ratios, and its NMSE curve is at the lowest position, indicating that the MABSBL estimation algorithm has the smallest error in the entire signal-to-noise ratio range. In contrast, the performance of the SBL estimation algorithm is better than that of the BOMP estimation algorithm, and the NMSE curve of SBL is between BOMP and MABSBL. The performance of the BOMP estimation algorithm is relatively poor under all signal-to-noise ratio conditions, and its NMSE curve is at the highest position in the figure, which reflects that its signal estimation error is the largest. Under the signal-to-noise ratio condition of 0dB, the NMSE of the MABSBL estimation algorithm is about 10^-1.5, while the NMSE of BOMP and SBL are 10^-0.7 and 10^-1, respectively, showing the superiority of the MABSBL estimation algorithm under low signal-to-noise ratio. As the signal-to-noise ratio increases, the NMSE of the three algorithms decreases, but the performance of the MABSBL estimation algorithm is still the best. For example, at a signal-to-noise ratio of 20 dB, the NMSE of the MABSBL estimation algorithm drops to 10^-3, while the NMSE of BOMP and SBL drops to 10^-2.1 and 10^-2.7, respectively. This result shows that in the high signal-to-noise ratio area, the NMSE performance of the MABSBL estimation algorithm exceeds that of the other two algorithms.
[0115] Figure 4The performance comparison of the bit error rate (BER) of the signal detected by the MMSE detection algorithm under QPSK modulation is shown using the channel state information obtained by the BOMP estimation algorithm, the SBL estimation algorithm and the MABSBL estimation algorithm, as well as the channel state information under ideal conditions. It can be clearly seen from the figure that the MABSBL estimation algorithm shows the lowest bit error rate under all signal-to-noise ratio conditions. Under the condition of 5dB signal-to-noise ratio, the bit error rate of the MABSBL estimation algorithm is about 10^-1.9, which is slightly better than the 10^-1.8 of the BOMP estimation algorithm and the 10^-1.9 of the SBL estimation algorithm. As the signal-to-noise ratio increases, the BER of the three algorithms shows a downward trend, but the speed and magnitude of the decline of the MABSBL estimation algorithm are more significant. Under a high signal-to-noise ratio of 20dB, the BER of the MABSBL estimation algorithm drops to about 10^-4.9, which is significantly better than the 10^-3.9 of the BOMP estimation algorithm and the 10^-4.6 of the SBL estimation algorithm. It even achieves an order of magnitude performance improvement compared to BOMP, and is closer to the BER performance under ideal conditions. This result shows that the performance advantage of the MABSBL estimation algorithm is more obvious in the high SNR area, and the difference from the ideal BER performance curve is smaller than that in the low SNR area. In contrast, the BOMP estimation algorithm has poor NMSE performance in the low SNR area, and its BER performance is relatively poor in the high SNR area, which is far from the ideal BER performance curve. In general, the MABSBL estimation algorithm shows better BER performance in the entire SNR range, especially in the high SNR area, which is closer to the ideal performance.
[0116] In summary, the improved estimation algorithm of the MIMO-OTFS system based on block sparse Bayesian in the present invention has greatly improved performance compared with the existing BOMP-based estimation algorithm and BML-based estimation algorithm.
[0117] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.
Claims
1. An improved estimation method for MIMO-OTFS system based on block sparse Bayes, characterized by: The specific steps include: Step S1. Using the lifting transformation matrix Convert the burst block sparse vector into the traditional block sparse vector D θi ; Step S2. Based on the traditional block sparse vector D θi Calculate the start and end positions of the burst block sparse vector, and find the distribution of non-zero blocks in the angle dimension Λ in combination with the angle burst block length D θi , block length d for the angle dimension i , the distribution of non-zero blocks in the angular dimension Λ θi And the corresponding columns of the block dictionary matrix Φ are updated; Step S3. According to step S1 and step S2, a block sparse Bayesian hierarchical model is used to model the channel vector, calculate the posterior probability density and likelihood function, accurately solve the burst block sparse channel vector, and realize block sparse channel vector estimation in a noisy environment.
2. The improved estimation method for MIMO-OTFS system based on block sparse Bayesian according to claim 1, characterized in that: The specific process of converting the burst block sparse vector into the traditional block sparse vector by using the lifting transformation matrix in step S1 is as follows: Step 1.1) Calculate the angle dimension related vector in, and y are the block dictionary matrix and the received signal respectively; Step 1.2) The angle dimension correlation vector x obtained in step 1.1) is i , using the lifting transformation matrix Map the burst block sparse vector to a high-dimensional block sparse vector, and increase it to N t A high-dimensional block sparse vector d of D×1 θi =Ω H μ i , which is then inversely quantized to obtain the traditional block sparse vector 3. The improved estimation method for MIMO-OTFS system based on block sparse Bayesian according to claim 1, characterized in that: In step S2, the start and end positions of the burst block sparse vector are calculated according to the traditional block sparse vector, and the distribution of the non-zero blocks in the angle dimension is found in combination with the angle burst block length D. The specific process of updating the angle dimension block length, the distribution of the non-zero blocks in the angle dimension and the block dictionary matrix is as follows: Step 2.1) The traditional block sparse vector D obtained by step 1.2) θi , calculate the traditional block sparse vector D θi The l2 norm of all row vectors is used to obtain the correlation vector g θi (r)=||D θi (r,:)||2, and then through the relevant vector g θi The maximum element of gets the start and end position of the burst block sparse vector t θi = argmax(g θi (r)); Step 2.2) Update the angle dimension correlation vector x according to the angle burst block length D i The corresponding angle dimension block length d i , and then the starting and ending positions t of the burst block sparse vector obtained in step 2.1) θi To update the distribution of non-zero blocks in the angle dimension Λ θi , and finally update the block dictionary matrix Φ. The update formula is as follows:
4. The improved estimation method for MIMO-OTFS system based on block sparse Bayesian according to claim 1, characterized in that: The specific process of step S3 is as follows: Step 3.1) Based on the block dictionary matrix Φ obtained in step 2.2), the channel vector is modeled using a block sparse Bayesian hierarchical model. The sparse relation is: y=Φh+ε Among them, y is the received signal; h is the block sparse signal, which is determined by the intra-block correlation parameter γ i and the intra-block correlation matrix B i Describe together; ε is a noise signal with a mean of zero and a covariance of λ -1 Complex Gaussian distribution of I, λ = δ -2 is the noise accuracy parameter; γ i , B i and λ is called the hyper-prior parameter; Step 3.2) By calculating the posterior probability density P(x|y,γ) of the block sparse signal h i ,B i ,λ) and the likelihood function P(y|γ i ,B i ,λ), and obtain the calculation formula of mean μ, covariance Σ and variance C of likelihood function: P(y|γ i ,B i ,λ)=CN(0,C) Among them, P(x|y,γ i ,B i ,λ) is a complex Gaussian distribution with mean μ and covariance Σ, P(y|γ i ,B i ,λ) is a complex Gaussian distribution with mean 0 and covariance C; By calculation, we get: μ=λΣΦ H y, S = (C) -1 +λΦ H F) -1 ,C=λ -1 I+FGG H In the above formula, the intra-block correlation diagonal matrix Γ = diag(γ1B1,γ2B2,…,γ g B g ); Step 3.3) For the hyper-prior parameters λ and B i Update; According to the likelihood function, the cost function can be obtained, and using the Woodbury matrix identity, we have: Find λ and B from the above formula i The partial derivatives of λ and B are obtained i The update formula is: Step 3.4) For the hyper-prior parameter γ i Update: Use the marginal optimization (BO) algorithm to calculate the hyper-prior parameters, find a substitute function as the lower bound of the cost function, and then find the extreme point of the substitute function to obtain the updated value of the hyper-prior parameters. The formula of the derived cost function is as follows: Among them, const is an irrelevant term; γ i Take the partial derivative and get the super prior parameter γ i The update formula is: Step 3.5) For the intra-block correlation matrix B i Add constraints: Take a single relevant parameter r i The symmetric positive definite matrix To the intra-block correlation matrix B i Replace; when the symmetric positive definite matrix is a Toeplitz matrix, it is considered a symmetric positive definite matrix It can well reflect the intra-block correlation of the block sparse signal h; In the above formula, the correlation parameter r of the block sparse signal h is i According to step 3.3) i By calculation: in, is the intra-block correlation matrix B i The average of the subdiagonal elements of For B i The average correlation parameter of the block sparse signal h is the correlation parameter r of each block sparse signal h. i The average is: Finally, the replaced intra-block correlation matrix of the block sparse signal h is obtained, that is, the symmetric positive definite matrix Step 3.6) Repeat the iteration to solve: Through steps 3.1) to 3.5), the burst block sparse channel vector h new To solve, if The algorithm ends and the result is output as burst block sparse channel vector h new (Λ θi )=μ, otherwise repeat the above steps 3.2)-3.5).
5. Application of an improved estimation method for MIMO-OTFS system based on block sparse Bayes, characterized in that: It is used in various complex application scenarios such as wireless communication systems under high-speed mobility in the 5G standard, in vehicle networking systems to provide a reliable data transmission basis for intelligent driving, and in underwater acoustic communication technology.
6. A system based on the improved estimation method of the MIMO-OTFS system based on the block sparse Bayesian as claimed in any one of claims 1 to 4, characterized in that: include: The block sparse vector conversion module, used in step S1, uses the lifting transformation matrix Convert the burst block sparse vector into the traditional block sparse vector D θi ; The traditional block sparse vector calculation module is used in step S2 to calculate the traditional block sparse vector D θi Calculate the start and end positions of the burst block sparse vector, and find the distribution of non-zero blocks in the angle dimension Λ in combination with the angle burst block length D θi , block length d for the angle dimension i , the distribution of non-zero blocks in the angular dimension Λ θi And the corresponding columns of the block dictionary matrix Φ are updated; The block sparse Bayes solution module is used in step S3. According to step S1 and step S2, the block sparse Bayes hierarchical model is used to model the channel vector, calculate the posterior probability density and the likelihood function, accurately solve the burst block sparse channel vector, and realize the block sparse channel vector estimation in a noisy environment.
7. A device according to any one of claims 1 to 4 for the improved estimation method of the MIMO-OTFS system based on the block sparse Bayesian approach, characterized in that: include: Memory for storing computer programs; A processor is used to implement the improved estimation method for the MIMO-OTFS system based on block sparse Bayesian when executing the computer program.
8. A computer-readable storage medium for storing a computer program, characterized in that: When the computer program is executed by a processor, it can solve the high-complexity channel estimation problem based on the improved estimation method for the MIMO-OTFS system based on block sparse Bayes as described in any one of claims 1 to 4.
Citation Information
Patent Citations
Block sparse compressive sensing-based infrared image reconstruction method and system thereof
CN106663316A
Low-complexity sparse Bayesian vector estimation method and system
CN111142065A
Improved block sparse Bayesian anti-interference method based on power characteristic value extraction
CN111796242A
Sparse Bayesian learning channel estimation method based on two-dimensional cluster structure
CN112887233A
Hyperspectral video reconstruction method based on multi-task blind compressed sensing
CN114245128A
