A metamaterial antenna architecture radar communication integrated system and waveform optimization method

Abstract

This invention relates to the field of radar-communication integration, specifically to a radar-communication integrated system with a metamaterial antenna architecture and a waveform optimization method. It utilizes a dynamic metamaterial antenna pre-encoder to fit an optimal pre-encoder for the dynamic metamaterial antenna, ultimately obtaining the optimal pre-encoder. This system optimizes radar transmission beamforming performance while satisfying communication user quality and constant power mode constraints. This invention considers dynamic metamaterial antennas, which saves costs and occupies a small area. The radar beam performance is similar to that of a phase-shifter-based hybrid antenna structure. Under reasonable communication quality constraints, the radar beam of the radar-communication integrated system with the dynamic metasurface antenna architecture approaches the performance of a radar beam without spectrum sharing.

Description

Technical Field

[0001] This invention relates to the field of radar-communication integration, specifically to a metamaterial antenna architecture radar-communication integrated system and waveform optimization method. Background Technology

[0002] With the advent of the 5G era, wireless devices have experienced explosive growth, and the global communications industry's demand for wireless spectrum is becoming increasingly urgent. Currently, in civilian applications, many scenarios require the joint design of sensing and communication, such as autonomous driving, smart cities, and smart homes. Simultaneously, with the continuous increase in wireless communication speeds, carrier frequencies have been pushed into the millimeter-wave frequency band traditionally allocated to radar systems. To limit electromagnetic interference between radar and communication, systems where radar and communication coexist need to be designed. Furthermore, in the millimeter-wave frequency band, the hardware structure, channel characteristics, and signal processing methods of radar and communication are very similar. Dual-functional radar-communication (DFRC) systems share the same hardware platform and spectrum resources between radar and communication, realizing both communication and radar sensing functions.

[0003] However, since the performance requirements of radar and communication functions are often conflicting, the transmission beam needs to be carefully designed to balance their respective needs. To improve radar performance while ensuring user communication quality, the transmit beamformer is optimized to match the required radar beam. To achieve even better radar performance, beam interference is also introduced. The antenna architecture used in this study is a dynamic metamaterial antenna architecture, where each antenna element must be configured with a corresponding phase shifter, resulting in high hardware costs. Currently, the most widely studied DFRC system is the hybrid beamformer antenna architecture based on phase shifters, which involves designing analog and digital beamformers to minimize the weighted sum of errors between them and the optimal communication precoder and the optimal radar beamformer; the research mainly focuses on radar beam constraints. Furthermore, the beam performance of the radar studied is directly determined by the signal-to-interference-plus-noise ratio (SINR) of the radar receiver.

[0004] Smart metasurfaces are generally used for blind zone coverage, physical layer-assisted secure communication, multi-stream transmission rank enhancement, edge coverage enhancement, large-scale D2D (Device-to-Device) communication, wireless power and information transmission in the Internet of Things (IoT), and indoor coverage. Indoor coverage relies on the refraction principle of smart metasurfaces, while other applications rely on their reflection principle. Besides passive reflection, metasurfaces can also be used as active transmitting antennas. Smart metasurfaces can function as active transmitting antennas because they can encode in real time. Baseband signals are coded and fed into the smart metasurface controller, and then the target frequency band RF carrier is transmitted onto the smart metasurface. Reflection modulates the baseband signal onto the carrier, creating a dynamic metasurface antenna (DMA) architecture. In DMA-based transceivers, each metamaterial antenna element can exhibit a wide frequency response with different combinations of amplitude and phase values, ranging from frequency selectivity to frequency-flat profiles. Each antenna element is controllable, meaning that DMA-based base stations can perform different attenuation and phase shift operations on transmitted and received signals. Therefore, DMA-based base stations can be viewed as hybrid A / D beamforming systems, but they do not require additional dedicated analog combinational circuitry, while offering greater flexibility. Since DMA can incorporate a large number of tunable metasurface antenna elements, and the spacing between its antenna elements can be smaller, the physical area required by DMA can be smaller, contributing to device miniaturization. Summary of the Invention

[0005] To apply DMA to integrated radar communication design, this invention discloses a metamaterial antenna architecture integrated radar communication system and waveform optimization method. This system optimizes the performance of radar transmission beamforming while satisfying communication user quality and power constant mode constraints.

[0006] The specific technical solution adopted in this invention is as follows:

[0007] A metamaterial antenna architecture radar-communication integrated system employs a dynamic metamaterial antenna, wherein the baseband signal s∈C of the system is... K×1 , where s i ~CN(0,1),i∈{1,...,K} represents the information symbol received by the i-th user.

[0008] The transmitted signal can be represented as:

[0009] y = F DMA F BB s

[0010] in Simulate the precoder matrix for the DMA antenna. For a digital pre-encoder matrix, the power constraint is: P max This is the maximum power allocation for the baseband. F DMA The matrix satisfies the following form:

[0011]

[0012] parameter Among them, non-zero items

[0013] The radar's transmitted power beam pattern in the θ-angle direction can be represented as:

[0014]

[0015] in The covariance matrix of the transmission beam:

[0016]

[0017] For a uniform linear antenna array with N antenna elements, its steering vector is:

[0018]

[0019] Where λ is the signal wavelength and d = λ / 2 is the spacing between antenna elements.

[0020] The beam cross-correlation of the radar between angles θ1 and θ2 can be expressed as:

[0021]

[0022] It can be seen that the radar's transmitted power beam pattern and beam cross-correlation are both determined by the covariance matrix R of the transmitted beam.

[0023] A loss function is constructed by weighting the beam direction error and beam cross-correlation, and the radar performance is evaluated using the loss function.

[0024] The first part can be evaluated using the mean square error between the received beam and the ideal beam:

[0025]

[0026] α is a scaling factor, d(θ) l ) is θ l Ideally oriented receiving beam.

[0027] The second part uses the mean square error of beam cross-correlation for evaluation:

[0028]

[0029] After weighting and summing the above two parts, the loss function of the radar beam pattern is expressed as:

[0030] L r (R,α)=L r,1 (R,α)+ωL r,2 (R)

[0031] Assuming each communication user has a single antenna, the signal received by the k-th user is:

[0032]

[0033] in This is the downlink channel between the base station and the k-th user. The noise is additive white Gaussian noise (AWGN) for the k-th user.

[0034] The SINR of the k-th user can be represented as:

[0035]

[0036] In summary, the problem can be expressed as:

[0037]

[0038] Where matrix W is the precoding matrix, h k For the downlink channel, Γ represents the signal-to-interference-plus-noise ratio threshold for communication users.

[0039] A waveform optimization method for a metamaterial antenna architecture radar-communication integrated system, using the aforementioned metamaterial antenna architecture radar-communication integrated system, specifically includes the following steps:

[0040] Step 1: Solve for the optimal precoding matrix using the all-digital antenna architecture radar-communication integrated system, and compare its radar beam with the ideal radar beam;

[0041] Step 2: Transform the radar beam design problem of the metamaterial antenna architecture radar-communication integrated system into the problem of fitting the optimal coding matrix;

[0042] Step 3: The problem of fitting the optimal encoding matrix for the metamaterial antenna architecture radar-communication integrated system is a non-convex problem, which is not easy to solve directly. It is decomposed into two sub-problems that are alternately minimized.

[0043] Step 4: When designing the optimal analog precoder with a fixed digital precoder matrix, the problem can be simplified to a vector problem with elements having a modulus of 1. The optimal analog precoder can be solved using the Riemann conjugate gradient.

[0044] Step 5: When designing the optimal digital precoder with a fixed analog precoder matrix, simplify the problem according to its form and reduce it to a quadratically constrained quadratic program (QCQP) problem using auxiliary variables. Since the constraints are non-convex, use the positive semidefinite relaxation technique to solve for the optimal digital precoder.

[0045] A further improvement to this invention is that, in step 2 above, the radar beam design problem of the metamaterial antenna architecture radar-communication integrated system is transformed into a problem of fitting the optimal coding matrix, with the specific function as follows:

[0046]

[0047] Where matrix F DMA With F BB For analog precoders and digital precoders, For the optimal dynamic all-digital antenna pre-encoder, P max For power constraints.

[0048] A further improvement of the present invention is that the problem of designing the optimal analog precoder with a fixed digital precoder matrix in step 4 is expressed as:

[0049]

[0050] Since the problem is in matrix form, which is inconvenient to solve, we will vectorize the matrix:

[0051] in

[0052] Because vec(F DMA The elements in ) are excluding phase-bound elements, i.e., zero elements. Since the exact location of the zero elements is known, they can be removed first. Let q be vec(F) DMA The vector A after removing zero elements is... Remove vec(F) DMA The column vector corresponding to the zero element. The problem then becomes:

[0053]

[0054] Due to the non-zero element q of the simulated preencoder i,l This can be represented as the center point being... radius is On the complex plane circle: Define vector b as: so Ultimately, the problem can be reduced to a problem about vector b:

[0055]

[0056] st|b k |=1∈b

[0057] This is a search space of N. T On a complex circle, there is a... The optimal solution b of the Riemannian submanifold can be obtained through the Riemann conjugate gradient. opt The Riemann gradient of this problem is:

[0058]

[0059] Due to F DMA Since the non-zero positions are known, the optimal solution b will be... opt Expanding this to matrix form yields the optimal analog precoder matrix.

[0060] A further improvement to this invention is that the design of the optimal digital precoder using the fixed analog precoder matrix in step 5 is represented as follows:

[0061]

[0062] Due to the second constraint F BB It is expanded column-wise, so the matrix F in the problem... BB and Also expand by column:

[0063]

[0064] The expanded problem is not easy to solve, so we introduce an auxiliary variable t. 2 =1, which can be resolved into a Quadratically Constrained Quadratic Programming (QCQP) problem:

[0065]

[0066] However, at this point, the second constraint of the original problem is non-convex, so this problem is also non-convex. Using the SDR technique to simplify the problem, let... The problem can be simplified to the standard SDR form:

[0067]

[0068] Due to the constraint term rank(y) k Since 1 = 1 is non-convex, we relax it first. The subsequent problem is convex, and we can use the CVX toolbox in MATLAB to find the optimal solution.

[0069] If the problem is solvable or bounded, then Furthermore, due to the SINR threshold limitation for each user, the optimal solution satisfies: Therefore, its optimal solution satisfies

[0070] Therefore, rank(y) is proved. k A relaxation of 1 is tight. It is the optimal solution to the original problem. yes Multiplying the largest eigenvector by the square root of the largest eigenvalue yields the optimal digital precoder matrix.

[0071] Compared with the prior art, the present invention has the following advantages:

[0072] This invention uses a dynamic metamaterial antenna, which saves costs and occupies a small area. Its radar beam performance is similar to that of a phase shifter-based hybrid antenna structure. Under reasonable communication quality constraints, the radar-communication integrated radar beam of the dynamic metasurface antenna architecture approaches the performance of radar beams without spectrum sharing. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the waveform optimization method for the metamaterial antenna architecture radar-communication integrated system of the present invention. Detailed Implementation

[0074] To facilitate understanding and implementation of the present invention by those skilled in the art, the present invention will be further described in detail below with reference to embodiments. It should be understood that the embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0075] Example: A waveform optimization method for a radar-communication integrated system with a metamaterial antenna architecture. The method uses a dynamic metamaterial antenna pre-encoder to fit the optimal pre-encoder of the all-digital antenna, and finally obtains the optimal pre-encoder of the dynamic metamaterial antenna. The system satisfies the communication user quality and power constant mode constraints, while optimizing the performance of radar transmission beamforming.

[0076] A waveform optimization method for a metamaterial antenna architecture integrated radar-communication system, the method comprising the following steps:

[0077] Step 1: Design an all-digital antenna architecture radar-communication integrated system. The optimal precoding matrix can be obtained by solving the problem, and its radar beam can be compared with the ideal radar beam.

[0078] Step 2: Set the precoding matrix of the dynamic metamaterial antenna as the optimal coding matrix. The radar beam design problem of the dynamic metasurface antenna architecture is transformed into the problem of fitting the optimal coding matrix.

[0079] Step 3: The problem of fitting the optimal encoding matrix for the dynamic metasurface antenna architecture is a non-convex problem, which is not easy to solve directly. Therefore, it is decomposed into two sub-problems that are alternately minimized.

[0080] Step 4: When designing the optimal analog precoder with a fixed digital precoder matrix, the problem can be simplified to a vector problem with elements having a modulus of 1. The optimal analog precoder can be solved using the Riemann conjugate gradient.

[0081] Step 5: When designing the optimal digital precoder with a fixed analog precoder matrix, simplify the problem according to its form and reduce it to a quadratically constrained quadratic program (QCQP) problem using auxiliary variables. Since the constraints are non-convex, use the positive semidefinite relaxation technique to solve for the optimal digital precoder.

[0082] In step 1, a fully digital antenna architecture integrating radar and communication is designed (e.g., Figure 1 As shown):

[0083]

[0084] Where matrix W is the precoding matrix, h k For the downlink channel, Γ represents the signal-to-interference-plus-noise ratio threshold for communication users.

[0085] In step 2, the radar beam design problem of the dynamic metasurface antenna architecture is transformed into the problem of fitting the optimal coding matrix of the all-digital antenna:

[0086]

[0087] Where matrix F DMA With F BB For analog precoders and digital precoders, For the optimal all-digital antenna pre-encoder, P max For power constraints.

[0088] In step 3, the problem is decomposed into two sub-problems that are designed to be solved iteratively. The two sub-problems are designed as digital and analog pre-encoders, respectively.

[0089] The problem of designing the optimal analog precoder with a fixed digital precoder matrix in step 4 can be expressed as:

[0090]

[0091] Since the problem is in matrix form, which is inconvenient to solve, we will vectorize the matrix:

[0092] in

[0093] Because vec(F DMA The elements in ) are excluding phase-bound elements, i.e., zero elements. Since the exact location of the zero elements is known, they can be removed first. Let q be vec(F) DMA The vector A after removing zero elements is... Remove vec(F) DMA The column vector corresponding to the zero element. The problem then becomes:

[0094]

[0095] Due to the non-zero element q of the simulated preencoder i,l This can be represented as the center point being... radius is On the complex plane circle: Define vector b as: so Ultimately, the problem can be reduced to a problem about vector b:

[0096]

[0097] st|b k |=1∈b

[0098] This is a search space of N. T On a complex circle, there is a... The optimal solution b of the Riemannian submanifold can be obtained through the Riemann conjugate gradient. opt The Riemann gradient of this problem is: Due to F DMA Since the non-zero positions are known, the optimal solution b will be... opt Expanding this to matrix form yields the optimal analog precoder matrix.

[0099] The problem of designing the optimal digital precoder with a fixed analog precoder matrix in step 5 can be expressed as:

[0100]

[0101] Due to the second constraint F BB It is expanded column-wise, so the matrix F in the problem... BB and Also expand by column:

[0102]

[0103] The expanded problem is not easy to solve, so we introduce an auxiliary variable t. 2 =1, which can be resolved into a Quadratically Constrained Quadratic Programming (QCQP) problem:

[0104]

[0105] However, since the second constraint of the original problem is non-convex, this problem is also non-convex. Using the SDR technique, the problem is simplified, letting... rank(y k If ) = 1, the problem can be simplified to the standard SDR form:

[0106]

[0107] Due to the constraint term rank(y) k Since 1 = 1 is non-convex, we relax it first. The subsequent problem is convex, and we can use the CVX toolbox in MATLAB to find the optimal solution.

[0108] If the problem is solvable or bounded, then Furthermore, due to the SINR threshold limitation for each user, the optimal solution satisfies: Therefore, its optimal solution satisfies Therefore, rank(y) is proved. k A relaxation of 1 is tight. It is the optimal solution to the original problem. yes Multiplying the largest eigenvector by the square root of the largest eigenvalue yields the optimal digital precoder matrix.

[0109] Based on the above example, perform data simulation:

[0110] In this section, a digital model is used to verify the DMA radar-communication integrated design algorithm. It is assumed that the antenna of the radar-communication integrated base station is a uniform linear antenna array with a total transmit power of 1 and a number of antennas of 24. It provides communication services to users and detects targets within the detection area. Three ideal targets with orientations of -40°, 0°, and 40° are set within the detection area, and their beam expressions are as follows:

[0111]

[0112] Δ is the width of the ideal beam, which is set to 2° here.

[0113] When the system design includes 12 DMA RF links and the signal-to-noise ratio is set to 20dB, simulations are conducted to optimize the radar beam under different antenna architectures while meeting user requirements. The optimal precoder is designed for a dynamic metamaterial antenna architecture, a DMA architecture, and a phase shifter-based hybrid architecture, with SINR thresholds of 6dB and 14dB for the four communication users.

[0114] In a DMA-based integrated radar system, the trade-off between user SINR threshold constraints and radar beam performance under different transmit power conditions is explored. At a constant transmit power, as the user signal-to-interference-plus-noise ratio (SINR) threshold increases, the mean square error (MSE) between the DMA antenna pre-encoder matrix and the dynamic metamaterial antenna pre-encoder matrix also increases. Furthermore, the MSE at a transmit power of 1 is significantly greater than that at a power of 1. This is because as communication quality requirements increase, more power is needed to provide sufficient user quality for transmission, while generating the radar beam requires less power, leading to a deterioration in radar beam performance. Therefore, reducing communication quality requirements can improve radar beam performance.

[0115] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A metamaterial antenna architecture radar-communication integrated system, characterized in that, Employing dynamic metamaterial antennas, The baseband signal of this system is: s∈C K×1 , where s i ~CN(0,1),i∈{1,…,K} represents the information symbol received by the i-th user; The transmitted signal is: y = F DMA F BB s, where Simulate the precoder matrix for the DMA antenna. For a digital pre-encoder matrix, the power constraint is: P max Maximum power allocation to baseband; F DMA The matrix satisfies the following form: parameter Among them, non-zero items The radar's transmitted power beam pattern in the θ-angle direction is as follows: in Here is the covariance matrix of the transmission beam: For a uniform linear antenna array with N antenna elements, its steering vector is: Where λ is the signal wavelength, and d = λ / 2 is the spacing between antenna elements. The beam cross-correlation of the radar between angles θ1 and θ2 is expressed as:

2. A waveform optimization method for a radar-communication integrated system based on the metamaterial antenna architecture described in claim 1, characterized in that, Specifically, the following steps are included: Step 1: Solve for the optimal precoding matrix using the all-digital antenna architecture radar-communication integrated system, and compare its radar beam with the ideal radar beam; Step 2: Transform the radar beam design problem of the metamaterial antenna architecture radar-communication integrated system into the problem of fitting the optimal coding matrix; Step 3: Decompose the problem of fitting the optimal coding matrix for the metamaterial antenna architecture radar-communication integrated system into two sub-problems that are alternately minimized; Step 4: When designing the optimal analog precoder with a fixed digital precoder matrix, the problem is simplified to a vector problem with elements having a modulus of 1. The optimal analog precoder can be solved using the Riemann conjugate gradient. Step 5: When designing the optimal digital precoder with a fixed analog precoder matrix, simplify the problem according to its form and reduce it to a quadratic constrained quadratic programming problem using auxiliary variables. Solve for the optimal digital precoder using the positive semidefinite relaxation technique.

3. The waveform optimization method for the metamaterial antenna architecture radar-communication integrated system according to claim 2, characterized in that, In step 2, the radar beam design problem of the metamaterial antenna architecture radar-communication integrated system is transformed into a problem of fitting the optimal coding matrix, and the specific function is as follows: Where matrix F DMA With F BB For analog precoders and digital precoders, For the optimal dynamic metamaterial antenna pre-encoder, P max For power constraints.

4. The waveform optimization method for the metamaterial antenna architecture radar-communication integrated system according to claim 3, characterized in that, In step 4, the problem of designing the optimal analog precoder with a fixed digital precoder matrix is ​​expressed as follows: Since the problem is in matrix form, which is inconvenient to solve, we will vectorize the matrix: in 5. The waveform optimization method for the metamaterial antenna architecture radar-communication integrated system according to claim 4, characterized in that, In step 4, let q be vec(F) DMA The vector A after removing zero elements is... Remove vec(F) DMA The problem then becomes: (The column vector corresponding to the zero element) Due to the non-zero element q of the simulated preencoder i,l Represented by the center point as radius is On the complex plane circle: Define vector b as: so Ultimately, the problem can be reduced to a problem about vector b:

6. The waveform optimization method for the metamaterial antenna architecture radar-communication integrated system according to claim 5, characterized in that, In step 5, the optimal digital precoder is designed with a fixed analog precoder matrix. The problem is expressed as: Due to the second constraint F BB It is expanded column-wise, so the matrix F in the problem... BB and Also expand by column: Introducing auxiliary variable t 2 =1, which reduces the problem to a quadratic constraint quadratic programming problem: The problem is simplified by using SDR technology, making rank(y k ) = 1, simplifying the problem to the standard SDR form: Finding the optimal solution using the CVX toolbox in MATLAB If the problem is solvable or bounded, then Furthermore, due to the SINR threshold limitation for each user, the optimal solution satisfies: Therefore, its optimal solution satisfies Therefore, rank(y) is proved. k A relaxation of 1 is tight. It is the optimal solution to the original problem. yes Multiplying the largest eigenvector by the square root of the largest eigenvalue yields the optimal digital precoder matrix.

Images