Hybrid metasurface auxiliary sensing integrated system for vehicle-to-vehicle communication

Abstract

The invention relates to a hybrid metasurface auxiliary sensing integrated system for vehicle-to-vehicle communication, and belongs to the technical field of communication. A hybrid metasurface auxiliary sensing integrated system for vehicle-to-vehicle communication comprises a BS (Base Station), an information receiving vehicle with L antennas and K target vehicles in a V2V (Vehicle-to-Vehicle) network, the system is characterized in that the BS is provided with a uniform linear array composed of M antennas; the system further comprises a hybrid RIS with N antennas, the hybrid RIS comprises Na active elements and N-Na passive elements, the position index set of the active elements is # imgabs0 #, # imgabs1 #, the V2V network comprises a signal sending model, a communication model, a radar model and a power consumption model, the spectrum efficiency is optimized through an AO algorithm, and the communication efficiency is improved.

Description

Technical Field

[0001] The present invention relates to a hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication, belonging to the technical field of communications. Background Art

[0002] Currently, there are numerous discussions on combining the advantages of RIS and ISAC to further enhance the utility of V2V networks. For example, some researchers have designed RIS reflector array elements that direct the beam toward the vehicle target while embedding communication symbols. Similarly, there are also efforts to use RIS and interawareness systems to detect the user's location, leveraging this information to reduce the computational complexity of communication channel estimation. However, in practical systems, due to the passive nature of RIS, its gain is limited by multiplicative fading. To address this, some have proposed active RIS, which utilizes amplifiers to amplify the incident signal. However, a large number of active elements results in excessive power consumption. Currently, few studies have examined the impact of RIS multiplicative fading on the interawareness function in V2V systems, and research on how to utilize limited active RIS elements to improve the communication perception performance of V2V networks remains unresolved. Balancing power consumption and multiplicative fading has become a major challenge in RIS applications. Summary of the Invention

[0003] In view of the problems existing in the prior art, a hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication is provided.

[0004] The present invention solves the above technical problems through the following technical solutions:

[0005] A hybrid metasurface-assisted interawareness integration algorithm for vehicle-to-vehicle communication (V2V) is proposed. In a V2V network (Vehicle to Vehicle, i.e., vehicle-to-vehicle communication technology), a base station (BS), an information receiving IR vehicle (IR) with L antennas, and K target vehicles are included. The special feature is that the BS is equipped with a uniform linear array consisting of M antennas. The algorithm also includes a reconfigurable intelligent surface (RIS) with N antennas, which contains Na active elements and N-Na passive elements. The position index set of the active elements is in The V2V network includes a signal transmission model, a communication model, a radar model, and a power consumption model. It uses a hybrid RIS to assist signal propagation and uses an alternating optimization (AO) algorithm to solve the optimization problem, thereby optimizing the system's spectral efficiency (SE) and improving communication efficiency.

[0006] On the basis of the above technical solution, this application also makes the following improvements and improvements to the above technical solution:

[0007] Furthermore, the signal transmission model is:

[0008] Definitions r is the detection signal used to perceive the target vehicle, s c To send communication symbols to IR vehicles; use communication precoder w c and perceptual precoder W r =[w r，1 ,...,w r，M ] respectively for s c and s r After precoding, the sum is performed; thus, the transmitted signal at the BS Expressed as:

[0009] x=W r s r +w c s c =W s (1),

[0010] Among them, define N s =M+1, indicating the total number of communication symbols; W=[W r , w c ] is defined as an M×N s The precoder matrix of size , Defined as N S elements of the transmission symbol vector; [·] T represents the transpose operation of the vector, and represents the column vector; it is assumed that these signals have unit power and have statistical independence characteristics, that is, the covariance of the complex baseband signal satisfies Indicates the expected operation, is an identity matrix.

[0011] Furthermore, the communication model is:

[0012] After the signal is transmitted from the BS, it reaches the IR vehicle through the RIS reflection channel;

[0013] Let G be the channel between BS and hybrid RIS, H r is the channel between the hybrid RIS and IR vehicles; the hybrid RIS coefficient matrix is recorded as p n Indicates the amplification factor of RIS to the signal, θ n represents the amplification and phase shift coefficient; when p n =1, otherwise p n>1; therefore, the signal received by the receiving vehicle is expressed as:

[0014] y I =H r ΥGx+H r Υz1+n (2),

[0015] Among them, the thermal noise of the hybrid RIS active component is Indicates that the noise obeys an independent complex Gaussian distribution with a covariance of 1. Unless otherwise specified, the thermal noise mentioned later should also obey this distribution; diag(·) indicates the construction of a diagonal matrix operation; represents the additive white Gaussian noise (AWGN) of the IR vehicle receiver;

[0016] Accordingly, the spectral efficiency of the receiving vehicle is:

[0017]

[0018] Among them, H I =H r ΥG represents the equivalent channel from BS to IR vehicle, (·) H Represents the conjugate transpose operation of a matrix.

[0019] Furthermore, the radar model is:

[0020] In addition to serving IR vehicles, the transmitted signal will also illuminate target vehicles after passing through the hybrid RIS;

[0021] set up is the channel between the hybrid RIS and the kth target vehicle; the entire target detection channel can be expressed as

[0022] Due to the presence of obstacles, the direct link is not considered. The echo signal received by the BS includes the signal of the BS-RIS-BS link and the signal transmitted by the BS-RIS-target-RIS-BS link, as well as the thermal noise of the BS receiver and the hybrid RIS active components. Accordingly, the target echo signal is expressed as:

[0023]

[0024] in:

[0025]

[0026] Both z1 and z2 represent the thermal noise introduced from the RIS active components; in addition, z rcis the AWGN at the BS receiver. Since the power is usually very small, this noise will be ignored in the following. The channel between the BS and the RIS and the reflection matrix of the RIS are known. Therefore, the signal H directly reflected by the RIS is ie x can be processed using self-interference cancellation technology; thus, the target echo signal is updated as:

[0027]

[0028] The performance of radar sensors is evaluated using the Signal-to-Interference-plus-Noise Ratio (SINR) of the received signal.

[0029]

[0030] Among them, tr(·) represents the matrix trace operation, represents the real part operation, where

[0031]

[0032] Since SINR directly affects the detection probability and parameter estimation accuracy, it is used as an indicator for target sensing.

[0033] Furthermore, the power consumption model is:

[0034] For the BS, the consumed transmit power is expressed as:

[0035] P BS =tr(WW H ) (9),

[0036] The power consumption of the RIS is due to the amplification of two reflections. The first reflection of the RIS reflects the signal transmitted by the BS to the target vehicle and the IR vehicle, and the second reflection reflects the echo signal of the target vehicle back to the BS. The expression of the first and second reflection signals of the mixed RIS is:

[0037]

[0038] Therefore, the power consumption of hybrid RIS is:

[0039]

[0040] Here, ||·||2 represents the L2 norm; thus, the total power consumption of the system is:

[0041] P=P BS +P RIS (13).

[0042] further,

[0043] By jointly designing the BS precoder E and the reflection matrix Y of the hybrid RIS to maximize the SE of the receiving vehicle, i.e., R, under the condition of limited power consumption of the BS and the hybrid RIS, and ensuring the minimum threshold ζ of the radar signal-to-interference-noise ratio to ensure target detection performance; the optimization problem is then written as:

[0044]

[0045] This optimization problem is a non-convex optimization problem.

[0046] Furthermore, the non-convex optimization problem is equivalently represented and divided into four sub-problems for solution;

[0047] First, the IR vehicle multiplies the equalizer matrix to recover the communication symbols Then the mean-square error (MSE) of communication symbol estimation is expressed as:

[0048]

[0049] After substituting (2) into (15), the MSE at the receiving vehicle can be further expressed. By introducing a positive auxiliary variable V, the objective function can be equivalently expressed by MSE:

[0050]

[0051] Among them, det(·) represents the determinant operation of the square matrix, tr(·) represents the trace operation of the matrix, and the optimal solution is V ＊ =[E1(C, Y, W)] -1 ; At a given V ＊ In this case, the optimal solution C ＊ Obtain

[0052]

[0053] With the solved equalizer, the minimum mean square error (MMSE) can be further expressed as:

[0054]

[0055] Therefore, combining formulas (10), (11) and (18), the optimization problem (14) can be transformed into:

[0056]

[0057] That is, under the four constraints of C1, C2, C3, and C4, the objective function affected by the four variables C, W, Y, and V is maximized.

[0058] Furthermore, the alternating optimization (AO) algorithm is used to solve the four subproblems:

[0059] (1) Optimize the auxiliary variable V while fixing the equalizer matrix C, precoding matrix W, and RIS reflection matrix Y:

[0060] Optimize V with fixed C, W, and Y; the optimization problem is simplified to an unconstrained optimization problem, and the optimal solution V is obtained directly by derivation. ＊ for:

[0061] V ＊ =[E1(C, Y, W)] -1 (20);

[0062] (2) Optimize the equalizer matrix C with fixed auxiliary variables V, precoding matrix W, and RIS reflection matrix Y:

[0063] Optimize C with fixed V, W and Y;

[0064] The optimization problem can be expressed as:

[0065]

[0066] in, The optimal solution is shown in formula (17);

[0067] (3) Optimize the precoding matrix W while fixing the equalizer matrix C, auxiliary variables V, and RIS reflection matrix Y:

[0068] Optimize W with C, V, and Y fixed; substitute formula (18) into the objective function and ignore irrelevant terms. The objective function can be reformulated as:

[0069]

[0070] in,

[0071] Constraints C1 and C3 are non-convex; to handle C1, C1 can be written as:

[0072]

[0073] Performing a first-order Taylor expansion on the second term in (23), the relaxed constraint C1 can be transformed into:

[0074]

[0075] Similarly, constraint C3 can also be transformed into a quadratic constraint:

[0076]

[0077] in:

[0078] J=G H Υ H A H ΥΥ H AΥG,

[0079] P=G H Υ H ΥG

[0080]

[0081] Therefore, the optimization problem can be transformed into:

[0082]

[0083] This problem is a convex QCQP problem and can be solved directly by the interior point method;

[0084] In addition, problem (26) can also be solved using the SDR algorithm; by introducing auxiliary variables, the problem is transformed into SDR form and can be solved directly;

[0085] (4) Optimize the RIS reflection matrix Y with fixed equalizer matrix C, auxiliary variables V, and precoding matrix W:

[0086] Ignoring irrelevant terms, the optimization problem can be simplified to:

[0087]

[0088] Where Z = GWVC H H r , P2=GWW H G H ;

[0089] Using the properties of the trace, the terms of the objective function are transformed into:

[0090]

[0091] in Z=GWVC H H r , z=[[Z] 1，1 ,…,[Z] N，N ] T And Ξ1=(P1)⊙(P2) T ,Ξ2=ησ 2 (P1)⊙I N ; where ⊙ represents the Hadamard product of the matrix.

[0092] The optimization problem is equivalent to:

[0093]

[0094] st C1, C3, C4 (33);

[0095] Where Ξ=Ξ1+Ξ2; since C1 and C3 are non-convex, the optimization problem is still non-convex;

[0096] To change the constraint into a convex constraint, first, we process the constraint C3: Substitute (10) and (11) into equation (13). The left side of the C3 inequality can be expressed as follows:

[0097]

[0098] In order to solve the constraint C3, the trace of the matrix and the properties of vectorization can be used to transform the terms on the left side of the constraint C3, and we can get:

[0099]

[0100] in:

[0101]

[0102] Among them, vec(·) represents the matrix vectorization operation, represents the Kronecker product; is a Hermitian matrix; using the first-order Taylor expansion and imaginary-real transformation, the equivalent expression is:

[0103]

[0104] Similarly, constraint C1 can be transformed into formula (24), focusing on simplifying the final term. By transforming each term in (24), the equivalent expression is finally obtained:

[0105]

[0106] in:

[0107]

[0108] in is a Hermitian matrix; through first-order Taylor expansion and complex transformation, the optimization problem can be transformed into:

[0109]

[0110] The problem can be transformed into SDR form for solution, and finally v can be obtained using rank-one decomposition.

[0111] Through step-by-step optimization, we can eventually find the optimal solution for the RIS-assisted V2V network equalizer matrix C, precoding matrix W, and RIS reflection matrix Y. This maximizes the spectrum efficiency of the communication network and improves the communication capabilities of the application system while balancing the spectrum resources occupied by communication and perception.

[0112] Based on this, the spectrum efficiency optimization problem of the hybrid RIS-assisted synaesthesia integrated system in V2V was solved, providing a feasible mathematical basis and specific implementation method for the system, indicating that the hybrid RIS-assisted synaesthesia integrated system can effectively improve the communication efficiency in the V2V network.

[0113] This paper pioneers the study of a hybrid RIS-assisted ISAC system in a V2V network, improving communication performance while maintaining radar detection performance. Under constraints such as vehicle perception performance and maximum power consumption, the RIS's precoding matrix and reflection matrix are jointly optimized to maximize the SE of the IR vehicle. Because this problem is complex and non-convex, we propose an AO algorithm that divides it into four subproblems. Simulation results demonstrate the effectiveness of the proposed algorithm and show that the low-power hybrid RIS-assisted ISAC system outperforms both perception and communication performance compared to systems based on either a single passive RIS or an active RIS.

[0114] To address the problem of low spectral efficiency in vehicle-to-vehicle communication systems, the present invention designs a hybrid RIS-assisted ISAC system for vehicle-to-vehicle communication to optimize the spectral efficiency of the receiving vehicle while satisfying the constraints of the target vehicle's perceived signal-to-noise ratio, transmission power, and RIS phase shift. To solve this problem, we decompose the optimization problem into four sub-problems and employ an alternating optimization (AO) algorithm based on semi-definite programming (SDR) and quadratic programming. Because the present invention utilizes a low-power hybrid RIS in the ISAC system, it can both adjust the phase shift and amplify the signal, whereas a passive RIS can only adjust the phase shift. Therefore, compared to passive RIS solutions, this invention offers significant advantages in improving the communication capacity of the ISAC system. BRIEF DESCRIPTION OF THE DRAWINGS

[0115] Figure 1 A hybrid RIS-assisted ISAC system in a V2V network;

[0116] Figure 2 It is the algorithm flow chart;

[0117] Figure 3 is the convergence behavior of the AO algorithm;

[0118] Figure 4 is the relationship between SE and active element ratio;

[0119] Figure 5 is the relationship between SE and residual self-interference power;

[0120] Figure 6 is the relationship between SE and radar SINR threshold. DETAILED DESCRIPTION

[0121] The following embodiments, in conjunction with the accompanying drawings, are only intended to illustrate the technical solutions described in the claims and are not intended to limit the scope of protection of the claims.

[0122] Attachment Figure 2 This is a flowchart for optimizing the spectrum effect of a hybrid RIS-assisted ISAC system on a receiving vehicle in a V2V network while satisfying the target vehicle's perceived signal-to-noise ratio and power consumption constraints. The specific steps are as follows: Step 1: Mathematically model the hybrid RIS-assisted ISAC system in the V2V network, analyze each model, and then derive the optimization problem to be solved.

[0123] As attached Figure 1 As shown in the figure, the present invention considers an application scenario of a hybrid RIS-assisted ISAC system in a V2V network. Among them, a DFRC base station (BS) provides communication services for an information receiving vehicle (IR) with L antennas and senses K target vehicles. The BS is equipped with a uniform linear array (ULA) composed of M antennas. Due to the presence of obstacles, the direct link between the BS and the target vehicle and the IR vehicle is blocked. In order to establish an effective link, a hybrid RIS with N antennas is deployed, which includes N a Active components, NN a passive components, and the position index set of active components is predetermined, where

[0124] Based on this, we established the signal transmission model, communication model, radar model and power consumption model for this scenario.

[0125] (1) Sending signal model:

[0126] definition is used to perceive the target signal of the vehicle. To send communication symbols to IR vehicles. Use communication precoder and perceptual precoder For s c and s rAfter precoding, the sum is performed. Thus, the transmitted signal at the BS is Expressed as:

[0127] x=W r s r +w c s c =Ws (1)

[0128] Among them, define N s =M+1, indicating the total number of communication symbols; W=[W r , w c ] is defined as an M×N s The precoder matrix of size , Defined as N S elements of the transmission symbol vector; [·] T represents the transpose operation of the vector, and represents the column vector; it is assumed that these signals have unit power and have statistical independence characteristics, that is, the covariance of the complex baseband signal satisfies Indicates the expected operation, is an identity matrix.

[0129] (2) Communication model:

[0130] After the signal is transmitted from the BS, it reaches the IR vehicle through the RIS reflection channel.

[0131] Let G be the channel between BS and hybrid RIS, H r is the channel between the hybrid RIS and IR vehicles; the hybrid RIS coefficient matrix is recorded as p n Indicates the amplification factor of RIS to the signal, θ n represents the amplification and phase shift coefficient; when p n =1, otherwise p n >1; therefore, the signal received by the receiving vehicle is expressed as:

[0132] y I = r ΥGx+H r Υz1+n (2)

[0133] Among them, the thermal noise of the hybrid RIS active component is Indicates that the noise obeys an independent complex Gaussian distribution with a covariance of 1. Unless otherwise specified, the thermal noise mentioned later should also obey this distribution; diag(·) indicates the construction of a diagonal matrix operation; represents the additive white Gaussian noise (AWGN) of the IR vehicle receiver.

[0134] Correspondingly, the spectrum efficiency SE of the receiving vehicle is

[0135]

[0136] Among them, H I =H r ΥG represents the equivalent channel from BS to IR vehicle, (·) H Represents the conjugate transpose operation.

[0137] (3) Radar model:

[0138] In addition to serving IR vehicles, the transmitted signal will also illuminate target vehicles after passing through the hybrid RIS.

[0139] set up is the channel between the hybrid RIS and the kth target vehicle. The entire target detection channel can be expressed as

[0140] Due to the presence of obstacles, the direct link is not considered. The echo signal received by the BS includes the signal of the BS-RIS-BS link and the signal transmitted by the BS-RIS-target-RIS-BS link, as well as the thermal noise of the BS receiver and the active components of the hybrid RIS. Accordingly, the target echo signal is expressed as:

[0141]

[0142] in

[0143]

[0144] Both z1 and z2 represent the thermal noise introduced from the RIS active components; in addition, z rc is the AWGN at the BS receiver. Since the power is usually very small, this noise will be ignored in the following. The channel between the BS and the RIS and the reflection matrix of the RIS are known. Therefore, the signal H directly reflected by the RIS is ie x can be processed using self-interference cancellation technology; thus, the target echo signal is updated as:

[0145]

[0146] The performance of radar sensors is evaluated using the Signal to Interference plus Noise Ratio (SINR) of the received signal.

[0147]

[0148] Among them, tr(·) represents the trace operation, represents the real part operation,

[0149]

[0150] Since SINR directly affects the detection probability and parameter estimation accuracy, it is used as an indicator for target sensing.

[0151] (4) Power consumption model:

[0152] For BS, the consumed transmit power is expressed as

[0153] P BS =tr(WW H ) (9)

[0154] The power consumption of the RIS is due to the amplification of two reflections. The first reflection of the RIS reflects the signal transmitted by the BS to the target vehicle and the IR vehicle, and the second reflection reflects the echo signal of the target vehicle back to the BS. The expression for the first and second reflection signals of the mixed RIS is:

[0155]

[0156]

[0157] Therefore, the power consumption of hybrid RIS is:

[0158]

[0159] Among them, ||·||2 represents the L2 norm. Therefore, the total power consumption of the system is:

[0160] P=P BS +P RIS (13)

[0161] (5) Raise questions:

[0162] In the above system, the BS precoder W and the hybrid RIS reflection matrix Y are jointly designed to maximize the SE of the receiving vehicle, i.e., R, under the condition of limited power consumption of the BS and hybrid RIS, and to ensure the minimum threshold ζ of the radar signal-to-interference-noise ratio to ensure target detection performance. The optimization problem can then be written as

[0163]

[0164] This optimization problem is a non-convex optimization problem. Step 2: Represent the non-convex optimization problem equivalently and divide it into four sub-problems for solution.

[0165] Considering that the optimization problem proposed in the present invention is a multivariable non-convex optimization problem, and the optimization objectives and constraints are very complex, the proposed optimization problem is equivalently transformed and the problem is decomposed into four sub-problems for solution.

[0166] The SE maximization problem is equivalent to the Minimum Mean Square Error (MMSE) problem.

[0167] First, the IR vehicle multiplies the equalizer matrix to recover the communication symbols The mean square error (MSE) of communication symbol estimation is then expressed as

[0168]

[0169] After substituting (2) into (15), the MSE at the receiving vehicle can be further expressed. By introducing a positive auxiliary variable V, the objective function can be equivalently expressed by MSE:

[0170]

[0171] Among them, det(·) represents the determinant operation of the square matrix, tr(·) represents the trace operation of the matrix, and the optimal solution is V ＊ =[E1(C, Y, W)] -1 ; At a given V ＊ In this case, the optimal solution C ＊ Obtain

[0172]

[0173] With the solved equalizer, the minimum mean square error (MMSE) can be further expressed as:

[0174]

[0175] Therefore, combining formulas (10), (11) and (18), the optimization problem (14) can be transformed into:

[0176]

[0177] That is, under the four constraints of C1, C2, C3, and C4, the objective function affected by the four variables C, W, Y, and V is maximized.

[0178] Step 3: Use the AO algorithm to solve the four sub-problems

[0179] (5) Optimize the auxiliary variable V while fixing the equalizer matrix C, precoding matrix W, and RIS reflection matrix Y:

[0180] Optimize V with fixed C, W, and Y; the optimization problem is simplified to an unconstrained optimization problem, and the optimal solution V is obtained directly by derivation. *for:

[0181] V ＊ =[E1(C, Y, W)] -1 (20);

[0182] (6) Optimize the equalizer matrix C with fixed auxiliary variables V, precoding matrix W, and RIS reflection matrix Y:

[0183] Optimize C with fixed V, W and Y;

[0184] The optimization problem can be expressed as:

[0185]

[0186] in, The optimal solution is shown in formula (17);

[0187] (7) Optimize the precoding matrix W while fixing the equalizer matrix C, auxiliary variables V, and RIS reflection matrix Y:

[0188] Optimize W with C, V, and Y fixed; substitute formula (18) into the objective function and ignore irrelevant terms. The objective function can be reformulated as:

[0189]

[0190] in,

[0191] Constraints C1 and C3 are non-convex; to handle C1, C1 can be written as:

[0192]

[0193] Performing a first-order Taylor expansion on the second term in (23), the relaxed constraint C1 can be transformed into:

[0194]

[0195] Similarly, constraint C3 can also be transformed into a quadratic constraint:

[0196]

[0197] in:

[0198] J=G H Υ H A H ΥΥ H AΥG,

[0199] P=G H Υ H ΥG

[0200]

[0201] Therefore, the optimization problem can be transformed into:

[0202]

[0203] This problem is a convex QCQP problem and can be solved directly by the interior point method;

[0204] In addition, problem (26) can also be solved using the SDR algorithm; by introducing auxiliary variables, the problem is transformed into SDR form and can be solved directly;

[0205] (8) Optimize the RIS reflection matrix Y with fixed equalizer matrix C, auxiliary variables V, and precoding matrix W:

[0206] Ignoring irrelevant terms, the optimization problem can be simplified to:

[0207]

[0208] Where Z = GWVC H H r , P2=GWW H G H ;

[0209] Using the properties of the trace, the terms of the objective function are transformed into:

[0210]

[0211] in Z=GWVC H H r , z=[[Z] 1，1 ,…,[Z] N，N ] T And Ξ1=(P1)⊙(P2) T ,Ξ2=ησ 2 (P1)⊙I N ; where ⊙ represents the Hadamard product of the matrix.

[0212] The optimization problem is equivalent to:

[0213]

[0214] st C1, C3, C4 (33);

[0215] Where Ξ=Ξ1+Ξ2; since C1 and C3 are non-convex, the optimization problem is still non-convex;

[0216] To change the constraint into a convex constraint, first, we process the constraint C3: Substitute (10) and (11) into equation (13). The left side of the C3 inequality can be expressed as follows:

[0217]

[0218] In order to solve the constraint C3, the trace of the matrix and the properties of vectorization can be used to transform the terms on the left side of the constraint C3, and we can get:

[0219]

[0220] in:

[0221]

[0222] Among them, vec(·) represents the matrix vectorization operation, represents the Kronecker product; is a Hermitian matrix; using the first-order Taylor expansion and imaginary-real transformation, the equivalent expression is:

[0223]

[0224] Similarly, constraint C1 can be transformed into formula (24), focusing on simplifying the final term. By transforming each term in (24), the equivalent expression is finally obtained:

[0225]

[0226] in:

[0227]

[0228] in is a Hermitian matrix; through first-order Taylor expansion and complex transformation, the optimization problem can be transformed into:

[0229]

[0230] The problem can be transformed into SDR form for solution, and finally v can be obtained using rank-one decomposition.

[0231] Based on this, the spectrum efficiency optimization problem of the hybrid RIS-assisted synaesthesia integrated system in V2V was solved. The alternating optimization algorithm was used to solve the four sub-problems, providing a mathematically feasible basis and specific implementation method for the system, indicating that in the V2V network, the hybrid RIS-assisted synaesthesia integrated system can effectively improve the communication efficiency.

[0232] To test the effectiveness of the present invention, we conducted a simulation experiment. The system assumes that the number of BS antennas is M = 8 and the number of target vehicle antennas is L = 8. A single target vehicle is detected with the aid of hybrid RIS, where the number of vehicles is K = 4 and the number of hybrid RIS antennas is N = 64. The number of active RIS elements N a is fixed. The total power budget P = 10 mW, including the power of the BS and the power of the hybrid RIS active components. The noise power of the IR vehicle and the target vehicle is Assume that the signal-to-noise ratio threshold of all target vehicles is the same, that is,

[0233] Regarding the convergence performance of the algorithm, Figure 3 The simulation results show the convergence of the proposed AO algorithm in different scenarios. We consider the number of active elements N a =0 and N a =1, the results are as follows Figure 3 As shown in Figure 2, it can be observed that the proposed algorithm converges to a stable point quickly. In addition, MM converges to a stable point faster than SDR.

[0234] Attachment Figure 4 The simulation results of Analyze the relationship between SE and the number of RIS components: We fix the total number of hybrid RIS components, but change the ratio of active components to total components, thereby analyzing the relationship between SE and the proportion of active components. Figure 4 As a result, SE increases with the increase of the proportion of active elements, which shows the effectiveness of the proposed algorithm. This is because as the number of active elements increases, more elements can amplify the signal, thereby improving the communication performance.

[0235] Attachment Figure 5 The relationship between SE and residual self-interference power is shown in Figure 2. Based on this, we analyze the relationship between SE and RIS thermal noise. The active components in the hybrid RIS may generate thermal noise, and the noise level may vary due to different manufacturing processes. Figure 5 In

[15] , we observed that SE decreases as the noise level of active components increases. This is because the interference from the hybrid RIS increases, resulting in a decrease in the vehicle's SINR. Furthermore, the impact of thermal noise becomes increasingly severe as the number of active components in the hybrid RIS increases.

[0236] Finally, Figure 6 The relationship between SE and the target vehicle’s SINR radar perception threshold is shown, with SE decreasing significantly as the perception SINR threshold increases. This illustrates the trade-off between the required target vehicle detection power and SE.

[0237] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hybrid metasurface-assisted interawareness integration algorithm for vehicle-to-vehicle communication, in a V2V network, including a base station, an information receiving IR vehicle with L antennas, and K target vehicles; characterized by: The BS is equipped with a uniform linear array consisting of M antennas; it also includes a RIS with N antennas, which contains Na active elements and N-Na passive elements. The position index set of the active elements is in The V2V network includes a signal transmission model, a communication model, a radar model, and a power consumption model. It uses a hybrid RIS to assist signal propagation and uses an alternating optimization algorithm to solve the optimization problem, thereby optimizing the system's spectrum efficiency and improving communication efficiency.

2. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 1, characterized in that: The signal transmission model is: Definitions r is the detection signal used to perceive the target vehicle, s c To send communication symbols to IR vehicles; use communication precoder w c and perceptual precoder W r =[w r，1 ,...,w r，M ] respectively for s c and s r Perform precoding processing and then sum up; Therefore, the transmitted signal at the BS Expressed as: x=W r s r +w c s c =Ws (1), Among them, define N s =M+1, indicating the total number of communication symbols; W=[W r , w c ] is defined as an M×N s The precoder matrix of size , Defined as N S elements of the transmission symbol vector; [·] T represents the transpose operation of the vector, and represents the column vector; it is assumed that these signals have unit power and have statistical independence characteristics, that is, the covariance of the complex baseband signal satisfies Indicates the expected operation, is an identity matrix.

3. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 1 or 2, characterized in that: The communication model is: After the signal is transmitted from the BS, it reaches the IR vehicle through the RIS reflection channel; Let G be the channel between BS and hybrid RIS, H r is the channel between the hybrid RIS and IR vehicles; the hybrid RIS coefficient matrix is recorded as p n Indicates the amplification factor of RIS to the signal, θ n represents the amplification and phase shift coefficient; when p n =1, otherwise p n >1; therefore, the signal received by the receiving vehicle is expressed as: y I =H r γGx+H r γz1+n (2), Among them, the thermal noise of the hybrid RIS active component is Indicates that the noise obeys an independent complex Gaussian distribution with a covariance of 1. Unless otherwise specified, the thermal noise mentioned later should also obey this distribution; diag(·) indicates the construction of a diagonal matrix operation; represents the additive white Gaussian noise (AWGN) of the IR vehicle receiver; Accordingly, the spectral efficiency of the receiving vehicle is: Among them, H I =H r ΥG represents the equivalent channel from BS to IR vehicle, (·) H Represents the conjugate transpose operation of a matrix.

4. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 3, characterized in that: The radar model is: In addition to serving IR vehicles, the transmitted signal will also illuminate target vehicles after passing through the hybrid RIS; set up is the channel between the hybrid RIS and the kth target vehicle; the entire target detection channel can be expressed as Due to the presence of obstacles, the direct link is not considered. The echo signal received by the BS includes the signal of the BS-RIS-BS link and the signal transmitted by the BS-RIS-target-RIS-BS link, as well as the thermal noise of the BS receiver and the hybrid RIS active components. Accordingly, the target echo signal is expressed as: in: Both z1 and z2 represent the thermal noise introduced from the RIS active components; in addition, z rc is the AWGN at the BS receiver. Since the power is usually very small, this noise will be ignored in the following. The channel between the BS and the RIS and the reflection matrix of the RIS are known. Therefore, the signal H directly reflected by the RIS is ie x can be processed using self-interference cancellation technology; thus, the target echo signal is updated as: The performance of radar sensors is evaluated using the Signal-to-Interference-plus-Noise Ratio (SINR) of the received signal. Among them, tr(·) represents the matrix trace operation, represents the real part operation, where Since SINR directly affects the detection probability and parameter estimation accuracy, it is used as an indicator for target sensing.

5. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 4, characterized in that: The power consumption model is: For the BS, the consumed transmit power is expressed as: P BS =tr(WW H ) (9), The power consumption of the RIS is due to the amplification of two reflections. The first reflection of the RIS reflects the signal transmitted by the BS to the target vehicle and the IR vehicle, and the second reflection reflects the echo signal of the target vehicle back to the BS. The expression of the first and second reflection signals of the mixed RIS is: Therefore, the power consumption of hybrid RIS is: Here, ||·||2 represents the L2 norm; thus, the total power consumption of the system is: P=P BS +P RIS (13)。 6. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 5, characterized in that: By jointly designing the BS precoder W and the hybrid RIS reflection matrix Y, we maximize the SE of the receiving vehicle, i.e., R, under the condition that the power consumption of the BS and the hybrid RIS is limited, and the radar signal-to-interference-noise ratio (ζ) is kept to a minimum threshold to ensure target detection performance. The optimization problem is then written as: This optimization problem is a non-convex optimization problem.

7. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 6, characterized in that: The non-convex optimization problem is expressed equivalently and divided into four sub-problems for solution; First, the IR vehicle multiplies the equalizer matrix to recover the communication symbols Then the mean square error of communication symbol estimation is expressed as: After substituting (2) into (15), the MSE at the receiving vehicle can be further expressed. By introducing a positive auxiliary variable V, the objective function can be equivalently expressed by MSE: Among them, det(·) represents the determinant operation of the square matrix, tr(·) represents the trace operation of the matrix, and the optimal solution is V ＊ =[E1(C,γ,W)] -1 ; At a given V ＊ In this case, the optimal solution C ＊ Obtain With the solved equalizer, the minimum mean square error (MMSE) can be further expressed as: Therefore, combining formulas (10), (11) and (18), the optimization problem (14) can be transformed into: That is, under the four constraints of C1, C2, C3, and C4, the objective function affected by the four variables C, W, γ, and V is maximized.

8. The hybrid metasurface-assisted synaesthesia integrated system for vehicle-to-vehicle communication according to claim 7, characterized in that: Use the alternating optimization algorithm to solve the four subproblems: (1) Optimize the auxiliary variable V while fixing the equalizer matrix C, precoding matrix W, and RIS reflection matrix γ: Optimize V with fixed C, W, and Y; the optimization problem is simplified to an unconstrained optimization problem, and the optimal solution V is obtained directly by derivation ＊ for: V ＊ =[E1(C,Υ,W)] -1 (20); (2) Optimize the equalizer matrix C with fixed auxiliary variables V, precoding matrix W, and RIS reflection matrix γ: Optimize C with fixed V, W and γ; The optimization problem can be expressed as: in, The optimal solution is shown in formula (17); (3) Optimize the precoding matrix W with fixed equalizer matrix C, auxiliary variable V and RIS reflection matrix γ: Optimize W with C, V, and γ fixed; substitute formula (18) into the objective function and ignore irrelevant terms. The objective function can be reformulated as: in, Constraints C1 and C3 are non-convex; to handle C1, C1 can be written as: Performing a first-order Taylor expansion on the second term in (23), the relaxed constraint C1 can be transformed into: Similarly, constraint C3 can also be transformed into a quadratic constraint: in: J=G H c H A H gg H AγG, PG H Y H YG Therefore, the optimization problem can be transformed into: This problem is a convex QCQP problem and can be solved directly by the interior point method; In addition, problem (26) can also be solved using the SDR algorithm; by introducing auxiliary variables, the problem is transformed into SDR form and can be solved directly; (4) Optimize the RIS reflection matrix γ when the equalizer matrix C, auxiliary variables V and precoding matrix W are fixed: Ignoring irrelevant terms, the optimization problem can be simplified to: stC1, C3, C4 (27); Where Z = GWVC H H r , P2=GWW H G H ; Using the properties of the trace, the terms of the objective function are transformed into: in Z=GWVC H H r , z=[[Z] 1，1 ,…,[Z] N，N ] T And Ξ1=(P1)⊙(P2) T ,Ξ2=ησ 2 (P1)⊙I N ;in, ⊙ represents the Hadamard product of matrices; The optimization problem is equivalent to: stC1, C3, C4 (33); Where Ξ=Ξ1+Ξ2; since C1 and C3 are non-convex, the optimization problem is still non-convex; To change the constraint into a convex constraint, first, we process the constraint C3: Substitute (10) and (11) into equation (13). The left side of the C3 inequality can be expressed as follows: In order to solve the constraint C3, the trace of the matrix and the properties of vectorization can be used to transform the terms on the left side of the constraint C3, and we can get: in: Among them, vec(·) represents the matrix vectorization operation, represents the Kronecker product; is a Hermitian matrix; using the first-order Taylor expansion and imaginary-real transformation, the equivalent expression is: Similarly, constraint C1 can be transformed into formula (24), focusing on simplifying the final term. By transforming each term in (24), the equivalent expression is finally obtained: in: in is a Hermitian matrix; through first-order Taylor expansion and complex transformation, the optimization problem can be transformed into: The problem can be transformed into SDR form for solution, and finally v can be obtained using rank-one decomposition.

Images