A Smart Metasurface-Assisted Method for Eliminating Zero-Forced Interference
Patent Information
- Application Number
- CN202410035366.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-09
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2044-01-09
Smart Images

Figure CN117879664B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of communications and relates to a wireless communication system for multiple-user multiple-input multiple-output (MIMO) assisted by a reconfigurable intelligent surface (RIS). Background Technology
[0002] In the field of wireless communication, MIMO (Multi-Input Multiple-Output) technology has been widely used. It can significantly improve the link performance and channel capacity of wireless communication systems while maintaining the existing transmit power and channel bandwidth. However, MIMO systems have some problems in both single-user and multi-user applications. In single-user applications, the number of antennas in a MIMO system is limited, thus limiting the system capacity. While in multi-user applications, although MIMO systems allow multiple users to transmit data simultaneously, thus achieving greater system capacity, they also face more complex challenges, such as interference management, hardware and computational complexity, acquisition of channel state information, and user scheduling.
[0003] With the development of 6G networks, new communication technologies and network architectures have been proposed, such as ultra-large-scale MIMO, ultra-dense heterogeneous networks, and cell-free architectures. While these offer higher data transmission rates and larger system capacity, they also bring more challenges, such as high cost, high energy consumption, complex heterogeneous networks, and severe co-channel interference. To address these issues, researchers have proposed using intelligent metasurfaces (RIS). RIS is an artificial electromagnetic surface structure with real-time programmable electromagnetic properties, composed of a large number of electromagnetic units. It possesses unique characteristics such as low cost, low energy consumption, programmability, and ease of deployment, which can alleviate the aforementioned problems and significantly improve communication performance. Due to the randomness of the wireless transmission environment, most existing wireless transmission technologies are applied at the receiver and transmitter levels. RIS can reconstruct the wireless transmission environment by controlling the angle and amplitude of the incident electromagnetic wave to form a three-dimensional passive beam that meets differentiated communication needs.
[0004] Existing research shows that RIS-assisted multi-user MIMO communication systems offer significant performance improvements. However, as the number of transmitters and receivers increases, the interference experienced by the receiver also increases, leading to a reduction in channel capacity. Traditional interference management schemes are mostly unable to achieve maximum channel capacity and degrees of freedom in scenarios with a large number of users.
[0005] The invention patent application with application number 202210212447.X provides a collaborative design method for active and passive beamforming in non-orthogonal multiple access secure transmission based on a smart reflector. It utilizes the characteristics of smart reflectors to reconstruct wireless channels, suppresses eavesdropping by internal eavesdroppers on secure user information, and ensures maximum security rate. It belongs to the security field of non-orthogonal multiple access technology.
[0006] Patent application number 202211724416.9 proposes a STAR-RIS-based method for optimizing the physical layer security of NOMA networks. The aim is to enhance the security of downlink MISONOMA networks under conditions of imperfect CSI (Content Security Indication) by eavesdroppers, primarily focusing on physical layer security optimization methods for NOMA networks. This method is also applicable to the field of wireless network security.
[0007] With the emergence of new communication technologies and network architectures, eliminating interference to communication systems and maximizing system channel capacity and speed while providing higher data transmission rates and greater system capacity has become a major research topic. Summary of the Invention
[0008] This invention addresses the shortcomings of existing technologies by proposing a zero-forcing interference cancellation method based on passive beamforming and a combination of active and passive beamforming. Building upon a RIS-assisted multi-user MIMO communication system, it utilizes a combination of active beamforming at the transmitter and passive beamforming provided by the RIS to eliminate interference, thereby maximizing system performance and data rate.
[0009] This invention focuses on a RIS-assisted multi-user MIMO system, employing a multi-antenna transmitter, a multi-antenna receiver, and the RIS to construct two specific system models: one with a line-of-sight path and the other without. Considering inter-user interference as an interference channel model, this invention combines active beamforming at the transmitter and passive beamforming design of the RIS to provide corresponding interference cancellation schemes. The goal of this invention is complete interference elimination, maximizing system performance and data rate. Due to the reciprocity of uplink and downlink communication systems, only the downlink transmission case is considered here, but this invention is not limited to downlink transmission networks.
[0010] Invention technical solution:
[0011] A smart metasurface-assisted zero-forcing interference cancellation method is proposed for RIS-assisted multi-user MIMO systems. The method comprises a multi-antenna transmitter, a multi-antenna receiver, and a RIS, forming two systems: one with a line-of-sight path and the other without. Inter-user interference is modeled using an interference channel model. An interference cancellation scheme is designed by combining active beamforming at the transmitter and passive beamforming at the RIS. The interference cancellation process includes the following steps:
[0012] S1. For a multi-user MIMO system with a line-of-sight path, construct a system model with a line-of-sight path.
[0013] S2, when there are enough RIS reflective units, an interference cancellation scheme is designed based on passive beamforming; a reflection coefficient matrix is designed at the smart metasurface, and precise beamforming is performed using sufficient reflective units to achieve complete interference cancellation.
[0014] S3. When the number of RIS reflection units is insufficient, an interference cancellation scheme is designed based on joint active and passive beamforming. A reflection coefficient matrix is designed at the smart metasurface. The rank of the interference channel matrix is reduced by passive beamforming, and then the active beamforming at the transmitter is used to force zero interference, so as to achieve the purpose of complete interference cancellation.
[0015] In the aforementioned intelligent metasurface-assisted zero-forcing interference elimination method, step S1 involves the following process for constructing a system model with a direct-looking path:
[0016] In a RIS-assisted multi-user MIMO system, there are K pairs of transmitters and receivers; where each transmitter has M antennas, each receiver has N antennas, and the number of RIS reflection units is Q.
[0017] In the communication system model described above, there are two communication methods: Method 1: The transmitted signal reaches the receiver through a direct line-of-sight link between the transmitter and the receiver; Method 2: The transmitted signal reaches the receiver through a multi-stage link relayed by a passive RIS.
[0018] Assumption Let represent the direct channel from the l-th transmitter to the k-th receiver. This represents the channel from RIS to the k-th receiver. Represents the channel from the l-th transmitter to the RIS; RIS reflection matrix in Represents the reflection coefficient vector, θ q ∈[0,2π] and ξ q ≥0 (q∈Q) represent the phase and amplitude of the q-th RIS reflector unit, respectively; and Let represent the signal transmitted by the k-th transmitter and the Gaussian white noise at the k-th receiver, respectively.
[0019] Given known channel state information, assuming the interference between users follows an interference channel model, if the k-th transmitter sends a signal to its corresponding receiver, this signal will interfere with the other k-1 receivers. The signal received by the k-th receiver in this case is as follows:
[0020]
[0021] in The matrix represents the desired channel. Indicates an interfering channel;
[0022] For multi-user interference networks, the network capacity is approximated by the degrees of freedom; these degrees of freedom can be equal to the number of user-transmitted data streams in parallel. Under a sufficiently high signal-to-noise ratio ρ, the degrees of freedom d... k It is the channel capacity R k The first-order approximation of is defined as:
[0023]
[0024] Among them, R k Let ρ represent the sum rate of the k-th link, and ρ represent the signal-to-noise ratio (SNR).
[0025] The specific formula for calculating the sum rate of the k-th link is as follows:
[0026]
[0027] in, P represents the transmission power, σ 2 U represents noise power. k This indicates interference leakage, specifically in the following forms:
[0028]
[0029] The intelligent metasurface-assisted zero-forcing interference cancellation method, in step S2, designs the following interference cancellation scheme based on passive beamforming:
[0030] First, based on the system model constructed in step S1, the passive beamforming matrix Θ is designed at RIS, and the specific interference cancellation conditions are as follows:
[0031]
[0032] To meet the interference cancellation requirements, the number of RIS reflection units must satisfy the following condition:
[0033] Q≥K(K-1)MN (6)
[0034] The interference channel matrix in formula (5) is vectorized, and its specific form is as follows:
[0035]
[0036] in, It is a matrix The 1st th (Q+2) th ,…,(Q 2 )th Composed of column vectors;
[0037] Solving the non-homogeneous linear system of equations (7) yields the reflection coefficient vector θ. Based on the system model parameters, the reflection matrix, through vectorization and data compression, can be transformed into the form vec(Θ) as follows:
[0038]
[0039] because The reflection matrix Θ is then obtained by performing an inverse vectorization operation on θ.
[0040] The intelligent metasurface-assisted zero-forcing interference cancellation method, step S3, the interference cancellation process based on joint active and passive beamforming design is as follows:
[0041] If the number of reflection units Q cannot meet the conditions of formula (6), then it is necessary to first design the reflection coefficient matrix to reduce the rank of the interference channel matrix, and then use the active beamforming matrix to perform zero-forcing interference elimination.
[0042] The active beamforming matrix design at the transmitting end is as follows:
[0043] s l =V l x l (9)
[0044] in, This represents the desired signal sent from the l-th transmitter to the l-th receiver. Active beamforming matrix at the l-th transmitter;
[0045] At this point, the conditions for complete elimination of interference are as follows:
[0046]
[0047] When Q < K(K-1)MN, the number of reflection units Q needs to satisfy the following condition:
[0048] Q≥K(K-1)max(M,N)(min(M,N)-p) (11)
[0049] Where (min(M,N)-p) represents the rank after the interference channel is reduced;
[0050] Next, in order to reduce the rank of the interference channel matrix, matrix vectorization is required. The rank of a matrix is equal to the minimum of its row rank and column rank. Therefore, the different relationships between the number of transmitting and receiving antennas will affect the rank of the interference channel matrix. Thus, the solution of the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0051] a) When N≥M: The rank of the interference channel matrix depends on N, so it is vectorized into a column vector, as follows:
[0052]
[0053] in, It is H kl The m-th column vector, yes The m-th column vector;
[0054] After vectorizing the interference channel matrix, the column vector group of the interference channel matrix can be obtained. To reduce the rank of the interference channel matrix, we need to discuss the largest linearly independent set of the vector group; let p be the number of the largest linearly independent vectors in the column vector group of the interference channel, then the column vector group... They are linearly independent; therefore, the rank of the interference channel matrix can be obtained as:
[0055]
[0056] This indicates that the rank of the interference channel matrix needs to be reduced to p; let t = p+1, p+2, ..., M, then It can be derived from a vector group Linear representation, meaning that the two are linearly dependent; the derivation process is as follows:
[0057]
[0058] To simplify the above results, let and The above derivation result can then be rewritten as:
[0059]
[0060] in, and
[0061] Finally, the non-homogeneous linear equation system (15) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0062] b) When N < M: The rank of the interference channel matrix depends on M, so it is vectorized into rows, as follows:
[0063]
[0064] in, It is by The 1st th (Q+2) th , ..., (Q 2 ) thRow vectors are composed of, H represents kl The nth row vector, express The nth row vector;
[0065] Let p be the maximum number of linearly independent vectors in the row vector group of the interference channel, then the row vector group... They are linearly independent; the rank of the interference channel matrix is:
[0066]
[0067] When t = p+1, p+2, ..., M, the specific form of the method for constructing a non-homogeneous linear system of equations is as follows:
[0068]
[0069] in, and and
[0070] Finally, the non-homogeneous linear equation system (18) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0071] The reflection coefficient matrix Θ obtained from cases a) and b) can only reduce the rank of the interfering channel; an active beamforming matrix V still needs to be designed. k Zero-forcing interference cancellation is performed; at the k-th transmitter, the active beamforming matrix V k The specific form is as follows:
[0072]
[0073] in, This represents all interference channels generated by the k-th transmitter.
[0074] Finally, the system of quasi-equations (19) is solved to obtain the active beamforming matrix V. k .
[0075] The aforementioned intelligent metasurface-assisted zero-forcing interference cancellation method constructs a system model without a direct line-of-sight path for reflection matrices where obstacles exist in the line-of-sight path between the receiver and transmitter. Since the signal can only be transmitted through a multi-stage link relayed by RIS, the received signal at the k-th receiver is:
[0076] The process of achieving interference cancellation based on passive beamforming design is as follows:
[0077] In communication links excluding the line-of-sight path, the condition for complete interference elimination becomes:
[0078]
[0079] To ensure the solvability of the homogeneous linear equations, the constraint on the number of RIS reflection units Q should be modified as follows:
[0080] Q≥K(K-1)MN+1 (22)
[0081] Finally, the reflection coefficient vector θ is obtained by solving the homogeneous linear equation system (21), and the reflection matrix Θ is obtained by matrix inverse vectorization.
[0082] Interference cancellation based on joint active and passive beamforming design:
[0083] When Q≤K(K-1)MN, first reduce the rank of the interference channel matrix, then perform zero-forcing interference through active beamforming; due to the relationship between the rank of the channel matrix and the number of transmitting and receiving antennas, the solution for the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0084] a) When N≥M, by using passive beamforming design to reduce the rank of the interference channel matrix, the following homogeneous linear equations can be obtained:
[0085]
[0086] b) When N < M, the specific form of the homogeneous linear equation is as follows:
[0087]
[0088] According to the solution conditions of the homogeneous linear equation system (24) and (25), the RIS reflection unit Q needs to satisfy the following conditions:
[0089] Q≥K(K-1)max(M,N)(min(M,N)-p)+1 (26)
[0090] By solving the homogeneous linear equations (24) and (25), the reflection coefficient vector θ can be calculated, and the reflection matrix Θ can be obtained by inverse vectorization.
[0091] Next, active beamforming design is performed to eliminate zero-forcing interference in the reduced-rank interference channel matrix, as follows:
[0092]
[0093] Finally, the system of sublinear equations (27) is solved to obtain the active beamforming matrix V. k .
[0094] Compared with the prior art, the positive effects of the present invention are as follows:
[0095] 1. This invention discloses a smart metasurface-assisted zero-forcing interference cancellation method. It constructs a model for a RIS-assisted multi-user MIMO system, employing an interference channel model to address inter-user interference. The goal is to maximize system performance and data rate when interference is completely eliminated. When the number of RIS reflection units is sufficient, passive beamforming is used to design the interference cancellation scheme. Simultaneously, when the number of RIS reflection units is insufficient, a method is adopted to reduce the rank of the interference channel matrix to design the reflection coefficient matrix, and both active and passive beamforming are combined to design the interference cancellation scheme. In both cases, this invention can maximize system performance and data rate while completely eliminating interference.
[0096] 2. The intelligent metasurface-assisted zero-forcing interference elimination method of this invention, compared with existing iterative interference elimination schemes, solves the reflection coefficient matrix of RIS using a closed-loop algorithm, resulting in lower algorithm complexity. Furthermore, compared with traditional interference management schemes and interference alignment schemes such as min-IL and Max-SINR, the system achievable speed of this invention is significantly improved.
[0097] 3. An intelligent metasurface-assisted method for eliminating zero-force interference was invented, which extends the application to situations where there are obstacles at the line of sight and verifies that its application scenarios are generalizable. Attached Figure Description
[0098] Figure 1 The figure shows a RIS-assisted K-user MIMO system model;
[0099] Figure 2 A K-user MIMO system model assisted by RIS (obstacles exist in the line-of-sight path);
[0100] Figure 3 A comparison chart of the system and speed of different algorithms;
[0101] Figure 4 A comparison of systems and rates with different numbers of RIS reflector units;
[0102] Figure 5 A comparison chart of system configurations and speeds for different communication networks;
[0103] Figure 6 A comparison chart of the system and speed of different algorithms (obstacles exist in the direct path);
[0104] Figure 7 The diagram shows the flowchart of the intelligent metasurface-assisted zero-force interference elimination method of the present invention. Detailed Implementation
[0105] To make the technical concept and advantages of the invention clearer, the technical solution of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the following embodiments are only used to explain and illustrate preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention.
[0106] Example 1
[0107] See Figure 7 This invention discloses a smart metasurface-assisted zero-forcing interference cancellation method for RIS-assisted multi-user MIMO systems. It utilizes a multi-antenna transmitter, a multi-antenna receiver, and the RIS to construct two systems: one with a line-of-sight path and the other without. Inter-user interference is addressed using an interference channel model. The method combines active beamforming at the transmitter with passive beamforming at the RIS to design an interference cancellation scheme. The interference cancellation process includes the following steps:
[0108] S1. For a multi-user MIMO system with a line-of-sight path, construct a system model with a line-of-sight path.
[0109] S2, when there are enough RIS reflective units, an interference cancellation scheme is designed based on passive beamforming; a reflection coefficient matrix is designed at the smart metasurface, and precise beamforming is performed using sufficient reflective units to achieve complete interference cancellation.
[0110] S3. When the number of RIS reflection units is insufficient, an interference cancellation scheme is designed based on joint active and passive beamforming. A reflection coefficient matrix is designed at the smart metasurface. The rank of the interference channel matrix is reduced by passive beamforming, and then the active beamforming at the transmitter is used to force zero interference, so as to achieve the purpose of complete interference cancellation.
[0111] This invention introduces a zero-forcing interference scheme for the first time in a RIS-assisted multi-user MIMO system. Whether the number of RIS reflection units is sufficient (based on passive beamforming to design the interference cancellation scheme) or the number of RIS reflection units is insufficient (by reducing the rank of the interference channel matrix to design the reflection coefficient matrix), this invention can maximize system performance and speed while completely eliminating interference. Moreover, compared with existing iterative interference cancellation schemes, the algorithm complexity of this invention is lower.
[0112] Example 2
[0113] The difference between the intelligent metasurface-assisted zero-forcing interference elimination method in this embodiment and Embodiment 1 is that the system model construction process with a direct-view path in step S1 is as follows:
[0114] In a RIS-assisted multi-user MIMO system, there are K pairs of transmitters and receivers; where each transmitter has M antennas, each receiver has N antennas, and the number of RIS reflection units is Q.
[0115] In the communication system model described above, there are two communication methods: Method 1: The transmitted signal reaches the receiver through a direct line-of-sight link between the transmitter and the receiver; Method 2: The transmitted signal reaches the receiver through a multi-stage link relayed by a passive RIS.
[0116] Assumption Let represent the direct channel from the l-th transmitter to the k-th receiver. This represents the channel from RIS to the k-th receiver. Represents the channel from the l-th transmitter to the RIS; RIS reflection matrix in Represents the reflection coefficient vector, θ q ∈[0,2π] and ξ q ≥0 (q∈Q) represent the phase and amplitude of the q-th RIS reflector unit, respectively; and Let represent the signal transmitted by the k-th transmitter and the Gaussian white noise at the k-th receiver, respectively.
[0117] Given known channel state information, assuming the interference between users follows an interference channel model, if the k-th transmitter sends a signal to its corresponding receiver, this signal will interfere with the other k-1 receivers. The signal received by the k-th receiver in this case is as follows:
[0118]
[0119] in The matrix represents the desired channel. Indicates an interfering channel;
[0120] For multi-user interference networks, the network capacity is approximated by the degrees of freedom; these degrees of freedom can be equal to the number of user-transmitted data streams in parallel. Under a sufficiently high signal-to-noise ratio ρ, the degrees of freedom d... k It is the channel capacity R k The first-order approximation of is defined as:
[0121]
[0122] Among them, R k Let ρ represent the sum rate of the k-th link, and ρ represent the signal-to-noise ratio (SNR).
[0123] The specific formula for calculating the sum rate of the k-th link is as follows:
[0124]
[0125] in, P represents the transmission power, σ 2 U represents noise power. k This indicates interference leakage, specifically in the following forms:
[0126]
[0127] Example 3
[0128] The intelligent metasurface-assisted zero-forcing interference cancellation method described in this embodiment differs from that in Embodiment 2 in that: step S2 involves designing the following interference cancellation scheme based on passive beamforming:
[0129] First, based on the system model constructed in step S1, the passive beamforming matrix Θ is designed at RIS, and the specific interference cancellation conditions are as follows:
[0130]
[0131] To meet the interference cancellation requirements, the number of RIS reflection units must satisfy the following condition:
[0132] Q≥K(K-1)MN (6)
[0133] The interference channel matrix in formula (5) is vectorized, and its specific form is as follows:
[0134]
[0135] in, It is a matrix The 1st th (Q+2) th ,…,(Q 2 ) th Composed of column vectors;
[0136] Solving the non-homogeneous linear system of equations (7) yields the reflection coefficient vector θ. Based on the system model parameters, the reflection matrix, through vectorization and data compression, can be transformed into the form vec(Θ) as follows:
[0137]
[0138] because The reflection matrix Θ is then obtained by performing an inverse vectorization operation on θ.
[0139] Example 4
[0140] The difference between the intelligent metasurface-assisted zero-forcing interference elimination method described in this embodiment and Embodiment 3 is that: Step S3, the interference elimination process based on the joint active and passive beamforming design, is as follows:
[0141] If the number of reflection units Q cannot meet the conditions of formula (6), then it is necessary to first design the reflection coefficient matrix to reduce the rank of the interference channel matrix, and then use the active beamforming matrix to perform zero-forcing interference elimination.
[0142] The active beamforming matrix design at the transmitting end is as follows:
[0143] s l =V l x l (9)
[0144] in, This represents the desired signal sent from the l-th transmitter to the l-th receiver. Active beamforming matrix at the l-th transmitter;
[0145] At this point, the conditions for complete elimination of interference are as follows:
[0146]
[0147] When Q < K(K-1)MN, the number of reflection units Q needs to satisfy the following condition:
[0148] Q≥K(K-1)max(M,N)(min(M,N)-p) (11)
[0149] Where (min(M,N)-p) represents the rank after the interference channel is reduced;
[0150] Next, in order to reduce the rank of the interference channel matrix, matrix vectorization is required. The rank of a matrix is equal to the minimum of its row rank and column rank. Therefore, the different relationships between the number of transmitting and receiving antennas will affect the rank of the interference channel matrix. Thus, the solution of the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0151] a) When N≥M: The rank of the interference channel matrix depends on N, so it is vectorized into a column vector, as follows:
[0152]
[0153] in, It is H kl The m-th column vector, yes The m-th column vector;
[0154] After vectorizing the interference channel matrix, the column vector group of the interference channel matrix can be obtained. To reduce the rank of the interference channel matrix, we need to discuss the largest linearly independent set of the vector group; let p be the number of the largest linearly independent vectors in the column vector group of the interference channel, then the column vector group... They are linearly independent; therefore, the rank of the interference channel matrix can be obtained as:
[0155]
[0156] This indicates that the rank of the interference channel matrix needs to be reduced to p; let t = p+1, p+2, ..., M, then It can be derived from a vector group Linear representation, meaning that the two are linearly dependent; the derivation process is as follows:
[0157]
[0158] To simplify the above results, let and The above derivation result can then be rewritten as:
[0159]
[0160] in, and
[0161] Finally, the non-homogeneous linear equation system (15) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0162] b) When N < M: The rank of the interference channel matrix depends on M, so it is vectorized into rows, as follows:
[0163]
[0164] in, It is by The 1st th (Q+2) th , ..., (Q 2 ) th Row vectors are composed of, H represents kl The nth row vector, express The nth row vector;
[0165] Let p be the maximum number of linearly independent vectors in the row vector group of the interference channel, then the row vector group... They are linearly independent; the rank of the interference channel matrix is:
[0166]
[0167] When t = p+1, p+2, ..., M, the specific form of the method for constructing a non-homogeneous linear system of equations is as follows:
[0168]
[0169] in, and and
[0170] Finally, the non-homogeneous linear equation system (18) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0171] The reflection coefficient matrix Θ obtained from cases a) and b) can only reduce the rank of the interfering channel; an active beamforming matrix V still needs to be designed. k Zero-forcing interference cancellation is performed; at the k-th transmitter, the active beamforming matrix V k The specific form is as follows:
[0172]
[0173] in, This represents all interference channels generated by the k-th transmitter.
[0174] Finally, the system of quasi-equations (19) is solved to obtain the active beamforming matrix V. k .
[0175] Compared to existing iterative interference cancellation schemes, this invention solves the reflection coefficient matrix of RIS using a closed-form algorithm, resulting in lower algorithm complexity. Furthermore, compared to traditional interference management schemes and interference alignment schemes such as min-IL and Max-SINR, the system achievable speed of this invention is significantly improved.
[0176] Example 5
[0177] The intelligent metasurface-assisted zero-forcing interference cancellation method described in this embodiment differs from the previous embodiments in that: for reflection matrices where there are obstacles in the line-of-sight path between the receiver and transmitter, a system model without a line-of-sight path is constructed. Since the signal can only be transmitted through a multi-stage link relayed by RIS, the received signal at the k-th receiver is:
[0178]
[0179] The process of achieving interference cancellation based on passive beamforming design is as follows:
[0180] In communication links excluding the line-of-sight path, the condition for complete interference elimination becomes:
[0181]
[0182] To ensure the solvability of the homogeneous linear equations, the constraint on the number of RIS reflection units Q should be modified as follows:
[0183] Q≥K(K-1)MN+1 (22)
[0184] Finally, the reflection coefficient vector θ is obtained by solving the homogeneous linear equation system (21), and the reflection matrix Θ is obtained by matrix inverse vectorization.
[0185] Interference cancellation based on joint active and passive beamforming design:
[0186] When Q≤K(K-1)MN, first reduce the rank of the interference channel matrix, then perform zero-forcing interference through active beamforming; due to the relationship between the rank of the channel matrix and the number of transmitting and receiving antennas, the solution for the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0187] a) When N≥M, by using passive beamforming design to reduce the rank of the interference channel matrix, the following homogeneous linear equations can be obtained:
[0188]
[0189] b) When N < M, the specific form of the homogeneous linear equation is as follows:
[0190]
[0191] According to the solution conditions of the homogeneous linear equation system (24) and (25), the RIS reflection unit Q needs to satisfy the following conditions:
[0192] Q≥K(K-1)max(M,N)(min(M,N)-p)+1 (26)
[0193] By solving the homogeneous linear equations (24) and (25), the reflection coefficient vector θ can be calculated, and the reflection matrix Θ can be obtained by inverse vectorization.
[0194] Next, active beamforming design is performed to eliminate zero-forcing interference in the reduced-rank interference channel matrix, as follows:
[0195]
[0196] Finally, the system of sublinear equations (27) is solved to obtain the active beamforming matrix V. k .
[0197] Example 6
[0198] This embodiment of the intelligent metasurface-assisted zero-forcing interference cancellation method targets a RIS-assisted multi-user MIMO system. It employs two system models: one with a line-of-sight path and the other without, comprised of a multi-antenna transmitter, a multi-antenna receiver, and the RIS. Inter-user interference is handled using an interference channel model. An interference cancellation scheme is designed by combining active beamforming at the transmitter and passive beamforming at the RIS. The specific implementation process is as follows:
[0199] 1) Model 1: Construction of a system model with a direct line of sight
[0200] In a RIS-assisted multi-user MIMO system, there are K pairs of transmitters and receivers. Each transmitter has M antennas, each receiver has N antennas, and the RIS has Q reflector elements. The communication system model has two communication modes: Mode 1: The transmitted signal reaches the receiver via a direct line-of-sight link between the transmitter and receiver; Mode 2: The transmitted signal reaches the receiver via a multi-stage link relayed by a passive RIS.
[0201] Assumption Let represent the direct channel from the l-th transmitter to the k-th receiver. This represents the channel from RIS to the k-th receiver. Represents the channel from the l-th transmitter to the RIS; RIS reflection matrix in Represents the reflection coefficient vector, θ q ∈[0,2π] and ξ q ≥0 (q∈Q) represent the phase and amplitude of the q-th RIS reflector unit, respectively; and Let represent the signal transmitted by the k-th transmitter and the Gaussian white noise at the k-th receiver, respectively.
[0202] Given known channel state information, assuming the interference between users follows an interference channel model, the signal transmitted by the k-th transmitter to its corresponding receiver will interfere with the other k-1 receivers. The signal received by the k-th receiver in this case is as follows:
[0203]
[0204] in The matrix represents the desired channel. Indicates an interference channel.
[0205] For multi-user interference networks, the network capacity can be approximated using degrees of freedom. The degrees of freedom can be equal to the number of users transmitting data streams in parallel. Under a sufficiently high signal-to-noise ratio ρ, the degrees of freedom d... k It is the channel capacity R kThe first-order approximation of is defined as:
[0206]
[0207] Among them, R k Let ρ represent the sum rate of the k-th link, and ρ represent the signal-to-noise ratio (SNR).
[0208] The specific formula for calculating the sum rate of the k-th link is as follows:
[0209]
[0210] in, P represents the transmission power, σ 2 U represents noise power. k This indicates interference leakage, specifically in the following forms:
[0211]
[0212] 2) Interference cancellation is achieved based on passive beamforming design.
[0213] First, based on the system model described in 1), a passive beamforming matrix Θ is designed at RIS, and the specific interference cancellation conditions are as follows:
[0214]
[0215] To meet the interference cancellation requirements, the number of RIS reflection units must satisfy the following condition:
[0216] Q≥K(K-1)MN (6)
[0217] The interference channel matrix in formula (5) is vectorized, and its specific form is as follows:
[0218]
[0219] in, It is a matrix The 1st th (Q+2) th ,…,(Q 2 ) th Composed of column vectors.
[0220] Solving the non-homogeneous linear system of equations (7) yields the reflection coefficient vector θ. Based on the system model parameters, the reflection matrix, through vectorization and data compression, can be represented by the form vec(Θ) as follows:
[0221]
[0222] because The reflection matrix Θ is then obtained by performing an inverse vectorization operation on θ. 3) Interference cancellation is designed based on joint active and passive beamforming.
[0223] If the number of reflection units Q cannot meet the conditions of formula (6), then it is necessary to first design the reflection coefficient matrix to reduce the rank of the interference channel matrix, and then use the active beamforming matrix to perform zero-forcing interference elimination.
[0224] The active beamforming matrix design at the transmitting end is as follows:
[0225] s l =V l x l (9)
[0226] in, This represents the desired signal sent from the l-th transmitter to the l-th receiver. Active beamforming matrix at the l-th transmitter.
[0227] At this point, the conditions for complete elimination of interference are as follows:
[0228]
[0229] When Q < K(K-1)MN, the number of reflection units Q needs to satisfy the following condition:
[0230] Q≥K(K-1)max(M,N)(min(M,N)-p) (11)
[0231] Where (min(M,N)-p) represents the rank after the interference channel is reduced.
[0232] Next, to reduce the rank of the interference channel matrix, matrix vectorization is required. The rank of a matrix is equal to the minimum of its row rank and column rank. Therefore, different relationships between the number of transmitting and receiving antennas will affect the rank of the interference channel matrix. Thus, solving for the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0233] a) When N≥M: The rank of the interference channel matrix depends on N, so it is vectorized into a column vector, as follows:
[0234]
[0235] in, It is H kl The m-th column vector, yes The m-th column vector.
[0236] After vectorizing the interference channel matrix, the column vector group of the interference channel matrix can be obtained. To reduce the rank of the interference channel matrix, we need to discuss the largest linearly independent set of the vector group. Let p be the number of the largest linearly independent vectors in the column vector group of the interference channel. They are linearly independent. Therefore, the rank of the interference channel matrix can be obtained as:
[0237]
[0238] This indicates that the rank of the interference channel matrix needs to be reduced to p. Let t = p+1, p+2, ..., M, then... It can be derived from a vector group Linear representation, meaning that the two are linearly dependent. The specific derivation process is as follows:
[0239]
[0240] To simplify the above results, let and The above derivation result can then be rewritten as:
[0241]
[0242] in, and
[0243] Finally, the non-homogeneous linear equation system (15) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0244] b) When N < M: The rank of the interference channel matrix depends on M, so it is vectorized into rows, as follows:
[0245]
[0246] in, It is by The 1st th (Q+2) th , ..., (Q 2 ) th Row vectors are composed of, H represents kl The nth row vector, express The nth row vector.
[0247] Let p be the maximum number of linearly independent vectors in the row vector group of the interference channel, then the row vector group... They are linearly independent. Therefore, the rank of the interference channel matrix can be obtained as:
[0248]
[0249] When t = p+1, p+2, ..., M, the method for constructing a non-homogeneous linear system of equations is as described in case a), and the specific form is as follows:
[0250]
[0251] in, and and
[0252] Finally, the non-homogeneous linear equation system (18) is solved to obtain the reflection coefficient vector θ, and then inversely vectorized to obtain the reflection coefficient matrix Θ.
[0253] The reflection coefficient matrix Θ obtained from cases a) and b) can only reduce the rank of the interfering channel; an active beamforming matrix V still needs to be designed. k Zero-forcing interference cancellation is performed. At the k-th transmitter, the active beamforming matrix V... k The specific form is as follows:
[0254]
[0255] in, This represents all interference channels generated by the k-th transmitter.
[0256] Finally, the system of quasi-equations (19) is solved to obtain the active beamforming matrix V. k .
[0257] In addition to the scenarios mentioned above, a design scheme for a reflection matrix is proposed where there are obstacles in the line-of-sight path between the receiver and transmitter. The details are as follows:
[0258] 1) Model 2: Construction of a system model without a direct line of sight
[0259] Model 2 uses the same network structure and parameters as Model 1, but there is no direct line-of-sight path between the transmitter and receiver. The signal in Model 2 can only be transmitted through a multi-stage link relayed by a RIS (Relay Link System). Therefore, the received signal at the k-th receiver is modified as follows:
[0260]
[0261] 2) Interference cancellation is achieved based on passive beamforming design.
[0262] In communication links excluding the line-of-sight path, the condition for complete interference elimination becomes:
[0263]
[0264] To ensure the solvability of the homogeneous linear equations, the constraint on the number of RIS reflection units Q should be modified as follows:
[0265] Q≥K(K-1)MN+1 (22)
[0266] Finally, the reflection coefficient vector θ is obtained by solving the homogeneous linear equation system (21), and the reflection matrix Θ is obtained by matrix inverse vectorization.
[0267] 3) Interference cancellation based on joint active and passive beamforming design.
[0268] The joint active and passive beamforming design proposed in Model 1 also applies to Model 2. When Q≤K(K-1)MN, the rank of the interference channel matrix is reduced first, and then zero-forcing interference is performed through active beamforming. Due to the relationship between the rank of the channel matrix and the number of transmitting and receiving antennas, the solution for the reflection coefficient matrix can be divided into two cases: N≥M and N<M.
[0269] a) When N≥M, by using passive beamforming design to reduce the rank of the interference channel matrix, the following homogeneous linear equations can be obtained:
[0270]
[0271] b) When N < M, the specific form of the homogeneous linear equation is as follows:
[0272]
[0273] According to the solution conditions of the homogeneous linear equation system (24) and (25), the RIS reflection unit Q needs to satisfy the following conditions:
[0274] Q≥K(K-1)max(M,N)(min(M,N)-p)+1 (26)
[0275] By solving the homogeneous linear equations (24) and (25), the reflection coefficient vector θ can be calculated, and the reflection matrix Θ can be obtained by inverse vectorization.
[0276] Next, active beamforming design is required to perform zero-forcing interference cancellation on the reduced-rank interference channel matrix, specifically as follows:
[0277]
[0278] Finally, the system of sublinear equations (27) is solved to obtain the active beamforming matrix V. k .
[0279] Figure 1The diagram shows an IRS-assisted multi-user MIMO system model, which has K pairs of multi-antenna transmitters and receivers. Each pair of multi-antenna transmitters is equipped with N transmit antennas, and each pair of multi-antenna receivers is equipped with M receive antennas. The number of reflection elements on the RIS is Q. During signal transmission, the system model has two paths: a direct line-of-sight transmission link and a RIS-assisted cascaded link. Figure 2 The communication network configuration of the model shown is as follows: Figure 1 The same approach applies, but it is assumed that the direct line-of-sight transmission link between the multi-antenna transmitter and receiver is a congested channel. The transmitted signal needs to be forwarded to the receiver via a RIS (Relay Interchange) with Q reflector elements as an intelligent passive repeater. Only cascaded links assisted by the RIS are used for signal transmission and reception. Since K pairs of multi-antenna transmitters / receivers operate simultaneously, for each receiver, all signals except those transmitted by the corresponding transmitter are considered interference signals. Active beamforming and passive beamforming designs are implemented at the transmitter and RIS respectively, effectively reducing the impact of interference on system performance, thereby achieving complete interference cancellation.
[0280] Figure 3 The graph shows the relationship between system performance and signal-to-noise ratio (SNR) under different interference cancellation schemes, with M=6 transmit / receive antennas, N=6 transmit / receive antennas, K=3 users, Q=216 RIS reflector units in passive beamforming, and Q=180 RIS reflector units in combined active and passive beamforming. First, the performance of the communication system with RIS assistance is significantly higher than that without RIS assistance, indicating that deploying RIS can greatly improve system performance. Then, in the iterative algorithm of IA and the TDMA scheme without RIS, the IA iterative algorithm performs better because it can achieve a higher degree of freedom d=9. The interference cancellation algorithm proposed in this invention can further increase the degree of freedom. When the number of RIS reflector units is sufficient, the proposed passive beamforming design method can achieve a maximum degree of freedom of d=6 per user. When the number of RIS reflector units Q decreases, the system's degree of freedom decreases to completely eliminate interference. The proposed joint active and passive beamforming method can achieve 4 degrees of freedom per user (d=4) with a reflector unit Q=180, which reduces system and rate compared to passive beamforming with Q=216. With the same number of RIS reflector units, the passive beamforming design still has a significant performance advantage over the TDMA scheme. In summary, the RIS-assisted interference nullification method proposed in this invention can effectively eliminate interference and achieve higher system and rate performance.
[0281] Figure 4The graph shows the relationship between system performance and signal-to-noise ratio (SNR) under different numbers of RIS reflector elements Q, with M=8 transmit / receive antennas, N=8 transmit / receive antennas, and K=3 users. According to the conditions of the proposed method, if the number of RIS reflector elements Q is less than K(K-1)MN, passive beamforming is needed to reduce the rank of the interference channel. Increasing p leads to an increase in the dimension of the interference subspace and a decrease in degrees of freedom. When Q=384, only passive beamforming design is needed, achieving the maximum system performance and speed under complete interference elimination. When the number of RIS reflector elements Q is insufficient, the system performance and speed decrease as the number of reflector elements Q decreases. Although the user degrees of freedom decrease, the overall system still maintains a good performance level and achieves complete interference elimination.
[0282] Figure 5 The graph shows the relationship between system sum rate and signal-to-noise ratio (SNR) for different communication network configurations with user numbers K of 3, 4, and 5. When the number of RIS-assisted reflection units Q is sufficient, the system sum rate is highest when K=5 users. Although K=4 has more users than K=3, its system sum degree of freedom is slightly lower than that of K=3 under the condition of complete interference elimination, resulting in a slightly lower system sum rate. When the number of RIS reflection units is insufficient, the degree of freedom of each user decreases to achieve complete interference elimination, and the overall sum rate is lower than when the number of RIS reflection units is sufficient. When K=3 users, each user has a large degree of freedom, and the system sum degree of freedom is the largest. At K=4 and K=5, each user has 1 degree of freedom, but K=5 has more users, so the system sum rate is higher than that of K=4. The proposed RIS-assisted interference cancellation method has good system performance and can be applied to generalized network models, demonstrating generality.
[0283] Figure 6 For application scenarios with obstacles in the line-of-sight path, with M=6 and N=6 transceiver antennas and K=3 users, the relationship between system sum and rate and signal-to-noise ratio for different interference cancellation methods is shown in the figure. Because the line-of-sight path between transceivers is blocked, the RIS design in passive beamforming requires Q=217 reflection units. When Q=216, the RIS reflection coefficient matrix is zero, and the system sum and rate of the passive beamforming design are zero, while the joint active-passive beamforming design still has a high system sum and rate. Furthermore, when the number of reflection units Q=216, the proposed joint active-passive beamforming design method still outperforms the orthogonal multiple access scheme. As the number of RIS reflection units Q increases, the degrees of freedom and system sum and rate can also be improved. This verifies that the proposed method can also achieve significant performance improvements in scenarios with obstacles in the line-of-sight path, making its application scenarios more generalizable.
[0284] This invention presents a smart metasurface-assisted zero-forcing interference cancellation method, specifically designed for RIS-assisted multi-user MIMO systems. Inter-user interference is modeled using an interference channel model. First, when the number of RIS reflector units is sufficient, passive beamforming is used to design the interference cancellation scheme. Simultaneously, when the number of RIS reflector units is insufficient, the rank of the interference channel matrix is reduced to design the reflection coefficient matrix, and both active and passive beamforming are combined to design the interference cancellation scheme. In both cases, this invention maximizes system performance and speed while achieving complete interference cancellation. Second, compared to existing iterative interference cancellation schemes, this invention uses a closed-form algorithm to solve the RIS reflection coefficient matrix, resulting in lower algorithm complexity. Furthermore, compared to traditional interference management schemes and interference-aligned min-IL and Max-SINR schemes, this invention significantly improves system performance and speed.
[0285] The above description is merely a preferred embodiment of the present invention and does not constitute a limitation thereof. Those skilled in the art, guided by existing technology, can make other modifications to the implementation of the present invention without creative effort. Any modifications made within the spirit and principles of the present invention, or simple substitutions or equivalent replacements using conventional techniques in the art, should be included within the scope of protection of the present invention.
Claims
1. A method for eliminating zero-forcing interference assisted by a smart metasurface, characterized in that, The process includes: Step S1: For a RIS-assisted multi-user MIMO system, construct a system model with a direct-look path; Step S2: When there are enough RIS reflective units, design an interference cancellation scheme based on passive beamforming; design a reflection coefficient matrix at the smart metasurface, and use sufficient reflective units to perform precise beamforming to achieve complete interference cancellation. Step S3: When the number of RIS reflection units is insufficient, an interference cancellation scheme is designed based on joint active and passive beamforming; a reflection coefficient matrix is designed at the smart metasurface, the rank of the interference channel matrix is reduced by passive beamforming, and then the active beamforming at the transmitter is used to force zero interference, so as to achieve the purpose of complete interference cancellation. In step S1, the system model construction process with a direct line of sight is as follows: In RIS-assisted multi-user MIMO systems, there exists K For transmitters and receivers; where each transmitter has the following number of antennas: M The number of antennas for each receiver is N The number of reflection units in RIS is ; In the communication system model described above, there are two communication methods: Method 1: The transmitted signal reaches the receiver through a direct line-of-sight link between the transmitter and the receiver; Method 2: The transmitted signal reaches the receiver through a multi-stage link relayed by a passive RIS. Assumption Indicates the first The transmitter to the first k Direct channel of each receiver Indicates RIS to the 1st k Channels of a receiver Indicates the first One transmitter to RIS channel; RIS reflection matrix ,in Represents the reflection coefficient vector. They represent the first q The phase and amplitude of each RIS reflector unit; They represent the first k The signal sent by the first transmitter and the first k Gaussian white noise at the receiver; Given that the channel state information is known, assuming the interference between users follows an interference channel model, then the... k Each transmitter sends a signal to its corresponding receiver, and this signal will affect other receivers. k -1 receivers generate interference; at this time, the -1st receiver... k The signals received by each receiver are represented as follows: ; in The matrix represents the desired channel. Indicates an interfering channel; For multi-user interference networks, the network capacity is approximated using degrees of freedom; these degrees of freedom can be equal to the number of users transmitting data streams in parallel, under sufficiently high signal-to-noise ratios. In this case, degrees of freedom Channel capacity The first-order approximation of is defined as: ; in, Indicates the first k The sum and rate of the links, Indicates the signal-to-noise ratio (SNR); No. k The specific formula for calculating the sum rate of the links is as follows: ; in, , Indicates the transmission power. Indicates noise power. This indicates interference leakage, specifically in the following forms: ; Step S2, based on passive beamforming, design the following interference cancellation scheme: First, based on the system model constructed in step S1, a passive beamforming matrix is designed at RIS. The specific interference cancellation conditions are as follows: ; To meet the interference cancellation requirements, the number of RIS reflection units must satisfy the following condition: ; The interference channel matrix in formula (5) is vectorized, and its specific form is as follows: ; in, , It is a matrix The Composed of column vectors; Solving the non-homogeneous linear system of equations (7) yields the reflection coefficient vector. According to the system model parameters, the reflection matrix can be obtained by vectorization and data compression. The format is as follows: ; because Then by means of Perform inverse vectorization to obtain the reflection matrix. ; Step S3, the interference cancellation process based on joint active and passive beamforming design is as follows: If the number of reflective units If the conditions of formula (6) cannot be met, it is necessary to first design the reflection coefficient matrix to reduce the rank of the interference channel matrix, and then use the active beamforming matrix to force zero interference elimination. The active beamforming matrix design at the transmitting end is as follows: ; in, Indicates the first The transmitter sent to the first The desired signal of each receiver, representing In the Active beamforming matrix at each transmitter; At this point, the conditions for complete elimination of interference are as follows: ; when At that time, the number of reflecting units Conditions to be met: ; in, Indicates the rank after the interference channel is reduced; Next, depending on the relationship between the number of transmitting and receiving antennas, the solution to the reflection coefficient matrix can be divided into... Two scenarios; a) when Time: The rank of the interference channel matrix depends on M Therefore, column vectorization is adopted, and the specific form is as follows: ; in, yes The m-th column vector, yes The m-th column vector; After vectorizing the interference channel matrix, the column vector group of the interference channel matrix can be obtained. Let the maximum number of linearly independent vectors in the column vector group of the interference channel be... Then the column vector group They are linearly independent; therefore, the rank of the interference channel matrix can be obtained as: ; This indicates that the rank of the interference channel matrix needs to be reduced to [value missing]. ;make ,but It can be derived from a vector group Linear representation, meaning that the two are linearly dependent; the derivation process is as follows: ; To simplify the above results, let and Therefore, the above derivation result can be rewritten as: ; in, ; Finally, solving the non-homogeneous linear system of equations (15) yields the reflection coefficient vector. And perform inverse vectorization to obtain the reflection coefficient matrix. ; b) When Time: The rank of the interference channel matrix depends on N Therefore, row vectorization is adopted, and the specific form is as follows: ; in, Row vectors are composed of, express The row vectors express The Row vectors; Let the maximum number of linearly independent vectors in the row vector group of the interference channel be... Then the row vector group They are linearly independent; the rank of the interference channel matrix is: ; when The specific form of the method for constructing a non-homogeneous linear system of equations is as follows: ; in, and , and ; Finally, solving the non-homogeneous linear system of equations (18) yields the reflection coefficient vector. And perform inverse vectorization to obtain the reflection coefficient matrix. ; The reflection coefficient matrix obtained from cases a) and b) It can only reduce the rank of the interference channel; an active beamforming matrix still needs to be designed. Perform zero-forcing interference cancellation; in the first k At each transmitter location, the active beamforming matrix The specific form is as follows: ; in, Indicates by the first k All interference channels generated by each transmitter; Finally, the system of quasi-equations (19) is solved to obtain the active beamforming matrix. .
2. The intelligent metasurface-assisted zero-forcing interference elimination method according to claim 1, characterized in that: For the reflection matrix of a line-of-sight path between the receiver and transmitter where there are obstacles, a system model without a line-of-sight path is constructed. Since the signal can only be transmitted through a multi-stage link relayed by RIS, then in the... k The received signal at each receiver is: ; The process of achieving interference cancellation based on passive beamforming design is as follows: In communication links excluding the line-of-sight path, the condition for complete interference elimination becomes: ; To ensure the solvability of homogeneous linear equations, the number of RIS reflection units... The restrictions should be modified as follows: ; Finally, the reflection coefficient vector is obtained by solving the homogeneous linear equation system (21). Then, the reflection matrix is obtained by matrix inverse vectorization. ; Interference cancellation based on joint active and passive beamforming design: when First, reduce the rank of the interference channel matrix, then perform zero-forcing interference through active beamforming. Since the rank of the channel matrix is related to the number of transmitting and receiving antennas, the solution for the reflection coefficient matrix can be divided into... and Two scenarios: a) when When the rank of the interference channel matrix is reduced through passive beamforming design, the following homogeneous linear equations can be obtained: ; b) When When the homogeneous linear equation takes the following form: ; Based on the solution conditions of the homogeneous linear equations (23) and (24), the RIS reflection unit The following conditions must be met: ; The reflection coefficient vector can be calculated by solving the homogeneous linear equations (23) and (24). And perform inverse vectorization to obtain the reflection matrix. ; Next, active beamforming design is performed to eliminate zero-forcing interference in the reduced-rank interference channel matrix, as follows: ; Finally, the system of quasi-linear equations (26) is solved to obtain the active beamforming matrix. .
Citation Information
Patent Citations
Active and passive beamforming collaborative design method for non-orthogonal multiple access secure transmission based on smart reflector
CN114630328B
NOMA network physical layer security optimization method based on STAR-RIS
CN116033424B
