A secure transmission method and device for a multi-cluster MIMO-NOMA system assisted by an intelligent reflector

Abstract

The present invention relates to a secure transmission method and device for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface, the method comprising: considering the scenario in which an eavesdropper exists, establishing a terahertz MIMO-NOMA secure communication system model based on IRS assistance; using NOMA technology for transmission, selecting cluster heads, and designing discrete phase-based analog precoding for transmit beamforming; using a low-complexity zero-forcing method to design digital precoding; under the constraints of system transmission power, user achievable rate, and IRS reflection coefficient, constructing a confidentiality and rate maximization optimization problem by jointly optimizing power allocation and IRS phase shift; using Taylor series expansion and semidefinite programming methods to decouple, converting the original non-convex problem into a convex problem, and proposing an efficient alternating iterative algorithm to obtain a suboptimal solution to the original problem. The present invention studies an IRS-assisted terahertz MIMO-NOMA system based on hybrid precoding in the presence of multiple eavesdroppers, maximizing confidentiality and rate, and improving the security and energy efficiency of the communication system.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and in particular relates to a secure transmission method and device for a multi-cluster MIMO-NOMA system assisted by an intelligent reflecting surface. Background Art

[0002] 6G aims to propose higher performance indicators and realize the vision of "stronger performance, smarter, wider coverage and more environmentally friendly". It has higher requirements for a series of indicators such as data transmission rate, number of connections, latency, etc. A notable feature of 6G is that it is moving towards the terahertz (THz) era. At the same time, in order to cope with the serious threats to the secure transmission of communication systems, the study of the physical layer security (PLS) of wireless communications from the perspective of channel propagation characteristics has gradually become a research hotspot in the field of wireless security.

[0003] IRS is a two-dimensional thin-layer artificial electromagnetic surface with programmable electromagnetic properties. It can effectively adjust the reflected amplitude and phase of the incident electromagnetic wave in real time, overcoming the negative effects of multipath fading and wireless propagation in the natural environment. The available spectrum resources of THz reach 10GHz, which can meet the spectrum requirements of extremely high transmission rates from 100Gbit / s to 1Tbit / s. NOMA can meet the large-scale connectivity, low latency and high-frequency spectrum efficiency of future networks. Compared with the traditional orthogonal multiple access (OMA) scheme, NOMA provides services for multiple users in a single resource block (time / frequency), which can significantly improve the spectrum efficiency. In order to alleviate the information security and power consumption problems brought about by the large-scale access of 6G networks, a terahertz MU-MIMO system with IRS-assisted OFDM access is established, and a sparse RF link antenna structure is used to reduce power consumption. The hybrid precoding and phase shift matrix are jointly optimized to maximize the weight and rate of the system.

[0004] Although many scholars have studied the optimization of IRS, THz transmission, MU-MIMO and NOMA separately, existing research seems to lack the joint optimization of the above technologies. Inspired by the advantages of these technologies, it is of great theoretical significance and practical value to integrate them to further improve the security, energy efficiency and coverage of wireless systems.

[0005] The present invention studies the secure transmission problem of IRS-assisted multi-cluster THz MIMO-NOMA systems. Under the conditions of optimizing power allocation and IRS phase shift constraints, a confidentiality and rate maximization optimization problem is constructed and an efficient alternating iterative algorithm is proposed to solve the original optimization problem. A secure transmission method and device design for multi-cluster MIMO-NOMA systems assisted by smart reflectors are given. Summary of the invention

[0006] The purpose of this invention is to further improve the security, energy efficiency and coverage of wireless systems, consider the scenario where eavesdroppers exist, and establish a terahertz MIMO-NOMA secure communication system model based on IRS assistance;

[0007] The object of the present invention is achieved by: a secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflector, comprising:

[0008] S1: Considering the scenario where an eavesdropper exists, a terahertz MIMO-NOMA secure communication system model based on IRS assistance is established;

[0009] S2: Two sparse RF antenna structures, fully connected and sub-connected, are used. Users are grouped according to channel correlation and gain. NOMA technology is used for transmission. Cluster heads are selected and discrete phase-based analog precoding is designed for transmit beamforming. Digital precoding is designed using a low-complexity zero-forcing method to eliminate inter-cluster interference.

[0010] S3: Under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift;

[0011]

[0012] Among them, p l,m represents the power of the mth user in the lth cluster, γ l,m represents the SINR of the mth user in the lth cluster, represents the SINR of the eavesdropping user; the maximum transmission power of the BS is P T , N RIS represents the number of reflection units of IRS. The reflection phase shift matrix of IRS is defined as in, θ n ∈(0, 2π]; (1a) is a non-convex optimization problem, (1b) is the maximum transmit power constraint, (1c) represents the minimum achievable rate requirement for each user, and (1d) represents the IRS phase shift constraint;

[0013] S4: Taylor series expansion and semidefinite programming method are used to decouple the original non-convex problem and transform it into a convex problem. An efficient alternating iterative algorithm is proposed to obtain the suboptimal solution of the original problem.

[0014] The step S1 specifically includes:

[0015] Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX The base station (BS) of the antenna is configured with N IRSThe base station transmits superimposed signal information on each beam through NOMA technology, groups users into L groups, improves spectrum efficiency and reduces system load. Assume that the communication link from BS to users is blocked by buildings, and all users can only receive the reflected signal from the base station via IRS.

[0016] The step S2 specifically includes:

[0017] The present invention adopts two sparse radio frequency chain antenna structures, namely, fully connected and sub-connected. To achieve multiplexing gain, it is assumed that the number of clusters is equal to the number of radio frequency chains (L = N RF ), first select the cluster head from the user according to the channel gain and its correlation, and then design the discrete phase-based analog precoding for transmit beamforming. The FC hybrid precoding structure requires N TX ×N RF phase shifters, simulating the precoding matrix F (full)

[0018]

[0019] in, is the simulated precoding vector associated with the qth RF chain. For the sake of generality, we assume that N SF =N TX / N RF is an integer, each RF chain is connected to N SF Phase shifter, for SC architecture, simulates the precoding matrix F (sub) for

[0020]

[0021] When using phase shifters to design analog precoding, it is usually necessary to quantize the diagonal elements. Considering a B-bit quantized phase shifter, the two structures of non-zero analog precoding matrices are

[0022]

[0023] in, g(l) is the channel state information of the lth cluster head. According to the channel gain and similarity, the cluster head selection algorithm is used to divide the users into L groups, and each cluster uses NOMA technology for transmission. Since each cluster contains multiple users, the low-complexity zero forcing (ZF) technology is used to design digital precoding to eliminate interference between cluster head users. It is assumed that the equivalent channel gain of all cluster head users is

[0024] G=

g 1,m ,g 2,m ,K,g L,m

[0025] Then, generate the ZF precoding matrix

[0026]

[0027] After normalization, the digital precoding vector of the lth beam can be written as

[0028] The step S3 specifically includes:

[0029] Since there is a coupling between the power allocation of different users and the phase shift of IRS, in the present invention, NOMA is used in each beam, intra-beam coverage coding is performed at the transmitting end, and SIC coding is performed at the receiving end to optimize the power allocation; under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing the power allocation and IRS phase shift.

[0030] The step S4 specifically includes:

[0031] In view of the non-convexity of the original problem, the present invention first initializes the reflection matrix, and uses the first-order Taylor expansion and semidefinite relaxation (SDR) technology to transform the power allocation subproblem into a tractable convex problem; on this basis, an efficient alternating iterative algorithm is proposed to iteratively design the reflection phase shift matrix to obtain the suboptimal solution of the original problem.

[0032] A secure transmission device for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface, comprising:

[0033] The model building module considers the scenario where an eavesdropper exists and builds a terahertz MIMO-NOMA secure communication system model based on IRS assistance;

[0034] The hybrid precoding module uses two sparse RF antenna structures: fully connected and sub-connected. It groups users according to channel correlation and gain, uses NOMA technology for transmission, selects cluster heads, and designs discrete phase-based analog precoding for transmit beamforming. It also uses a low-complexity zero-forcing method to design digital precoding to eliminate inter-cluster interference.

[0035] The equation construction module constructs the confidentiality and rate maximization optimization problem by jointly optimizing power allocation and IRS phase shift under the constraints of system transmission power, user achievable rate and IRS reflection coefficient;

[0036]

[0037] Among them, p l,m represents the power of the mth user in the lth cluster, γ l,m represents the SINR of the mth user in the lth cluster, represents the SINR of the eavesdropping user; the maximum transmission power of the BS is P T , N RIS represents the number of reflection units of IRS. The reflection phase shift matrix of IRS is defined as in, θ n ∈(0, 2π]; (8a) is a non-convex problem, (8b) is the maximum transmit power constraint, (8c) represents the minimum achievable rate requirement for each user, and (8d) represents the IRS phase shift constraint;

[0038] The iterative solution module uses Taylor series expansion and semidefinite programming method to decouple, transforms the original non-convex problem into a convex problem, and proposes an efficient alternating iterative algorithm to obtain the suboptimal solution of the original problem.

[0039] The model building module specifically includes:

[0040] Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX The base station (BS) of the antenna is configured with N IRS The base station transmits superimposed signal information on each beam through NOMA technology, groups users into L groups, improves spectrum efficiency and reduces system load. Assume that the communication link from BS to users is blocked by buildings, and all users can only receive the reflected signal from the base station via IRS.

[0041] The hybrid precoding module specifically includes:

[0042] Two sparse RF antenna structures, sub-connected structure and fully connected structure, are adopted. Users are grouped according to channel correlation and gain. NOMA technology is used for transmission, cluster heads are selected, and discrete phase-based analog precoding is designed for transmit beamforming. Then, a low-complexity zero-forcing method is used to design digital precoding to eliminate inter-cluster interference.

[0043] The equation construction module specifically includes:

[0044] Equation construction module, since there is coupling between the power allocation of different users and the phase shift of IRS, in the present invention, by using NOMA in each beam, intra-beam coverage coding is performed at the transmitting end, and SIC coding is performed at the receiving end to optimize power allocation; under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift.

[0045] The iterative solution module specifically includes:

[0046] Iterative solution module, in view of the non-convexity of the original problem, the present invention first initializes the reflection matrix, and uses the first-order Taylor expansion and semidefinite relaxation (SDR) technology to transform the power allocation subproblem into a tractable convex problem; on this basis, an efficient alternating iterative algorithm is proposed to iteratively design the reflection phase shift matrix to obtain the suboptimal solution of the original problem. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a structural schematic diagram of a secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface provided by the present invention.

[0048] Figure 2 This is a schematic diagram of the structure of IRS-assisted terahertz NOMA secure communication.

[0049] Figure 3 A performance graph comparing FC and SC at different transmit powers is shown;

[0050] Figure 4 A graph showing the relationship between confidentiality and rate and the number of reflective elements;

[0051] Figure 5 A graph showing the relationship between confidentiality energy efficiency and transmission power;

[0052] Figure 6 It is a structural schematic diagram of a secure transmission device of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface provided by the present invention; DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] The present invention provides a secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface. The base station (BS) adopts a hybrid precoding (HP) architecture of a sparse radio frequency chain, including an FC architecture and an SC architecture; considering the existence of eavesdroppers, a joint optimization problem of power allocation and IRS reflection matrix is ​​proposed to maximize confidentiality and rate. Figure 1 As shown, the method comprises the steps of:

[0055] S1: Considering the scenario where an eavesdropper exists, a terahertz MIMO-NOMA secure communication system model based on IRS assistance is established;

[0056] S2: Two sparse RF antenna structures, fully connected and sub-connected, are used. Users are grouped according to channel correlation and gain. NOMA technology is used for transmission. Cluster heads are selected and discrete phase-based analog precoding is designed for transmit beamforming. Digital precoding is designed using a low-complexity zero-forcing method to eliminate inter-cluster interference.

[0057] S3: Under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift;

[0058] S4: Taylor series expansion and semidefinite programming method are used to decouple the original non-convex problem and transform it into a convex problem. An efficient alternating iterative algorithm is proposed to obtain the suboptimal solution of the original problem.

[0059] like Figure 2 As shown in the figure, the method described in this embodiment is applied to the performance of the IRS-assisted THz MIMO-NOMA system. Due to the serious path loss in THz, considering the short-distance communication scenario, the distances from the BS to the IRS and Eve are set to 15m and 5m respectively. The distance from the IRS to all users and Eve is 5m, and the users of each cluster are located in a circle with a radius of 2m. In the simulation, M users are grouped into N RF beams, with multiple users in each group. Security Energy Efficiency (SEE) is defined as the ratio between the achievable confidentiality and rate and the total power consumption

[0060]

[0061] Where P RF is the power consumption of each RF chain, P PS represents the power consumption of each phase shifter, P B Indicates baseband power consumption. Set the typical value P RF =300mW, P PS =40mW (4 phase shifters), baseband power P B =200mW. PSF and N PSS are the number of phase shifters for FC and SC respectively. In addition, the random phase shift method is adopted and the OMA scheme is used for performance comparison. Other simulation parameters are shown in Table 1:

[0062] Table 1 System simulation parameters

[0063]

[0064] In this embodiment, the specific process of step S1 is as follows:

[0065] Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX The base station (BS) of the antenna is configured with N IRS The base station transmits superimposed signal information on each beam through NOMA technology, groups users into L groups, improves spectrum efficiency and reduces system load. Assume that the communication link from BS to users is blocked by buildings, and all users can only receive the reflected signal from the base station via IRS.

[0066] User clusters can help achieve good performance of NOMA systems. Assuming that the number of clusters is the same as the number of RF chains (L = N RF ), and selects cluster heads from the users based on channel gain and correlation. Then the users are divided into L groups using a cluster head selection algorithm. By using NOMA, each beam is able to support multiple users, and SIC in NOMA should be used continuously.

[0067] Two typical architectures of HP are used to reduce the number of RF chains required for THz massive MIMO systems. For the FC structure, F (full) From the following formula we get

[0068]

[0069] in For the SC architecture, the simulated precoding matrix F (sub) for

[0070]

[0071] When designing analog precoding using phase shifters, the corner elements usually need to be quantized. Considering a B-bit quantized phase shifter, the analog precoding matrix with non-zero elements is given by

[0072]

[0073] in, And g(l) is the channel state information of the lth cluster head.

[0074] Assume that the equivalent channel gain of all cluster head users is

[0075] G=[g 1,m ,g 2,m ,K,g L,m ] (14)

[0076] Generate ZF precoding matrix

[0077]

[0078] After normalization, the digital precoding vector of the lth beam can be written as and

[0079] The received signal of the mth user in the lth cluster is

[0080]

[0081] Similarly, the signal received by Eve is expressed as

[0082]

[0083] Among them, G t and G r denote the transmit and receive antenna gains respectively, represents the channel vector between BS and IRS, represents the equivalent channel from IRS to the mth user in the lth cluster, is the channel vector from IRS to Eve; the reflection phase shift matrix of IRS is defined as Θ=diag{θ1,K,θ n}, θ n ∈(0,2π]s l,m =Fv l x l,j ;x l,j represents the transmission signal of the jth user in the lth cluster, is the digital precoding vector of the lth cluster, is the simulated beamforming of all clusters, where n l,m and is equal to additive white Gaussian noise (AWGN). In addition, terahertz waves are easily absorbed by water when propagating in the air, resulting in the scattered component being much lower than the line-of-sight component. Ignoring the influence of the scattered component, H is expressed as

[0084]

[0085] in represents the path loss composed of free diffusion loss and molecular absorption loss in terahertz communication, c is the speed of light, f is the system frequency, τ(f) represents the absorption coefficient of water molecules in the air, and d is the distance from BS to IRS. in and are the antenna array response vectors of the transmitter and receiver respectively,

[0086]

[0087] in, and Where d0 represents the distance between antennas, φi ∈[-π / 2,π / 2], i∈[r,t] is the angle of departure (AOD) and angle of arrival (AOA). Similarly, g l,m Can be written as

[0088]

[0089] in d l,m It represents the distance from IRS to the mth user in the lth cluster.

[0090] In addition, you can get

[0091]

[0092] in and and d E Represents the distance between IRS and Eve.

[0093] For convenience, let β l,m =G t G r q(f,d)q(f,d l,m ) and β E =G t G r q(f,d)q(f,d E ), where η is the path compensation factor. After the hybrid precoding design, the SINR of the mth user of the lth cluster and Eve can be expressed as

[0094]

[0095] in Then, the security rate of the mth user in the lth cluster can be written as

[0096]

[0097] The power allocation and phase shift matrix are jointly optimized to maximize confidentiality and rate. The optimization problem is stated as follows:

[0098]

[0099] Where P T is the transmission power of BS.

[0100] A. Optimize power distribution

[0101] By using NOMA in each beam, intra-beam coverage coding is performed at the transmitting end, and SIC coding is performed at the receiving end. and Assume that the order of the effective channel gain satisfies That is, the SIC decoding sequence of NOMA users. Next, the security rate is further written as

[0102]

[0103] B1, A1, D1 and C1 in (29) are respectively expressed as

[0104]

[0105] Due to the existence of the logarithmic term -log2(B1)-log2(C1), problem (28) is still non-convex. To solve these problems, at the point Apply a first-order Taylor expansion on . Set and Then -log2(B1)-log2(C1) are transformed into convex forms respectively,

[0106]

[0107] in and It can be found that (32) and (33) are both linear. Then, substituting them into constraint (26c) transforms them into convex constraints:

[0108]

[0109] Through the above transformation, problem (28) can be re-read as

[0110]

[0111] st(26b),(34)(35b)

[0112] When Θ is given, problem (35) can be computed by applying a convex problem solver such as the CVX toolbox.

[0113] B. Solving the phase shift of RIS

[0114] After obtaining the power allocation p for all users l,m After that, the original problem (19) becomes a feasibility check problem.

[0115] set up and Then reconsider the reflection matrix Θ of IRS and transform the safety rate into the following form

[0116]

[0117] express A2, B2, C2 and D2 in (38) can be expressed as

[0118]

[0119] where (b) is obtained by calling the identity a H Ta=Tr(Taa H ) is obtained. Similarly, -log2(B2)-log2(C2) is a non-convex term, and a first-order Taylor expansion is applied at point Φ. To facilitate the following operations, set and Then, -log2(B2)-log2(C2) can be converted into convex forms respectively as shown below.

[0120]

[0121] in, Then inserting (41) and (42) into constraint (36), constraint (26c) can be written as

[0122]

[0123] Therefore, problem (28) can be rewritten as

[0124]

[0125] rank(Φ)=1,Φ≥0(44c)

[0126] (43)(44d)

[0127] Ignoring rank 1, rank(Φ) = 1, the problem can be solved efficiently by applying the CVX toolbox. In general, the solution to Φ computed using eigenvalue decomposition does not necessarily satisfy rank 1. A high-quality rank-one solution to the problem is then recovered by standard Gaussian randomization.

[0128] It can be seen from the above technical solution that the present invention provides a secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflecting surface. Under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, two sparse RF chain antenna structures are adopted to maximize confidentiality and rate by jointly optimizing power allocation and IRS phase shift.

[0129] Figure 3 Compare the performance of FC and SC at different transmit powers. Regardless of the FC structure or the SC structure, confidentiality and rate increase monotonically with the increase of transmit power. Under the same structure, the confidentiality and rate after optimizing the IRS phase shift are significantly higher than the random phase shift scheme. At the same time, optimizing the phase of IRS can also improve the beamforming gain. And under the same conditions, the confidentiality and rate under the FC structure are higher than those under the SC structure.

[0130] Figure 4 The influence of the number of reflection elements on confidentiality and rate is given. The more reflection elements there are in the IRS, the higher the confidentiality and rate are. This shows that the present invention can make full use of the advantages of IRS to improve channel conditions, that is, the spatial DoF and beamforming gain provided by the IRS will gradually increase with the increase in the number of reflection elements. At the same time, more reflection elements increase the number of paths for the reflected signal, thereby bringing higher beamforming gain to the user. In addition, when the base station has more antennas, a higher beamforming gain can be obtained.

[0131] Figure 5 The effect of transmission power on confidentiality energy efficiency is given. Obviously, the SEE of all schemes decreases slowly with the increase of transmission power. In addition, it can be verified that the proposed NOMA scheme can obtain a higher SEE than the traditional OMA scheme with the same HP structure. At the same time, the SC structure can achieve a higher SEE than the FC structure.

[0132] Figure 6 It is a structural schematic diagram of a secure transmission device of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface provided by the present invention;

[0133] The model building module is used to consider the scenario where an eavesdropper exists and to establish a terahertz MIMO-NOMA secure communication system model based on IRS assistance;

[0134] The hybrid precoding module uses two sparse RF antenna structures: fully connected and sub-connected. It groups users according to channel correlation and gain, uses NOMA technology for transmission, selects cluster heads, and designs discrete phase-based analog precoding for transmit beamforming. It also uses a low-complexity zero-forcing method to design digital precoding to eliminate inter-cluster interference.

[0135] The equation construction module constructs the confidentiality and rate maximization optimization problem by jointly optimizing power allocation and IRS phase shift under the constraints of system transmission power, user achievable rate and IRS reflection coefficient;

[0136]

[0137]

[0138] Among them, p l,m represents the power of the mth user in the lth cluster, γ l,m represents the SINR of the mth user in the lth cluster, represents the SINR of the eavesdropping user; the maximum transmission power of the BS is P T , N RIS represents the number of reflection units of IRS. The reflection phase shift matrix of IRS is defined as in, θ n ∈(0,2π];

[0139] (8a) is a non-convex problem, (8b) is the maximum transmit power constraint, (8c) represents the minimum achievable rate requirement for each user, and (8d) represents the IRS phase shift constraint;

[0140] The iterative solution module uses Taylor series expansion and semidefinite programming method to decouple, transforms the original non-convex problem into a convex problem, and proposes an efficient alternating iterative algorithm to obtain the suboptimal solution of the original problem.

[0141] In this embodiment, the model building module specifically includes:

[0142] Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX The base station (BS) of the antenna is configured with N IRS The base station transmits superimposed signal information on each beam through NOMA technology, groups users into L groups, improves spectrum efficiency and reduces system load. Assume that the communication link from BS to users is blocked by buildings, and all users can only receive the reflected signal from the base station via IRS.

[0143] In this embodiment, the hybrid precoding establishment module specifically includes:

[0144] Two sparse RF antenna structures, sub-connected structure and fully connected structure, are adopted. Users are grouped according to channel correlation and gain. NOMA technology is used for transmission, cluster heads are selected, and discrete phase-based analog precoding is designed for transmit beamforming. Then, a low-complexity zero-forcing method is used to design digital precoding to eliminate inter-cluster interference.

[0145] In this embodiment, the equation construction module specifically includes:

[0146] Equation construction module, since there is coupling between the power allocation of different users and the phase shift of IRS, in the present invention, by using NOMA in each beam, intra-beam coverage coding is performed at the transmitting end, and SIC coding is performed at the receiving end to optimize power allocation; under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift.

[0147] In this embodiment, the iterative solution module specifically includes:

[0148] Iterative solution module, in view of the non-convexity of the original problem, the present invention first initializes the reflection matrix, and uses the first-order Taylor expansion and semidefinite relaxation (SDR) technology to transform the power allocation subproblem into a tractable convex problem; on this basis, an efficient alternating iterative algorithm is proposed to iteratively design the reflection phase shift matrix to obtain the suboptimal solution of the original problem.

Claims

1. A secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface, characterized in that: include: S1: Considering the scenario where Eve exists, a terahertz MIMO-NOMA secure communication system model based on IRS assistance is established; Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX BS with antennas, a configuration of N IRS The BS uses the NOMA technology to transmit superimposed signal information on each beam, and uses the cluster head selection algorithm to divide the users into L groups. Each cluster uses the NOMA technology for transmission. Each cluster contains multiple users. The communication link from the BS to the user is blocked by the building, and all users can only receive the reflected signal from the BS via the IRS. S2: Two sparse RF antenna structures, fully connected and sub-connected, are used. Users are grouped according to channel correlation and gain. NOMA technology is used for transmission. Cluster heads are selected and discrete phase-based analog precoding is designed for transmit beamforming. Digital precoding is designed using a low-complexity zero-forcing method to eliminate inter-cluster interference. S3: Under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift; Among them, p l,m represents the power of the mth user in the lth cluster, γ l,m represents the SINR of the mth user in the lth cluster, represents Eve’s SINR; the maximum transmission power of BS is P T , N RIS represents the number of reflection units of IRS. The reflection phase shift matrix of IRS is defined as in, θ n ∈(0, 2π]; (1a) is a non-convex optimization problem, (1b) is the maximum transmit power constraint, (1c) represents the minimum achievable rate requirement for each user, and (1d) represents the IRS phase shift constraint; S4: Taylor series expansion and semidefinite programming method are used to decouple the original non-convex problem and transform it into a convex problem. The alternating iterative algorithm is used to obtain the suboptimal solution of the original problem.

2. According to claim 1, a secure transmission method for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface is characterized in that: The step S2 specifically includes: Two sparse RF chain antenna structures, fully connected and sub-connected, are used. To achieve multiplexing gain, the number of clusters is set equal to the number of RF chains, i.e. L = N. RF , according to the channel gain and its correlation, the cluster head is first selected from the users, and then the discrete phase-based analog precoding is designed for transmit beamforming. The FC hybrid precoding structure requires N TX ×N RF phase shifters, simulating the precoding matrix F (full) in, is the simulated precoding vector associated with the qth RF chain, and N is set SF =N TX N RF is an integer, each RF chain is connected to N SF Phase shifter, for SC architecture, simulates the precoding matrix F (sub) for When using phase shifters to design analog precoding, it is usually necessary to quantize the diagonal elements. Considering a B-bit quantized phase shifter, the two structures of non-zero analog precoding matrices are Among them, g(l) is the channel state information of the lth cluster head; according to the channel gain and similarity, the cluster head selection algorithm is used to divide the users into L groups, and each cluster uses NOMA technology for transmission; since each cluster contains multiple users, the low-complexity ZF technology is used to design digital precoding to eliminate interference between cluster head users. The equivalent channel gain of all cluster head users is G=[g 1,m ,g 2,m ,K,g L,m ],(6) Then, generate the ZF precoding matrix After normalization, the digital precoding vector of the lth beam can be written as 3. The secure transmission method of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface according to claim 1, characterized in that: The step S3 specifically includes: Since there is a coupling between the power allocation of different users and the phase shift of IRS, NOMA is used in each beam, intra-beam coverage coding is performed at the transmitter, and SIC coding is performed at the receiver to optimize the power allocation. Under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing power allocation and IRS phase shift.

4. The secure transmission method of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface according to claim 1, characterized in that: The step S4 specifically includes: In view of the non-convexity of the original problem, the reflection matrix is ​​first initialized, and the first-order Taylor expansion and SDR technology are used to transform the power allocation subproblem into a tractable convex problem; on this basis, the alternating iterative algorithm is used to iteratively design the reflection phase shift matrix to obtain the suboptimal solution of the original problem.

5. A secure transmission device for a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface, characterized in that: include: The model building module considers the scenario where Eve exists and builds a terahertz MIMO-NOMA secure communication system model based on IRS assistance; Establish an IRS-assisted terahertz NOMA secure communication system, which includes an N TX BS with antennas, a configuration of N IRS The BS uses the NOMA technology to transmit superimposed signal information on each beam, and uses the cluster head selection algorithm to divide the users into L groups. Each cluster uses the NOMA technology for transmission. Each cluster contains multiple users. The communication link from the BS to the user is blocked by the building, and all users can only receive the reflected signal from the BS via the IRS. The hybrid precoding module uses two sparse RF antenna structures: fully connected and sub-connected. It groups users according to channel correlation and gain, uses NOMA technology for transmission, selects cluster heads, and designs discrete phase-based analog precoding for transmit beamforming. It also uses a low-complexity zero-forcing method to design digital precoding to eliminate inter-cluster interference. The equation construction module constructs the confidentiality and rate maximization optimization problem by jointly optimizing power allocation and IRS phase shift under the constraints of system transmission power, user achievable rate and IRS reflection coefficient; Among them, p l,m represents the power of the mth user in the lth cluster, γ l,m represents the SINR of the mth user in the lth cluster, represents Eve’s SINR; the maximum transmission power of BS is P T , N RIS represents the number of reflection units of IRS. The reflection phase shift matrix of IRS is defined as in, θ n ∈(0, 2π]; (8a) is a non-convex problem, (8b) is the maximum transmit power constraint, (8c) represents the minimum achievable rate requirement for each user, and (8d) represents the IRS phase shift constraint; The iterative solution module applies Taylor series expansion and semidefinite programming decoupling to transform the original non-convex problem into a convex problem, and uses an alternating iterative algorithm to obtain a suboptimal solution to the original problem.

6. The secure transmission device of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface according to claim 5, characterized in that: The equation construction module specifically includes: Equation construction module, since there is coupling between the power allocation of different users and the phase shift of IRS, by using NOMA in each beam, intra-beam coverage coding is performed at the transmitting end, and SIC coding is performed at the receiving end to optimize the power allocation; under the constraints of system transmission power, user achievable rate and IRS reflection coefficient, the confidentiality and rate maximization optimization problem is constructed by jointly optimizing the power allocation and IRS phase shift.

7. The secure transmission device of a multi-cluster MIMO-NOMA system assisted by an intelligent reflective surface according to claim 5, characterized in that: The iterative solution module specifically includes: In the iterative solution module, in view of the non-convexity of the original problem, the reflection matrix is ​​first initialized, and the first-order Taylor expansion and SDR technology are used to transform the power allocation subproblem into a tractable convex problem; on this basis, the alternating iterative algorithm is used to iteratively design the reflection phase shift matrix to obtain the suboptimal solution of the original problem.

Images