Cooperative rate-splitting network transmission method with untrusted relay threat

Abstract

The present invention provides a collaborative rate-splitting network secure transmission method in the presence of untrusted relay threats, which belongs to the field of secure transmission optimization of downlink communications and relates to a design scheme for preventing relay users from eavesdropping through precoding optimization. While the base station transmits information, the untrusted user relay attempts to eavesdrop on the private message of another legitimate user. By using beamforming technology, the private stream of the legitimate user is placed in the null space of the relay user as much as possible, thereby improving the security of the user. Based on this model, the present invention provides a design method for jointly optimizing the precoding matrix, direct transmission time slot, and the public message rate of each user. The method can generate the optimal precoding matrix and direct transmission information of the base station according to the model parameters to maximize the user's secure transmission rate.

Description

Technical Field

[0001] The present invention belongs to the field of downlink network security rate optimization based on a rate splitting framework, and relates to a secure transmission scheme with the assistance of an untrusted relay user. Specifically, when a base station adopts a rate splitting framework to serve users with the assistance of an untrusted relay user, linear precoding and direct transmission time slots are jointly optimized to maximize the user's secure transmission rate. Background Art

[0002] The rate split (RS) scheme is a downlink multi-antenna multiple access framework that divides user information into a public part and a private part. At the transmitter, the public message and each user's private message are superimposed and transmitted in the same time-frequency resource block through linear precoding technology. Adding the multicast public message part strengthens the interference management between users, and can also reduce the inter-user interference caused by the imperfect transmitter channel information (CSIT) and reduce the feedback overhead. At the receiver, the serial interference cancellation (SIC) technology is used to decode part of the interference and treat the remaining interference as noise. The performance achieved by the serial interference cancellation receiver depends on the precoding technology. The rate split strategy is a framework that can integrate multiple multiple access technologies, such as orthogonal multiple access (OMA), non-orthogonal multiple access (NOMA), and space division multiplexing access (SDMA). Since the rate split strategy allocates more power to the public stream, it can hide the private stream in the public stream when the eavesdropper cannot successfully decode the public stream, thereby improving the security of the system.

[0003] Cooperative relay between users can effectively improve the reliability and throughput of the system. The cooperative relay broadcast channel strategy is to use users with better channel status as user relay nodes and forward cooperative information to other users through relay auxiliary links, which can improve the capacity and reliability of the system. In the context of non-orthogonal multiple access, the cooperative relay broadcast channel strategy can take advantage of user channel differences, improve interruption performance, and improve spectrum and energy efficiency.

[0004] Due to the openness of wireless communication transmission media, the broadcast nature of wireless channels, and the complexity of users, there may be eavesdroppers, which poses a great threat to the privacy and security of users. The traditional method to improve security is to encrypt and decrypt information through keys at the network layer, encrypt information through mathematical algorithms, and prevent eavesdroppers from decrypting information. However, complex algorithms have poor real-time encryption and decryption, and with the improvement of computer computing power, eavesdroppers may be able to crack the key, causing message leakage. In order to solve this problem, the characteristics of wireless channels can be used at the physical layer to increase the difference in channel conditions between legitimate users and eavesdroppers, and reduce the amount of information of the transmitted signal obtained by the eavesdropper while ensuring that the legitimate user can correctly receive the information, so that it tends to zero. Physical layer security can improve the security performance of the system through beamforming, adding artificial noise, etc. Combining physical layer security technology with upper layer security technology can ensure the security of legitimate user information to a greater extent.

[0005] The use of user relay collaboration in the rate splitting system can effectively utilize the channel differences of different users, improve the information transmission rate, and ensure the stability of the system. Since the public stream can be decoded by untrusted relay users, it is crucial to ensure the secure transmission of private messages of other legitimate users in an open wireless communication network. While ensuring the quality of service (QoS) of untrusted relay users, the present invention aims to maximize the sum of the secure transmission rate and the public message transmission rate, and reasonably designs the precoding matrix and the direct transmission time slot. Summary of the invention

[0006] The purpose of the present invention is to solve the anti-eavesdropping problem of a legitimate user at a distance in a rate-splitting system with untrusted relay users assisting in transmission. A base station serves two users at a long distance and a short distance. The user at a short distance acts as an untrusted user relay to assist the base station in forwarding public messages to the user at a long distance. The specific scheme is as follows: Figure 1 Based on this model, the present invention provides a solution for jointly optimizing the precoding matrix and the direct transmission time slot, thereby improving the sum of the user's secure transmission rate and the public message transmission rate.

[0007] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0008] A cooperative rate-splitting network security transmission method with an untrusted relay threat comprises the following steps:

[0009] The first step is to build a system model:

[0010] 1) In a multiple-input single-output (MISO) downlink, there is a T Antennas (N T≥2) and two single-antenna users (U1, U2). The distance between U1 and the base station is closer than that between U2 and the base station. U1, as an untrusted user relay, assists the base station in forwarding public messages to U2 by decoding and forwarding. U1 works in half-duplex (HD) mode. In the direct transmission time slot, the base station transmits messages to the two users in a rate-splitting manner, and U1 eavesdrops on U2's private messages. In the cooperative transmission time slot, U1 forwards the public message to U2.

[0011] 2) In this model, the channels from the base station to U1, the base station to U2, and U1 to U2 are represented as h1, h2, and h3, respectively, which are Rayleigh fading channels. According to the rate division principle, U k The message is divided into a public part W c,k and a private section W p,k (k=1,2). c,1 and W c,2 Combined to form a super public message W c , encoded into a public stream s by a public codebook that can be decoded by all users c , W p,1 and W p,2 Encoded into private streams s1 and s2 respectively. The precoding matrix is ​​P = [p c ,p1,p2], where p c is the public stream precoding vector, p1, p2 are private stream precoding vectors. Signal stream s = [s c ,s1,s2] T Linear precoding is performed by the precoding matrix P, and the transmitted signal is:

[0012] x=Ps=p c s c +p1s1+p2s2, (1)

[0013] in(·) T represents the transpose of a vector or matrix. Assume that E[ss H ]=I, where E[·] represents the expectation and I is the unit matrix. The base station transmit power is given by tr(PP H )≤P t Restriction, where tr(·) is the trace of the matrix, (·) H Represents the conjugate transpose of a vector or matrix, P t It is the maximum power that the base station can transmit.

[0014] 3) In the direct transmission time slot θ, the base station transmits at the maximum power P t Send messages to two users. In the cooperative transmission time slot 1-θ, U1 forwards the decoded and re-encoded public message to U2, and the relay forwarding power is P r .

[0015] During direct transmission, U k The received signal is:

[0016]

[0017] Among them, n k is additive white Gaussian noise (AWGN), with a distribution mean of 0 and a variance of σ k 2 The complex Gaussian distribution of

[0018] Since most of the transmission power is allocated to the public stream, all user private streams are considered as noise when decoding the public stream. k In decoding c The signal to interference plus noise ratio (SINR) at this time is:

[0019]

[0020] In the direct transmission time slot θ, U k Decode public streams c The rate is:

[0021]

[0022] When decoding the private stream, first use serial interference cancellation detection to separate the public message signal, and then treat the signal of another user as interference to decode the private stream. k In decoding k The signal-to-interference-noise ratio is:

[0023]

[0024] During the entire time slot, U k Decoding k The rate is:

[0025]

[0026] U1 can remove the common stream s through serial interference cancellation c Using U1's private stream s2 and its own private stream s1 to decode U2's private stream s2, the signal-to-noise ratio of U1 when eavesdropping on s2 is:

[0027]

[0028] During the entire direct transmission time slot, the eavesdropping rate of U1 decoding s2 is:

[0029]

[0030] The safe transmission rate of U2 is:

[0031]

[0032] where [x] + =max{x,0}.

[0033] During the collaborative transmission process, relay user U1 re-encodes the public message and sends it to U2. The signal received by U2 is:

[0034]

[0035] Where n3 is additive white Gaussian noise, with a distribution mean of 0 and a variance of σ3 2 The complex Gaussian distribution of U2 in decoding s c The signal-to-noise ratio is:

[0036]

[0037] In the cooperative transmission time slot 1-θ, U2 decodes the public stream s c The rate is:

[0038]

[0039] During the entire time slot, U1 decodes the public stream s c The rate is U2 decoding public streams c The rate is To ensure that both users can successfully decode the public stream, the public message transmission rate needs to meet the following requirements:

[0040] R c =min{R c,1 ,R c,2}. (13)

[0041] For possible external eavesdroppers, when they cannot decode public messages, public messages can be used as artificial noise to prevent them from eavesdropping on private messages of legitimate users. Since the base station allocates more power to public messages in the rate splitting strategy, private messages can be hidden in public messages, thereby improving the security of private messages. At this time, the rate at which external eavesdroppers eavesdrop on U2 private messages is:

[0042]

[0043] Among them, h e is the channel information from the base station to the external eavesdropper, satisfying Rayleigh fading, σ e 2 is the power of the additive white Gaussian noise.

[0044] The second step is to determine the objective function and propose an optimization problem:

[0045] Since the public stream is also relayed and decoded by untrusted users in the rate-splitting cooperative transmission system, the secure transmission of the U2 private stream must be guaranteed to ensure the security of the U2 information. The secure transmission rate R2 of the U2 private stream should be increased as much as possible. sec . At the same time, the public message transmission rate R is increased as much as possible c , reducing the interruption probability of public message reception and improving the reliability of the system. At the same time, public messages can be used as artificial noise to increase the difficulty for external eavesdroppers to eavesdrop on internal user messages. In this regard, the proposed optimization problem is as follows:

[0046]

[0047] stc1+c2≤R c ,(15b)

[0048]

[0049]

[0050] c1≥0,c2≥0, (15e)

[0051]

[0052] tr(PP H )≤P t , (15g)

[0053] 0<θ≤1, (15h) Where u is R2 sec and R c The weight of R2 can be changed by adjusting u sec and R c The optimization importance of θ is the direct transmission time slot, P is the linear precoding matrix, c = [c1, c2], c k Represents U k Public message W c,k The rate of. Formula (15a) means to find the optimal θ, P and c to maximize the optimization objective. Formula (15b) is the rate limit of the public message part of the two users. Formulas (15c) and (15d) guarantee the service quality of the user. R1 th and R2 th is the service quality threshold of U1 and U2, formula (15e) represents that the rate of user in public message allocation is non-negative, and formula (15f) means that the private message transmission rate of U1 is greater than the decoding threshold r1 th, Formula (15g) is the base station transmission power limit, and Formula (15h) is the direct transmission time slot allocation limit. At this time, since the public stream must be decoded by all users, more U1 messages can be allocated to public messages, which can improve the fairness of the system without causing U2 information leakage.

[0054] The third step is to design an algorithm to solve the optimization problem:

[0055] Since the optimization problem (15) is non-convex and difficult to solve, the continuous convex approximation (SCA) method is used to solve the above optimization problem. p,1 ,ν p,2 ,ν c,1 ,ν c,2 ,ν e,1,2 ], ν p,k Indicates that in U k The rate of decoding private messages at c,k Indicates that in U k The rate at which public messages are decoded, ν e,1,2 represents the eavesdropping rate of U1 on U2's private message, and α=[α p,1 ,α p,2 ,α c,1 ,α c,2 ,α e,1,2 ],ε=[ε p,1 ,ε p,2 ,ε c,1 ,ε c,2 ,ε e,1,2 ] and β=[β p,1 ,β p,2 ,β c,1 ,β c,1 ], which represent the achievable rate (excluding time slot θ), SINR and noise plus interference respectively.

[0056] 1) Non-convex problem transformation

[0057] By introducing the variable ν, the optimization objective can be transformed into At the same time, the restriction conditions are also transformed, adding restriction R p,k ≥ν p,k ,,R e,1,2 ≤ν e,1,2 , R c,k ≥ν c,k , the first restriction can be transformed into:

[0058] Γ [n] (θ,α p,k )≥ν p,k , (16a)

[0059]

[0060] log2(1+ε p,k )≥α p,k , (16c)

[0061] Θ [n] (p k ,β p,k )≥ε p,k , (16d)

[0062] in [n] represents the nth iteration, which can be obtained by The Taylor expansion of is obtained. The first-order Taylor expansion at the point and the linear lower bound approximation can be obtained Re(·) indicates taking the real part.

[0063] The second restriction can be translated into:

[0064] Φ [n] (θ,α e,1,2 )≤ν e,1,2 , (17a)

[0065]

[0066] 1+ε e,1,2 -Ψ [n] (α e,1,2 )≤0, (17c) in, pass exist The first-order Taylor expansion of the point gives, By Perform a first-order Taylor expansion at to obtain .

[0067] The third restriction can be translated into:

[0068] Γ [n] (θ,α c,1 )≥ν c,1 , (18a)

[0069]

[0070] log2(1+ε c,k )≥α c,k , (18c)

[0071] Θ [n] (p k ,β c,k )≥ε c,k , (18d)

[0072]

[0073] Where k = 1, 2, It is the achievable rate of the public flow in the cooperative transmission time slot.

[0074] The non-convex optimization problem is transformed into the following convex optimization problem:

[0075]

[0076] stc1+c2≤ν c,k ,k=1,2,(19b)

[0077]

[0078]

[0079]

[0080] (16),(17),(18),(15e),(15g),(15h). (19f) Based on this, the original problem has been transformed into a convex optimization problem that is easy to solve based on the idea of ​​continuous convex approximation.

[0081] 2) Algorithm design based on continuous convex approximation

[0082] The flow of Algorithm 1 based on continuous convex approximation is as follows:

[0083] 2.1) Initialize n←0, v [n] ←0,θ [n] , P [n] , c [n] , [n] , α [n] , ε [n] , β [n] , error tolerance e

[0084] 2.2) Repeat

[0085] 2.3)n←n+1

[0086] 2.4) Substitute θ [n-1] , P [n-1] , c [n-1] , [n-1] , α [n-1] , ε [n-1] , β [n-1] Solve problem (19) and get the optimal solution θ * , P * , c * , * , α * , ε *, β * , the optimal value is v *

[0087] 2.5) Update the optimal solution and optimal value, v [n] ←v * ,θ [n] ←θ * , P [n] ←P * , c [n] ←c * , [n] ←υ * , α [n] ←α * , ε [n] ←ε * , β [n] ←β *

[0088] 2.6) Until |v [n] -v [n-1] | <e

[0089] 2.7) Output v,θ,P,c.

[0090] In this algorithm, the optimization value of the objective function increases with the number of iterations, and the maximum power limits the upper limit of the optimization target. Therefore, the algorithm based on continuous convex approximation can converge to a stable point.

[0091] In this regard, we maximize the sum of the U2 secure transmission rate and the public message rate by jointly optimizing the precoding matrix, transmission time slots and rate allocation, thereby improving the security of more distant legitimate users in a collaborative rate-splitting network with the threat of untrusted relay users, and preventing the information of more distant legitimate users from being leaked to untrusted relay users.

[0092] The beneficial effects of the present invention are as follows: the present invention provides a solution for protecting the security of private messages of legitimate users in a cooperative rate division multiple access network with untrusted relay users, and by reasonably designing base station precoding, direct transmission time slot allocation, and user message rate allocation in public messages, the security of legitimate users and the robustness of the system are guaranteed while satisfying the service quality of untrusted relay users. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 For user relay two-user cooperative rate division multiple access system.

[0094] Figure 2 Convergence analysis of cooperative rate partitioning for algorithms based on continuous convex approximation.

[0095] Figure 3Comparison of the achievable rates for cooperative rate splitting and non-cooperative rate splitting at different base station transmit powers and antenna numbers.

[0096] Figure 4 Comparison of transmission time slots between cooperative rate splitting and non-cooperative rate splitting at different base station transmit powers and antenna numbers.

[0097] Figure 5 Comparison of the achievable rates of cooperative rate splitting and non-cooperative rate splitting under different user forwarding powers and user distances.

[0098] Figure 6 Comparison of transmission time slots between cooperative rate splitting and non-cooperative rate splitting under different user forwarding powers and user distances.

[0099] Figure 7 This is the effect of different base station transmission power and external eavesdropper distance on the external eavesdropper's eavesdropping rate. DETAILED DESCRIPTION

[0100] In order to better understand the above technical solution, a specific analysis is given below in conjunction with the accompanying drawings and specific implementation methods.

[0101] The first step is to build a system model:

[0102] 1) If Figure 1 For the rate-splitting cooperative transmission system model, in this example, the number of base station transmit antennas is set to N T =3, the base station serves two users, where the distance between U1 and the base station is farther than that between U2 and the base station. U1, as an untrusted relay user, decodes the public stream transmitted by the base station to obtain the public message and re-encodes it with a different codebook and sends it to U2. In the direct transmission time slot, the base station transmits messages to the two users in a rate-splitting manner, and U1 eavesdrops on U2's private message. In the cooperative transmission time slot, U1 forwards the public message to U2.

[0103] 2) In this model, the channels from the base station to U1, the base station to U2, and U1 to U2 are represented by h1, h2, and h3, respectively, which are Rayleigh fading channels. as well as Where β is the channel gain at the reference distance d=1, and β=10 -4 , d k For BS and U k The distance between them, d3 is the distance between users, r is the path loss index, take r = 2.6, g k N is randomly generated T ×1 normalized complex Gaussian distribution vector, where each element is independent, g3 is a randomly generated normalized complex Gaussian distribution factor, and the noise average power is set to σ1 2 =σ22 =σ3 2 =0.01mW.

[0104] In the second step, according to the specific parameter settings of the first step and formulas (3)-(13), determine the objective function, list the constraints, and list the optimization problem.

[0105]

[0106] stc1+c2≤R c , (15b)

[0107]

[0108]

[0109] c1≥0,c2≥0, (15e)

[0110]

[0111] tr(PP H )≤P t , (15g)

[0112] 0<θ≤1, (15h)

[0113] where u is R2 sec and R c The weight is set to 0.5, R1 th and R2 th is the service quality requirement threshold of U1 and U2, set to 2bit / s / Hz, r1 th It is the decoding threshold of U1 private stream, which is set to 1 bit / s / Hz.

[0114] The third step is to design an algorithm to solve the optimization problem:

[0115] 1) Based on the idea of ​​continuous convex approximation, non-convex optimization problems are transformed into convex optimization problems.

[0116]

[0117] stc1+c2≤ν c,k ,k=1,2,(19b)

[0118]

[0119]

[0120]

[0121] (16),(17),(18),(15e),(15g),(15h).(19f)

[0122] 2) Design an iterative algorithm to solve problem (19). Convex problems can be solved using MATLAB, for example, using the CVX toolbox. The flow of Algorithm 1 based on continuous convex approximation is as follows:

[0123] 2.1) Initialize n←0, v [n] ←0,θ [n] , P [n] , c [n] , [n] , α [n] , ε [n] , β [n] , error tolerance e

[0124] 2.2) Repeat

[0125] 2.3)n←n+1

[0126] 2.4) Substitute θ [n-1] , P [n-1] , c [n-1] , [n-1] , α [n-1] , ε [n-1] , β [n-1] Solve problem (19) and get the optimal solution θ * , P * , c * , * , α * , ε * , β * , the optimal value is v *

[0127] 2.5) Update the optimal solution and optimal value, v [n] ←v * ,θ [n] ←θ * , P [n] ←P * , c [n] ←c * , [n] ←υ * , α [n] ←α * , ε [n] ←ε * , β [n] ←β *

[0128] 2.6) Until |v [n] -v [n-1] | <e

[0129] 2.7) Output v,θ,P,c.

[0130] in [n] Represents the nth iteration.

[0131] Figure 2 The convergence analysis of the cooperative rate splitting scheme based on the continuous convex approximation algorithm under different base station transmission powers is shown, d1 = 70m, d2 = 80m, d3 = 15m. The optimization target U2 security transmission rate and the sum of the public flow rate increase steadily with the number of iterations, and finally tend to be constant, which shows that the optimization algorithm based on continuous convex approximation is feasible. When the number of iterations increases from 1 to 6, the optimization target rises rapidly, and the rising trend gradually slows down from 7 to 10. After the number of iterations increases to 10, it remains almost unchanged, and the algorithm converges. In addition, when the transmission power increases from 10mW to 100mW, the sum of the U2 security transmission rate and the public flow rate increases from 8.0bit / s / Hz to 11.3bit / s / Hz.

[0132] Figure 3 The comparison of the achievable rates of the cooperative and non-cooperative schemes under different base station transmit power and number of antennas is shown, d1 = 70m, d2 = 80m, d3 = 12m. The base station transmit power is uniformly set from 10mW to 60mW, and the number of antennas is 3, 4, and 5 respectively. It can be seen that the performance of cooperative rate splitting is better than that of non-cooperative rate splitting when the base station transmit power and the number of antennas are changed. As the base station transmit power increases, the sum of the U2 security transmission rate and the public flow rate of the two strategies increases, and the performance gap between the two strategies continues to narrow. Relay cooperative transmission plays a good compensation role when the base station transmit power is small. In general, when the user acts as a relay, the power of the relay forwarding is smaller than the base station transmit power, which reduces the energy consumption of the system and improves energy efficiency. When the number of antennas increases, the sum of the U2 security transmission rate and the public flow rate of the two schemes is significantly improved. This is because as the antenna dimension increases, the scope of the solution during optimization increases, and the optimization effect on the transmission rate is better.

[0133] Figure 4 The comparison of direct transmission time slots of cooperative and non-cooperative schemes at different base station transmit power and number of antennas is shown. As the base station transmit power increases, the direct transmission time slots gradually increase, direct transmission becomes dominant, and the role of cooperative transmission decreases. As the number of antennas increases, the optimization effect of direct transmission is better.

[0134] Figure 5The comparison of the achievable rates of the cooperative and non-cooperative schemes at different user forwarding powers and user distances is shown. The distance d1 is changed to 75m, d3 is uniformly selected from 6m to 30m, and the user relay forwarding powers are 1mW, 2mW, and 4mW respectively. It can be seen that the cooperative rate splitting scheme performs better than the non-cooperative rate splitting scheme when the distance between users and the relay forwarding power are changed, which can improve the system performance. No matter how the distance between users and the user relay forwarding power are changed, the sum of the U2 security transmission rate and the public flow rate of the non-cooperative scheme remains unchanged. This is because the scheme does not adopt user relay cooperative forwarding, and the change of relay forwarding conditions will not affect its performance. For the cooperative rate splitting strategy, as the distance between users increases, the channel quality between users deteriorates, and the cooperative transmission effect deteriorates. Therefore, the sum of the U2 security transmission rate and the public flow rate that can be achieved by the scheme gradually decreases. As the user relay forwarding power increases, the cooperative rate splitting performance improves, which is because the public flow transmission rate of the cooperative transmission time slot increases due to the increase in relay forwarding power.

[0135] Figure 6 The comparison of direct transmission time slots of cooperative and non-cooperative schemes at different user forwarding powers and user distances is shown. It can be seen that as the distance between users increases, the direct transmission time slot gradually increases, that is, the direct transmission time slot becomes longer. This is because the channel quality between users deteriorates and the role of cooperative transmission deteriorates, so direct transmission occupies a dominant position. When the distance between users is too large, the cooperative rate splitting strategy will degenerate into a non-cooperative rate splitting strategy, that is, the direct transmission time slot approaches 1. In addition, when the user relay forwarding power increases, the role of cooperative transmission can be more fully exerted and the direct transmission time slot is reduced.

[0136] Figure 7 The influence of different base station transmission power and external eavesdropper distance on the external eavesdropper's eavesdropping rate is shown. d1 = 70m, d2 = 80m, d3 = 15m, and the distance from the eavesdropper to the base station is 10m, 50m, 100m, and 200m. After optimizing the precoding matrix and transmission time slot, the eavesdropper channel is randomly generated according to the Rayleigh distribution, and then the eavesdropping rate of U2's private message eavesdropped by the external eavesdropper is calculated according to equation (14). It can be seen from the figure that the eavesdropper's eavesdropping rate of U2's private message is very small compared with U2's private message rate. When the eavesdropper is closer to the base station than the legitimate user, the eavesdropping rate will not reach 0.4bit / s / Hz. This shows that the rate splitting strategy can hide private messages in public messages through power allocation, thereby ensuring the security of users in the system. In addition, the eavesdropping rate decreases with the increase of the distance from the eavesdropper to the base station.

[0137] The above-described embodiments merely express the implementation methods of the present invention, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for secure transmission of a cooperative rate-splitting network in the presence of an untrusted relay threat, characterized in that: The following steps are involved: The first step is to build a system model: 1) In a multiple-input single-output (MISO) downlink, there is a T A base station with 1 antenna and two single-antenna users (U1, U2), where N T ≥2, the distance between U1 and the base station is closer than that between U2 and the base station. U1, as an untrusted user relay, assists the base station in forwarding public messages to U2 by decoding and forwarding. In the direct transmission time slot, the base station transmits messages to the two users in a rate-splitting manner, and U1 eavesdrops on U2's private messages. In the cooperative transmission time slot, U1 forwards the public message to U2. 2) In this model, the channels from the base station to U1, the base station to U2, and U1 to U2 are represented as h1, h2, and h3, respectively, all of which are Rayleigh fading channels. According to the rate division principle, U k The message is divided into a public part W c,k and a private section W p,k , where k = 1, 2; W c,1 and W c,2 Combined to form a super public message W c , encoded into a public stream s by a public codebook that can be decoded by all users c , W p,1 and W p,2 Encoded into private streams s1 and s2 respectively; the precoding matrix is ​​P = [p c ,p1,p2], where p c is the public stream precoding vector, p1, p2 are private stream precoding vectors; signal stream s = [s c ,s1,s2] T Linear precoding is performed by the precoding matrix P, and the transmitted signal is: x=Ps=p c s c +p1s1+p2s2, (1) In which, assuming E[ss H ]=I, where E[·] represents the expectation and I is the unit matrix; The base station transmission power is given by tr(PP H )≤P t Limit, P t is the maximum power that the base station can transmit; 3) In the direct transmission time slot θ, the base station transmits at the maximum power P t Send messages to two users; in the cooperative transmission time slot 1-θ, U1 forwards the decoded and re-encoded public message to U2, and the relay forwarding power is P r ; During direct transmission, U k The received signal is: Among them, n k is additive white Gaussian noise, with a distribution mean of 0 and a variance of σ k 2 The complex Gaussian distribution of When decoding the public stream, all user private streams are considered as noise. k In decoding c The signal to interference plus noise ratio is: In the direct transmission time slot θ, U k Decode public streams c The rate is: When decoding the private stream, first use serial interference cancellation detection to separate the public message signal, and then treat the signal of another user as interference to decode the private stream; k In decoding k The signal-to-interference-noise ratio is: During the entire time slot, U k Decoding k The rate is: U1 can remove the common stream s through serial interference cancellation c Using U1's private stream s2 and its own private stream s1 to decode U2's private stream s2, the signal-to-noise ratio of U1 when eavesdropping on s2 is: During the entire direct transmission time slot, the eavesdropping rate of U1 decoding s2 is: The safe transmission rate of U2 is: where [x] + =max{x,0}. During the collaborative transmission process, relay user U1 re-encodes the public message and sends it to U2; the signal received by U2 is: Where n3 is additive white Gaussian noise, with a distribution mean of 0 and a variance of σ3 2 Complex Gaussian distribution; U2 in decoding s c The signal-to-noise ratio is: In the cooperative transmission time slot 1-θ, U2 decodes the public stream s c The rate is: During the entire time slot, U1 decodes the public stream s c The rate is U2 decoding public streams c The rate is To ensure that both users can successfully decode the public stream, the public message transmission rate needs to meet the following requirements: R c =min{R c,1 ,R c,2 }. (13) For possible external eavesdroppers, when they cannot decode public messages, public messages can be used as artificial noise to prevent them from eavesdropping on private messages of legitimate users. Since the base station allocates more power to public messages in the rate splitting strategy, private messages can be hidden in public messages to improve the security of private messages. At this time, the rate at which external eavesdroppers eavesdrop on U2 private messages is: Among them, h e is the channel information from the base station to the external eavesdropper, satisfying Rayleigh fading, σ e 2 is the power of additive white Gaussian noise; The second step is to determine the objective function and propose an optimization problem: The proposed optimization problem is as follows: s.t.c1+c2≤R c ,(15b) c1≥0,c2≥0,(15e) tr(PP H )≤P t ,(15g) 0<θ≤1,(15h)where u is R2 sec and R c The weight of R2 can be changed by adjusting u sec and R c The optimization importance of θ is the direct transmission time slot, P is the linear precoding matrix, c = [c1, c2], c k Represents U k Public message W c,k rate; Formula (15a) means to find the optimal θ, P and c to maximize the optimization objective; Formula (15b) is the rate limit of the public message part of the two users, and Formulas (15c) and (15d) guarantee the user's service quality, R1 th and R2 th is the service quality threshold of U1 and U2, formula (15e) represents that the rate of user in public message allocation is non-negative, and formula (15f) means that the private message transmission rate of U1 is greater than the decoding threshold r1 th , formula (15g) is the base station transmission power limit, and formula (15h) is the direct transmission time slot allocation limit; At this time, since the public stream must be decoded by all users, more U1 messages can be allocated to the public messages, which improves the fairness of the system without causing U2 information leakage; The third step is to design an algorithm to solve the optimization problem: The above optimization problem is solved by using the continuous convex approximation method; the variable ν = [ν p,1 ,ν p,2 ,ν c,1 ,ν c,2 ,ν e,1,2 ], ν p,k Indicates that in U k The rate of decoding private messages at c,k Indicates that in U k The rate at which public messages are decoded, ν e,1,2 represents the eavesdropping rate of U1 on U2's private message, and α=[α p,1 ,α p,2 ,α c,1 ,α c,2 ,α e,1,2 ],ε=[ε p,1 ,ε p,2 ,ε c,1 ,ε c,2 ,ε e,1,2 ] and β=[β p,1 ,β p,2 ,β c,1 ,β c,1 ], which represent the achievable rate, SINR, and noise plus interference excluding time slot θ, respectively; 1) Non-convex problem transformation By introducing the variable ν, the optimization objective is transformed into At the same time, the restriction conditions are also transformed, adding restriction R p,k ≥ν p,k , R e,1,2 ≤ν e,1,2 , R c,k ≥ν c,k ; The first restriction translates into: C [n] (will p,k )≥ν p,k , (16a) log2(1+ε p,k )≥a p,k ,(16c) Θ [n] (p k ,β p,k )≥ε p,k ,(16d) The second restriction translates into: F [n] (will e,1,2 )≤ν e,1,2 ,(17a) 1+e e,1,2 -P [n] (a e,1,2 )≤0,(17c) The third restriction translates into: C [n] (will c,1 )≥ν c,1 ,(18a) log2(1+ε c,k )≥a c,k ,(18c) I [n] (p k ,b c,k )≥e c,k ,(18d) Where k = 1, 2, is the achievable rate of the public flow in the cooperative transmission time slot; The non-convex optimization problem is transformed into the following convex optimization problem: s.t.c1+c2≤ν c,k ,k=1,2,(19b) (16a), (16b), (16c), (16d), (17a), (17b), (17c), (18a), (18b), (18c), (18d), (18e), (15e), (15g), (15h)(19f)2) Algorithm design based on continuous convex approximation 2.1) Initialize n←0, v [n] ←0,θ [n] , P [n] , c [n] , [n] , α [n] , ε [n] , β [n] , error tolerance e 2.2) Repeat 2.3)n←n+1 2.4) Substitute θ [n-1] , P [n-1] , c [n-1] , [n-1] , α [n-1] , ε [n-1] , β [n-1] Solve problem (19) and get the optimal solution θ * , P * , c * , * , α * , ε * , β * , the optimal value is v * 2.5) Update the optimal solution and optimal value, v [n] ←v * ,θ [n] ←θ * , P [n] ←P * , c [n] ←c * , [n] ←υ * , α [n] ←α * , e [n] ←e * ,b [n] ←b * 2.6) Until |v [n] -v [n-1] | <e 2.7) Output v, θ, P, c; In this algorithm, the algorithm based on continuous convex approximation can converge to a stable point; By jointly optimizing the precoding matrix, transmission time slot and rate allocation, the sum of the U2 secure transmission rate and the public message rate is maximized, thereby improving the security of more distant legitimate users in a collaborative rate-splitting network with the threat of untrusted relay users and preventing the information of more distant legitimate users from being leaked to untrusted relay users.

2. According to claim 1, a method for secure transmission of a cooperative rate-splitting network in the presence of an untrusted relay threat, characterized in that: In the first step 1), the U1 works in half-duplex HD mode.

Images