Secure transmission scheme design method in rate division multiple access system

Abstract

The invention discloses a method for designing a secure transmission scheme in a rate division multiple access system, which aims at a scene that part of messages sent to users need to be kept secret among the users, only non-confidential messages are transmitted in a public stream, and the non-confidential messages and confidential messages are transmitted in a time-sharing manner in a private stream. On the premise of ensuring the confidential message transmission rate, joint optimization is carried out on precoding vector sending, rate segmentation, private stream non-confidential and confidential message transmission duration distribution and the like of each message stream, and non-confidential message transmission and rate are maximized. In the optimization problem solving process, an original optimization problem is decomposed into two layers of optimization problems, and then the decomposed optimization problems are converted and solved through binary search, slack variables, continuous convex approximation and other methods. Simulation results show that compared with comparison schemes such as rate division multiple access, time-sharing space division multiple access and the like that private streams only transmit confidential messages, the scheme provided by the invention can obtain a higher non-confidential message transmission rate.

Description

Technical Field

[0001] The present invention relates to the field of information communication, and in particular to a design method for a secure transmission scheme in a rate division multiple access system. Background Art

[0002] One of the main goals of the development of mobile communication systems is to achieve higher data transmission rates and higher spectrum efficiency. Multiple access technology is one of the important factors affecting data transmission rates and spectrum efficiency [Mao Yijie, Dizdar O, Clerckx B, et al. Rate-splitting multiple access: Fundamentals, survey, and future research trends [J]. IEEE Communications Surveys & Tutorials, 2022, 24 (4): 2073-2126.]. The first to fifth generation cellular mobile communication systems all use orthogonal multiple access (OMA) technology. OMA allocates orthogonal wireless resources to users to avoid interference between users, but the number of users supported at the same time is limited by the available orthogonal resources. Non-orthogonal multiple access (NOMA) uses non-orthogonal superposition of the symbols of multiple users on the same time-frequency resource and then transmits them. It uses technologies such as successive interference cancellation (SIC) to eliminate interference between users [Liu Yuanwei, Qin Zhijin, Elkashlan M, et al. Nonorthogonal multiple access for 5G and beyond [J]. Proceedings of the IEEE, 2017, 105 (12): 2347-2381.], making resource scheduling more flexible and supporting more users to access the network at the same time. Space division multiple access (SDMA) uses the spatial freedom obtained by equipping the base station with multiple antennas to provide access services to multiple users using the same time-frequency resource. Since user channels are not completely orthogonal in space, SDMA also has interference between users. Unlike NOMA, SDMA treats interference as noise and does not use SIC.Rate-Splitting Multiple Access (RSMA) is a new multiple access technology proposed in recent years that can achieve higher spectrum efficiency and higher flexibility [Clerckx B, Mao Yijie, Jorswieck EA, et al. A primer on rate-splitting multiple access: Tutorial, myths, and frequently asked questions [J]. IEEE Journal on Selected Areas in Communications, 2023, 41 (5): 1265-1308.]. Compared with SDMA and NOMA, RSMA can better adapt to different levels of interference. RSMA divides the message sent to the user into a public part and a private part, and encodes them into a public stream and a private stream using different codebooks, and then linearly pre-codes them and superimposes them for transmission. At each receiving end, the private stream is first regarded as noise to decode the public stream message, and then the SIC technology is used to eliminate the public stream signal in the received signal, and then the private stream message is decoded. During decoding, the private stream signal sent to other users is regarded as interference. In RSMA, some interference is eliminated and other interference is regarded as noise. Flexible interference control and management are achieved by adjusting the rate and signal power allocation of public and private streams [Clerckx B, Mao Yijie, Schober R, et al. Rate-splitting unifying SDMA, OMA, NOMA, and multicasting in MISO broadcast channel: A simple two-user rate analysis [J]. IEEE Wireless Communications Letters, 2019, 9(3): 349-353.].

[0003] At present, there have been many studies on the application of RSMA in various communication systems. Generally, the base station precoding vector and rate splitting are optimized with the goal of improving the transmission rate. The focus of the research is how to solve more complex optimization problems, such as using semidefinite relaxation algorithms [[Liu Penglu, Li Yong, Cheng Wei, et al. Active intelligent reflecting surface aided RSMAfor millimeter-wave hybrid antenna array[J]. IEEE Transactions on Communications, 2023, 71(9): 5287-5302.], continuous convex approximation algorithms [Lei Weijia, Zhang Zhi, Lei Hongjiang, et al. Cooperative relay transmission strategy and optimization in rate-splitting multiple access systems[J]. Journal of Beijing University of Posts and Telecommunications, 2024, 47(02): 58-65.], and alternating iterative algorithms [Si ZW, Yin Longfei, Clerckx B. Rate-splitting multiple access for multigateway multibeam satellite systems with feeder link interference[J]. IEEE Transactions on Communications, 2022, 70(3): 2147-2162.], etc. These studies show that RSMA can be applied to a variety of communication systems, with high spectrum efficiency, good flexibility and low latency. In addition, when the channel state information (CSI) is incomplete or inaccurate, RSMA also has good robustness [Papazafeiropoulos A, Ratnarajah T. Rate-splitting robustness in multi-pairmassive MIMO relay systems[J]. IEEE Transactions on Wireless Communications, 2018, 17(8): 5623-5636.].

[0004] The physical layer security technology in RSMA system is also an important issue in RSMA research. In existing related research, two scenarios of confidential message transmission requirements are mainly considered: one is that part of the message to be transmitted is confidential, while the other part does not need to be confidential; the other is that all the transmitted messages need to be confidential. For the scenario where some messages are kept confidential, existing related literatures all adopt the scheme of transmitting non-confidential messages through public streams and only transmitting confidential messages through private streams, such as [Tong Yuqiao, Li Dongdong, Yang Zhutian, et al. Cooperative rate splitting secure transmission with an untrusted user relay[J]. IEEE Transactions on Vehicular Technology, 2022, 72(2): 2667-2671.], [Tong Yuqiao, Li Dongdong, Yang Zhutian, et al. Outage analysis of rate splitting networks with an untrusted user[J]. IEEE Transactions on Vehicular Technology, 2022, 72(2): 2626-2631.], [Xia Huiyuan, Mao Yijie, Clerckx B, et al. Weighted sum-rate maximization for rate-splitting multiple access based secure communication[C] / / IEEE Wireless Communications and Networking Conference, Austin, TX, USA, 2022: 19-24.], the confidentiality rate can be improved or the confidentiality interruption probability can be reduced by jointly optimizing the precoding matrix and the public flow rate division.For scenarios where all messages need to be kept confidential, existing literature adopts a scheme of allocating confidential messages to public and private streams for transmission, such as [Fu Hao, Feng Suili, Tang Weijun, et al. Robust secure beamforming design for two-user downlink MISO rate-splitting systems[J]. IEEE Transactions on Wireless Communications, 2020, 19(12): 8351-8365.], [Xia Huiyuan, Zhou Xiaokang, Han Shuai, et al. Security-reliability tradeoff in RSMA-based communications against eavesdropper collusion[J]. IEEE Wireless Communications Letters, 2023, 12(9): 1504-1507.], [Xia Huiyuan, Han Shuai, Li Cheng. Max-min fair optimization in RSMA-assisted secure communications with artificial noise[J]. IEEE Communications Letters, 2023, 27(12): 3181-3184.], the confidential transmission rate is improved by optimizing the base station precoding matrix and public rate splitting, and treating the public stream signal as artificial noise or introducing Gaussian artificial noise. Summary of the invention

[0005] The purpose of the present invention is to provide a design and optimization problem of a physical layer security transmission scheme in a rate division multiple access system, considering that some messages sent to users need to be kept confidential between users and other messages do not need to be kept confidential. The specific steps are as follows:

[0006] (1) Construct a communication system model: The system includes a multi-antenna base station and two single-antenna users. The messages sent to the two users are split into a public stream and a private stream. The messages that need to be kept confidential are transmitted in the private stream. The transmission of the private stream is divided into two stages, which transmit non-confidential messages and messages that need to be kept confidential between users respectively.

[0007] (2) Taking the maximization of the system non-confidential transmission rate as the optimization goal, an optimization mathematical model for the user non-confidential transmission rate is constructed under the constraints of each user's confidential message transmission rate, the user's minimum rate, and the total transmission power of the base station;

[0008] (3) Decompose the optimization problem into two-level optimization problems: solving the duration allocation and non-confidential transmission rate;

[0009] (4) Use binary search method to obtain the optimal duration allocation coefficient;

[0010] (5) When the duration allocation coefficient is given, the non-confidential transmission rate is solved using slack variables and continuous convex approximation.

[0011] The present invention also relates to a secure transmission system based on a rate division multiple access system, in which a computer program is stored, and when the program is executed, the above method is implemented to obtain the system's time allocation coefficient, non-confidential transmission and rate.

[0012] Compared with existing research, the system model of the present invention is different from the literature [Tong Yuqiao, Li Dongdong, Yang Zhutian, et al. Cooperative rate splitting secure transmission with an untrusted user relay [J]. IEEE Transactions on Vehicular Technology, 2022, 72 (2): 2667-2671.], [Tong Yuqiao, Li Dongdong, Yang Zhutian, et al. Outage analysis of rate splitting networks with an untrusted user [J]. IEEE Transactions on Vehicular Technology, 2022, 72 (2): 2626-2631.], but is similar to the literature [Xia Huiyuan, Mao Yijie, Clerckx B, et al. Weighted sum-rate maximization for rate-splitting multiple access based secure communication [C] / / IEEE Wireless Communications and Networking Conference, Austin, TX, USA, 2022: 19-24.]. The system of the present invention consists of a multi-antenna base station and two single-antenna users. The messages sent to the two users are split into a public stream and a private stream, and the messages that need to be kept confidential are transmitted in the private stream. However, unlike the literature [Xia Huiyuan, Mao Yijie, Clerckx B, et al. Weighted sum-rate maximization for rate-splitting multiple access based secure communication [C] / / IEEE Wireless Communications and Networking Conference, Austin, TX, USA, 2022: 19-24.], in the transmission scheme of the present invention, not all messages transmitted in the private stream need to be kept confidential. The present invention divides the transmission of the private stream into two stages, which transmit non-confidential messages and messages that need to be kept confidential between users.Under the condition of ensuring the transmission rate of confidential messages, the transmission and rate of non-confidential messages are maximized by optimizing the transmission precoding vectors, rate splitting, confidential and non-confidential transmission time allocation of public streams, confidential private streams and non-confidential private streams. Since the non-confidential messages in the scheme can be transmitted in both public and private streams, by scheduling the rate splitting of public stream messages, non-confidential messages in private streams and confidential messages, interference management can be performed more flexibly and better performance can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A communication system model of the present invention;

[0014] Figure 2 Algorithm convergence process;

[0015] Figure 3 How the transmission rate changes with the time-length distribution coefficient;

[0016] Figure 4 How non-confidentiality and rate vary with base station transmission power;

[0017] Figure 5 How the non-confidential and rate vary with the confidential transmission rate;

[0018] Figure 6 How non-confidentiality and rate vary with the number of base station antennas;

[0019] Figure 7 Transmission rate changes with error factor. DETAILED DESCRIPTION

[0020] The model of the two-user downlink RSMA secure transmission system studied in this invention is as follows: Figure 1 As shown in the figure, the system consists of a base station with N antennas and two single-antenna users. Since both are internal users of the system, the transmitter knows the CSI of the channel between it and the two users. and are the channel coefficient vectors between the base station and the two users respectively. The base station sends messages to two users at the same time, and some of the messages need to be kept secret from the other user. The message sent to user k (k = 1, 2) is composed of the message that needs to be kept secret W s,k and no need to keep secret message W k Composition, W k Further divided into public messages W c,k and private message W p,k Two parts. c,1 and W c,2 Combined into a public message W c , and encoded as a common stream signal s c , private message Wp,k and confidential message W s,k Encoded as signal s p,k and s,k . Assume that the duration of a complete transmission process is T, and the public stream signal s c Transmitted throughout the process, rather than keeping private signals confidential p,k and the signal s carrying the confidential message s,k Time-sharing transmission together constitutes the private stream signal in the RSMA system. In order to avoid excessive beamforming switching and reduce the complexity of system optimization, a time allocation scheme in which two users sample the same time is considered. The time allocation coefficient is θ, s p,k and s,k It is transmitted in two phases with durations of θT and (1-θ)T respectively, as Figure 1 Shown in the right half.

[0021] The transmission signal of the base station in the two stages can be expressed as

[0022]

[0023] The superscripts “[1]” and “[2]” represent the first and second stages, respectively. They are the public flow signals of phases 1 and 2 respectively. are the precoding vectors of the public stream signal, the non-confidential private signal of user 1 and user 2, and the precoding vectors of the confidential signal. The precoding matrix is ​​denoted as F = [f c ,f p,1 ,f p,2 ,f s,1 ,f s,2 ].

[0024] The received signal of user k (k = 1, 2) in the i-th (i = 1, 2) stage is

[0025]

[0026] in, is additive white Gaussian noise, represents the noise power of user k. The user first regards all private stream signals as noise, detects and decodes the public stream signals to obtain the public message. The Signal-to-Interference-plus-Noise Ratio (SINR) when user k decodes the public stream message in the two stages is

[0027]

[0028] The achievable rate at which user k can decode public stream messages is

[0029]

[0030] To ensure that all users can successfully decode the public stream, the transmission rate of the public stream message should be

[0031] R c =min{R c,1 ,R c,2}

[0032] The public stream message consists of the public messages of two users, so R c =c1+c2, where c k (k=1,2) represents the public message rate allocated to user k, and the rate partition vector is c=[c1,c2].

[0033] After completing the detection of the public stream signal, the two users each use the SIC technology to cancel the public stream signal in the received signal, and then regard the private stream signal sent to the other user as noise, and then detect and decode the private stream signal. The SINR when decoding the non-confidential private message and the confidential message at user k can be expressed as

[0034]

[0035] In the formula, when k = 1, When k = 2, γ j,k Where j is p and s is the SINR for decoding non-confidential and confidential private messages, respectively. The achievable rates for decoding non-confidential and confidential private messages at user k are

[0036] R p,k =θlog2(1+γ p,k )

[0037] R s,k =(1-θ)log2(1+γ s,k )

[0038] Assume user When trying to decode a secret message sent to user k, first cancel out the secret signal sent to him in the received signal In this way, the SINR when decoding the secret message of user k is

[0039]

[0040] Then the user The rate of eavesdropping on the confidential messages of user k is

[0041]

[0042] In this way, the achievable confidentiality rate of the base station sending confidential messages to user k is

[0043]

[0044] Where [x] + represents max{x,0}. The transmission rate of the non-confidential message of user k is R p,k +c k .

[0045] Considering that confidential messages are relatively more important in the system and the transmission rate is relatively fixed, when optimizing the system, the confidential message transmission rate of each user is prioritized, and on this basis, the non-confidential message transmission rate of the system is maximized. At the same time, in order to ensure the basic service quality of any user, it is also necessary to ensure that the non-confidential message transmission rate of each user cannot be lower than a certain lower limit during the optimization process. Based on the above considerations, the optimization problem of the present invention is:

[0046]

[0047] c1+c2≤R ck ,k=1,2(c)

[0048]

[0049] c k ≥0,k=1,2(e)

[0050] 0<θ<1(f)

[0051] The optimized variables are the duration allocation coefficient θ, the precoding matrix F (including the power of each signal) and the rate split vector c. Constraints (a) and (b) ensure that the transmission rate of confidential and non-confidential messages of user k is not less than and in is the confidentiality rate value of user k, is the minimum constraint value of the non-confidential message transmission rate of user k; constraint (c) ensures that every user can decode the public stream; (d) constrains the total transmission power not to exceed P t , P t is the base station transmit power.

[0052] The constraint (d) in the optimization problem contains the optimization variable θ, which is a non-convex constraint that is difficult to handle directly. First, θ is separated and the optimization problem is transformed into a two-level optimization problem, where the first level is to find the optimal θ, and the second level is to optimize other optimization variables when θ is given.

[0053] By observing the expressions of the transmission rates of private stream non-confidential and confidential messages, it is easy to see that, when other factors such as power allocation remain unchanged, the larger θ is, the greater the non-confidential message transmission rate is, and the smaller the confidential transmission rate is. Since the optimization problem is to maximize the non-confidential transmission rate under the premise of ensuring the confidential transmission rate, the optimal θ should be the maximum θ value that meets the confidential transmission rate requirements. The optimal θ can be searched using the binary search method. At each binary point of the search for θ, try to solve the second-level optimization problem. If the problem is solvable (that is, the confidential transmission rate requirements can be met under this θ), continue to increase θ; if the problem is unsolvable, it means that the confidential transmission rate can no longer be guaranteed and θ needs to be reduced. Repeat this process until the accuracy meets the preset requirements. For the detailed process of the search algorithm, please refer to Algorithm 1 and Algorithm 2 below. The second-level optimization problem and its solution are discussed below.

[0054] The second-level optimization problem is to jointly optimize the precoding matrix and the public flow rate allocation to maximize the non-confidential transmission and rate when θ is given. At this time, the problem is non-convex and needs to be transformed. First, the slack variable ν is introduced p,k ,ν sec,k and ν c,k They represent the lower bounds of the transmission rates of non-confidential, confidential, and public stream messages of user k’s private stream, respectively, and the optimization problem is transformed into

[0055]

[0056] c1+c2≤ν c,k ,k=1,2(c)

[0057] R p,k ≥ν p,k ,k=1,2(d)

[0058] R c,k ≥ν c,k ,k=1,2(e)

[0059]

[0060] Where ν is ν p,k ,ν sec,k ,ν c,k represents the set of slack variables that represent the upper or lower bounds of each transmission rate. Constraints (d), (e), and (f) in the above equation are still non-convex. Rate R p,k Determined by SINRγ p,k , introduce the slack variable β p,k As γ p,k The lower bound of Transform constraint (d) into

[0061]

[0062] Where (a) is still non-convex, and a relaxation variable ρ is introduced to represent the upper bound of the interference power plus noise power in the SINR p,k , which is converted into

[0063]

[0064] In the above formula, the left side of (a) is neither convex nor concave. The first-order Taylor expansion of the expression of the form gives the lower bound

[0065]

[0066] in, and is the Taylor expansion point, Indicates containing Its value determines the tightness of the lower bound. The optimization process needs to adopt an iterative method to update the result of the last iterative optimization. and Continuously improve the tightness of the lower bound and convert it to a convex constraint

[0067]

[0068] For the constraint R c,k ≥ν c,k , introduce the slack variable ν c,k,i and β c,k,i (k=1,2,i=1,2) are the lower bounds of the public stream message rate and corresponding SINR decoded by user k in the i-th stage, respectively.

[0069]

[0070] After the transformation, only constraint (b) is still non-convex. Introducing the slack variable ρ c,k,i Convert (b) to a convex constraint

[0071]

[0072] For non-convex constraints Introducing the slack variable ν s,k and to represent the lower bound of the confidential message transmission rate in user k’s private flow and the user The upper bound of the eavesdropping rate of eavesdropping on the confidential message of user k is obtained

[0073]

[0074] Among them, (b) and (c) are still non-convex. By introducing the slack variable β s,k and ρ s,k , convert (b) into a convex constraint

[0075]

[0076] For non-convex constraints By introducing representative users Slack variables of the SINR upper bound when eavesdropping on the confidential message of user k First convert it into

[0077]

[0078] Where (b) is still non-convex, the expression on the right side of the expression is replaced by its first-order Taylor expansion to obtain the lower bound and converted into a convex constraint

[0079]

[0080] In the formula is the Taylor expansion point, which needs to be iteratively updated during optimization. So far, the non-convex constraint has been converted to a convex constraint.

[0081] At this point, the objective function and constraints in the original optimization problem have been transformed into convex ones, and the final convex problem is

[0082]

[0083] Where β represents the slack variable β p,k ,β s,k ,β c,k,i , The set of components, ρ represents the slack variable ρ p,k ,ρ s,k ,ρ c,k,i The optimization problem can be solved using algorithms for solving convex problems, or using convex problem solvers such as the CVX toolbox.

[0084] Algorithm 1: Algorithm for solving optimization problems

[0085] (1) Initialize the number of iterations n = 0, the convergence factor ε,t {0} ,f c {0} ,

[0086] (2) while

[0087] (3)n=n+1

[0088] (4) The optimization problem and Set to and Solve the problem and get the optimal solution fc * ,, and t *

[0089] (5) Update t {n} = t *

[0090] (6)

[0091] (7) Output: t * , f c * , and

[0092] Optimized Solution of the Duration Allocation Coefficient θ of Algorithm 2

[0093] (1) l < θ < u, initialize l = 0, u = 1, convergence factor δ

[0094] (2) while

[0095] (3) θ = (l + u) / 2

[0096] (4) Call Algorithm 1 to solve the optimization problem of the second layer

[0097] (5) If the problem has a feasible solution under the current θ, then l = θ; otherwise u = θ

[0098] (6) until u - l ≤ δ

[0099] (7) Output the optimal θ * = θ

[0100] The performance of the proposed scheme of the present invention is verified by simulation. In the simulation, all channels are Rayleigh flat fading channels, and the channel fading includes path loss (large-scale fading) and small-scale fading. The channel coefficient vector from the base station to user k is where ζ = 10 -4 is the channel gain at a reference distance of 1m, d1 = d2 = 100m are the distances between the base station and user 1 and user 2 respectively, η = 3.3 is the path loss exponent, is the small-scale fading coefficient, and its elements are independent complex Gaussian random variables with a mean of 0 and a variance of 1; the noise variance The minimum value of the non-secret message transmission rate is The secrecy message transmission rate requirements of the two users are the same, that is When there is no special instruction, the transmit power P of the base station t=30 dBm, the number of base station antennas N = 4. The convergence factors ε and δ of Algorithm 1 and Algorithm 2 are both set to 10 -3 .

[0101] The present invention simulates the performance of two comparative schemes for comparison. Comparative Scheme 1: No rate division is performed, and the common stream precoding vector f is set c =0, non-confidential and confidential messages are transmitted in two time slots respectively, and an algorithm similar to the present invention is used to optimize the duration allocation coefficient, the precoding vector of confidential and non-confidential messages. Comparison Scheme 2: Similar to the scheme proposed in the literature [XiaHuiyuan, Mao Yijie, Clerckx B, et al. Weighted sum-rate maximization for rate-splitting multiple access based secure communication [C] / / IEEE Wireless Communications and Networking Conference, Austin, TX, USA, 2022: 19-24.], all non-confidential messages are carried by public stream signals, while messages transmitted in private streams are all confidential. This can be regarded as a special case of the present invention when θ = 0. The precoding vector and rate splitting are optimized using the above-mentioned literature 0 scheme.

[0102] Figure 2 The convergence process of Algorithm 2 at θ = 0.5 under different confidentiality transmission rate requirements and two sets of random channel samples is given. It can be seen that the non-confidentiality and rate increase rapidly in the early stage of iterative optimization and basically converge after 3 to 4 iterative optimizations.

[0103] Figure 3 The following table shows how the non-confidential sum rate and confidential transmission rate change with θ under different base station transmit powers under the same channel sample. The black solid line in the figure is the non-confidential sum rate, and the blue dotted line is the confidential rate of user 1. Since the confidential rate value and change of user 2 are exactly the same as those of user 1, they are not shown in the figure. In the simulation, θ increases with a step size of 0.1, and the non-confidential sum rate increases monotonically with θ, which verifies the monotonic relationship between the non-confidential sum rate and θ. The confidential rate value always maintains the constraint value of 0.6 bit / (s·Hz) when the optimization problem is solvable. The base station transmit power P t =36, 33 and 30 dBm, the confidentiality transmission rate constraint can no longer be met after θ increases to 0.9, 0.8 and 0.7, and the layer 2 optimization problem has no solution. The higher the base station transmission power, the higher the power that can be allocated to the confidentiality message signal, and the shorter the confidentiality message transmission time required. Therefore, the private flow non-confidential message transmission time can be longer, that is, a larger θ can be taken.

[0104] Figure 4 It shows that the non-confidentiality and rate of all schemes increase with the increase of base station transmission power. The non-confidential transmission rate of comparative scheme 2 is significantly lower than that of the scheme of the present invention and comparative scheme 1, and the gap increases with the increase of transmission power. This is because the non-confidential message of comparative scheme 2 can only be transmitted in the public stream, and there is interference formed by the private stream signals of two users when the user is detecting, and the precoding of the public stream needs to take into account both users; in the scheme of the present invention, some non-confidential messages are transmitted in the private stream, different users can use different precoding, and only the private stream signal of another user will be interfered with during detection, so a higher non-confidential message transmission rate can be obtained than scheme 2. The confidential message and non-confidential message of comparative scheme 1 are time-sharing transmission, so compared with comparative scheme 2, the interference during the transmission of non-confidential messages is lower, and the non-confidential transmission rate is also higher than that of comparative scheme 2. Further, when the transmission power increases, the transmission time allocated to the confidential message in the private stream of comparative scheme 1 and the scheme of the present invention can be shortened, and the transmission time and power of the non-confidential message are increased, while the transmission power of comparative scheme 2 only increases, so the growth rate of non-confidential messages of comparative scheme 1 and the scheme of the present invention is higher than that of scheme 2. There is no public flow in Comparative Scheme 1, which can be regarded as an extreme case where the public flow signal power of the scheme of the present invention is 0. The scheme of the present invention can optimize the power and transmission rate allocation of the public flow signal, and the interference management is more effective. Therefore, the non-confidential rate is slightly higher than that of Comparative Scheme 1.

[0105] Figure 5 Given the non-confidential and rate along with the confidential transmission rate requirement value R s,th When a higher secure transmission rate value is set, more power and transmission time need to be allocated to the signal carrying the confidential message, and the power and transmission time of the signal carrying the non-confidential message will be reduced accordingly. Therefore, the non-confidentiality and rate of all schemes decrease as the confidentiality transmission rate requirement increases. In addition, as the confidentiality transmission rate requirement value increases, the difference between the confidentiality transmission rate of comparison scheme 2 and other schemes decreases. This is because when the confidentiality transmission rate requirement increases, the transmission time and power of the non-confidential message signal in the scheme of the present invention and comparison scheme 1 are reduced, while only the transmission power is reduced in comparison scheme 2.

[0106] Figure 6The variation of non-confidentiality and rate with the number of base station antennas is given. The more transmitting antennas, the greater the spatial freedom of the transmitter, the better the control of beamforming, the shorter the transmission power and transmission time required to meet the transmission rate of confidential messages, and thus the longer time and higher power to transmit non-confidential messages. At the same time, the more antennas there are, the better the mutual interference between the signals of each message stream can be suppressed, so the transmission rate of non-confidential messages of all schemes increases. Comparative Scheme 2 cannot adjust the transmission time of confidential and non-confidential messages, and can only increase the transmission rate of non-confidential messages by increasing the SINR brought about by the increase of transmitting antennas, while the scheme of the present invention and Comparative Scheme 1 can also increase the transmission rate of non-confidential messages by increasing the transmission time of non-confidential messages, so the growth is faster.

[0107] In order to verify the robustness of the scheme in the presence of CSI errors, the confidential and non-confidential transmission rates of each scheme are simulated when there are errors in the small-scale fading coefficients in the user channel coefficients. The mathematical model of the channel error is where λ is the error factor, and They represent the estimated value and estimation error of the small-scale fading coefficient, respectively, and are independent and identically distributed complex Gaussian random variables with mean 0 and variance 1. The transmitter is optimized based on the obtained erroneous channel coefficients, so the optimization result is not optimal under the actual channel, and the system performance will be reduced, and the degree of reduction is proportional to the channel coefficient error factor. Figure 7 The average values ​​of the confidentiality rate and the non-confidentiality sum rate are given as the error factor changes. The black dashed line in the figure is the confidentiality rate, and the blue solid line is the non-confidentiality sum rate. Since the channel statistical characteristics of the two users are the same, the simulation results of the confidentiality rate average value are also the same. Figure 7Only the simulation results of the confidentiality rate of user 1 are given in the figure. Compared with the non-confidential sum rate when there is no error (i.e., λ=0), when λ=0.2, the non-confidential sum rate of the scheme of the present invention, comparative scheme 1 and comparative scheme 2 decreases by 15.2%, 16.6% and 16.1% respectively. The decrease degree of the scheme of the present invention and comparative scheme 2 using RSMA is slightly lower than that of comparative scheme 1 using SDMA, indicating that the robustness of RSMA is better than that of SDMA. In addition, when there is no error in the obtained channel coefficient, the achievable confidentiality rate values ​​of all schemes can reach the required 0.1 bit / (s·Hz); when there is an error in the channel coefficient, the achievable confidentiality rate value decreases, and the decrease degree of the scheme of the present invention and comparative scheme 2 is significantly lower than that of comparative scheme 1, further indicating that RSMA has better robustness. In addition, when there is an error in the channel coefficient, the achievable confidentiality rate value of the scheme of the present invention is lower than that of comparative scheme 2, because the scheme of the present invention optimizes more variables than comparative scheme 2 and is more affected by the channel coefficient error. However, the non-confidential sum rate value of the scheme of the present invention is always higher than that of comparative scheme 2, indicating that the performance of the scheme of the present invention is better.

Claims

1. A method for designing a secure transmission scheme in a rate division multiple access system, characterized in that: The following steps are involved: (1) Construct a communication system model: The system includes a multi-antenna base station and two single-antenna users. The messages sent to the two users are split into a public stream and a private stream. The messages that need to be kept confidential are transmitted in the private stream. The transmission of the private stream is divided into two stages, which transmit non-confidential messages and messages that need to be kept confidential between users respectively. (2) Taking the maximization of the system non-confidential transmission and rate as the optimization goal, an optimization mathematical model for the user non-confidential transmission and rate is constructed under the constraints of each user's confidential message transmission rate, the user's minimum rate, and the total transmission power of the base station; (3) Decompose the optimization problem into two-level optimization problems: solving the duration allocation and non-confidential transmission rate; (4) Use binary search method to obtain the optimal duration allocation coefficient; (5) When the duration allocation coefficient is given, the non-confidential transmission rate is solved using slack variables and continuous convex approximation.

2. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 1, characterized in that: The communication system model in step (1) is a multiple-input single-output system in which a base station is equipped with N antennas and two users at the receiving end are each equipped with a single antenna. The base station knows the channel state information of the channel between the base station and the two users. and are the channel coefficient vectors between the base station and the two users respectively. The base station sends messages to the two users at the same time, and some of the messages need to be kept secret from the other user. Suppose the duration of a complete transmission process is T, and the public stream signal s c During the whole process, non-confidential private signals are transmitted p,k (k=1,2) and the signal s carrying the confidential message s,k Time-sharing transmission, together forming the private stream signal in the rate division multiple access system, the duration allocation coefficient is θ, s p,k and s,k It is transmitted in two phases with durations of θT and (1-θ)T respectively.

3. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 1 or 2, characterized in that: The transmission signals of the base stations in the two stages of the communication system model are expressed as The superscripts "[1]" and "[2]" represent stage 1 and stage 2, respectively. are the common flow signals of the 1st and 2nd stages, respectively, and f c ,f p,1 ,f p,2 ,f s,1 ,f s,2 They are the precoding vectors of the public stream signal, the precoding vectors of the non-confidential private signals of user 1 and user 2, and the precoding vectors of the confidential signals of user 1 and user 2. The precoding matrix is ​​denoted as F = [f c ,f p,1 ,f p,2 ,f s,1 ,f s,2 ]; The received signal of user k (k = 1, 2) in the i-th (i = 1, 2) stage is in, is additive white Gaussian noise, represents the noise power of user k.

4. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 3, characterized in that: The step (2) includes modeling the signal-to-interference-noise ratio and transmission rate when user k decodes public stream messages, private stream confidential and non-confidential messages in two stages, specifically: The signal-to-noise ratio when user k decodes the public stream message in two stages is The achievable rate at which user k can decode public stream messages is To ensure that all users can successfully decode the public stream, the transmission rate of the public stream message is R c =min{R c,1 ,R c,2 } The public stream message consists of the public messages of two users, so R c =c1+c2, where c k (k=1,2) represents the public message rate assigned to user k, and the rate partition vector c=[c1,c2]; After completing the detection of the public stream signal, the two users each use the serial interference cancellation technology to cancel the public stream signal in the received signal, and then regard the private stream signal sent to the other user as noise, and then detect and decode the private stream signal; the signal-to-noise ratio when decoding the non-confidential private message and the confidential message at user k is expressed as In the formula, when k = 1, When k = 2, γ j,k Where j is p and s is the signal-to-interference-noise ratio when decoding non-confidential and confidential private messages, respectively.

5. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 4, characterized in that: It also includes the confidential transmission rate and the transmission rate of non-confidential messages of user k, specifically including: The achievable rates for decoding non-confidential and confidential private messages at user k are R p,k =θlog2(1+γ p,k ) R s,k =(1-θ)log2(1+γ s,k ) user When trying to decode a secret message sent to user k, first cancel out the secret signal sent to him in the received signal In this way, the signal-to-interference-noise ratio when decoding the secret message of user k is Then the user The rate of eavesdropping on the confidential messages of user k is In this way, the achievable confidentiality rate of the base station sending confidential messages to user k is Where [x] + represents max{x,0}, and the transmission rate of the non-confidential message of user k is R p,k +c k .

6. The method for designing a secure transmission scheme in a rate division multiple access system according to any one of claims 1 to 5, characterized in that: The optimization mathematical model in step (2) is specifically: The optimized variables are the duration allocation coefficient θ, the precoding matrix F, and the rate split vector c. The constraints (a) and (b) ensure that the transmission rate of confidential and non-confidential messages of user k is not less than is the confidentiality rate value of user k, is the minimum constraint value of the non-confidential message transmission rate of user k; constraint (c) ensures that every user can decode the public stream; (d) constrains the total transmission power not to exceed P t , P t is the base station transmit power.

7. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 6, characterized in that: In step (3), the optimization problem is decomposed into two layers of optimization problems for solving duration allocation and non-confidential transmission and rate. Specifically, the first layer is to find the optimal θ, and the second layer is to optimize other optimization variables when θ is given.

8. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 1 or 7, characterized in that: The binary search method in step (4) specifically includes: The optimal θ is obtained by searching using the binary search method. At each binary point of θ, an attempt is made to solve the second-level optimization problem. If the problem is solvable, that is, the confidentiality transmission rate requirement can be met under this θ, θ is increased continuously. If the problem is not solvable, θ needs to be decreased. This process is repeated until the accuracy reaches the preset requirement.

9. The method for designing a secure transmission scheme in a rate division multiple access system according to claim 8, characterized in that: In the step (5), the second layer optimization problem of the non-confidential transmission and rate is solved by using the method of slack variables and continuous convex approximation, which specifically includes: the second layer optimization problem is to jointly optimize the precoding matrix and the common flow rate allocation when θ is given, and to maximize the non-confidential transmission and rate by using the method of slack variables and continuous convex approximation.

10. A secure transmission system based on a rate division multiple access system, characterized in that: The secure transmission scheme and algorithm in the rate division multiple access system proposed in the present invention are verified and performance evaluated by means of computer program simulation. When the program is executed, the method described in any one of claims 1 to 9 is implemented, thereby obtaining the system's time allocation coefficient, non-confidential transmission and rate.

Images