A method and system for improving security and robustness of RSMA system with non-circular signals

Abstract

The application discloses a method and system for improving the security and robustness of an RSMA system by using a non-circular signal, and specifically comprises the following steps: a base station divides the message of each user by rate according to the channel state information of the user to obtain a private message of each user; a public message is constructed by mapping an initial circular signal to a non-circular constellation diagram, and the private message of each user is mapped to a traditional circular Gaussian signal; the public signal and the private signal of each user are superimposed and coded, and the public stream transmission power and the private stream transmission power are allocated to form a transmission signal; after receiving the signal, a user decodes the public message, calculates the reachable rate of the public stream of the RSMA system, then decodes the private message of the user, and calculates the reachable rate of the private stream of the user k; and the non-circularity coefficient, the public stream power and the private stream power are jointly optimized with the system performance index as the target. The application provides a new dimension for non-ideal factors such as imperfect SIC, and enhances the robustness and reliability of the system.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a method for improving the security and robustness of RSMA systems using non-circular signals. Background Technology

[0002] RSMA (Rate-Splitting Multiple Access) is a novel multiple access technology for 5G and future communication systems. It aims to address the performance bottlenecks of traditional multiple access technologies in complex scenarios (such as high-density users, heterogeneous services, and non-ideal channels). Its core principle is flexible multi-user interference management through signal segmentation. Signal segmentation refers to dividing user messages into two parts: common messages and private messages. Common messages are decoded by all users and then removed from the total message using serial interference cancellation technology, while private messages are decoded only by the target user.

[0003] In real-world communication scenarios, incomplete serial interference cancellation (SIC) is often caused by hardware defects, channel estimation errors, and decoding errors. Although recent research has begun to incorporate imperfect SIC into RSMA analysis, the focus remains primarily on achievable rates and throughput.

[0004] Furthermore, the security aspects of RSMA have attracted considerable attention. Increasing the power allocated to the public stream typically improves the overall rate, but at the cost of exposing more user messages, making private messages more susceptible to eavesdropping and thus reducing the confidentiality rate. Conversely, allocating more power to the private stream can enhance confidentiality, but may reduce the overall rate due to insufficient power in the public stream. This trade-off requires a balanced design. However, most existing research relies on the conventional Appropriate Gaussian Signal (PGS), which has limited freedom in managing the complex interaction between imperfect SIC and confidentiality constraints. Summary of the Invention

[0005] Purpose of the invention: In order to solve the problems existing in the prior art, the present invention provides a method for improving the security and robustness of RSMA system using non-circular signals.

[0006] Technical solution: This invention provides a method for improving the security and robustness of RSMA systems using non-circular signals, specifically as follows:

[0007] The base station performs rate segmentation on each user's message based on the user's channel state information to obtain each user's private message;

[0008] A common message is constructed by mapping the initial circular signal through a non-circular constellation diagram. The public message has a non-circularity coefficient. Mapping each user's private messages to traditional circular Gaussian signals , Index for users;

[0009] Public signals Private signals of each user Perform overlay coding and allocate common stream transmit power. and private stream transmit power This generates a transmitted signal x.

[0010] After receiving the signal, the user decodes the common message and calculates the reachable rate of the common flow in the RSMA system. Then, decode its own private message and calculate the achievable rate of user k's private stream. ;

[0011] With system performance indicators as the objective, the non-circularity coefficient is jointly optimized. Common stream transmit power and the transmit power of each private current .

[0012] Furthermore, public information The expression is:

[0013] ;

[0014] Where s is the original circular Gaussian signal from the M-QAM constellation, To control the constellation diagram parameters for non-circularity, , To control the rotation angle of the constellation's minimum Euclidean distance, ;

[0015] The non-circularity coefficient This is the ratio of the absolute value of the complementary variance to the variance of the signal, and .

[0016] Furthermore, the expression for the transmitted signal is:

[0017] ;

[0018] Where K is the total number of users.

[0019] Furthermore, the signal received by user k is ;in For users The channel coefficient, It is additive white Gaussian noise;

[0020] If user k treats the private message as interference and decodes the public message, then the received signal of user k is rewritten as: ;

[0021] in, For interference and noise, The reachable rate of the common flow for user k is obtained. The expression is:

[0022] ;

[0023] in, For user k, the channel noise ratio is... ,in The average power of the noise;

[0024] The system's common flow reachable rate: ;

[0025] After user k decodes and deletes the public message, user k decodes its own private message. At this time, the signal received by user k is:

[0026] ;

[0027] in, This is the sum of residual common interference, interference from other users, and noise. ; The SIC imperfection coefficient;

[0028] The achievable rate of user k's private flow The expression is:

[0029] ;

[0030] Where K is the total number of users.

[0031] Furthermore, with the goal of maximizing system performance and speed, the non-circularity coefficients are jointly optimized. Common stream transmit power and the transmit power of each private stream :

[0032] ;

[0033] in, For system and rate, Let K be the minimum achievable rate of the private stream among K users. Where P is the minimum common current power and P is the total power;

[0034] The soft actor-critic reinforcement learning algorithm is used to solve the problem.

[0035] Furthermore, the reward function when using the soft actor-critic reinforcement learning algorithm is:

[0036] ;

[0037] in, A positive penalty coefficient, To constrain violations of indicator functions, if any but It is 1 if it is 1, otherwise it is 0. Indicates state, For action.

[0038] Furthermore, if there are only two users, the non-circularity coefficient can be jointly optimized with the goal of maximizing the reachability of the private flow. Common current power and the power of each private current ,:

[0039] ;

[0040] in, Let K be the private stream reachable rate for user k, and K be the total number of users. Let P be the minimum common current power, and P be the total power.

[0041] Furthermore, if there are only two users, the non-circularity coefficient can be jointly optimized with the goal of maximizing the reachability rate of the common flow. Common stream transmit power and the transmit power of each private stream :

[0042] ;

[0043] in, Let K be the minimum achievable rate of the private stream among K users. Where P is the minimum common current power and P is the total power;

[0044] Calculate the common flow transmit power based on the constraints. and relational formula :

[0045] ;

[0046] in,

[0047] ;

[0048] ; ; in, This represents the user ID corresponding to the minimum reachable rate of the private stream, where j represents the j-th user. ;

[0049] Calculate separately and Common stream transmit power at time and ;like or If no feasible solution exists under the current power allocation constraint, the process terminates.

[0050] like The subscripts 1 and 2 represent user 1 and user 2, respectively; in this case, the larger one should be selected if possible. However, it is necessary to ensure Not exceeding the available power budget ,therefore,

[0051] like The optimal solution is ;

[0052] like Then the optimal non-circularity The expression is: ;

[0053] like Then you should choose the smallest possible value. The optimal solution is .

[0054] A system for improving the security and robustness of RSMA systems using non-circular signals includes:

[0055] The rate segmentation module is used to split user messages into private and public parts based on channel information at the base station; the non-circular signal generation module constructs a public message from an initial circular signal through a non-circular constellation diagram mapping. The public message has a non-circularity coefficient. ;

[0056] Precoding and power distribution module, common signal Private signals of each user Perform overlay coding and allocate common stream transmit power. and private stream transmit power This generates a transmitted signal x.

[0057] The superposition encoding module is used to superimpose the power-scaled signals to form the transmitted signal;

[0058] The decoding module decodes the common message after the user receives the signal and calculates the reachable rate of the common flow in the RSMA system. Then, decode its own private message and calculate the achievable rate of user k's private stream. ;

[0059] The joint optimization module, with system performance indicators as the objective, jointly optimizes the non-circularity coefficient. Common stream transmit power and the transmit power of each private stream .

[0060] A computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for improving the security and robustness of an RSMA system using non-circular signals, as described above.

[0061] Beneficial Effects: By introducing non-circular Gaussian signals into the RSMA framework and jointly optimizing their non-circularity and system power resources, this invention brings the following significant beneficial effects:

[0062] 1. Enhanced security performance: When imperfect SICs exist, using the maximum non-circular signal can effectively offset the influence of residual common interference and make it meet certain private flow rate requirements, thereby enhancing the physical layer security of information.

[0063] 2. Flexible trade-offs in system performance: Under the constraint of ensuring the minimum safe rate of the private flow, the non-circularity is adjusted. This allows for a flexible balance between increasing the rate of public flows and minimizing interference with private flows. (Optimized version) The value can be different from 0 or 1, thus achieving a performance trade-off that traditional circular signals cannot reach.

[0064] 3. Improve system efficiency and speed: By using reinforcement learning algorithms to perform global joint optimization of non-circularity and power, complex non-convex constraints can be effectively handled, and an approximate optimal solution for the overall system speed can be found, thereby improving the overall spectral efficiency.

[0065] 4. Enhanced system robustness: The design of non-circular signals provides a new dimension for the system to resist interference and cope with non-ideal factors such as imperfect SiC, thereby enhancing the robustness and reliability of the system. Attached Figure Description

[0066] Figure 1 This is a schematic diagram of a non-circular constellation mapping in an embodiment of the present invention, where (a) is the original 32QAM constellation diagram, and (b) is... , For the non-circular constellation diagram below, (c) the minimum distance varies Change curve graph.

[0067] Figure 2 The diagram shows an RSMA system model based on a non-circular Gaussian signal, provided for an embodiment of the present invention. Figure 3 This is a flowchart of the soft actor-critic algorithm in an embodiment of the present invention.

[0068] Figure 4This is a graph showing the reachable rate of a private flow as a function of SNR.

[0069] Figure 5 shows the variation of reachability rate of public flow with non-circularity.

[0070] Figure 6 shows the different The graph shows the variation of the sum and rate with SNR, where (a) is... The graph shows the variation of the rate with SNR under the given conditions, and (b) is... The graph shows the variation of the rate with SNR under certain conditions.

[0071] Figure 7 The graph shows the sum of gains and rate versus SNR for different channel gain ratios, where (a) is... The graph shows the variation of the rate with SNR under the given conditions, and (b) is... The graph shows the variation of the rate with SNR under certain conditions.

[0072] Figure 8 For different The optimal non-circularity varies with SNR (comparison under different imperfection SIC coefficients), where (a) is... The optimal non-circularity as a function of SNR is shown in (b). The optimal non-circularity varies with SNR under the given condition.

[0073] Figure 9 The graph shows the optimal non-circularity as a function of SNR under different channel gain ratios (comparison under different imperfect SIC coefficients), where (a) is... The optimal non-circularity as a function of SNR is shown in (b). The optimal non-circularity varies with SNR under the given condition. Detailed Implementation

[0074] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0075] This invention provides a rate division multiple access transmission method based on non-circular Gaussian signals, applied to a communication system including one base station and at least two users, the method comprising:

[0076] Step S1: The base station performs rate segmentation on each user's message based on the user's channel state information to obtain each user's private message and a public message shared by all users.

[0077] Step S2: Construct the common message from the initial circular signal through a non-circular constellation diagram mapping. It has a non-circularity coefficient Mapping each user's private messages to traditional circular Gaussian signals ,in Index for users.

[0078] The public message is: Where s is the original circular Gaussian signal from a uniformly distributed M-QAM constellation, Indicates the conjugate of the s-signal. To control the parameters of non-roundness, The rotation angle for controlling the minimum Euclidean distance of the constellation; the non-circularity coefficient Defined as the ratio of the absolute value of the complementary variance of the signals to the variance, and .

[0079] Step S3, the base station will transmit public signals Private signals of each user Perform overlay coding and allocate transmit power. and This generates a transmission signal.

[0080] ;

[0081] Where K is the total number of users.

[0082] Step S4: After receiving the signal, the user terminal first decodes the public message; performs imperfect continuous interference cancellation to remove some interference from the public signal, and then decodes the respective private messages.

[0083] The signal received by user k is ;in For users The channel coefficient, It is additive white Gaussian noise;

[0084] If user k treats the private message as interference and decodes the public message, then the received signal of user k is rewritten as: ;

[0085] in, For interference and noise, The reachable rate of the common flow for user k is obtained. The expression is:

[0086] ;in, For user k, the signal-to-noise ratio is... ,in The average power of the noise;

[0087] The system's common flow reachable rate: ;

[0088] After user k decodes and deletes the public message, user k decodes its own private message. At this time, the signal received by user k is:

[0089] ;

[0090] in, This is the sum of residual common interference, interference from other users, and noise. ;

[0091] Obtain the achievable rate of the private stream for user k. The expression is:

[0092]

[0093] Step S5: Optimize the non-circularity coefficient jointly, with system performance indicators as the target. Common current power and the power of each private current .

[0094] The system performance indicators and optimization problems mentioned in step S5 include the following three modes:

[0095] Mode 1 (applicable to two users): Maximize the security and rate of private flows. The optimization problem is to maximize the sum of the private rates of all users under the constraints of total power and minimum power of the common flow. The optimization problem is as follows:

[0096] ;

[0097] In this mode, the optimal non-circularity coefficient The maximum value is 1.

[0098] Mode 2 (applicable to two users): Maximize the common flow rate. The optimization problem is to ensure that the total power constraint, the minimum power constraint of the common flow, and the safe flow rate of each private flow is not lower than a threshold are met. Maximize the common rate under constraints By analyzing the tightness of constraints and common rates about The monotonicity of the coefficients yields the optimal non-circularity coefficient. The analytical solution or piecewise function expression. The optimization problem is as follows:

[0099] ;

[0100] in, It represents the minimum achievable rate of the private stream among K users; For minimum common current power , This represents the total power.

[0101] Mode 3 (applicable to two or more users): Maximize system performance and rate. The optimization problem is to ensure that the total power is constrained, the minimum power of the common flow is constrained, and the safe rate of each private flow does not fall below a threshold. Maximize system and speed under constraints The optimization problem is as follows:

[0102] ;

[0103] Since the problem is non-convex, a soft actor-commentator (SAC) algorithm based on reinforcement learning is used for solving it. The state space of the algorithm includes system states such as channel parameters, noise power, and total power, while the action space includes... , , , The design of the reward function is related to the system, rate, and degree of constraint violation.

[0104] This invention provides a rate-division multiple access transmission system based on non-circular Gaussian signals, used to implement the method described in the first aspect, the system comprising: The rate segmentation module is used to split user messages into private and public parts based on channel information at the base station.

[0105] The non-circular signal generation module is used to generate signals based on a given non-circularity coefficient. and optimized rotation angle The symbols of public messages are mapped to generate non-circular Gaussian constellation points.

[0106] The precoding and power allocation module is used to apply the power allocation coefficients obtained from the optimization algorithm. and Power scaling is applied to both public and private signals.

[0107] The superposition encoding module is used to superimpose the power-scaled signals to form the transmitted signal.

[0108] The joint optimization module is used to run the optimization algorithm and solve for the optimal non-circularity coefficient. Common current power and private flow power .

[0109] The receiving and decoding module, located on the user side, is used to perform decoding of public messages, imperfect interference cancellation, and decoding of private messages.

[0110] This invention embodiment considers a downlink single-base station, two-user SISO-RSMA system, such as... Figure 2As shown. The base station is equipped with a single antenna, and each user is equipped with a single antenna. The channel exhibits quasi-static flat fading, and the base station has a known perfect CSI.

[0111] Generation of non-circular Gaussian signals:

[0112] Non-circular Gaussian signal Through formula Generate. For example... Figure 1 As shown, taking a 32-QAM constellation as an example, by selecting a suitable... (corresponding to a specific )and This can generate a non-circular constellation with a larger minimum Euclidean distance, thereby improving decoding performance. Optimal rotation angle By maximizing the minimum Euclidean distance of the transformed constellation diagram get, The calculation formula is:

[0113] .

[0114] The system model and the transmitted signal of the achievable rate system are as follows: .

[0115] in For the transmission power of the public message stream, The sending power for user 1's private messages. Power for sending private messages for user 2.

[0116] user The received signal is ,in For users The channel coefficient, It is additive white Gaussian noise.

[0117] user First, decode the public message, treating the private message as interference. The received signal can be written as:

[0118] ;

[0119] in Add noise to the interference. Obtain the achievable rate of the common flow. The expression is: ,in For users The signal-to-noise ratio. After successful decoding and partial elimination of common messages (residual factor is...) ),user Decode its private message; the received signal at this time is ,in This is the sum of residual common interference, interference from another user, and noise. The achievable rate of the private flow is obtained. The expression is: ,in The system's common rate is Assuming Then there is .

[0120] Optimization Problems and Solutions

[0121] Mode 1: Maximizing private flow and rate (providing a closed-form solution with optimal non-circularity to visually observe the role of non-circularity in increasing the rate of public flow in the system):

[0122] The optimization problem is: Under this problem, it can be proven that... yes It is a monotonically increasing function, therefore the optimal solution is in Obtained from [location]. Further proof can be made when... hour, Since pc is monotonically decreasing, it is optimal. The problem simplifies to... Maximize under constraints This is a convex optimization problem that can be solved efficiently using standard KKT conditions.

[0123] Mode 2: Maximizing the common flow rate (provide a closed-form solution with optimal non-circularity to visually observe the role of non-circularity in increasing the common flow rate of the system). The optimization problem is: ;

[0124] in Due to the constraint of the private stream rate, the private stream is only decoded by its corresponding user. Therefore, the private stream rate is related to the security of the entire system. In the presence of eavesdroppers, the confidentiality rate can be calculated as follows: ,in For private flow rate, For the eavesdropper's rate. Because... ,and Depends solely on the total power of the private stream Rather than its allocation, the analysis shows that at least one private rate constraint is compact at the optimal solution (equality sign). Taking a tight constraint as an example, the common flow power required to satisfy the constraint can be solved inversely. and relational formula :

[0125] ;

[0126] Among them, coefficient , Determined by system parameters:

[0127] ; ; ;

[0128] in For user 2.

[0129] Will Substituting the expression into the common rate In the formula. Through analysis, it can be seen that, Follow The monotonicity is determined by a key condition: ,in When this expression is greater than 0, Follow Increases when it increases; conversely, decreases when it is less than 0.

[0130] Next, determine the optimal non-circularity coefficient.

[0131] Analyze the trend of common rate with non-circularity.

[0132] Will Substituting the expression into the common rate In the formula. Through analysis, it can be seen that, Follow The monotonicity is determined by a key condition: ,in When this expression is greater than 0, Follow Increases when it increases; conversely, decreases when it is less than 0.

[0133] Next, determine the optimal non-circularity coefficient.

[0134] First, calculate the feasibility: Calculate and Time corresponding and .like or If so, then the problem has no feasible solution.

[0135] If the problem has a feasible solution, then determine the monotonicity case:

[0136] like ,but yes An increasing function. In this case, a function with the largest possible value should be chosen. However, it is necessary to ensure Not exceeding the available power budget .therefore,

[0137] like The optimal solution is .

[0138] like Then the optimal non-circularity The expression is:

[0139] ;

[0140] like Then you should choose the smallest possible value. The optimal solution is .

[0141] Mode 3: The problem of maximizing system and speed optimization is as follows: ;

[0142] This problem is non-convex and is solved using the Soft Actor-Critic (SAC) reinforcement learning algorithm. Figure 3 As shown, the algorithm flow is as follows:

[0143] (1) Define the state space S: containing the current channel state , Imperfect SIC coefficient minimum common current power noise power Total power .

[0144] (2) Define the action space : Contains variables to be optimized , , , Before executing the action, the power needs to be normalized to ensure that the sum is 1, and then multiplied by the total power P.

[0145] (3) Design the reward function: ,in To constrain violations of indicator functions, if The corresponding value is 1 if the value is 1, otherwise it is 0. A positive penalty coefficient.

[0146] (4) Initialize SAC network parameters and experience replay pool.

[0147] (5) In each training round, the agent adjusts its strategy according to the current policy. Choose an action, interact with the environment to receive rewards and the next state, and replay the experience in the experience pool.

[0148] (6) Sample small batches of data from the experience replay pool and update the critic network (minimize temporal difference error), the actor network (maximize expected reward and policy entropy), and the temperature parameter in sequence. .

[0149] (7) Soft update the target network parameters.

[0150] Through sufficient training, the agent learns a near-optimal joint optimization strategy.

[0151] Experiment 1 is set up to verify the theoretical analysis of Mode 1 (maximizing private flow and rate). System parameters are set as follows: noise power normalized to... Channel coefficients for weak users (user 2) Common current power threshold Channel gain ratio .

[0152] Figure 4 Private flow and rate were plotted under different signal-to-noise ratios (SNR). The curves in the graph represent different imperfection SIC coefficients. Different colored curves in the graph represent different SIC coefficients. Different line types represent different non-circularity coefficients. The results show that using a non-circular Gaussian signal (IGS), Compared to traditional circular Gaussian signals (PGS), This can improve private flow and rate. With the SIC imperfection coefficient... Increase, different The performance difference corresponding to these values ​​is more pronounced. For example, when When the value is large, the maximum non-circular signal is used ( )compared to or It can bring more significant gains.

[0153] The experimental results are consistent with the theoretical analysis conclusions of Mode 1 (i.e., when imperfect SIC exists, the optimal...). The results (and IGS can improve private speed) are completely consistent, proving the effectiveness of the IGS scheme proposed in this invention in scenarios that maximize privacy and speed.

[0154] Experiment 2: This experiment is designed to verify the theoretical analysis of Mode 2 (maximizing the common flow rate). The experimental parameters are fixed as follows: , , This embodiment evaluates different SIC imperfection coefficients λ and minimum rate constraints. Below, common rate With non-circularity The changing relationship.

[0155] Figure 5 The simulation results are shown. The figure illustrates the theoretical optimal value calculated using the formula. Values ​​and their corresponding values Use an asterisk ( Mark it. Also, for each The optimal common power is obtained by searching using the Soft Actor-Critic (SAC) algorithm. Corresponding Plotted with solid lines. It can be observed that the theoretically calculated optimal point largely coincides with the optimal point obtained through numerical search, verifying the optimality derived theoretically in Mode 2. The correctness of the piecewise solution. Follow The monotonicity of change does indeed depend on and The specific value of . Under certain parameters (such as when λ is large). Follow Increases with increasing; under other parameters (such as λ being smaller and...) (larger) Possibly Increase and decrease. The experimental results show that, under the constraint of ensuring the safe rate of the private flow, non-circularity can be reduced through joint optimization. This can effectively increase the common flow rate and achieve better performance than using a fixed circular signal.

[0156] Experiment 3: The Soft Actor-Critic (SAC) algorithm is used to solve the non-convex optimization problem of Mode 3 (maximizing system and rate). Key parameters of the SAC algorithm are shown in Table 1. This embodiment combines the proposed IGS scheme with the traditional PGS (i.e., fixed-point optimization) scheme. The two schemes were compared.

[0157] Table 1 Key Parameter Settings for SAC Algorithm

[0158]

[0159] Figure 6 Demonstrates different minimum rate requirements Below, system and rate Curves showing SNR variation ( It can be seen that in the high SNR region, the sum rate obtained by the IGS scheme is higher than that of the PGS scheme.

[0160] like Figure 6 As shown in (a) and (b) in the figure, when At lower speeds, the advantages of the IGS scheme are more significant. This indicates that when security constraints are relatively relaxed, the additional degrees of freedom provided by IGS can more effectively improve the overall throughput.

[0161] Figure 7 Figures (a) and (b) show the sum rate performance of IGS and PGS under different channel gain ratios. The results indicate that when the channel difference is large ( When channel conditions are similar, IGS shows a more significant advantage over PGS; however, when channel conditions are similar, their performance is close. This indicates that the IGS scheme proposed in this invention is particularly suitable for scenarios with significant differences in channel quality between users, and can better utilize the channel advantages of strong users.

[0162] Furthermore, this embodiment also analyzes the optimal non-circularity coefficient under different parameters. , Figure 8 and Figure 9 show:

[0163] The larger the SIC imperfection coefficient λ, the more optimal... The overall value is also larger. This indicates that the greater the degree of imperfect SiC, the more necessary it is to utilize the non-circular characteristics of the signal to cancel out residual interference.

[0164] In the high SNR region, optimal There is a tendency for it to increase with increasing SNR. However, due to the nonconvexity of the system and the rate maximization problem, The variation with SNR is not a strictly monotonic function.

[0165] These experimental results collectively demonstrate that the RSMA joint optimization scheme based on IGS proposed in this invention can effectively improve the overall sum rate performance of the system under different channel conditions, different interference levels, and different security requirements by adaptively adjusting the non-circularity of the signal. Its gain is particularly prominent in scenarios with severe imperfect SIC, large differences in user channels, or moderate security constraints.

[0166] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. A method for improving the security and robustness of an RSMA system using non-circular signals, characterized in that, Specifically: The base station performs rate segmentation on each user's message based on the user's channel state information to obtain each user's private message; A common message is constructed by mapping the initial circular signal through a non-circular constellation diagram. The public message has a non-circularity coefficient. Mapping each user's private messages to traditional circular Gaussian signals , Index for users; Public signals Private signals of each user Perform overlay coding and allocate common stream transmit power. and private stream transmit power This generates a transmitted signal x. After receiving the signal, the user decodes the common message and calculates the reachable rate of the common flow in the RSMA system. Then, decode its own private message and calculate the achievable rate of user k's private stream. ; With system performance indicators as the objective, the non-circularity coefficient is jointly optimized. Common stream transmit power and the transmit power of each private current .

2. The method for improving the security and robustness of an RSMA system using non-circular signals according to claim 1, characterized in that, Public Message The expression is: ； Where s is the original circular Gaussian signal from the M-QAM constellation, To control the constellation diagram parameters for non-circularity, , To control the rotation angle of the constellation's minimum Euclidean distance, ; The non-circularity coefficient This is the ratio of the absolute value of the complementary variance to the variance of the signal, and .

3. The method for improving the security and robustness of an RSMA system using non-circular signals according to claim 1, characterized in that, The expression for the transmitted signal is: ； Where K is the total number of users.

4. The method for improving the security and robustness of an RSMA system using non-circular signals according to claim 1, characterized in that, The signal received by user k is ;in For users The channel coefficient, It is additive white Gaussian noise; If user k treats the private message as interference and decodes the public message, then the received signal of user k is rewritten as: ; in, For interference and noise, The reachable rate of the common flow for user k is obtained. The expression is: ； in, For user k, the channel noise ratio is... ,in The average power of the noise; The system's common flow reachable rate: ; After user k decodes and deletes the public message, user k decodes its own private message. At this time, the signal received by user k is: ； in, This is the sum of residual common interference, interference from other users, and noise. ; The SIC imperfection coefficient; The achievable rate of user k's private flow The expression is: ； Where K is the total number of users.

5. A method for improving the security and robustness of an RSMA system using non-circular signals according to claim 4, characterized in that, With the goal of maximizing system performance and speed, the non-circularity coefficients are jointly optimized. Common stream transmit power and the transmit power of each private stream : ； in, For system and rate, Let K be the minimum achievable rate of the private stream among K users. Where P is the minimum common current power and P is the total power; The soft actor-critic reinforcement learning algorithm is used to solve the problem.

6. The method for improving the security and robustness of an RSMA system using non-circular signals according to claim 5, characterized in that, The reward function when using the soft actor-critic reinforcement learning algorithm is: ； in, A positive penalty coefficient, To constrain violations of indicator functions, if any but It is 1 if it is 1, otherwise it is 0. Indicates state, For action.

7. A method for improving the security and robustness of an RSMA system using non-circular signals according to claim 4, characterized in that, If there are only two users, the non-circularity coefficient can be jointly optimized with the goal of maximizing the reachability of the private flow. Common current power and the power of each private current ,: ； in, Let K be the private stream reachable rate for user k, and K be the total number of users. Let P be the minimum common current power, and P be the total power.

8. A method for improving the security and robustness of an RSMA system using non-circular signals according to claim 4, characterized in that, If there are only two users, the non-circularity coefficient can be jointly optimized with the goal of maximizing the reachability of the common flow. Common stream transmit power and the transmit power of each private stream : ； in, Let K be the minimum achievable rate of the private stream among K users. Where P is the minimum common current power and P is the total power; Calculate the common flow transmit power based on the constraints. and relational formula : ； in, ； ; ; in, This represents the user ID corresponding to the minimum reachable rate of the private stream, where j represents the j-th user. ; Calculate separately and Common stream transmit power at time and ;like or If no feasible solution exists under the current power allocation constraint, the process terminates. like The subscripts 1 and 2 represent user 1 and user 2, respectively; in this case, the larger one should be selected if possible. However, it is necessary to ensure Not exceeding the available power budget ,therefore, like The optimal solution is ; like Then the optimal non-circularity The expression is: ; like Then you should choose the smallest possible value. The optimal solution is .

9. A system for improving the security and robustness of an RSMA system using non-circular signals, characterized in that, include: The rate segmentation module is used on the base station side to split user messages into private and public parts based on channel information; The non-circular signal generation module constructs a common message from the initial circular signal through a non-circular constellation diagram mapping. The public message has a non-circularity coefficient. ; Precoding and power distribution module, common signal Private signals of each user Perform overlay coding and allocate common stream transmit power. and private stream transmit power This generates a transmitted signal x. The superposition encoding module is used to superimpose the power-scaled signals to form the transmitted signal; The decoding module decodes the common message after the user receives the signal and calculates the reachable rate of the common flow in the RSMA system. Then, decode its own private message and calculate the achievable rate of user k's private stream. ; The joint optimization module, with system performance indicators as the objective, jointly optimizes the non-circularity coefficient. Common stream transmit power and the transmit power of each private stream .

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements a method for improving the security and robustness of an RSMA system using non-circular signals as described in any one of claims 1 to 8.

Images