Air-ground collaborative secure transmission and confidential communication method based on rate splitting multiple access
By splitting and transmitting information in a drone-assisted air-to-ground integrated network, using rate splitting multiple access technology and optimization algorithms, the problems of air-to-ground channel interference and secure transmission of multi-user communications in complex urban market scenarios are solved. Efficient and secure communication is achieved.
Patent Information
- Application Number
- CN202411070249.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-06
AI Technical Summary
The prior art is susceptible to interference in complex urban market scenarios, resulting in a decline in channel quality and failing to effectively consider the problem of secure transmission in multi-user communication services and eavesdropping scenarios, especially in the integration of air-to-ground assisted by drones. In the network.
The rate splitting multiple access technology is used to split the information into public and private information. By optimizing the phase shift of the intelligent reflection surface and the drone position, the phase shift of the intelligent reflection surface is optimized by using continuous convex approximation and semi-positive slack technology, combined with Deep deterministic policy gradient algorithm optimizes drone location and maximizes the total confidential capacity of collaborative communications to ensure secure transmission when eavesdroppers exist.
The challenges of channel interference and multi-user communication are effectively overcome, the security and efficiency of air-to-ground collaborative communication are improved, the secure transmission of information when the eavesdropper exists, and the total confidentiality capacity of the system is significantly improved.
Smart Images

Figure CN118659805B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of air-ground collaborative communication, and in particular to an air-ground collaborative secure transmission and confidential communication method based on rate splitting multiple access. Background Art
[0002] With the continuous development of communication technology, the number and performance requirements of user terminals are becoming increasingly diverse. Future 6G scenarios such as massive connections, dense and urgent connections, air-ground integration, and large-capacity communications will all pose higher challenges and requirements for multiple access technologies.
[0003] One of the multiple access technologies is Rate-Splitting Multiple Access (RSMA). By splitting and reconstructing user information at the transmitter and using serial interference cancellation technology at the user end, RSMA enables users to obtain partial decoding interference, thereby significantly improving spectrum efficiency and energy efficiency. It can also meet the low-cost, high-reliability and low-latency access requirements in the sixth-generation (6G) communication-intensive scenarios.
[0004] At the same time, facing the urgent need for the sixth generation (6G) communication technology to have efficient, reliable and seamless coverage, in complex urban scenarios, the air-to-ground (A2G) channel is easily interfered by buildings or environmental fading, resulting in a decrease in channel quality. The air-to-ground integrated network assisted by unmanned aerial vehicles (UAVs) has attracted widespread attention due to its advantages such as low cost, high throughput and flexible network architecture. Compared with the limitations of ground base stations, drones have high mobility and deployment flexibility. As aerial relays, they can achieve efficient data transmission and comprehensive communication coverage, and are widely used in emergency rescue, intelligent transportation and other fields.
[0005] Due to the broadcast characteristics of wireless channels in communication technology, information is easily eavesdropped. Considering the problem of secure transmission of communication systems, the first prior art designs a comprehensive optimization algorithm that combines UAV path planning with user joint scheduling to minimize system energy consumption while ensuring real-time transmission of communication services, pointing out a promising direction for 6G communications. The second prior art proposes a two-way training scheme to detect active eavesdropping scenarios, and studies the problem of maximizing the secure rate of joint optimization of transmit beamforming and RIS phase shift. Although the above work proposes an optimization scheme and model for RIS-assisted enhanced transmission, there are still some limitations. The first prior art implies the assumption that the UAV cooperative communication network based on RIS is secure; at the same time, both the first and second prior arts do not take into account multi-user communication services. With the rise of the Internet of Things and the surge in communication devices, the current communication network faces the challenge of limited multi-user spectrum resources. Summary of the invention
[0006] The purpose of the present invention is to provide an air-to-ground collaborative secure transmission and confidential communication method based on rate splitting multiple access, which uses rate splitting multiple access technology to split the transmitted information into public information and private information, optimizes the phase shift of the intelligent reflecting surface and the position of the drone, and cooperates in communication according to the optimized total confidentiality capacity, thereby ensuring the secure transmission of the collaborative communication based on the downlink drone-rate splitting multiple access technology of the intelligent reflecting surface when an eavesdropper is present.
[0007] The technical solution for achieving the purpose of the present invention is as follows:
[0008] On the one hand, the present invention provides an air-ground collaborative secure transmission method based on rate split multiple access, wherein a UAV carrying an intelligent reflective surface ensures collaborative communication between a source base station and a ground user when a direct communication link is interrupted, wherein the direct communication link is a direct communication link between the source base station and the ground user and an eavesdropper; the air-ground collaborative secure transmission method comprises:
[0009] The sent information is divided into public information and private information, wherein the public information includes non-private information required by all ground users, and the private information includes private information required by all ground users;
[0010] Based on the total confidentiality capacity of cooperative communication, the source base station transmits the public information and the private information to the ground user through the intelligent reflection surface, and the total confidentiality capacity is jointly determined by the phase shift of the intelligent reflection surface, the position of the UAV, and the information security rate of the ground user; wherein the phase shift of the intelligent reflection surface is obtained based on continuous convex approximation and semi-positive relaxation technology optimization, the position of the UAV is obtained based on a deep deterministic policy gradient algorithm, and the total confidentiality capacity is jointly optimized using an alternating iterative optimization model;
[0011] The ground user first decodes the public information and then decodes the private information; during the decoding of the private information, the private information of other ground users sent by the source base station is used as interference information.
[0012] Based on one aspect, in one embodiment of the present invention, the drone carrying the intelligent reflective surface ensures cooperative communication between the source base station and the ground user when the direct communication link is interrupted, including:
[0013] The smart reflective surface is in the yoz plane and has M reflective elements, the M reflective elements are arranged in a rectangular shape and adopt a uniform plane array;
[0014] The location of the source base station is expressed as L s = [ x s , y s ] , represents the two-dimensional coordinates of the source base station;
[0015] The position of the UAV carrying the smart reflective surface is expressed as L [ n ] = [ x [ n ] , y [ n ] ] , x [ n ] , y [ n ] Represents the two-dimensional coordinates of the drone;
[0016] The locations of the ground users and eavesdroppers are expressed as L k = [ x k , y k ] , k ∈ K and L e = [ x e , y e ] ,in Represents the two-dimensional coordinates of the ground user, Represents the two-dimensional coordinates of the eavesdropper.
[0017] Based on one aspect, in one embodiment of the present invention, the source base station divides the transmitted information into public information and private information, including:
[0018] In the downlink communication network of cooperative communication, based on the rate split multiple access technology, the source base station divides the transmitted information into public information and private information of each ground user, and performs linear precoding through the precoder. When the ground user receives the received signal, it first decodes the public information and then decodes the private information, where:
[0019] The received signal received by the kth ground user It is expressed as:
[0020] Formula (1)
[0021] In formula (1), represents the channel coefficient matrix from the smart reflection surface to the kth ground user, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the common information, Indicates the public information flow sent by the source base station after division, The beamforming vector representing the private information of the kth ground user, represents the private information flow sent by the source base station to the kth ground user, represents the noise received by ground users, express terrestrial users.
[0022] Based on one aspect, in one embodiment of the present invention, the information security rate of the ground user includes the information rate of the ground user decoding public information and the information rate of the ground user decoding private information;
[0023] Among them, the information rate of public information is expressed as:
[0024] Formula (2)
[0025] In formula (2), represents the channel coefficient matrix from the smart reflection surface to the kth ground user, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the common information, express terrestrial users, Indicates The beamforming vector of the private information of the ground user, represents the signal bandwidth, represents the noise power;
[0026] The information rate and eavesdropping rate of the kth ground user decoding private information are expressed as:
[0027] Formula (3)
[0028] In formula (3), represents the channel coefficient matrix from the smart reflection surface to the kth ground user, represents the channel coefficient matrix from the smart reflection surface to the eavesdropper, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the private information of the kth ground user, Indicates The beamforming vector of the private information of the ground user, , represents the signal bandwidth, represents the noise power;
[0029] The information security rate of the kth ground user is expressed as:
[0030] R s , k = [ R p , k − R e , k ] + Formula (4)
[0031] In formula (4), represents the information rate at which ground users decode private information, represents the information rate at which an eavesdropper decodes private information, [ ⋅ ] + express .
[0032] Based on one aspect, in one embodiment of the present invention, the total confidentiality capacity is expressed as:
[0033] Formula (5)
[0034] In formula (5), represents the information security rate of the first ground user, represents the information security rate of the second ground user, Indicates The information security rate of each ground user, express terrestrial users.
[0035] Based on one aspect, in one embodiment of the present invention, the phase shift of the smart reflective surface is optimized based on continuous convex approximation and semi-positive relaxation technology, wherein the phase shift optimization problem of the smart reflective surface is expressed as:
[0036] max v ∑ k = 1 K B [ log 2 ( 1 + C k ) − log 2 ( 1 + C e , k ) ] s . t . | v m | = 1 , m ∈ M Formula (6)
[0037] In formula (6), v = [ v 1 , v 2 , … , v M ] H ∈ ℂ M × 1 , , , , is an auxiliary variable, and there is and , represents the reflection phase shift of the mth element of the smart reflective surface, Indicates the number of reflective elements of the smart reflective surface. , represents the signal bandwidth, represents the information rate of the kth terrestrial user, The beamforming vector representing the private information of the kth ground user, Indicates The beamforming vector of the private information of the ground user, express terrestrial users, represents a complex matrix of dimension M×1;
[0038] Relaxing constraints based on continuous convex approximation and semi-positive definite relaxation techniques When , the relaxation of formula (6) is:
[0039] Formula (7)
[0040] definition ,in v ˜ = [ v 1 ] T ∈ C ( M + 1 ) × 1 , introduce auxiliary variables G e , i = [ g e oh i oh i H g e H 0 0 0 ] , i ∈ { 0 , 1 , 2 ,..., K } , G k , i = [ g k oh i oh i H g k H 0 0 0 ] , k ∈ K , i ∈ K , express dimensional complex matrix, Indicates The beamforming vector of the private information of the ground user, express There are terrestrial users, so the information security rate of terrestrial users in formula (7) is expressed as:
[0041] R sk ' = B ln 2 × { ln [ 1 s 2 Tr ( ∑ i = 1 K G k , i V ) + 1 ] + ln Y k + 1 − Y k ⋅ [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G k , i V ) + 1 ] + ln [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G e , i V ) + 1 ] − Y e ⋅ [ 1 s 2 Tr ( ∑ i = 1 K G e , i V ) + 1 ] + ln Y e + 1 } Formula (8)
[0042] In formula (8), Y k = [ 1 s 2 T r ( ∑ i = 1 , i ≠ k K G k , i V ( r ) ) + 1 ] − 1 , Y e = [ 1 s 2 T r ( ∑ i = 0 K G e , i V ( r ) ) + 1 ] − 1 , represents the signal bandwidth, represents the V value obtained at the rth iteration, Represents the noise power.
[0043] Based on one aspect, in one embodiment of the present invention, the position of the drone is obtained based on a deep deterministic policy gradient algorithm, wherein the position of the drone is expressed as:
[0044] Formula (9)
[0045] In formula (9), Indicates the position of the drone. represents the total confidentiality capacity of the system, Indicates the range of motion of the drone;
[0046] The position of the drone is obtained based on a deep deterministic policy gradient algorithm, wherein the deep deterministic policy gradient algorithm includes environment, action, state and reward; specifically, the environment includes: obtaining the location information of the source base station and the ground user of the cooperative communication, and obtaining the channel state information according to the position of the drone itself; the action includes: defining the action as the angle dimension and the displacement dimension during the movement of the drone, and the position information of the drone at the next moment is expressed as: x [ n + 1 ] = x [ n ] + u ⋅ E ⋅ cos ψ y [ n + 1 ] = y [ n ] + u ⋅ E ⋅ sin ψ , is the UAV flight speed, is the displacement of the drone from time slot n to time slot n+1, is the angle between the UAV's motion direction and the positive x-axis; the state includes: a set of coordinates of all possible locations that the UAV can reach in all time slots; the reward includes: calculating the current moment The corresponding rewards are: Indicates the current state. Represents the action selected at the current moment, and sums the rewards corresponding to the different drone positions in all time slots;
[0047] The optimized position of the drone is obtained by interacting with the environment based on the deep deterministic policy gradient algorithm.
[0048] Based on one aspect, in one embodiment of the present invention, the method of jointly optimizing the total confidentiality capacity using an alternating iterative optimization model includes:
[0049] The initial position of the UAV is fixed, and the phase shift of the smart reflective surface is optimized based on the initial value after initialization based on the continuous convex approximation and semi-positive relaxation technology. According to the phase shift matrix of the smart reflective surface obtained after optimization, the position of the UAV is optimized using the deep deterministic policy gradient algorithm. It is determined whether the difference between the information security rate calculated by the optimization result and the information security rate of the corresponding initial position meets the convergence threshold. If so, the phase shift of the optimal smart reflective surface and the position of the UAV are output. Otherwise, the iterative solution is continued until the convergence threshold or the set maximum number of iterations is reached, and the phase shift matrix of the optimal smart reflective surface and the position of the UAV are obtained.
[0050] The phase shift optimization of the intelligent reflective surface based on continuous convex approximation and semi-definite relaxation technology and the position optimization of the drone based on the deep deterministic policy gradient algorithm are used to maximize the total confidentiality capacity of the cooperative communication through alternating iterative optimization models.
[0051] On the other hand, the present invention also provides an air-ground collaborative secure and confidential communication method based on rate splitting multiple access. Based on the above-mentioned air-ground collaborative secure transmission method based on rate splitting multiple access, the air-ground collaborative secure and confidential communication method includes:
[0052] Optimize the phase shift of smart reflective surface in cooperative communication based on continuous convex approximation and semi-definite relaxation technology;
[0053] Obtain the position of the UAV in cooperative communication based on the deep deterministic policy gradient algorithm;
[0054] Determine the total confidentiality capacity of cooperative communication by using the phase shift of the intelligent reflective surface, the position of the UAV, and the information security rate of the ground user, and jointly optimize the total confidentiality capacity by using an alternating iterative optimization model;
[0055] Air-to-ground collaborative secure and confidential communications are performed based on the total confidentiality capacity.
[0056] Based on another aspect, in one embodiment of the present invention, the method of optimizing and acquiring the phase shift of the smart reflection surface in cooperative communication based on continuous convex approximation and semi-positive relaxation technology includes:
[0057] Relaxing constraints based on continuous convex approximation and semi-positive definite relaxation techniques When , the relaxation of formula (6) is:
[0058] Formula (7)
[0059] definition ,in v ˜ = [ v 1 ] T ∈ C ( M + 1 ) × 1 , introduce auxiliary variables G e , i = [ g e oh i oh i H g e H 0 0 0 ] , i ∈ { 0 , 1 , 2 ,..., K } , G k , i = [ g k oh i oh i H g k H 0 0 0 ] , k ∈ K , i ∈ K , express dimensional complex matrix, Indicates The beamforming vector of the private information of the ground user, express There are terrestrial users, so the information security rate of terrestrial users in formula (7) is expressed as:
[0060] R sk ' = B ln 2 × { ln [ 1 s 2 Tr ( ∑ i = 1 K G k , i V ) + 1 ] + ln Y k + 1 − Y k ⋅ [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G k , i V ) + 1 ] + ln [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G e , i V ) + 1 ] − Y e ⋅ [ 1 s 2 Tr ( ∑ i = 1 K G e , i V ) + 1 ] + ln Y e + 1 } Formula (8)
[0061] In formula (8), Y k = [ 1 s 2 T r ( ∑ i = 1 , i ≠ k K G k , i V ( r ) ) + 1 ] − 1 , Y e = [ 1 s 2 T r ( ∑ i = 0 K G e , i V ( r ) ) + 1 ] − 1 , represents the signal bandwidth, represents the V value obtained at the rth iteration, represents the noise power;
[0062] The method of obtaining the position of the UAV in cooperative communication based on the deep deterministic policy gradient algorithm includes:
[0063] The position of the drone is represented as:
[0064] Formula (9)
[0065] In formula (9), Indicates the position of the drone. represents the total confidentiality capacity of the system, Indicates the range of motion of the drone;
[0066] The deep deterministic policy gradient algorithm includes environment, action, state and reward; the environment includes: obtaining the location information of the source base station and ground users of cooperative communication, and obtaining the channel state information according to the position of the drone itself; the action includes: defining the action as angle dimension and displacement dimension during the movement of the drone, and the position information of the drone at the next moment is expressed as: x [ n + 1 ] = x [ n ] + u ⋅ E ⋅ cos ψ y [ n + 1 ] = y [ n ] + u ⋅ E ⋅ sin ψ , is the UAV flight speed, is the displacement of the drone from time slot n to time slot n+1, is the angle between the UAV's motion direction and the positive x-axis; the state includes: a set of coordinates of all possible locations that the UAV can reach in all time slots; the reward includes: calculating the current moment The corresponding rewards are: Indicates the current state. Represents the action selected at the current moment, sums the rewards corresponding to the different drone positions in all time slots, and obtains the optimized drone position through interaction with the environment based on the deep deterministic policy gradient algorithm;
[0067] The method of determining the total confidentiality capacity of cooperative communication by using the phase shift of the intelligent reflective surface, the position of the drone, and the information security rate of the ground user includes:
[0068] The initial position of the UAV is fixed, and the phase shift of the smart reflective surface is optimized based on the initial value after initialization based on the continuous convex approximation and semi-positive relaxation technology. According to the phase shift matrix of the smart reflective surface obtained after optimization, the position of the UAV is optimized using the deep deterministic policy gradient algorithm. It is determined whether the difference between the information security rate calculated by the optimization result and the information security rate corresponding to the initial position meets the convergence threshold. If so, the phase shift of the optimal smart reflective surface and the position of the UAV are output. Otherwise, the iterative solution is continued until the convergence threshold or the set maximum number of iterations is reached, and the phase shift matrix of the optimal smart reflective surface and the position of the UAV are obtained.
[0069] The phase shift optimization of the intelligent reflective surface based on continuous convex approximation and semi-definite relaxation technology and the position optimization of the drone based on the deep deterministic policy gradient algorithm are used to maximize the total confidentiality capacity of the cooperative communication by alternating the iterative optimization model.
[0070] Compared with the prior art, the present invention has the following beneficial effects:
[0071] The present invention adopts rate splitting multiple access technology to split the transmitted information into public information and private information. By optimizing the phase shift of the intelligent reflecting surface and the position of the drone, collaborative communication is performed according to the optimized total confidentiality capacity, ensuring the secure transmission of the collaborative communication of the downlink drone-rate splitting multiple access technology based on the intelligent reflecting surface in the presence of an eavesdropper. BRIEF DESCRIPTION OF THE DRAWINGS
[0072] Figure 1 A flow chart of a method for air-ground coordinated secure transmission based on rate splitting multiple access provided by an embodiment of the present invention;
[0073] Figure 2 A network model diagram of collaborative communication provided by an embodiment of the present invention;
[0074] Figure 3 A comparison diagram of the effects of the air-ground collaborative secure transmission solution provided in an embodiment of the present invention, the NOMA solution, and the random phase shift solution;
[0075] Figure 4 A diagram showing the effect of different numbers of reflective elements on the total confidentiality capacity of the system provided by an embodiment of the present invention;
[0076] Figure 5 A block diagram of a deep deterministic policy gradient algorithm provided by an embodiment of the present invention;
[0077] Figure 6 A block diagram of an alternating iterative optimization model provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0078] The present invention is described in detail below in conjunction with the various embodiments shown in the accompanying drawings, but it should be noted that these embodiments are not limitations of the present invention, and any equivalent transformations or substitutions in functions, methods, or structures made by ordinary technicians in the field based on these embodiments are all within the scope of protection of the present invention.
[0079] The embodiment of the present invention provides an air-ground collaborative secure transmission method based on rate split multiple access, wherein a drone carrying an intelligent reflective surface ensures collaborative communication between a source base station and a ground user when a direct communication link is interrupted, and the direct communication link is a direct communication link between the source base station and the ground user and an eavesdropper; the drone carrying an intelligent reflective surface ensures collaborative communication between the source base station and the ground user when a direct communication link is interrupted, including:
[0080] See also Figure 2 In the embodiment of the present invention, the cooperative communication network includes a source basestation (SBS), an unmanned aerial vehicle (UAV) carrying a reconfigurable intelligent surface (RIS), a legitimate communication user, i.e., a ground user, and an eavesdropper (Eve); due to the obstruction of ground buildings and the influence of environmental fading, the direct communication link between the source basestation and the ground user and the eavesdropper is interrupted, so the UAV carrying the RIS is used as an air relay to ensure the communication between the source basestation and the ground user.
[0081] The intelligent reflective surface of the embodiment of the present invention is in the yoz plane and has M reflective elements, having horizontal elements and vertical elements ( ), M reflective elements are arranged in a rectangular shape and use a uniform planar array; ground users and eavesdroppers are equipped with a single antenna; the location of the source base station is represented by L s = [ x s , y s ] , represents the two-dimensional coordinates of the source base station; the position of the drone carrying the intelligent reflective surface is represented by L [ n ] = [ x [ n ] , y [ n ] ] , x [ n ] , y [ n ] represents the two-dimensional coordinates of the drone; the position of the ground user is represented by L k = [ x k , y k ] , k ∈ { 1 , 2 , e } , Represents the two-dimensional coordinates of the ground user.
[0082] In complex urban scenarios, air-to-ground (A2G) channels are easily interfered by buildings or environmental fading, resulting in channel quality degradation. The intelligent reflecting surface (RIS) of the embodiment of the present invention is composed of low-cost electromagnetic materials, and reconfigures the channel through beamforming and intelligent parameter adjustment, thereby effectively enhancing backscatter communication. The embodiment of the present invention uses an air-ground integrated network assisted by an unmanned aerial vehicle (UAV), which has the advantages of low cost, high throughput and flexible network architecture.
[0083] Compared with the traditional method of fixing RIS on the surface of buildings, UAV-RIS cooperative communication takes advantage of the flexible mobility of UAV and the channel reconstruction capability of RIS, which can play a greater role in improving system throughput and energy efficiency. The UAV-RIS cooperative communication network has lower path loss and higher line-of-sight (LoS) link probability, can provide seamless and flexible coverage through UAV location deployment, and the intelligent reflective surface with low hardware cost and low power consumption can achieve enhanced transmission through intelligent configuration of wireless propagation environment.
[0084] Furthermore, the embodiment of the present invention proposes an iterative framework for maximizing the total confidentiality capacity for secure transmission of the downlink UAV-RSMA cooperative communication network based on RIS in the presence of eavesdroppers, and performs secure information transmission by optimizing the RIS phase shift and UAV position. Specifically, a UAV equipped with RIS is first deployed to overcome the effects of link blocking and environmental fading, and the transmitted information is split into public information and private information based on RSMA technology; then, by optimizing the RIS phase shift and UAV position, the total confidentiality capacity of the downlink UAV-RSMA communication network based on RIS is maximized, the RIS phase shift is optimized using SCA and SDR technology, and the UAV position is optimized using the deep deterministic policy gradient algorithm (DDPG); next, the optimization variables are solved in turn through an alternating iterative optimization framework with polynomial complexity; finally, the simulation results show that the total confidentiality capacity of the embodiment of the present invention is significantly better than the comparative scheme of the background technology.
[0085] Therefore, in the embodiments of the present invention, (1) an air-ground cooperative communication network model consisting of a source base station, a UAV equipped with an intelligent reflective surface, an eavesdropper and a communication user is constructed; (2) a cooperative communication model is constructed based on the aerial intelligent reflective surface and rate split multiple access, and the air-ground cooperative communication network security problem is modeled; (3) the phase shift of the intelligent reflective surface is optimized based on continuous convex approximation and semi-definite relaxation technology; (4) the position of the UAV is obtained based on the deep deterministic policy gradient algorithm; (5) based on the optimization of the phase shift of the intelligent reflective surface and the position of the UAV, the alternating iterative optimization framework is adopted to jointly optimize and obtain the optimal phase shift and the position of the UAV.
[0086] According to the above air-ground cooperative communication network model, please refer to Figure 1 The air-ground coordinated secure transmission method based on rate splitting multiple access of an embodiment of the present invention comprises the following steps:
[0087] Step 1: Divide the transmitted information into public information and private information. The public information contains the non-private information required by all ground users, and the private information contains the private information required by all ground users. Based on the communication network model constructed above, in the downlink Rate Splitting Multiple Access (RSMA) communication network, the source base station first linearly precodes the information through the precoder, that is, divides the transmitted information into public information and the private information of each user; since the public information contains part of the non-private information required by all users, all ground users need to decode the public information first, and then decode the private information sent to themselves.
[0088] Step 1: The source base station divides the transmitted information into public information and private information, including:
[0089] In the downlink communication network of cooperative communication, based on the rate split multiple access technology, the source base station divides the transmitted information into public information and private information of each ground user, and performs linear precoding through the precoder. When the ground user receives the received signal, it first decodes the public information and then decodes the private information, where:
[0090] The received signal received by the kth ground user It is expressed as:
[0091] Formula (1)
[0092] In formula (1), represents the channel coefficient matrix from the smart reflection surface to the kth ground user, represents the phase shift matrix of the smart reflective surface, ,in represents the phase shift of the i-th reflective element, represents the channel coefficient matrix from the source base station to the smart reflection surface, , The beamforming vector representing the common information, Indicates the public information flow sent by the source base station after division, The beamforming vector representing the private information of the kth ground user, represents the private information flow sent by the source base station to the kth ground user, that is, the signal sent from the source base station to the ground user, which has zero mean and normalized power and is independent of each other, and ; represents the noise received by ground users, represents the noise received by the i-th ground user, ,in represents the noise power, express terrestrial users.
[0093] The eavesdropper's received signal is:
[0094]
[0095] in, is the channel coefficient matrix from RIS to the eavesdropper, ; is the noise received by the eavesdropper, where , The beamforming vector representing the common information, The beamforming vector representing the first user's private information, The beamforming vector representing the second user's private information.
[0096] Step 2: Based on the total confidentiality capacity of cooperative communication, the source base station transmits public information and private information to ground users through the intelligent reflection surface. The total confidentiality capacity is jointly determined by the phase shift of the intelligent reflection surface, the position of the UAV, and the information security rate of the ground user; among them, the phase shift of the intelligent reflection surface is obtained based on continuous convex approximation and semi-definite relaxation technology optimization, the position of the UAV is obtained based on the deep deterministic policy gradient algorithm, and the total confidentiality capacity is jointly optimized using an alternating iterative optimization model.
[0097] Step 2: The above-mentioned information security rate of the ground user includes the information rate of the ground user decoding public information and the information rate of the ground user decoding private information;
[0098] Among them, the information rate of public information is expressed as:
[0099] Formula (2)
[0100] In formula (2), represents the channel coefficient matrix from the smart reflection surface to the ground user, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the common information, express terrestrial users, Indicates The beamforming vector of the private information of the ground user, represents the signal bandwidth, represents the noise power;
[0101] In order to ensure that public information can be successfully acquired and decoded by all ground users, the information transmission rate of public information should not exceed ; Then the ground user decodes the public information by using the continuous interference cancellation technology, removes the decoded public information and starts to decode the private information sent to itself by the source base station, and at the same time uses the private information of other users as interference. Then, the information rate and eavesdropping rate of the kth ground user decoding the private information are expressed as:
[0102] Formula (3)
[0103] In formula (3), represents the channel coefficient matrix from the smart reflection surface to the kth ground user, represents the channel coefficient matrix from the smart reflection surface to the eavesdropper, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the private information of the kth ground user, Indicates The beamforming vector of the private information of the ground user, , represents the signal bandwidth, represents the noise power;
[0104] Private information is generally stored in the private information stream sent to each user. When an eavesdropper eavesdrops on the private information of each user, public information can be used as noise to interfere with the eavesdropping. Therefore, the information rate at which the eavesdropper decodes the private information is expressed as:
[0105] .
[0106] The information security rate of the k-th ground user in the embodiment of the present invention is expressed as:
[0107] R s , k = [ R p , k − R e , k ] + Formula (4)
[0108] In formula (4), represents the information rate at which ground users decode private information, represents the information rate at which an eavesdropper decodes private information, [ ⋅ ] + express .
[0109] By optimizing the RIS phase shift and UAV location , thereby maximizing the total confidentiality capacity of the cooperative communication network ; The UAV flight area is specified as , the total confidentiality capacity maximization problem is modeled as:
[0110] The total confidentiality capacity in step 2 of the embodiment of the present invention is expressed as:
[0111] Formula (5)
[0112] In formula (5), represents the information security rate of the first ground user, represents the information security rate of the second ground user, Indicates The information security rate of each ground user, express terrestrial users.
[0113] refer to , define the semi-positive definite matrix , the information security rate of ground legal communication users is rewritten using the sequential convex approximation algorithm (SCA) and semi-definite relaxation (SDR).
[0114] Step 2: The phase shift of the above-mentioned smart reflector is optimized based on continuous convex approximation and semi-positive relaxation technology, which specifically includes:
[0115] The phase shift optimization problem of the smart reflective surface in the embodiment of the present invention is expressed as:
[0116] max v ∑ k = 1 K B [ log 2 ( 1 + C k ) − log 2 ( 1 + C e , k ) ] s . t . | v m | = 1 , m ∈ M Formula (6)
[0117] In formula (6), v = [ v 1 , v 2 , … , v M ] H ∈ C M × 1 , , , , is an auxiliary variable, and there is and , represents the reflection phase shift of the mth element of the smart reflective surface, Indicates the number of reflective elements of the smart reflective surface. , represents the signal bandwidth, represents the information rate of the kth terrestrial user, The beamforming vector representing the private information of the kth ground user, Indicates The beamforming vector of the private information of the ground user, express terrestrial users, represents a complex matrix of dimension M×1;
[0118] Formula (6) of the embodiment of the present invention is rewritten by the following calculation: Assuming that the UAV position Given, the RIS phase shift optimization problem is expressed as:
[0119]
[0120] make ,in ; Similarly, the eavesdropping channel is rewritten as ,and ; Introduce auxiliary variables v = [ v 1 , v 2 , … , v M ] H ∈ C M × 1 and .
[0121] Among them, the relaxation constraints are based on continuous convex approximation and semi-positive definite relaxation techniques. When , the relaxation of formula (6) is:
[0122] Formula (7)
[0123] definition ,in v ˜ = [ v 1 ] T ∈ C ( M + 1 ) × 1 , introduce auxiliary variables G e , i = [ g e oh i oh i H g e H 0 0 0 ] , i ∈ { 0 , 1 , 2 ,..., K } , G k , i = [ g k oh i oh i H g k H 0 0 0 ] , k ∈ K , i ∈ K , express dimensional complex matrix, Indicates The beamforming vector of the private information of the ground user, express There are terrestrial users, so the information security rate of terrestrial users in formula (7) is expressed as:
[0124] R sk ' = B ln 2 × { ln [ 1 s 2 Tr ( ∑ i = 1 K G k , i V ) + 1 ] + ln Y k + 1 − Y k ⋅ [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G k , i V ) + 1 ] + ln [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G e , i V ) + 1 ] − Y e ⋅ [ 1 s 2 Tr ( ∑ i = 1 K G e , i V ) + 1 ] + ln Y e + 1 } Formula (8)
[0125] In formula (8), Y k = [ 1 s 2 T r ( ∑ i = 1 , i ≠ k K G k , i V ( r ) ) + 1 ] − 1 , Y e = [ 1 s 2 T r ( ∑ i = 0 K G e , i V ( r ) ) + 1 ] − 1 , represents the signal bandwidth, represents the V value obtained at the rth iteration, Represents the noise power.
[0126] The above embodiment of the present invention is a convex optimization problem. The phase shift obtained by using SDR to relax the constraints cannot always satisfy ; Therefore, a Gaussian randomization process is used to obtain an approximate solution of rank one , that is, the phase shift obtained above is of (M+1)×(M+1) dimensions, so Gaussian randomization is required to obtain the original phase shift of M×M dimensions while satisfying the rank-one constraint.
[0127] Step 2: The position of the above-mentioned drone is obtained based on the deep deterministic policy gradient algorithm, where the position of the drone is expressed as:
[0128] Formula (9)
[0129] In formula (9), Indicates the position of the drone. represents the total confidentiality capacity of the system, Indicates the range of motion of the drone;
[0130] The position of the drone is obtained based on the deep deterministic policy gradient algorithm, where the deep deterministic policy gradient algorithm includes environment, action, state and reward; specifically, the environment includes: obtaining the location information of the source base station and ground users of cooperative communication, and obtaining the channel state information according to the position of the drone itself; the action includes: defining the action as the angle dimension and the displacement dimension during the movement of the drone, and the position information of the drone at the next moment is expressed as: x [ n + 1 ] = x [ n ] + u ⋅ E ⋅ cos ψ y [ n + 1 ] = y [ n ] + u ⋅ E ⋅ sin ψ , is the UAV flight speed, is the displacement of the drone from time slot n to time slot n+1, is the angle between the UAV's motion direction and the positive x-axis; the state includes: a set of coordinates of all possible locations that the UAV can reach in all time slots; the reward includes: calculating the current moment The corresponding rewards are: Indicates the current state. Represents the action selected at the current moment, and sums the rewards corresponding to the different drone positions in all time slots;
[0131] The optimized position of the drone is obtained by interacting with the environment based on the deep deterministic policy gradient algorithm.
[0132] See also Figure 5 The deep deterministic policy gradient (DDPG) of the embodiment of the present invention is a deep reinforcement learning algorithm based on the Actor-Critic network architecture, which adopts a dual neural network, wherein the actor network interacts with the environment to make action decisions, and the critic network is used to evaluate the quality of the action; in the DDPG algorithm, the actor online network interacts with the environment to obtain state information, takes the state as input, and generates corresponding actions under the support of random noise; the state and action corresponding to the time slot are used as input parameters of the Critic online network, and the Q value is generated according to the Bellman equation:
[0133] Q π ( s t , a t ) = E [ r t + ς Q π ( s t + 1 , π ( s t + 1 ) ) ]
[0134] in ς ∈ ( 0 , 1 ] represents the reward discount factor; then the Critic network updates by minimizing the error between the evaluation value and the target value , the Actor network is updated by maximizing the cumulative expectation, i.e., the policy gradient , the error function and policy gradient are expressed as:
[0135] Loss = E [ ( r t + ς Q π ( s t + 1 , π ( s t + 1 ) ) − Q π ( s t , a t ) ) 2 ] ∇ i π J = E t ∼ p i ( t ) [ R ( t ) ∇ p i ( t ) ]
[0136] in is the probability distribution function, For expectations and.
[0137] For the two target networks, they do not participate in the environment interaction during the training and learning process, and are not updated in real time with the online network. Instead, they are soft updated at regular intervals while ensuring accuracy and algorithm convergence. The update method is expressed as:
[0138]
[0139] It should be noted that the DDPG algorithm framework of the embodiment of the present invention includes four parts: environment, action, state and reward;
[0140] Environment: Based on the first step of the air-ground cooperative communication network model, the location information of the ground base station and the user is obtained, and then the channel state information is obtained according to the UAV's own position;
[0141] Action: During the UAV motion process, the action selection is defined as two action dimensions: angle and displacement. The displacement depends on the UAV flight speed and the length of each time slot. The angle is defined as the angle with the positive x-axis. Therefore, the position information of the UAV at the next moment is expressed as:
[0142] x [ n + 1 ] = x [ n ] + u ⋅ E ⋅ cos ψ y [ n + 1 ] = y [ n ] + u ⋅ E ⋅ sin ψ
[0143] in is the UAV flight speed, is the displacement of the UAV from time slot n to time slot n+1, is the angle between the UAV motion direction and the positive x-axis;
[0144] State: The current two-dimensional position information of the UAV is used as the state information of the neural network, that is, the set of all possible position coordinates of the UAV in all time slots is used as the state information;
[0145] Reward: The action generated by the state and actor network is used as the input parameter of the Critic network to calculate the moment The corresponding reward is then summed up for the different UAV positions in all time slots to get the reward value for the episode;
[0146] Based on the DDPG algorithm, training and data collection are performed by interacting with the environment, and finally the optimized UAV position is obtained. The embodiment of the present invention is based on the phase shift optimization of SCA and SDR and the UAV position optimization based on DDPG, and the total confidentiality capacity of the communication system is maximized through the designed alternating iterative optimization framework.
[0147] The above-mentioned alternate iterative optimization model is used in the embodiment of the present invention to jointly optimize the total confidentiality capacity, including:
[0148] The initial position of the UAV is fixed, and the phase shift of the smart reflective surface is optimized based on the initial value after initialization based on the continuous convex approximation and semi-positive relaxation technology. According to the phase shift matrix of the smart reflective surface obtained after optimization, the position of the UAV is optimized using the deep deterministic policy gradient algorithm. It is determined whether the difference between the information security rate calculated by the optimization result and the information security rate of the corresponding initial position meets the convergence threshold. If so, the phase shift of the optimal smart reflective surface and the position of the UAV are output. Otherwise, the iterative solution is continued until the convergence threshold or the set maximum number of iterations is reached, and the phase shift matrix of the optimal smart reflective surface and the position of the UAV are obtained.
[0149] The phase shift optimization of the intelligent reflective surface based on continuous convex approximation and semi-definite relaxation technology and the position optimization of the drone based on the deep deterministic policy gradient algorithm are used to maximize the total confidentiality capacity of the cooperative communication through the alternating iterative optimization model. The alternating iterative optimization model of the embodiment of the present invention is as follows: Figure 6 shown.
[0150] Step 3: The ground user first decodes the public information and then decodes the private information; during the decoding of the private information, the private information of other ground users sent by the source base station is used as interference information.
[0151] In addition, an embodiment of the present invention further provides an air-ground collaborative secure and confidential communication method based on rate splitting multiple access. Based on the above-mentioned air-ground collaborative secure transmission method based on rate splitting multiple access, the air-ground collaborative secure and confidential communication method based on rate splitting multiple access of an embodiment of the present invention includes:
[0152] Optimize the phase shift of smart reflective surface in cooperative communication based on continuous convex approximation and semi-definite relaxation technology;
[0153] Obtain the position of the UAV in cooperative communication based on the deep deterministic policy gradient algorithm;
[0154] The total confidentiality capacity of cooperative communication is determined by the phase shift of the intelligent reflective surface, the position of the UAV, and the information security rate of the ground user, and the alternating iterative optimization model is used to jointly optimize the total confidentiality capacity.
[0155] Conduct air-ground collaborative secure and confidential communications based on total confidentiality capacity.
[0156] Among them, the phase shift of the smart reflection surface in cooperative communication is optimized based on continuous convex approximation and semi-positive relaxation technology, including:
[0157] Relaxing constraints based on continuous convex approximation and semi-positive definite relaxation techniques When , the relaxation of formula (6) is:
[0158] Formula (7)
[0159] definition ,in v ˜ = [ v 1 ] T ∈ C ( M + 1 ) × 1 , introduce auxiliary variables G e , i = [ g e oh i oh i H g e H 0 0 0 ] , i ∈ { 0 , 1 , 2 ,..., K } , G k , i = [ g k oh i oh i H g k H 0 0 0 ] , k ∈ K , i ∈ K , express dimensional complex matrix, Indicates The beamforming vector of the private information of the ground user, express There are terrestrial users, so the information security rate of terrestrial users in formula (7) is expressed as:
[0160] R sk ' = B ln 2 × { ln [ 1 s 2 Tr ( ∑ i = 1 K G k , i V ) + 1 ] + ln Y k + 1 − Y k ⋅ [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G k , i V ) + 1 ] + ln [ 1 s 2 Tr ( ∑ i = 1 , i ≠ k K G e , i V ) + 1 ] − Y e ⋅ [ 1 s 2 Tr ( ∑ i = 1 K G e , i V ) + 1 ] + ln Y e + 1 } Formula (8)
[0161] In formula (8), Y k = [ 1 s 2 T r ( ∑ i = 1 , i ≠ k K G k , i V ( r ) ) + 1 ] − 1 , Y e = [ 1 s 2 T r ( ∑ i = 0 K G e , i V ( r ) ) + 1 ] − 1 , represents the signal bandwidth, represents the V value obtained at the rth iteration, represents the noise power;
[0162] The method of obtaining the position of the UAV in cooperative communication based on the deep deterministic policy gradient algorithm includes:
[0163] The position of the drone is represented as:
[0164] Formula (9)
[0165] In formula (9), Indicates the position of the drone. represents the total confidentiality capacity of the system, Indicates the range of motion of the drone;
[0166] The deep deterministic policy gradient algorithm includes environment, action, state and reward; the environment includes: obtaining the location information of the source base station and ground users of cooperative communication, and obtaining the channel state information according to the position of the drone itself; the action includes: defining the action as angle dimension and displacement dimension during the movement of the drone, and the position information of the drone at the next moment is expressed as: x [ n + 1 ] = x [ n ] + u ⋅ E ⋅ cos ψ y [ n + 1 ] = y [ n ] + u ⋅ E ⋅ sin ψ , is the UAV flight speed, is the displacement of the drone from time slot n to time slot n+1, is the angle between the UAV's motion direction and the positive x-axis; the state includes: a set of coordinates of all possible locations that the UAV can reach in all time slots; the reward includes: calculating the current moment The corresponding rewards are: Indicates the current state. Represents the action selected at the current moment, sums the rewards corresponding to the different drone positions in all time slots, and obtains the optimized drone position through interaction with the environment based on the deep deterministic policy gradient algorithm;
[0167] The total confidentiality capacity of cooperative communication is determined by the phase shift of the intelligent reflective surface, the position of the drone, and the information security rate of the ground user, including:
[0168] The initial position of the UAV is fixed, and the phase shift of the smart reflective surface is optimized based on the initial value after initialization based on the continuous convex approximation and semi-positive relaxation technology. According to the phase shift matrix of the smart reflective surface obtained after optimization, the position of the UAV is optimized using the deep deterministic policy gradient algorithm. It is determined whether the difference between the information security rate calculated by the optimization result and the information security rate corresponding to the initial position meets the convergence threshold. If so, the phase shift of the optimal smart reflective surface and the position of the UAV are output. Otherwise, the iterative solution is continued until the convergence threshold or the set maximum number of iterations is reached, and the phase shift matrix of the optimal smart reflective surface and the position of the UAV are obtained.
[0169] The phase shift optimization of the intelligent reflective surface based on continuous convex approximation and semi-definite relaxation technology and the position optimization of the drone based on the deep deterministic policy gradient algorithm are used to maximize the total confidentiality capacity of the cooperative communication by alternating the optimization model.
[0170] The technical effects of the embodiments of the present invention are described in detail below in conjunction with simulation.
[0171] This experiment conducted a simulation experiment to evaluate the performance of the proposed total confidentiality capacity maximization scheme and verify the superiority of this method. The specific simulation parameters are as follows: L s = [ 10 , 50 ] , L 1 = [ 130 , 60 ] , L 2 = [ 140 , 70 ] , L e = [ 160 , 45 ] , , , The initial position of the UAV is (140,30,50) and its flight area is 200×200m.
[0172] Figure 3 The performance comparison of the total confidentiality capacity of the system under different schemes is shown. It is clearly observed that the total secure channel capacity maximization scheme proposed in the embodiment of the present invention is better than other comparison schemes. In the same scenario, compared with the non-orthogonal multiple access (NOMA) scheme (the user end uses NOMA technology for decoding) and the random RIS phase shift scheme, the total secure channel capacity of the proposed scheme can be increased by 29.7% and 71.9%. Figure 4 The comparison of the total security channel capacity under different numbers of reflection elements is plotted, and it can be observed that the total security channel capacity increases with the number of reflection elements.
[0173] In summary, the embodiment of the present invention proposes an iterative framework for maximizing the total confidentiality capacity for the secure transmission problem of the downlink UAV-RSMA cooperative communication network based on RIS in the presence of eavesdroppers, targeting the problems existing in the prior art. First, a UAV equipped with RIS is deployed to overcome the effects of link blocking and environmental fading, and the transmitted information is split into public information and private information based on RSMA technology; then, by optimizing the RIS phase shift and UAV position, the problem of maximizing the total confidentiality capacity of the downlink UAV-RSMA communication network based on RIS is proposed, and the RIS phase shift is optimized using SCA and SDR technologies, and the UAV position is optimized using the deep deterministic policy gradient algorithm (DDPG); next, an alternating iterative optimization framework with polynomial complexity is designed to solve the optimization variables in turn; finally, the simulation results show that the total confidentiality capacity of the embodiment of the present invention is significantly better than the comparative scheme of the background technology.
[0174] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. Any equivalent implementation methods or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.
[0175] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.
[0176] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.
Claims
1. A method for air-ground coordinated secure transmission based on rate splitting multiple access, characterized in that: The UAV carrying the intelligent reflective surface ensures cooperative communication between the source base station and the ground user in the event of a direct communication link interruption, wherein the direct communication link is a direct communication link between the source base station and the ground user and the eavesdropper; The air-ground collaborative safety transmission method includes: The sent information is divided into public information and private information, wherein the public information includes non-private information required by all ground users, and the private information includes private information required by all ground users; Based on the total confidentiality capacity of cooperative communication, the source base station transmits the public information and the private information to the ground user through the intelligent reflection surface, and the total confidentiality capacity is jointly determined by the phase shift of the intelligent reflection surface, the position of the UAV, and the information security rate of the ground user; wherein the phase shift of the intelligent reflection surface is obtained based on continuous convex approximation and semi-positive relaxation technology optimization, the position of the UAV is obtained based on a deep deterministic policy gradient algorithm, and the total confidentiality capacity is jointly optimized using an alternating iterative optimization model; The ground user first decodes the public information and then decodes the private information; during the decoding of the private information, the private information of other ground users sent by the source base station is used as interference information; The phase shift of the smart reflective surface is optimized based on continuous convex approximation and semi-positive relaxation technology, wherein the phase shift optimization problem of the smart reflective surface is expressed as: Formulas (6) In formula (6), , , , , is an auxiliary variable, and there is and , represents the reflection phase shift of the mth element of the smart reflective surface, Indicates the number of reflective elements of the smart reflective surface. , represents the signal bandwidth, Indicates k The information rate of a terrestrial user, Indicates k The beamforming vector of the private information of the ground user, Indicates The beamforming vector of the private information of the ground user, express terrestrial users, express -dimensional complex matrix; Relaxing constraints based on continuous convex approximation and semi-positive definite relaxation techniques When , the relaxation of formula (6) is: Formulas (7) definition ,in , introduce auxiliary variables , , represents a complex matrix of (M+1)×1 dimensions, Indicates The beamforming vector of the private information of the ground user, express There are terrestrial users, so the information security rate of terrestrial users in formula (7) is expressed as: Formulas (8) In formula (8), , , represents the signal bandwidth, Indicates r The V value obtained by the iteration is represents the noise power; The position of the drone is obtained based on a deep deterministic policy gradient algorithm, where the position of the drone is expressed as: Formulas (9) In formula (9), Indicates the position of the drone. represents the total confidentiality capacity of the system, Indicates the range of motion of the drone; The position of the drone is obtained based on a deep deterministic policy gradient algorithm, wherein the deep deterministic policy gradient algorithm includes environment, action, state and reward; specifically, the environment includes: obtaining the location information of the source base station and the ground user of the cooperative communication, and obtaining the channel state information according to the position of the drone itself; the action includes: defining the action as the angle dimension and the displacement dimension during the movement of the drone, and the position information of the drone at the next moment is expressed as: , is the UAV flight speed, is the displacement of the drone from time slot n to time slot n+1, is the angle between the UAV's motion direction and the positive x-axis; the state includes: a set of coordinates of all possible locations that the UAV can reach in all time slots; the reward includes: calculating the current moment The corresponding rewards are: Indicates the current state. Represents the action selected at the current moment, and sums the rewards corresponding to the different drone positions in all time slots; The position of the drone is optimized by interacting with the environment based on the deep deterministic policy gradient algorithm; The method of jointly optimizing the total confidentiality capacity by using an alternating iterative optimization model includes: The initial position of the UAV is fixed, and the phase shift of the smart reflective surface is optimized based on the initial value after initialization based on the continuous convex approximation and semi-positive relaxation technology. According to the phase shift matrix of the smart reflective surface obtained after optimization, the position of the UAV is optimized using the deep deterministic policy gradient algorithm. It is determined whether the difference between the information security rate calculated by the optimization result and the information security rate of the corresponding initial position meets the convergence threshold. If so, the phase shift of the optimal smart reflective surface and the position of the UAV are output. Otherwise, the iterative solution is continued until the convergence threshold or the set maximum number of iterations is reached, and the phase shift matrix of the optimal smart reflective surface and the position of the UAV are obtained. The phase shift optimization of the intelligent reflective surface based on continuous convex approximation and semi-definite relaxation technology and the position optimization of the drone based on the deep deterministic policy gradient algorithm are used to maximize the total confidentiality capacity of the cooperative communication through alternating iterative optimization models.
2. The air-ground coordinated secure transmission method based on rate splitting multiple access according to claim 1 is characterized in that: The UAV carrying the intelligent reflective surface ensures cooperative communication between the source base station and the ground user when the direct communication link is interrupted, including: Smart reflective surface is yoz Planar and has M A reflective element, M The reflective elements are arranged in a rectangular shape and adopt a uniform planar array; The location of the source base station is expressed as , represents the two-dimensional coordinates of the source base station; The position of the UAV carrying the smart reflector is expressed as , Represents the two-dimensional coordinates of the drone; The locations of the ground users and eavesdroppers are expressed as and ,in Represents the two-dimensional coordinates of the ground user, Represents the two-dimensional coordinates of the eavesdropper.
3. The air-ground coordinated secure transmission method based on rate splitting multiple access according to claim 1 is characterized in that: The source base station divides the transmitted information into public information and private information, including: In the downlink communication network of cooperative communication, based on the rate split multiple access technology, the source base station divides the transmitted information into public information and private information of each ground user, and performs linear precoding through the precoder. When the ground user receives the received signal, it first decodes the public information and then decodes the private information, where: No. k The received signal received by the ground user It is expressed as: Formula (1) In formula (1), Indicates that the smart reflective surface is k The channel coefficient matrix of the terrestrial user is: represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the common information, Indicates the public information flow sent by the source base station after division, express k The beamforming vector of the user's private information, Indicates that the source base station sends k Private information streams sent by ground users, represents the noise received by ground users, express terrestrial users.
4. The air-ground coordinated secure transmission method based on rate splitting multiple access according to claim 1, characterized in that: The information security rate of the ground user includes the information rate of the ground user decoding public information and the information rate of the ground user decoding private information; Among them, the information rate of public information is expressed as: Formula (2) In formula (2), Indicates that the smart reflective surface is k The channel coefficient matrix of the terrestrial user is: represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, The beamforming vector representing the common information, express terrestrial users, Indicates The beamforming vector of the private information of the ground user, represents the signal bandwidth, represents the noise power; No. k The information rate and eavesdropping rate of a ground user decoding private information are expressed as: Formula (3) In formula (3), Indicates that the smart reflective surface is k The channel coefficient matrix of the terrestrial user is: represents the channel coefficient matrix from the smart reflection surface to the eavesdropper, represents the phase shift matrix of the smart reflective surface, represents the channel coefficient matrix from the source base station to the smart reflection surface, Indicates k The beamforming vector of the private information of the ground user, Indicates The beamforming vector of the private information of the ground user, represents the signal bandwidth, represents the noise power; No. k The information security rate of a ground user is expressed as: Formula (4) In formula (4), represents the information rate at which ground users decode private information, represents the information rate at which an eavesdropper decodes private information, express .
5. The air-ground coordinated secure transmission method based on rate splitting multiple access according to claim 4 is characterized in that: The total confidentiality capacity is expressed as: Formula (5) In formula (5), represents the information security rate of the first ground user, represents the information security rate of the second ground user, Indicates k The information security rate of each ground user, express terrestrial users.
6. A method for air-ground collaborative secure communication based on rate splitting multiple access, characterized in that: Based on the air-ground collaborative secure transmission method based on rate splitting multiple access according to any one of claims 1 to 5, the air-ground collaborative secure and confidential communication method includes: Optimize the phase shift of smart reflective surface in cooperative communication based on continuous convex approximation and semi-definite relaxation technology; Obtain the position of the UAV in cooperative communication based on the deep deterministic policy gradient algorithm; Determine the total confidentiality capacity of cooperative communication by using the phase shift of the intelligent reflective surface, the position of the UAV, and the information security rate of the ground user, and jointly optimize the total confidentiality capacity by using an alternating iterative optimization model; Air-to-ground collaborative secure and confidential communications are performed based on the total confidentiality capacity.
