A robust and secure beamforming method in air-ground integrated networks based on queue stability constraints
Through a robust secure beamforming method based on queue stability constraints in the integrated space-space and earth network, the cloud computing and quantum computing challenges faced by traditional satellite communication network security protocols are solved, efficient and robust secure transmission is achieved, and communication is improved Safety and reliability, and optimize system energy efficiency while ensuring service quality.
Patent Information
- Application Number
- CN202410935399.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-12
AI Technical Summary
In the integrated air-space and earth network, the traditional satellite communication network security protocol is based on computing cryptography and faces the challenges of new technologies such as cloud computing and quantum computing. How to achieve efficient and robust secure transmission has become a difficult point, especially in the signal-to-noise ratio and there are strict requirements on interference suppression.
Using a robust secure beamforming method based on queue stability constraints, the beamforming rights are optimized by obtaining channel state information, building data queues, and establishing joint optimization problems. Vector, realizes the system's long-term average power minimization, while ensuring user service quality and data queue stability.
While ensuring the dynamic stability of the system, it improves the security performance of the integrated air-space and earth network, achieves a compromise between power consumption and queue stability, improves the security and reliability of communications, and optimizes system energy efficiency.
Smart Images

Figure CN118473510B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of wireless communications and discloses a robust security beamforming method in an air-ground integrated network based on queue stability constraints. Background Art
[0002] Satellite communication has the advantage of wide-area coverage that is difficult to match with ground wireless communication. Combining the advantages of satellite communication that is not restricted by geographical conditions and has a wide coverage area with the advantages of UAV communication with flexible coverage and low latency with ground wireless communication, we study the mobile communication system of Space-Air-Ground Integrated Network (SAGIN), taking into account the capacity advantages of ground systems and the coverage advantages of satellite systems, and constructing a new generation of mobile information network covering the air, land and sea to provide users with all-weather, all-round, high-quality and diversified services. This is the inevitable development trend of future mobile information networks and the inevitable choice for my country to realize global information services.
[0003] Compared with traditional terrestrial cellular networks, in order to improve spectrum utilization, SAGIN adopts full frequency reuse technology. However, this technology is often accompanied by more stringent requirements, such as higher signal-to-noise ratio and greater interference suppression capability. Therefore, how to explore how to achieve efficient and robust secure transmission under SAGIN from multiple dimensions has become a pain point that needs to be solved urgently. Traditional satellite communication network security protocols are mainly based on computational cryptography methods, and the computational complexity required to crack the key determines the effectiveness of the encryption algorithm. However, with the emergence of new technologies such as cloud computing and quantum computing, this key security system based on computational complexity faces huge challenges. In this case, the physical layer security (PLS) technology that uses the diversity and randomness of wireless channels to achieve secure information transmission has shown broad application prospects in the field of satellite communications. Summary of the invention
[0004] Purpose of the invention: To provide a robust and secure beamforming method in an air-ground integrated network based on queue stability constraints to solve the above-mentioned problems existing in the prior art.
[0005] Technical solution: A robust and secure beamforming method in an air-ground integrated network based on queue stability constraints, comprising the following steps:
[0006] S1. Obtain channel state information, and based on the channel state information, use multicast technology to build a signal transmission solution for users within the coverage of satellites, drones, and ground communication systems;
[0007] S2. Based on the signal transmission scheme, obtain the buffer status data of the satellite and the base station before transmission and build a data queue;
[0008] S3. Based on the signal transmission scheme and the data queue, a joint optimization problem is constructed. The objective function of the joint optimization problem is to minimize the long-term average power of the system. The constraints of the joint optimization problem include user service quality, eavesdropping probability, system transmission power and data queue stability.
[0009] S4, using Lyapunov optimization technology, the joint optimization problem is transformed into a single time slot power minimization problem;
[0010] S5. Bernstein inequality and continuous convex approximation algorithm are used in sequence to solve the minimization problem and obtain the beamforming weight vectors of the satellite, UAV and base station.
[0011] Beneficial effects: The present invention improves the security performance of the integrated air-space-ground network system while ensuring the dynamic stability of the system through a robust and secure beamforming algorithm; the queue stability condition is derived through a queue model, and a compromise between power consumption and queue stability can be achieved while ensuring security. BRIEF DESCRIPTION OF THE DRAWINGS
[0012] Figure 1 It is a flow chart of the present invention.
[0013] Figure 2 This is a model diagram of the air-ground-space integrated network downlink communication system of the present invention.
[0014] Figure 3 This is a flow chart of step S1 of the present invention.
[0015] Figure 4 Graph showing the relationship between queue length and cycle number under different parameters V in an embodiment of the present invention.
[0016] Figure 5 The system power consumption diagram under different schemes.
[0017] Figure 6 is the normalized eavesdropper threshold distribution map.
[0018] Figure 7 A flowchart of a robust and secure beamforming method in an air-ground integrated network based on queue stability constraints provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0019] like Figure 1 As shown, the present application proposes a robust security beamforming method in an air-ground integrated network based on queue stability constraints, comprising the following steps:
[0020] S1. Obtain channel state information, and based on the channel state information, use multicast technology to build a signal transmission solution for users within the coverage of satellites, drones, and ground communication systems;
[0021] S2. Based on the signal transmission scheme, obtain the buffer status data of the satellite and the base station before transmission and build a data queue;
[0022] S3. Based on the signal transmission scheme and the data queue, a joint optimization problem is constructed. The objective function of the joint optimization problem is to minimize the long-term average power of the system. The constraints of the joint optimization problem include user service quality, eavesdropping probability, system transmission power and data queue stability.
[0023] S4, using Lyapunov optimization technology, the joint optimization problem is transformed into a single time slot power minimization problem;
[0024] S5. Bernstein inequality and continuous convex approximation algorithm are used in sequence to solve the minimization problem and obtain the beamforming weight vectors of the satellite, UAV and base station.
[0025] In this embodiment, a signal transmission scheme is constructed by adopting multicast technology, and resources are dynamically allocated to users in satellites, drones and ground communication systems, thereby improving the efficiency and quality of signal transmission; by constructing a data queue, data caching and transmission are effectively managed and optimized, ensuring the stability of data flow, reducing transmission delay and packet loss rate; by considering multiple factors such as power, service quality, security and stability, the system can minimize power consumption and improve safety performance while ensuring user service quality; by using Lyapunov optimization technology, long-term joint optimization problems can be converted into single-time slot problems, simplifying the complexity of the problem and making real-time optimization possible; through the application of Bernstein inequality and continuous convex approximation algorithm, a more accurate and robust beamforming weight vector can be obtained.
[0026] This embodiment comprehensively considers factors such as channel status, data queues, and system power, and adopts optimization technology to provide an effective solution for secure communications in an integrated air-ground-space network. It can not only improve the security and reliability of communications, but also optimize the energy efficiency of the system while ensuring the quality of service.
[0027] In a further embodiment, a Space-Air-Ground Integrated Network (SAGIN) downlink communication scenario in which there are multiple eavesdroppers is discussed, such as Figure 2As shown in Figure 1, the space-based network consists of a geostationary satellite and an earth station; the air-based network consists of a drone; and the ground-based network consists of a base station and multiple ground users. In the case of shared spectrum resources, the satellite configuration has M s A single reflector antenna with 1 feed source provides services to N earth stations in the presence of K potential eavesdroppers (Eve) eavesdropping on private signals; the base station and the drone are equipped with N T and N s The uniform planar arrays (UPA) of array elements provide services for M ground users and send green interference to the satellite network. In order to suppress co-channel interference, satellites, base stations and drones all use beamforming technology. The data arrival of satellites and base stations changes dynamically over time. Satellites and base stations have a certain size buffer to store data before transmission, and use multicast technology to transmit data to service users in each time slot.
[0028] Firstly, when satellites, UAVs and base stations all use multicast technology to serve earth stations and ground users and realize spectrum resource sharing, a joint optimization problem is established with the goal of maximizing the achievable safety rate threshold of the earth station and the constraints of user service quality, earth station confidentiality interruption probability and transmission power meeting requirements; secondly, in order to solve this non-convex problem, Bernstein's inequality is used to deal with the probability constraints, and a robust security beamforming method based on semi-positive programming and Dinkelbach algorithm is further proposed to achieve safe and reliable transmission in the air-ground integrated network.
[0029] like Figure 3 As shown, step S1 is further as follows:
[0030] S11, obtaining channel state information, including a channel vector between the satellite and the user, a channel vector between the drone and the user, and a channel vector between the base station and the user;
[0031] S12, based on the channel state information, using beamforming technology to construct the earth station receiving signal, the eavesdropper receiving signal and the ground user receiving signal respectively;
[0032] S13, based on the earth station receiving signal, the eavesdropper receiving signal and the ground user receiving signal, a signal transmission scheme for the users within the coverage of the satellite, the UAV and the ground communication system is formed, and the earth station reachable rate, the eavesdropping rate and the ground user reachable rate are calculated at the same time;
[0033] In a specific embodiment, the user signals of the satellite, the UAV and the base station are represented as s(t), e(t) and x(t), respectively, and satisfy E[|s(t)| 2 ]=1, E[|e(t)| 2] = 1 and E[|x(t)| 2 ] = 1. Before sending the signal, the satellite, drone and base station all use beamforming (BF) technology to process the signal, and their BF weight vectors can be expressed as w∈C Ms×1 、v u ∈C Ns×1 and v b ∈C NT×1 .
[0034] The channel vector between the satellite and the i-th user is h i H (t)∈C Ms×1 , i∈{(s,n),(e,k)}, where C represents a set of complex numbers, M s represents the number of satellite feeds, (s, n) represents the communication between the sth satellite and the nth earth station, (e, k) represents the communication between the e-th eavesdropper and the k-th eavesdropper; the channel vector between the UAV and the k-th user is f j H (t)∈C Ns×1 , j∈{(s,n),(e,k),(b,m)}, where N s represents the number of UAV feed sources, (b, m) represents the communication between the bth ground base station and the mth ground user; the channel vector between the base station and the jth user is g j H (t)∈C NT×1 , j∈{(s,n),(e,k),(b,m)}, where N T Indicates the number of base station feed sources, n s,i (t) is the zero-mean additive white Gaussian noise of the earth station.
[0035] The satellite sends a signal s(t) to the earth station in the presence of a potential eavesdropper. Considering the interference signal from the base station, the received signal of the earth station can be expressed as:
[0036] y s,n (t)=h s,n H (t)w(t)s(t)+ f s,n H (t)v u (t)e(t)+ g s,n H (t)v b (t)x(t)+n s,n (t)
[0037] Among them, h s,n (t)∈C Ms×1 、f s,n (t)∈CNs×1 and g s,n (t)∈C NT×1 are the channel vectors between satellite, UAV, base station and earth station respectively; n s,n (t) is the zero-mean additive white Gaussian noise of the earth station, w∈C Ms×1 、v u ∈C Ns×1 and v b ∈C NT×1 are the beamforming weight vectors of satellite, UAV and base station, s(t), e(t) and x(t) represent the user signals of satellite, UAV and base station respectively.
[0038] Similar to the earth station, the signal received by the kth eavesdropper can be expressed as:
[0039] y e,k (t)=h e,k H (t)w(t)s(t)+ f e,k H (t)v u (t)e(t)+ g e,k H (t)v b (t)x(t)+n e,k (t)
[0040] Among them, h e,k (t), f e,k (t) and g e,k (t) are the channel vectors between the satellite, UAV, base station and eavesdropper respectively; n e,k (t) is the zero-mean additive white Gaussian noise of the kth eavesdropper.
[0041] In actual situations, satellite earth stations generally use high-gain antennas to compensate for the path loss caused by long-distance transmission, while ground mobile terminals generally use low-gain omnidirectional antennas. Therefore, without loss of generality, this application does not consider the interference of satellite signals to ground terminals, and the ground user receiving signal can be expressed as:
[0042] y b,m (t) = f b,m H (t)v u (t)e(t)+ g b,m H (t)v b (t)x(t)+n b,m (t)
[0043] where f b,m (t) and g b,m(t) are the channel vectors between the UAV and the base station and the ground user, n b,m (t) is the zero-mean additive white Gaussian noise of the ground user.
[0044] According to the above formula, the achievable rate of the earth station can be expressed as:
[0045] R s,n =log2(1+γ s,n )=log2(1+|h s,n H (t)w(t)∣ 2 / (|f s,n H (t)v u (t)∣ 2 +|g s,n H (t)v b (t)∣ 2 +σ s,n 2 ))
[0046] where γ s,n represents the output signal-to-interference-noise ratio of the nth earth station, σ s,n 2 represents the corresponding noise power;
[0047] The achievable rate of ground users can be expressed as:
[0048] R b,m =log2(1+γ b,m )=log2(1+(|f b,m H (t)v u (t)∣ 2 +|g b,m H (t)v b (t)∣ 2 ) / σ b,m 2 )
[0049] where γ b,m represents the output signal-to-interference-noise ratio of the mth ground user, σ b,m 2 represents the corresponding noise power;
[0050] The eavesdropper's eavesdropping rate can be expressed as:
[0051] R e,k =log2(1+γ e,k )=log2(1+|h e,k H (t)w(t)∣2 / (|f e,k H (t)v u (t)∣ 2 +|g e,k H (t)v b (t)∣ 2 +σ e,k 2 ))
[0052] where γ e,k represents the output signal-to-interference-noise ratio of the kth eavesdropper, σ e,k 2 represents the corresponding noise power.
[0053] Furthermore, in the space-based network, considering the effects of path loss, rain attenuation and beam gain, the downlink channel between the satellite and the user can be expressed as:
[0054] h i =sqrt(G i,t G i,r C i )⊙b* 1 / 2 ⊙r -1 / 2 ⊙h i '
[0055] Among them G i,t represents the satellite transmission gain, G i,r is the parabolic antenna gain of the earth station, C i represents the free space loss between the satellite and the user, r∈C Ms×1 represents the rain attenuation vector; b* represents the spot beam gain, ⊙ represents the Hadamard product, h i ' is composed of the phase difference from the earth station to different satellite feed sources, and its elements are h m '=e' -j ' 2πdm / λ ', e' represents the basis of the exponential function, j' represents the imaginary unit, d m represents the distance to the mth satellite feed, and λ' represents the wavelength.
[0056] For air-based and ground-based networks, the base station and the drone are equipped with N T and N s The ground downlink channel is a uniform array of elements. Due to the high directional and quasi-optical properties of millimeter wave transmission, the ground downlink channel can be regarded as the superposition of the main line-of-sight (LoS) transmission component and the sparse single non-line-of-sight (NLoS) transmission component. Therefore, the ground downlink channel matrix can be expressed as:
[0057] h = sqrt(g(θ0, φ0))ρ0a a (θ0, φ0)θa e (θ0, φ0)+ sqrt(1 / N)∑ n=1 N sqrt(g(θ n ,φ n ))ρ n a a (θ n , φ n )Θa e (θ n , φ n )
[0058] Where N is the number of NLoS paths, ρ0 and ρ n (n=1,…N) denote the complex channel gains associated with the LoS path and the nth NLoS path, respectively. g(θ, φ) denotes the co-directional directivity pattern of the antenna unit, where θ and φ are the vertical and horizontal channel angles, respectively. a (θ, φ)∈C N1×1 and a e (θ, φ)∈C N2×1 are the azimuth and elevation direction vectors of the UPA, respectively, and Θ represents the Kronecker product.
[0059] In practical applications, due to the existence of estimation errors and quantization errors, it is difficult for the fusion network to obtain perfect channel state information. Therefore, the actual channel vector can be expressed as:
[0060] h i =h* i +Δh i , i∈{(s,n),(e,k)}
[0061] f i =f* i +Δf i ,i∈{(s,n),(b,m),(e,k)}
[0062] g i =g* i +Δg i ,i∈{(s,n),(b,m),(e,k)}
[0063] Among them, h* i 、f* i , g* i and Δh i , Δf i , Δg i denote the estimated channel part and the channel estimation error part respectively, and Δh i~CN(0,E i ), Δf i ~CN(0,E i ), Δg i ~CN(0,E i ), error matrix E i =ε i I i , where ε represents the channel estimation error factor, I i Represents the identity matrix.
[0064] Since the arriving data cannot be transmitted to the user immediately, the already arrived data is stored in a buffer with infinite capacity, and these newly arrived data can only be served in the time slot and later. Assuming that the data arriving at the satellite and base station follow the Poisson distribution, the downlink transmission of the satellite and base station can be conceptualized as a queuing process. Y(t) and Q(t) represent the queue lengths at the satellite and base station at time t, and the dynamic characteristics of the queue can be written as:
[0065] Y(t+1)=max[Y(t)-R s (t), 0]+A s (t)
[0066] Q(t+1)=max[Q(t)-R b (t), 0]+A b (t)
[0067] Among them, the data arrival process in the data queue is subject to the parameters λ s and λ b Poisson distribution, A s (t) and A b (t) represents the data arrival rate at the satellite and base station sides, which is used to describe the increase in the number of data packets in the satellite and base station queues, respectively. s (t)=min n R s,n (t) and R b (t)=min m R b,m (t) represents the packet departure rate of the satellite and base station, which is used to describe the amount of packets reduced in the satellite and base station queues, respectively.
[0068] Taking the minimization of the long-term average power of the system as the optimization goal, considering the user service quality, eavesdropping probability, transmission power budget and data queue stability constraints, the secure transmission of information in SAGIN is achieved by jointly optimizing the BF weight vectors of the satellite and base station. The above optimization problem can be expressed as:
[0069]
[0070] Where T represents the total duration, E[ ] represents the expected value, Pr{} represents the probability, and I e,th represents the eavesdropping threshold, p e,out represents the probability of eavesdropping, p b,out represents the ground user outage probability, p s,out represents the earth station outage probability, Indicates all, P s,th Indicates the maximum satellite transmission power, P b,th Indicates the maximum transmission power of the base station, P u,th represents the maximum transmit power of the UAV; P1 represents the minimization of the long-term average power of the system; C1 and C2 represent queue stability constraints; C3 represents the eavesdropping probability constraint; C4 and C5 represent the service quality requirements of the mth ground user and the nth earth station respectively; C6 represents the transmit power budget of the satellite, base station and UAV.
[0071] Through Lyapunov optimization technology, the long-term optimization problem is transformed into a single-time slot minimization problem:
[0072]
[0073] Where V represents the control parameter, which is a non-negative weight parameter, such as Figure 4 As shown, it indicates the importance of penalty relative to the stability of the data queue, and reflects the priority of optimization during the execution of the algorithm. When it is large enough, the time-averaged power consumption can be arbitrarily close to the optimal value.
[0074] Substituting the earth station achievable rate, eavesdropping rate and terrestrial user achievable rate into the minimization problem, the minimization problem is transformed into:
[0075]
[0076] Where W = ww H , V u =v u v u H , V b =v b v b H , Tr( ) represents the trace of the matrix, and rank( ) represents the rank of the matrix.
[0077] Furthermore, due to the existence of channel estimation errors, the probability constraint form of C1'~C3' is difficult to solve, and an iterative algorithm combined with Bernstein's inequality is proposed.
[0078] Using Bernstein's inequality type I and transforming h e,k =h* e,k +Δh e,k 、fe,k =f* e,k +Δf e,k and g e,k =g* e,k +Δg e,k Substituting into C1', the C1' constraint can be transformed into the following form:
[0079] Pr{x g,k H A 1,k x h,k -x g,k H B 1,k x g,k +2Re (x h,k H a 1,k )-2Re (x g,k H b 1,k ) ≤c 1,k}≥1-p e,out
[0080] Among them, x g,k , A 1,k , x h,k , B 1,k , a 1,K and b 1,k All represent matrices used in constraint transformation;
[0081] c 1,k =(2 Ie,th -1) σ e,k 2 -h* e,k H W h* e,k + f* e,k H ( (2 Ie,th -1) V u ) f* e,k + g* e,k H ( (2 Ie ,th -1) V b ) g* e,k ;
[0082] Δh e,k =E h,k 1 / 2 x h,k , x h,k ~CN(0,I Ms );
[0083] Δf e,k =Ef,k 1 / 2 x f,k , x f,k ~CN(0,I Ns );
[0084] Δg e,k =E g,k 1 / 2 x g,k , x g,k ~CN(0,I NT );
[0085] A 1,k = E h,k 1 / 2 WE h,k 1 / 2 ;
[0086] a 1,k = E h,k 1 / 2 W h* e,k ;
[0087] B 1,k = E g,k 1 / 2 ( (2 Ie,th -1) V b ) E g,k 1 / 2 ;
[0088] b 1,k = E g,k 1 / 2 ( (2 Ie,th -1) V b ) g* e,k ;
[0089] D 1,k = E f,k 1 / 2 ( (2 Ie,th -1) V u ) E f,k 1 / 2 ;
[0090] d 1,k =E f,k 1 / 2 ( (2 Ie,th -1) V u ) f* e,k .
[0091] Among them, E h,k , E f,k , E g,k represents the identity matrix, D 1,k , d 1,krepresents the matrix used in the probability constraint,
[0092] Let x k =[ x h,k ;x g,k ;x f,k ],m 1,k =[ a 1,k ; -b 1,k ; -d 1,k ],
[0093]
[0094] In order to facilitate the solution, an auxiliary variable λ is introduced 1,k and μ 1,k , the problem can be transformed into:
[0095] Among them, M 1,k represents the matrix used to transform the probability constraint of the eavesdropper into the convex optimization problem, λ 1,k and μ 1,k represents auxiliary variables, c 1,k The value used in the constraint transformation of the eavesdropper, m 1,k represents the vector used in the probability constraint of the eavesdropper, and I represents the identity matrix;
[0096] Similar to the above problem, by introducing the auxiliary variable λ 2,m and μ 2,m , then the C2' constraint can be transformed into the following form:
[0097] Among them, M 2,m represents the matrix used to transform the probability constraints of ground users into the convex optimization problem, λ 2,m and μ 2,m represents auxiliary variables, c 2,m Represents the value in the constraint conversion for ground users, m 2,m represents the vector in the probability constraint for ground users;
[0098] Using Bernstein inequality type II, by introducing the auxiliary variable λ 3,n and μ 3,n , the C3' constraint can be transformed into:
[0099] Among them, M 3,n represents the matrix used to transform the probability constraints of the earth station into the convex optimization problem, λ 3,n and μ 3,n represents auxiliary variables, c 3,n Indicates the value used in the constraint transformation, m 3,n represents the vector used in the probability constraint;
[0100] Then, without considering the rank-one constraint, the problem is transformed into a convex problem. The optimization problem can be expressed as:
[0101]
[0102] The convex optimization problem is solved iteratively to obtain the beamforming weight vectors of the satellite, UAV and base station, and complete the transmission scheme design of the entire system. The process includes:
[0103] Set the total duration T and set t=0;
[0104] Set the initial queue lengths Q(0) and Y(0), ∨m, n, control parameters V, and achievable rate R s,n , R b,m and R e,k ;
[0105] When t≤T, the packet arrival rate A is obtained s (t) and A b (t), and solve to obtain R s (t) and R b (t), update the queue lengths Q(t) and Y(t) according to the corresponding formula;
[0106] Update the number of iterations t=t+1;
[0107] If rank (W) = 1, the BF weight vector is obtained by eigenvalue decomposition, otherwise the BF weight vector is obtained by Gaussian randomization method;
[0108] Simulation parameter settings:
[0109] Satellite orbit: GEO;
[0110] Satellite feed number M s : M s =7;
[0111] Number of base station feed sources N T : N T =8×8;
[0112] Number of drone feed sources N s : N s =8×8;
[0113] Carrier frequency f: f=20GHz;
[0114] Noise bandwidth B: B=50MHz;
[0115] Noise temperature O: O=300K;
[0116] 3dB angle: Φ3dB =0.4°;
[0117] Satellite antenna gain b max : b max =52dB;
[0118] NLoS path number L n : L n =5;
[0119] In another embodiment, using Bernstein inequality type I, an auxiliary variable λ is introduced 1,k and μ 1,k , transform constraint C1 into:
[0120] Tr(M 1,k )+sqrt(-2In(p e,out )) sqrt(||vec(M 1,k ) ∣∣+2∣∣m 1,k ∣∣ 2 )- In(p e,out )s + (M 1,k )≤c 1,k , ∨k
[0121] Where ∨ represents all;
[0122] Introducing auxiliary variable λ 2,m and μ 2,m , transform constraint C2 into:
[0123] Tr(M 2,m -sqrt(-2In(p b,out )) sqrt(||vec(M 2,m ) ∣∣+2∣∣m 2,m ∣∣ 2 )- In(p b,out )s - (M 2,m )≤c 2,m , ∨m
[0124] Similarly, we introduce the auxiliary variable λ 3,n and μ 3,n , using Bernstein inequality type II, the C3 constraint can be transformed into:
[0125] Pr{ x h,n H A n x h,n -x f,n H B n x f,n +2Re (xh,n H a n )-2Re (x f,n H b n ) ≥c n}≥1-p s,out
[0126] Further, such as Figure 5 As shown, the relationship between power consumption and the number of cycles under different schemes is shown. It can be observed that as the number of cycles increases, the scheme proposed in this application gradually tends to the power minimization algorithm. This observation shows that the scheme proposed in this application can fully address the impact of the number of random data, thereby promoting long-term power minimization. In contrast, the ZF algorithm and the MRT algorithm exhibit poor performance. This is because the ZF scheme does not involve the joint optimization of beamforming BF and power, while the MRT scheme does not consider network interference. Therefore, both methods fail to fully suppress interference and require higher power levels to meet QoS requirements. In addition, the queue minimization scheme always adopts maximum power transmission to ensure the queue length, resulting in the highest power consumption. The above observations demonstrate the superiority of the proposed scheme.
[0127] like Figure 6 As shown, the distribution of the normalized eavesdropper threshold is shown. As expected, the scheme proposed in this application takes the influence of channel error into account during the design process and can effectively meet the confidentiality requirements. In the non-robust scheme, about 50% of the values exceed the expected threshold. This confirms that the scheme proposed in this application can greatly reduce the influence of channel estimation error, thereby significantly improving the confidentiality rate of satellite communications.
[0128] In another embodiment, Figure 7 As shown, a robust and secure beamforming method based on imperfect channel state information is provided for air-ground integration, specifically:
[0129] Step 1: Using the acquired channel state information, the satellite communication system, the UAV communication system and the ground communication system use multicast technology to transmit signals to users within their coverage areas;
[0130] Step 2: Based on the time-varying characteristics of the network topology and considering the impact of random arrival of data packets, data queues and virtual queues are constructed according to the buffer states of the satellite and base station before transmission;
[0131] Step 3: In the case where satellites, drones, and base stations share spectrum resources, a joint optimization problem is established with the minimization of long-term average power as the objective function and the user service quality, eavesdropping probability, system transmission power, and data queue stability meeting the requirements as constraints;
[0132] Step 4: Use Lyapunov optimization technology to transform the original problem into a single-slot power minimization problem, and further combine Bernstein inequality and continuous convex approximation algorithm to obtain the beamforming weight vectors of satellites, drones and base stations, and complete the design of the entire system transmission plan.
[0133] This application uses multicast technology to serve earth stations and ground users while taking into account the sharing of spectrum resources by satellites, drones and base stations, and suppresses various interferences within and between networks through beamforming technology. At the same time, it considers the impact of random arrival of data packets based on the time-varying characteristics of the network topology, significantly improves the confidentiality performance, and achieves a good compromise between power consumption and queue stability while ensuring security.
[0134] The preferred embodiments of the present invention are described in detail above; however, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all belong to the protection scope of the present invention.
Claims
1. A robust secure beamforming method in an air-ground integrated network based on queue stability constraints, characterized in that: The steps include: S1. Obtain channel state information, and based on the channel state information, use multicast technology to build a signal transmission solution for users within the coverage of satellites, drones, and ground communication systems; S2. Based on the signal transmission scheme, obtain the buffer status data of the satellite and the base station before transmission and build a data queue; S3, construct a joint optimization problem based on the signal transmission scheme and data queue; The objective function of the joint optimization problem is to minimize the long-term average power of the system. The constraints of the joint optimization problem include user service quality, eavesdropping probability, system transmission power and data queue stability. S4, using Lyapunov optimization technology, the joint optimization problem is transformed into a single time slot power minimization problem; S5. The minimization problem is solved by using Bernstein inequality and continuous convex approximation algorithm in sequence to obtain the beamforming weight vectors of the satellite, UAV and base station; Step S1 is further as follows: S11, obtaining channel state information, including a channel vector between the satellite and the user, a channel vector between the drone and the user, and a channel vector between the base station and the user; S12, based on the channel state information, using beamforming technology to construct the earth station receiving signal, the eavesdropper receiving signal and the ground user receiving signal respectively; S13, based on the earth station receiving signal, the eavesdropper receiving signal and the ground user receiving signal, a signal transmission scheme for the users within the coverage of the satellite, the UAV and the ground communication system is formed, and the earth station reachable rate, the eavesdropping rate and the ground user reachable rate are calculated at the same time; The channel vector between the satellite and the i-th user is h i H (t)∈C Ms×1 , i∈{(s,n),(e,k)}, where C represents a set of complex numbers, M s represents the number of satellite feeds, (s, n) represents the communication between the s-th satellite and the n-th earth station, and (e, k) represents the communication between the e-th eavesdropper and the k-th eavesdropper; The channel vector between the drone and the kth user is f j H (t)∈C Ns×1 , j∈{(s,n),(e,k),(b,m)}, where N s represents the number of UAV feed sources, (b, m) represents the communication between the bth ground base station and the mth ground user; The channel vector between the base station and the jth user is g j H (t)∈C NT×1 , j∈{(s,n),(e,k),(b,m)}, where N T Indicates the number of base station feed sources; The signal received by the earth station is: y s,n (t)=h s,n H (t)w(t)s(t)+ f s,n H (t)v u (t)e(t)+ g s,n H (t)v b (t)x(t)+n s,n (t) Among them, h s,n (t)∈C Ms×1 、f s,n (t)∈C Ns×1 and g s,n (t)∈C NT×1 are the channel vectors between satellite, UAV, base station and earth station respectively; n s,n (t) is the zero-mean additive white Gaussian noise of the earth station, w∈C Ms×1 、v u ∈C Ns×1 and v b ∈C NT×1 are the beamforming weight vectors of satellite, UAV and base station, s(t), e(t) and x(t) represent the user signals of satellite, UAV and base station respectively; The eavesdropper receives the signal: y e,k (t)=h e,k H (t)w(t)s(t)+ f e,k H (t)v u (t)e(t)+ g e,k H (t)v b (t)x(t)+n e,k (t) Among them, h e,k (t), f e,k (t) and g e,k (t) are the channel vectors between the satellite, UAV, base station and eavesdropper respectively; n e,k (t) is the zero-mean additive white Gaussian noise of the kth eavesdropper; The signal received by the ground user is: y b,m (t)= f b,m H (t)v u (t)e(t)+ g b,m H (t)v b (t)x(t)+n b,m (t) where f b,m (t) and g b,m (t) are the channel vectors between the UAV and the base station and the ground user, n b,m (t) is the zero-mean additive white Gaussian noise of the ground user; The achievable rate of the earth station is: R s,n =log2(1+γ s,n )=log2(1+∣ h s,n H (t)w(t)∣ 2 / (∣f s,n H (t)v u (t)∣ 2 +∣g s,n H (t)v b (t)∣ 2 +σ s,n 2 )) where γ s,n represents the output signal-to-interference-noise ratio of the nth earth station, σ s,n 2 represents the corresponding noise power; The achievable rate for ground users is: R b,m =log2(1+γ b,m )=log2(1+(∣f b,m H (t)v u (t)∣ 2 +∣g b,m H (t)v b (t)∣ 2 ) / σ b,m 2 ) where γ b,m represents the output signal-to-interference-noise ratio of the mth ground user, σ b,m 2 represents the corresponding noise power; The eavesdropping rate is: R e,k =log2(1+γ e,k )=log2(1+∣ h e,k H (t)w(t)∣ 2 / (∣f e,k H (t)v u (t)∣ 2 +∣g e,k H (t)v b (t)∣ 2 +σ e,k 2 )) where γ e,k represents the output signal-to-interference-noise ratio of the kth eavesdropper, σ e,k 2 represents the corresponding noise power; The data queue is: Y(t+1)=max[Y(t)-R s (t),0]+A s (t) Q(t+1)=max[Q(t)-R b (t),0]+A b (t) Among them, the data arrival process in the data queue is subject to the parameters λ s and λ b Poisson distribution, A s (t) and A b (t) represents the data arrival rate at the satellite and base station sides, which is used to describe the increase in the number of data packets in the satellite and base station queues, respectively. s (t)=min n R s,n (t) and R b (t)=min m R b,m (t) represents the packet departure rate of the satellite and base station, which is used to describe the amount of packets reduced in the satellite and base station queues, respectively. Y(t) and Q(t) represent the queue lengths at the satellite and base station at time t; The joint optimization problem is specifically: ; Where T represents the total duration, E[ ] represents the expected value, Pr{} represents the probability, and I e,th represents the eavesdropping threshold, p e,out represents the probability of eavesdropping, p b,out represents the ground user outage probability, p s,out represents the earth station outage probability, Indicates all, P s,th Indicates the maximum satellite transmission power, P b,th Indicates the maximum transmission power of the base station, P u,th represents the maximum transmission power of the UAV; P1 represents the minimization of the long-term average power of the system; C1 and C2 represent the queue stability constraints; C3 represents the eavesdropping probability constraint; C4 and C5 represent the service quality requirements of the mth ground user and the nth earth station respectively; C6 represents the transmission power budget of the satellite, base station and UAV; The single-slot power minimization problem is: ; Where V represents the control parameter; Substituting the earth station achievable rate, eavesdropping rate and terrestrial user achievable rate into the minimization problem, the minimization problem is transformed into: ; Where W = ww H , V u =v u v u H , V b =v b v b H , Tr() represents the trace of the matrix, and rank() represents the rank of the matrix; Step S5 is further as follows: Using the Bernstein inequality iterative algorithm, the C1' constraint is transformed into: ; Among them, M 1,k represents the matrix used to transform the probability constraint of the eavesdropper into the convex optimization problem, λ 1,k and μ 1,k represents auxiliary variables, c 1,k represents the value used in the constraint transformation of the eavesdropper, m 1,k represents the vector used in the probability constraint of the eavesdropper, and I represents the identity matrix; The C2' constraint is converted to: ; Among them, M 2,m represents the matrix used to transform the probability constraints of ground users into the convex optimization problem, λ 2,m and μ 2,m represents auxiliary variables, c 2,m Represents the value in the constraint conversion for ground users, m 2,m represents the vector in the probability constraint for ground users; The C3' constraint is converted to: ; Among them, M 3,n represents the matrix used to transform the probability constraints of the earth station into the convex optimization problem, λ 3,n and μ 3,n represents auxiliary variables, c 3,n Indicates the value used in the constraint transformation, m 3,n represents the vector used in the probability constraint; Convert the minimization problem to a convex optimization problem: ; The convex optimization problem is iteratively solved to obtain the beamforming weight vectors of the satellite, UAV and base station.
Citation Information
Patent Citations
Artificial noise assisted beam forming scheme with robustness for related main stealing channel
CN113630163A
Scheduling and beam forming method for improving energy efficiency of high-reliability low-delay communication system
CN116709550A
Space-air-ground integrated UAV-assisted IoT data collectioncollection method based on aoi
US20230239037A1
