CN117424633A - Secure communication transmission strategy and system under deep learning auxiliary active ARIS - Google Patents

Abstract

The invention discloses a secure communication transmission strategy and a secure communication transmission system under deep learning auxiliary active ARIS, which jointly consider the horizontal position of the ARIS and a beam forming matrix of the ARIS to maximize the security rate of a network. In addition, compared with the simple consideration of the paired phase shift circuit and the power amplifier in the ARIS, the power consumption of the active ARIS is fully considered, the total power consumption of the network is greatly reduced, and the whole network has minimum safety energy consumption.

Description

Secure communication transmission strategy and system under deep learning auxiliary active ARIS

Technical Field

The invention belongs to the technical field of communication, and particularly relates to a secure communication transmission strategy and system under deep learning auxiliary active ARIS.

Background

To meet the increasing demand for ultra-reliable low-latency communications in the fifth generation (5G) and future mobile networks, reconfigurable smart surfaces (reconfigurable intelligent surface, RIS) have become a promising revolutionary technology. RIS plays a vital role in improving wireless network coverage, especially in typical scenarios where communication distance is long, there are obstacles on the ground, and transmit power is insufficient. In addition, from the point of view of physical layer security (physical layer security, PLS), RIS can amplify the channel difference between legitimate and eavesdropped channels, thereby enhancing privacy.

In general, a RIS with passive loading is considered passive, requiring the deployment of some of the necessary control and switching circuitry to design the beam forming. The active RIS is composed of a large number of active cells, each of which can reflect an incoming signal through an adjustable phase shift and amplifier.

However, current work with passive or active RIS networks always deploys RIS at fixed locations, such as the exterior walls of a building, which limits the application of RIS. Meanwhile, due to the agility and flexibility of the unmanned aerial vehicle (unmanned aerial vehicle, UAV), the UAV can be rapidly deployed to a target area, the network topology structure is dynamically changed, and a reliable communication link is established. Furthermore, the deployment of UAVs is economical, with the cost of UAVs continually decreasing.

Therefore, deploying RIS, that is, air RIS (ARIS), on the UAV can greatly improve the communication performance of the UAV, and is of great interest. However, the conventional optimization method relies on a mathematical model, which is very complex for handling the ARIS communication network. The artificial intelligence technology is adopted for ARS, so that the problems of network topology, deployment scale, deployment mode and the like of ARIS can be greatly simplified.

Disclosure of Invention

The technical problem to be solved by the invention is to provide a secure communication transmission strategy and system under deep learning auxiliary active ARIS for solving the technical problem of low secure rate in the existing ARIS-UAV network and effectively improving the security and quality of wireless communication.

The invention adopts the following technical scheme:

a secure communication transmission strategy under deep learning assisted active ARIS, comprising the steps of:

s1, constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

s2, constructing a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converting the problem P1 into a problem P2, and then decoupling the problem P2 into 2 sub-problems;

and S3, solving an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem through the 2 sub-problems obtained in the alternate iteration step S2, obtaining an ARIS optimal horizontal position and an ARIS optimal reflection coefficient matrix, and realizing the safety rate of the maximum combination method user and minimizing the total power consumption of the system.

Specifically, step S1 specifically includes:

constructing an ARIS-UAV secure communication network model comprising 1 ground access point, 1 UAV, 1 active RIS, 1 legal user and 1 eavesdropping user; the horizontal position of the ground access point is expressed asThe horizontal position of the legitimate user is denoted +.>The level of eavesdropping user is expressed asThe beamforming matrix of the active ARIS is +.>The UAV is continuously fixed at the height h above the ground, the total time T required by the unmanned aerial vehicle to execute the optimization task is divided into T time slots, and the horizontal position of the UAV at the T time slot is expressed as +. >T is more than or equal to 0 and less than or equal to T; establishing a communication model of a ground access point, an ARIS and a legal user/eavesdropping user; definition of legal subscribersAn achievable safe rate; calculating the total power P consumed by ARIS-UAV network _total [t]。

Further, a communication channel h between the ground access point and the ARIS _bu [t]Modeling is as follows:

wherein h is _bu [t]Responsible for large-scale fading effects such as path loss and shadowing, d _bu [t]For euclidean distance between the access point and the ARIS, β is the reference channel gain when distance d=1, and k is the path loss index;

the antenna array response vector of the RIS at the receiving end is expressed as:

wherein n=n _x N _y ,N _x And N _y The number of elements of the active RIS in the x-axis and the y-axis respectively; q _ris And lambda denote the element spacing and the communication wavelength of the RIS respectively,θ ^r respectively representing azimuth angle and pitch angle of the RIS representative receiving end;

the CSI model between RIS and eavesdropping user is expressed as:

wherein,for the channel estimation between the ARIS and eavesdropping user, delta _e In order to estimate the error vector(s),for the channel estimation error from the ith active RIS element to the eavesdropping user in time slot t, is>Thermal noise for eavesdropping on the user;

channel h between ARIS-legitimate users _ud [t]And ARIS-eavesdropping userModeling is as follows:

wherein,and->All representing a determined line of sight component, d _ud For the Euclidean distance between ARIS and legitimate users, d _ue [t]For the Euclidean distance between ARIS and eavesdropping user, < >>Antenna response vector when RIS is used as transmitting end, < >>And theta ^t Respectively representing azimuth angle and pitch angle of the transmitting end by ARIS, wherein K is a Lais factor, and ++>And->Is a non-line-of-sight component, isRandom small scale fading and obeys a complex gaussian distribution with a mean of 0 and a variance of 1.

Further, the security rate reachable by the legitimate user is:

R _act,s [t]＝W[log ₂ (1+Υ _act,d [t])-log ₂ (1+Υ _act,e [t])] ⁺

wherein [ x ]] ⁺ The operation shows that x is more than or equal to 0, W is the total bandwidth, gamma _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]To tap the signal-to-interference-and-noise ratio of the user.

Further, the ARIS-UAV network consumes the total power of:

wherein v is the inverse of the energy conversion coefficient of the active ARIS, P _UAV Maintaining hover and flight state total power consumption, W for UAVs _PS And W is _PA Static power consumption, W, of the phase shift circuit and the power amplifier, respectively _U And W is _E The power consumption of the mobile terminal is used for legal users and eavesdropping users respectively.

Specifically, step S2 specifically includes:

jointly considering a beam forming matrix and a horizontal position of an active ARIS, and modeling an SEE maximization problem of an ARIS-UAV network as a problem P1; converting the partial objective function in the problem P1 into a form of parameter subtraction by using a Dinkelbach algorithm to obtain a problem P2; the problem P2 is decoupled into an ARIS horizontal position optimization sub-problem and an ARIS beamforming matrix design sub-problem.

Further, the decoupling of the problem P2 into the ARIS horizontal position optimization sub-problem and the ARIS beamforming matrix design sub-problem is specifically as follows:

only constraints on the ARIS horizontal position are retained, and the ARIS horizontal position optimization sub-problem is expressed as follows:

only the constraints on the ARIS beamforming matrix are preserved, and the ARIS beamforming matrix design sub-problem is expressed as follows:

wherein R is _act,s [t]For the safe rate of legal users in time slot t, q [ t ]]For the horizontal position of ARIS in time slot t, h _bu [t]H is the channel between the access point and the legal user in time slot t _ud [t]For the channel between the ARIS-legitimate users at time slot t,for thermal noise of legal users, N is the number of ARIS phase shift elements, W _PS For static power consumption of the phase shift circuit, L is the number of ARIS power amplifiers, W _PA For the static power consumption of the power amplifier, +.>For ARIS maximum transmit power, +.>For the width of the UAV's flyable area, +.>Is the length of the UAV flyable area, T is the total UAV flight time, gamma _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]To eavesdrop on the signal-to-interference-and-noise ratio of the user, P _AP For the transmitting power of the ground access point, v is the reciprocal of the energy conversion coefficient of the active ARIS, a _l Is the first power amplifier, θ _n Is the phase shift of the nth element.

Specifically, the step S3 specifically includes:

solving a horizontal position optimization sub-problem of the UAV by using a DDPG algorithm;

the process for solving the ARIS beam forming matrix design sub-problem by using ADMM algorithm is as follows: firstly, giving initial values of all variables, carrying out solution on split variables under the k+1th iteration, then solving a beam forming matrix under the k+1th iteration, then solving auxiliary variables under the k+1th iteration, then solving Lagrangian multipliers under the k+1th iteration, and continuously increasing iteration times until reaching a maximum iteration times ending algorithm to obtain an optimal solution of the ARIS beam forming matrix;

solving a safety energy consumption maximization problem by using Dinkelbach algorithm: initializing variables, obtaining an optimal solution of the UAV horizontal position under the ith iteration, bringing the approximate solution of the UAV horizontal position under the ith iteration into a RIS beamforming matrix design sub-problem, solving an ARIS beamforming matrix approximate solution under the ith iteration, solving a safe energy consumption value under the ith iteration, continuously increasing the iteration times until the difference between P2 objective function values of the two iterations is within a certain range, obtaining the maximum safe energy consumption, and ending the algorithm.

Further, the solution of the UAV horizontal position optimization sub-problem by using the DDPG algorithm is specifically as follows:

Initializing various variables; resetting the environmental state; the intelligent agent obtains the current state, obtains a strategy according to an actor-online network, and obtains the current action after adding noise; executing actions, interacting with the environment, and obtaining instant rewards and new states; storing the status, actions, instant rewards, and new status to an experience buffer; if the capacity of the experience buffer zone reaches a set threshold value, extracting small batch data from the experience buffer zone to serve as data for training the neural network parameters, and updating the actor and critic network parameters; ending the algorithm when the maximum training round number is reached; if the capacity of the experience buffer zone does not reach the set threshold, the intelligent agent obtains the current state again, then obtains the strategy according to the actor-online network, and obtains the current action after adding noise.

In a second aspect, an embodiment of the present invention provides a secure communication transmission policy system under deep learning auxiliary active ARIS, including:

the construction module is used for constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

the decoupling module constructs a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converts the problem P1 into a problem P2 and then decouples the problem P2 into 2 sub-problems;

And the output module is used for solving the ARIS horizontal position optimization sub-problem and the ARIS beam forming matrix design sub-problem through the 2 sub-problems obtained in the alternate iteration step S2, obtaining the optimal horizontal position of the ARIS and the optimal reflection coefficient matrix of the ARIS, and realizing the safety rate of the maximum combination method user and the minimum total power consumption of the system.

Compared with the prior art, the invention has at least the following beneficial effects:

a secure communication transmission strategy under deep learning auxiliary active ARIS jointly considers the horizontal position of the ARIS and the beam forming matrix of the ARIS, and maximizes the security rate of the network. In addition, compared with the simple consideration of the paired phase shift circuit and the power amplifier in the ARIS, the power consumption of the active ARIS is fully considered, the total power consumption of the network is greatly reduced, and the whole network has minimum safety energy consumption.

Furthermore, an active ARIS-assisted downlink safety communication model is provided, and the problem of safety energy efficiency maximization is established by jointly optimizing the reflection matrix and the horizontal position of the ARIS.

Further, the Dinkelbach algorithm can convert the partial problem P1 into a parameter subtraction form of the problem P2.

Furthermore, the problem P2 is non-convex, and the variables are coupled, so that the problem P2 is difficult to directly solve, and therefore, the problem P2 is decoupled into an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem, and sub-optimal solutions are obtained through alternate iteration.

Further, the quadratic programming problem of the quadratic inequality constraint is obtained by searching the lower bound function of the non-convex problem P4.1, then the ADMM algorithm is utilized for iterative solution, and the closed solution is obtained in each iterative process, so that excellent convergence is provided.

Further, by utilizing the mobility and flexibility of the UAV, the DDPG model is trained by constructing a state space, an action space and a reward function to solve the Markov decision problem to obtain an optimal decision, and the horizontal position of the ARIS is adjusted more flexibly and effectively.

It will be appreciated that the advantages of the second aspect may be found in the relevant description of the first aspect, and will not be described in detail herein.

In summary, the invention uses the ARIS to direct the reflected signal beam to the legitimate user, so as to maximize the downlink security energy consumption.

The technical scheme of the invention is further described in detail through the drawings and the embodiments.

Drawings

FIG. 1 is a flow chart of a secure communication transmission method under deep learning auxiliary active ARIS in the scheme of the invention;

FIG. 2 is a diagram of an ARIS-UAV secure communication network model in the scheme of the invention;

FIG. 3 is a graph of the security energy consumption versus the transmit power of a ground access point for the present and baseline schemes;

FIG. 4 is a graph of the safe energy consumption versus legal user x-axis position for the present and baseline schemes;

fig. 5 is a graph of the safe energy consumption versus maximum power amplified for the inventive and baseline schemes.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In the description of the present invention, it will be understood that the terms "comprises" and "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

It is also to be understood that the terminology used in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations, e.g., a and/or B, may represent: a exists alone, A and B exist together, and B exists alone. In the present invention, the character "/" generally indicates that the front and rear related objects are an or relationship.

It should be understood that although the terms first, second, third, etc. may be used to describe the preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish one preset range from another. For example, a first preset range may also be referred to as a second preset range, and similarly, a second preset range may also be referred to as a first preset range without departing from the scope of embodiments of the present invention.

Depending on the context, the word "if" as used herein may be interpreted as "at … …" or "at … …" or "in response to a determination" or "in response to detection". Similarly, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determination" or "when detected (stated condition or event)" or "in response to detection (stated condition or event), depending on the context.

Various structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the accompanying drawings. The figures are not drawn to scale, wherein certain details are exaggerated for clarity of presentation and may have been omitted. The shapes of the various regions, layers and their relative sizes, positional relationships shown in the drawings are merely exemplary, may in practice deviate due to manufacturing tolerances or technical limitations, and one skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions as actually required.

The invention provides a safe communication transmission strategy under deep learning auxiliary active ARIS, which jointly considers the horizontal position of the ARIS and the wave beam forming matrix of the ARIS to maximize the safe rate of a network. In addition, compared with the simple consideration of the paired phase shift circuit and the power amplifier in the ARIS, the power consumption of the active ARIS is fully considered, the total power consumption of the network is greatly reduced, and the whole network has minimum safety energy consumption.

Referring to fig. 1, the secure communication transmission strategy under deep learning auxiliary active ARIS of the present invention includes the following steps:

s1, constructing an ARIS-UAV safety network communication model, wherein the ARIS-UAV safety network communication model comprises a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

The specific steps for constructing the ARIS-UAV safety network model are as follows:

s101, constructing an ARIS-UAV safety communication network model comprising 1 ground access point, 1 UAV, 1 active RIS, 1 legal user and 1 eavesdropping user; the horizontal position of the ground access point is expressed asThe horizontal position of the legitimate user is denoted +.>The level of eavesdropping user is expressed as

The beamforming matrix of the active ARIS isDividing the active ARIS evenly into M subarrays and noting that L=N/M represents the required amplification factor number, +.>For the ARIS phase diagonal matrix, for +.>j and phi _n Representing the imaginary unit and the phase of each active element respectively,amplification factor vector for active ARIS, < ->An index matrix representing the connection between the matrix Θ and the vector a. In a passive RIS, the +.>However, in the active RIS, due to the self-amplifying property, < > a->It is only feasible; in addition to this phi _n Can be optimized to any possible phase angle, with a feasible set of 0,2 pi.

The UAV is kept fixed at a height h above the ground, the total time T required by the unmanned aerial vehicle to perform the optimization task is divided into T time slots, and therefore the horizontal position of the UAV at the T time slot is expressed asT is more than or equal to 0 and less than or equal to T; when 1 slot->For a sufficient time, the UAV is set to remain stationary during the time slot; after 1 slot, the UAV is moved from position q [ t ] ]Fly to a new horizontal position q [ t+1 ]]：

Wherein v [ t ]]∈[0,v _max ]For flying unmanned aerial vehicle in time slot tLine speed, v _max For maximum flight speed, εt]∈[0,2π]The flight angle of the unmanned aerial vehicle in the time slot t is shown.

S102, establishing a communication model of a ground access point, an ARIS and a legal user/eavesdropping user;

since the line-of-sight path exists between the ground access point and the ARIS, the communication channel h between the ground access point and the ARIS _bu [t]Is modeled as:

wherein h is _bu [t]Responsible for large-scale fading effects such as path loss and shadowing, expressed asFor euclidean distance between the access point and the ARIS, β is the reference channel gain at distance d=1, and k is the path loss index.

Assuming that the RIS is composed of a uniform matrix array, then the antenna array response vector at the receiving end of the RIS is expressed as:

wherein n=n _x N _y ,N _x And N _y The number of elements of the active RIS in the x-axis and the y-axis respectively; q _ris And lambda denote the element spacing and the communication wavelength of the RIS respectively,respectively representing azimuth angle and pitch angle of the RIS representative receiving end;

for an ARIS-terrestrial link, one set of communication signals will be received in line-of-sight conditions and the other set of communication signals will be received by strong reflection and diffraction in non-line-of-sight conditions; since the channel between the RIS and the eavesdropping user is very difficult to estimate, it is assumed that the channel state information (channel state information, CSI) between the ARIS and the eavesdropping user is partly known; assuming that the CSI has an estimation error, the CSI model between the RIS and the eavesdropping user is expressed as:

Wherein,for the channel estimation between the ARIS and the eavesdropping user,/and/or>For estimating the error vector>For the channel estimation error from the ith active RIS element to the eavesdropping user in time slot t. />To eavesdrop on the thermal noise of the user. Channel h between ARIS-legitimate users _ud [t]And ARIS-eavesdropping user>Modeled as rice channel models, expressed as follows:

wherein,and->All representing a determined line of sight component, +.>For the Euclidean distance between ARIS and legitimate user, < >>For the Euclidean distance between ARIS and eavesdropping user, < >>Antenna response vector when RIS is used as transmitting end, < >>And theta ^t Respectively representing azimuth angle and pitch angle of the transmitting end by ARIS, wherein K is a Lais factor, and ++>And->Is a non-line-of-sight component, is a random small-scale fading and follows a composite gaussian distribution with a mean of 0 and a variance of 1.

S103, defining the reachable security rate of the legal user as follows:

R _act,s [t]＝W[log ₂ (1+Υ _act,d [t])-log ₂ (1+Υ _act,e [t])] ⁺

wherein [ x ]] ⁺ The operation shows that x is more than or equal to 0, W is the total bandwidth, gamma _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]Signal-to-interference-and-noise ratio for eavesdropping users;

according to shannon's formula, the signal-to-interference-and-noise ratios of the downlinks between the ARIS and legitimate users and the ARIS and eavesdropping users are expressed as:

wherein P is _AP Is the transmit power of the ground access point, And->Respectively representing the receiving noise of legal users and eavesdropping users; />

S104, calculating the total power consumed by the ARIS-UAV network as follows:

wherein v is the inverse of the energy conversion coefficient of the active ARIS, P _UAV Maintaining hover and flight state total power consumption, W for UAVs _PS And W is _PA Static power consumption, W, of the phase shift circuit and the power amplifier, respectively _U And W is _E The power consumption of the mobile terminal is used for legal users and eavesdropping users respectively.

S2, constructing a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converting the problem P1 into a problem P2, and decoupling the problem P2 into 2 sub-problems;

s201, jointly considering a beam forming matrix and a horizontal position of an active ARIS, and modeling an SEE maximization problem of an ARIS-UAV network as a problem P1:

wherein C1 represents a safe rate constraint for an authorized user, and it is noted that the active RIS amplifies both the received signal and the received noise of each reflective element, so that C2 represents a limit on the maximum transmit power of the RIS, C3 ensures that the active RIS operates in signal amplification mode, C4 represents a phase angle constraint for the RIS, and C5 represents a flight area of the drone.

S202, converting a partial objective function in a problem P1 into a form of parameter subtraction by using a Dinkelbach algorithm, and then expressing the problem P2 as follows:

S203, decoupling the problem P2 into 2 sub-problems, including an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem.

The problem P2 is decoupled into 2 sub-problems, which are specifically as follows:

a. only constraints on the ARIS horizontal position are retained, and the ARIS horizontal position optimization sub-problem is expressed as follows:

b. only the constraints on the ARIS beamforming matrix are preserved, and the ARIS beamforming matrix design sub-problem is expressed as follows:

s3, solving an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem through the problem P2 obtained in the alternate iteration step S2, obtaining an ARIS optimal horizontal position and an ARIS optimal reflection coefficient matrix, and simultaneously achieving the safety rate of the maximum combination method user and minimizing the total power consumption of the system.

S301, solving a UAV horizontal position optimization sub-problem by using a DDPG algorithm, wherein the process comprises the following steps:

(1) Initializing various variables;

(2) Resetting the environmental state;

(3) The intelligent agent obtains the current state and obtains the current action after adding noise according to the actor-online network obtaining strategy;

(4) Executing actions, interacting with the environment, and obtaining instant rewards and new states;

(5) Storing the status, actions, instant rewards, and new status to an experience buffer;

(6) If the capacity of the experience buffer reaches the set threshold value, executing (7), otherwise executing (3);

(7) Extracting small batches of data from the experience buffer area as data for training the neural network parameters, and updating the actor and critic network parameters;

(8) And ending the algorithm when the algorithm reaches the maximum training round number.

S302, solving an ARIS beam forming matrix design sub-problem by using an ADMM algorithm, wherein the process comprises the following steps: firstly, giving initial values of all variables, carrying out solution on split variables under the k+1th iteration, then solving a beam forming matrix under the k+1th iteration, then solving auxiliary variables under the k+1th iteration, then solving Lagrangian multipliers under the k+1th iteration, and continuously increasing iteration times until reaching a maximum iteration times ending algorithm to obtain an optimal solution of the ARIS beam forming matrix;

s303, solving a safety energy consumption maximization problem by using a Dinkelbach algorithm: initializing variables, obtaining an optimal solution of the UAV horizontal position under the ith iteration, bringing the approximate solution of the UAV horizontal position under the ith iteration into a RIS beamforming matrix design sub-problem, solving an ARIS beamforming matrix approximate solution under the ith iteration, solving a safe energy consumption value under the ith iteration, continuously increasing the iteration times until the difference between P2 objective function values of the two iterations is within a certain range, obtaining the maximum safe energy consumption, and ending the algorithm.

In still another embodiment of the present invention, a secure communication transmission policy system under deep learning auxiliary active ARIS is provided, where the system can be used to implement the secure communication transmission policy under deep learning auxiliary active ARIS described above, and specifically, the secure communication transmission policy system under deep learning auxiliary active ARIS includes a construction module, a decoupling module, and an output module.

The system comprises a construction module, a communication module and a communication module, wherein the construction module is used for constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

the decoupling module constructs a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converts the problem P1 into a problem P2 and then decouples the problem P2 into 2 sub-problems;

and the output module is used for solving the ARIS horizontal position optimization sub-problem and the ARIS beam forming matrix design sub-problem through the 2 sub-problems obtained in the alternate iteration step S2, obtaining the optimal horizontal position of the ARIS and the optimal reflection coefficient matrix of the ARIS, and realizing the safety rate of the maximum combination method user and the minimum total power consumption of the system.

In yet another embodiment of the present invention, a terminal device is provided, the terminal device including a processor and a memory, the memory for storing a computer program, the computer program including program instructions, the processor for executing the program instructions stored by the computer storage medium. The processor may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processor, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc., which are the computational core and control core of the terminal adapted to implement one or more instructions, in particular to load and execute one or more instructions to implement the corresponding method flow or corresponding functions; the processor in the embodiment of the invention can be used for deep learning to assist the operation of the secure communication transmission strategy under the active ARIS, and comprises the following steps:

Constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user; constructing a problem P1 with maximized safety energy consumption according to an ARIS-UAV safety communication network model, converting the problem P1 into a problem P2, and then decoupling the problem P2 into 2 sub-problems; the ARIS horizontal position optimization sub-problem and the ARIS beam forming matrix design sub-problem are solved through alternately iterating the 2 sub-problems, the optimal horizontal position of the ARIS and the optimal reflection coefficient matrix of the ARIS are obtained, and the safety rate of the maximum combination method user and the total power consumption of the system are realized.

In a further embodiment of the present invention, the present invention also provides a storage medium, in particular, a computer readable storage medium (Memory), which is a Memory device in a terminal device, for storing programs and data. It will be appreciated that the computer readable storage medium herein may include both a built-in storage medium in the terminal device and an extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also stored in the memory space are one or more instructions, which may be one or more computer programs (including program code), adapted to be loaded and executed by the processor. The computer readable storage medium may be a high-speed RAM Memory or a Non-Volatile Memory (Non-Volatile Memory), such as at least one magnetic disk Memory.

One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to implement the corresponding steps in the above embodiments with respect to secure communication transmission policies under deep learning assisted active ARIS; one or more instructions in a computer-readable storage medium are loaded by a processor and perform the steps of:

constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user; constructing a problem P1 with maximized safety energy consumption according to an ARIS-UAV safety communication network model, converting the problem P1 into a problem P2, and then decoupling the problem P2 into 2 sub-problems; the ARIS horizontal position optimization sub-problem and the ARIS beam forming matrix design sub-problem are solved through alternately iterating the 2 sub-problems, the optimal horizontal position of the ARIS and the optimal reflection coefficient matrix of the ARIS are obtained, and the safety rate of the maximum combination method user and the total power consumption of the system are realized.

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Examples

The deep learning assisted active ARIS secure communication network comprises the following specific steps:

firstly, constructing an ARIS-UAV safety network communication model, wherein the ARIS-UAV safety network communication model comprises a ground access point, a UAV, an active RIS, legal users and eavesdropping users;

step (1.1), constructing an ARIS-UAV secure communication network model comprising 1 ground access point, 1 UAV, 1 active RIS, 1 legal user and 1 eavesdropping user, as shown in figure 2. The horizontal position of the ground access point is expressed asThe horizontal position of the legitimate user is denoted +.>The level position of the eavesdropping user is denoted +.>

The beamforming matrix of active ARIS is defined asDividing the active ARIS evenly into M subarrays and noting that L=N/M represents the required amplification factor number, +.>For ARIS phase pairsAngular line matrix for->j and phi _n Representing the imaginary unit and the phase of each active element respectively,amplification factor vector for active ARIS, < ->An index matrix representing the connection between the matrix Θ and the vector a. In a passive RIS, the +.>However, in the active RIS, due to the self-amplifying property, < > a->It is only feasible. In addition to this we assume phi _n Can be optimized to any possible phase angle, with a feasible set of 0,2 pi.

The UAV is continuously fixed at the height h above the ground, the total time T required by the UAV to execute the optimization task is divided into T time slots, and the horizontal position of the UAV at the T time slot is expressed asWhen 1 slot->For a sufficient time, it can be assumed that the UAV remains stationary during that time slot; after 1 slot, the UAV is moved from position q [ t ]]Fly to a new horizontal position q [ t+1 ]]：

Wherein v [ t ]]∈[0,v _max ]For the flying speed of the unmanned aerial vehicle in the time slot t, v _max For maximum flight speed, εt]∈[0,2π]The flight angle of the unmanned aerial vehicle in the time slot t is shown.

Step (1.2), establishing communication modes of the ground access point and the ARIS and the legal user/eavesdropping user, wherein a line-of-sight path exists between the ground access point and the ARIS, and a communication channel h between the ground access point and the ARIS _bu [t]Is modeled as:

wherein h is _bu [t]Responsible for large-scale fading effects such as path loss and shadowing, expressed asFor euclidean distance between the access point and the ARIS, β is the reference channel gain when distance d=1, and k is the path loss index; let us assume that the RIS consists of a uniform matrix array, then the antenna array response vector at the receiving end of the RIS is expressed as:

wherein n=n _x N _y ,N _x And N _y The number of elements of the active RIS in the x-axis and the y-axis respectively; q _ris And lambda denote the element spacing and the communication wavelength of the RIS respectively, Respectively representing azimuth angle and pitch angle of the RIS representative receiving end;

for an ARIS-terrestrial link, one set of communication signals will be received in line-of-sight conditions and the other set of communication signals will be received by strong reflection and diffraction in non-line-of-sight conditions; since the channel between the RIS and the eavesdropping user is very difficult to estimate, it is assumed that the channel state information (channel state information, CSI) between the ARIS and the eavesdropping user is partly known; assuming that CSI has estimation errors, the CSI model between RIS and eavesdropping user is expressed as:

wherein,for the channel estimation between the ARIS and the eavesdropping user,/and/or>For estimating the error vector>For the channel estimation error from the ith active RIS element to the eavesdropping user in time slot t. />To eavesdrop on the thermal noise of the user. Channel h between ARIS-legitimate users _ud [t]And ARIS-eavesdropping user>Modeled as rice channel models, expressed as follows: />

Wherein,and->All representing a determined line of sight component, +.>For the Euclidean distance between ARIS and legitimate user, < >>For the Euclidean distance between ARIS and eavesdropping user, < >>Antenna response vector when RIS is used as transmitting end, < >>And theta ^t Respectively representing azimuth angle and pitch angle of the transmitting end by ARIS, wherein K is a Lais factor, and ++ >And->Is a non-line-of-sight component, is a random small-scale fading and obeys a composite Gaussian distribution with a mean value of 0 and a variance of 1;

step (1.3), defining the security rate reachable by legal users as follows:

R _act,s [t]＝W[log ₂ (1+Υ _act,d [t])-log ₂ (1+Υ _act,e [t])] ⁺

wherein [ x ]] ⁺ The operation shows that x is more than or equal to 0, W is the total bandwidth, gamma _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]Signal-to-interference-and-noise ratio for eavesdropping users;

wherein, according to shannon's formula, the signal-to-interference-and-noise ratio of the downlinks between the ARIS and legitimate users and the ARIS and eavesdropping users are expressed as:

wherein P is _AP Is the transmit power of the ground access point,and->Respectively representing the receiving noise of legal users and eavesdropping users; />

Step (1.4), calculating the total power consumed by the ARIS-UAV network as follows:

wherein v is the inverse of the energy conversion coefficient of the active ARIS, P _UAV Maintaining hover and flight state total power consumption, W for UAVs _PS And W is _PA Static power consumption, W, of the phase shift circuit and the power amplifier, respectively _U And W is _E The power consumption of the mobile terminal is used for legal users and eavesdropping users respectively.

Secondly, constructing a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the first step, converting the problem P1 into a problem P2, and decoupling the problem P2 into 2 sub-problems;

step (2.1), jointly considering the beamforming matrix and the horizontal position of the active ARIS, and modeling the SEE maximization problem of the ARIS-UAV network as a problem P1:

Wherein C1 represents a safe rate constraint for an authorized user, and it is noted that the active RIS amplifies both the received signal and the received noise of each reflective element, so that C2 represents a limit on the maximum transmit power of the RIS, C3 ensures that the active RIS operates in signal amplification mode, C4 represents a phase angle constraint for the RIS, and C5 represents a flight area of the drone.

Step (2.2), converting the partial objective function in the problem P1 into a form of parameter subtraction by using a Dinkelbach algorithm, and then the problem P2 is expressed as follows:

step (2.3), the problem P2 is decoupled into 2 sub-problems including an ARIS horizontal position optimization sub-problem and an ARIS beamforming matrix design sub-problem, which are specifically as follows:

a) Only constraints on the ARIS horizontal position are retained, and the ARIS horizontal position optimization sub-problem is expressed as follows:

the ARIS horizontal position optimization sub-problem is non-convex, and a DDPG algorithm is adopted to train the intelligent agent to acquire the optimal action of the intelligent agent; firstly, modeling an ARIS-UAV network environment as a Markov model, enabling an UAV carrying RIS to serve as an agent, and taking the safe rate of a legal user as a target and the horizontal position of the UAV as the agent, wherein the method comprises the following steps of:

(1) Initializing parameters of an intelligent agent and a neural network and environment information of the intelligent agent;

The agent information includes: the initial position of the UAV, the total time for executing tasks, the maximum flight speed and the total time slot of flight;

the neural network parameters include: parameters of an actor network and parameters of a critic network; the agent environment information includes: an agent action space, a state space, and an experience buffer;

(2) The intelligent agent selects actions according to the action policy network through the current state, updates actions after adding noise and constraint, executes actions to interact with the environment, and returns instant rewards and new states;

(3) Storing the state, the action, the rewards and the new state into an experience buffer zone, judging whether the quantity of experiences cached in the experience buffer zone reaches a threshold value of a set capacity, if so, executing (4), otherwise, executing (2) the intelligent and environment continuous interaction;

(4) Extracting small batch data in the experience buffer area as data for training and evaluating network parameters and target network parameters; calculating a critic-online network target value, calculating a mean square error loss function value, and updating a critic-online network parameter; updating an actor-online network parameter by using a gradient method; soft update of the actor-target network and critic-target network parameters;

(5) And when the algorithm reaches the maximum training round number, the algorithm is terminated, and the most action of the intelligent agent is obtained.

b) Only the constraints on the ARIS beamforming matrix are preserved, and the ARIS beamforming matrix design sub-problem is expressed as follows:

the ARIS beamforming matrix design sub-problem is a non-convex problem, which we first converts to the qqp problem as follows:

wherein,

/>

then, because constraint condition C4 is non-convex, it is proposed to use ADMM algorithm to carry out, loop the following 5 steps, solve and get ARIS optimal beam forming matrix under the current iteration number, including:

(1)

(2)Φ ^k+1 [t]＝Z[t] ^-1 X[t]；

(3)

(4)

(5)v ^k+1 [t]＝v ^k [t]+Γ ^k [t]-Φ ^k+1 [t]。

wherein,Z[t]＝(2A[t]+ρ ₁ Ω[t]Γ ^k+1 [t](Γ ^k+1 ) ^H [t]Ω ^H [t]+ρ ₂ I _N )，ρ ₁ ＞0，ρ ₂ > 0 is penalty coefficient, Γ=Φ is the introduced split variable, ++>s and v are Lagrangian multipliers.

To sum up, obtain the optimal solution phi of ARIS beam forming matrix ^opt The optimal solutions of the diagonal phase shift matrix and the method factor vector of the ARIS are expressed as:

Θ ^opt ＝diag(exp(jarg(Φ ^opt )))

a ^opt ＝Λdiag(exp(-jarg(Φ ^opt )))Φ ^opt

thirdly, solving an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem through the problem P2 obtained in the second step of alternate iteration, obtaining an ARIS optimal horizontal position and an ARIS optimal reflection coefficient matrix, and simultaneously realizing the safety rate of a maximum combination method user and minimizing the total power consumption of a system, wherein the method specifically comprises the following steps:

step (3.1), the process of solving the UAV horizontal position optimization sub-problem by using the DDPG algorithm is as follows: (1) initializing variables; (2) resetting the environmental state; (3) The intelligent agent obtains the current state and obtains the current action after adding noise according to the actor-online network obtaining strategy; (4) Executing actions, interacting with the environment, and obtaining instant rewards and new states; (5) Storing the status, actions, instant rewards, and new status to an experience buffer; (6) If the capacity of the experience buffer reaches the set threshold value, executing (7), otherwise executing (3); (7) Extracting small batches of data from the experience buffer area as data for training the neural network parameters, and updating the actor and critic network parameters; (8) And ending the algorithm when the algorithm reaches the maximum training round number.

Step (3.2), the process of solving the ARIS beam forming matrix design sub-problem by using the ADMM algorithm is as follows: giving initial values of variables, and carrying out solution on split variable Γ under the k+1th iteration ^k+1 [t]Solving the wave beam forming matrix phi under the k+1th iteration ^k+1 [t]Solving the auxiliary variable P under the k+1th iteration _o ^k+1 [t]Solving Lagrangian multiplier s at k+1th iteration ^k+1 [t]Solving Lagrangian multiplier v at the (k+1) th iteration ^k+1 [t]And continuously increasing the iteration times until the maximum iteration times are reached, ending the algorithm, and obtaining the optimal solution of the ARIS beam forming matrix.

Step (3.3), solving a safety energy consumption maximization problem by using a Dinkelbach algorithm: initializing variables, obtaining an optimal solution of the UAV horizontal position under the ith iteration, bringing the approximate solution of the UAV horizontal position under the ith iteration into a RIS beamforming matrix design sub-problem, solving an ARIS beamforming matrix approximate solution under the ith iteration, solving a safe energy consumption value under the ith iteration, continuously increasing the iteration times until the difference between P2 objective function values of the two iterations is within a certain range, obtaining the maximum safe energy consumption, and ending the algorithm.

The technical effects of the present invention will be described in detail with reference to the following.

The simulation is in a network environment with fixed numbers of ground access points, RIS, unmanned aerial vehicles, legal users and eavesdropping users, and two performance indexes of safety rate and safety energy consumption are counted.

Main network simulation parameters:

the ground access point has a transmit power of 23dBm, a bandwidth resource of 1MHz, a UAV flight height of 200m, and a number of reflective elements of ARIS of 256.

The present invention is compared with a baseline scheme one (active ARIS with 256 amplifiers under DDPG algorithm), a baseline scheme two (active ARIS with 128 amplifiers under DDPG algorithm), a baseline scheme three (active ARIS with 64 amplifiers under DDPG algorithm), a baseline scheme four (active ARIS with 1 amplifier under DDPG algorithm), a baseline scheme five (active ARIS with 256 amplifiers under random algorithm) and a baseline scheme six (passive ARIS under DDPG algorithm), as shown in fig. 3, 4 and 5.

In summary, according to the secure communication transmission strategy and system under deep learning auxiliary active ARIS, the RIS is carried on the UAV, and the amplifier of the RIS is partitioned, so that the RIS can be endowed with extra degrees of freedom, the flexibility of the RIS for deployment in a network is effectively improved, meanwhile, the security of the network is fully considered, and the risk of eavesdropping when a ground access point transmits confidential information to a legal user is greatly reduced by adopting a physical layer security technology. Compared with the existing mechanism, the invention can effectively improve the safety energy consumption of the network.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/terminal and method may be implemented in other manners. For example, the apparatus/terminal embodiments described above are merely illustrative, e.g., the division of the modules or units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RandomAccess Memory, RAM), an electrical carrier wave signal, a telecommunications signal, a software distribution medium, etc., it should be noted that the computer readable medium may contain content that is appropriately increased or decreased according to the requirements of jurisdictions and patent practices, such as in certain jurisdictions, according to the jurisdictions and patent practices, the computer readable medium does not contain electrical carrier wave signals and telecommunications signals.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above is only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by this, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the claims of the present invention.

Claims (10)

1. A secure communication transmission strategy under deep learning assisted active ARIS, comprising the steps of:

s1, constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

s2, constructing a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converting the problem P1 into a problem P2, and then decoupling the problem P2 into 2 sub-problems;

And S3, solving an ARIS horizontal position optimization sub-problem and an ARIS beam forming matrix design sub-problem through the 2 sub-problems obtained in the alternate iteration step S2, obtaining an ARIS optimal horizontal position and an ARIS optimal reflection coefficient matrix, and realizing the safety rate of the maximum combination method user and minimizing the total power consumption of the system.

2. The secure communication transmission strategy under deep learning-assisted active ARIS according to claim 1, wherein step S1 is specifically:

constructing an ARIS-UAV secure communication network model comprising 1 ground access point, 1 UAV, 1 active RIS, 1 legal user and 1 eavesdropping user; the horizontal position of the ground access point is expressed asThe horizontal position of the legitimate user is denoted +.>The level of eavesdropping user is expressed asThe beamforming matrix of the active ARIS is +.>The UAV is continuously fixed at the height h above the ground, the total time T required by the unmanned aerial vehicle to execute the optimization task is divided into T time slots, and the horizontal position of the UAV at the T time slot is expressed as +.>Establishing a communication model of a ground access point, an ARIS and a legal user/eavesdropping user; defining a safe rate reachable by a legal user; calculating the total power P consumed by ARIS-UAV network _total [t]。

3. According to claim The deep learning-assisted active ARIS secure communication transmission strategy of claim 2, wherein a communication channel h between a ground access point and the ARIS _bu [t]Modeling is as follows:

wherein h is _bu [t]Responsible for large-scale fading effects such as path loss and shadowing, d _bu [t]For euclidean distance between the access point and the ARIS, β is the reference channel gain when distance d=1, and k is the path loss index;

the antenna array response vector of the RIS at the receiving end is expressed as:

wherein n=n _x N _y ,N _x And N _y The number of elements of the active RIS in the x-axis and the y-axis respectively; q _ris And lambda denote the element spacing and the communication wavelength of the RIS respectively,θ ^r respectively representing azimuth angle and pitch angle of the RIS representative receiving end;

the CSI model between RIS and eavesdropping user is expressed as:

wherein,for ARIS and eavesdroppingChannel estimation value, delta between users _e For estimating the error vector>For the channel estimation error from the ith active RIS element to the eavesdropping user in time slot t, is>Thermal noise for eavesdropping on the user;

channel h between ARIS-legitimate users _ud [t]And ARIS-eavesdropping userModeling is as follows:

wherein,and->All representing a determined line of sight component, d _ud For the Euclidean distance between ARIS and legitimate users, d _ue [t]For the Euclidean distance between ARIS and eavesdropping user, < > >Antenna response vector when RIS is used as transmitting end, < >>And theta ^t Respectively representing azimuth angle and pitch angle of the transmitting end by ARIS, wherein K is a Lais factor, and ++>And->Is a non-line-of-sight component, is a random small-scale fading and follows a composite gaussian distribution with a mean of 0 and a variance of 1.

4. The secure communication transmission strategy under deep learning assisted active ARIS according to claim 2, characterized in that the security rate reachable by legitimate users is:

R _act,s [t]＝W[log ₂ (1+Υ _act,d [t])-log ₂ (1+Υ _act,e [t])] ⁺

wherein [ x ]] ⁺ The operation shows that x is more than or equal to 0, W is the total bandwidth, gamma _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]To tap the signal-to-interference-and-noise ratio of the user.

5. The secure communication transmission strategy under deep learning assisted active ARIS according to claim 2, characterized in that the total power consumed by the ARIS-UAV network is:

wherein v is the inverse of the energy conversion coefficient of the active ARIS, P _UAV Maintaining hover and flight state total power consumption, W for UAVs _PS And W is _PA Static power consumption, W, of the phase shift circuit and the power amplifier, respectively _U And W is _E The power consumption of the mobile terminal is used for legal users and eavesdropping users respectively.

6. The secure communication transmission strategy under deep learning-assisted active ARIS according to claim 1, wherein step S2 is specifically:

Jointly considering a beam forming matrix and a horizontal position of an active ARIS, and modeling an SEE maximization problem of an ARIS-UAV network as a problem P1; converting the partial objective function in the problem P1 into a form of parameter subtraction by using a Dinkelbach algorithm to obtain a problem P2; the problem P2 is decoupled into an ARIS horizontal position optimization sub-problem and an ARIS beamforming matrix design sub-problem.

7. The deep learning assisted active ARIS based secure communication transmission strategy of claim 6, wherein the decoupling of the problem P2 into the ARIS horizontal position optimization sub-problem and the ARIS beamforming matrix design sub-problem is specified as follows:

only constraints on the ARIS horizontal position are retained, and the ARIS horizontal position optimization sub-problem is expressed as follows:

s.t.C1:R _act,s [t]≥0

only the constraints on the ARIS beamforming matrix are preserved, and the ARIS beamforming matrix design sub-problem is expressed as follows:

s.t.C1:R _act,s [t]≥0

wherein R is _act,s [t]For the safe rate of legal users in time slot t, q [ t ]]For the horizontal position of ARIS in time slot t, h _bu [t]H is the channel between the access point and the legal user in time slot t _ud [t]For the channel between the ARIS-legitimate users at time slot t,for thermal noise of legal users, N is the number of ARIS phase shift elements, W _PS For static power consumption of the phase shift circuit, L is the number of ARIS power amplifiers, W _PA For the static power consumption of the power amplifier, +.>For ARIS maximum launch power, χ is the width of the UAV flyable region, y is the length of the UAV flyable region, T is the total UAV flight time, y _act,d [t]Gamma, for signal-to-interference-and-noise ratio of legal user _act,e [t]To eavesdrop on the signal-to-interference-and-noise ratio of the user, P _AP For the transmitting power of the ground access point, v is the reciprocal of the energy conversion coefficient of the active ARIS, a _l Is the first power amplifier, θ _n Is the phase shift of the nth element.

8. The secure communication transmission strategy under deep learning-assisted active ARIS according to claim 1, wherein step S3 is specifically:

solving a horizontal position optimization sub-problem of the UAV by using a DDPG algorithm;

the process for solving the ARIS beam forming matrix design sub-problem by using ADMM algorithm is as follows: firstly, giving initial values of all variables, carrying out solution on split variables under the k+1th iteration, then solving a beam forming matrix under the k+1th iteration, then solving auxiliary variables under the k+1th iteration, then solving Lagrangian multipliers under the k+1th iteration, and continuously increasing iteration times until reaching a maximum iteration times ending algorithm to obtain an optimal solution of the ARIS beam forming matrix;

solving a safety energy consumption maximization problem by using Dinkelbach algorithm: initializing variables, obtaining an optimal solution of the UAV horizontal position under the ith iteration, bringing the approximate solution of the UAV horizontal position under the ith iteration into a RIS beamforming matrix design sub-problem, solving an ARIS beamforming matrix approximate solution under the ith iteration, solving a safe energy consumption value under the ith iteration, continuously increasing the iteration times until the difference between P2 objective function values of the two iterations is within a certain range, obtaining the maximum safe energy consumption, and ending the algorithm.

9. The secure communication transmission strategy under deep learning assisted active ARIS according to claim 8, wherein solving a UAV horizontal position optimization sub-problem using a DDPG algorithm is specifically:

initializing various variables; resetting the environmental state; the intelligent agent obtains the current state, obtains a strategy according to an actor-online network, and obtains the current action after adding noise; executing actions, interacting with the environment, and obtaining instant rewards and new states; storing the status, actions, instant rewards, and new status to an experience buffer; if the capacity of the experience buffer zone reaches a set threshold value, extracting small batch data from the experience buffer zone to serve as data for training the neural network parameters, and updating the actor and critic network parameters; ending the algorithm when the maximum training round number is reached; if the capacity of the experience buffer zone does not reach the set threshold, the intelligent agent obtains the current state again, then obtains the strategy according to the actor-online network, and obtains the current action after adding noise.

10. A secure communication transmission policy system under deep learning assisted active ARIS, comprising:

the construction module is used for constructing an ARIS-UAV safety network communication model comprising a ground access point, a UAV, an active RIS, a legal user and an eavesdropping user;

The decoupling module constructs a problem P1 with maximized safety energy consumption according to the ARIS-UAV safety communication network model obtained in the step S1, converts the problem P1 into a problem P2 and then decouples the problem P2 into 2 sub-problems;

and the output module is used for solving the ARIS horizontal position optimization sub-problem and the ARIS beam forming matrix design sub-problem through the 2 sub-problems obtained in the alternate iteration step S2, obtaining the optimal horizontal position of the ARIS and the optimal reflection coefficient matrix of the ARIS, and realizing the safety rate of the maximum combination method user and the minimum total power consumption of the system.