A method for realizing integrated sensing and communication waveform and unmanned aerial vehicle trajectory design for secure mobile edge computing

Abstract

The application discloses a sensing-integrated waveform and unmanned aerial vehicle (UAV) trajectory design method for realizing secure mobile edge computing, and aims at the case that potential eavesdroppers exist in a mobile edge computing network architecture in which multiple ground users perform uplink information transmission to a UAV, optimizes and designs a sensing-integrated waveform and a UAV trajectory with the minimum total user energy consumption as the target. First, an optimization problem conforming to the scene is constructed, then the original optimization problem is decoupled into sub-problems about the sensing-integrated waveform and the UAV flight trajectory based on a Block Coordinate Descent (BCD) method, and a convex approximation fitting is performed on each non-convex sub-problem, finally, an iterative algorithm is adopted to solve the problem, so that a global suboptimal solution of the whole optimization problem and optimal sensing-integrated waveform and UAV flight trajectory are found.

Description

Technical Field

[0001] This invention relates to the field of integrated communication and sensing, specifically to a method for designing waveforms and drone trajectories that integrates communication and sensing to achieve secure mobile edge computing. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure, and these statements may constitute prior art. In the process of developing this invention, the inventors discovered at least the following problems in the prior art.

[0003] Wireless sensing performance is a crucial development direction for future 6G, with communication and sensing technologies gradually converging. Mobile edge computing is a classic technology for addressing latency-sensitive tasks where user resources are limited. Drones, as core nodes in future mobile communication networks, have a wide range of applications and are attracting significant attention from various industries.

[0004] In the literature [C. Chen, J. Yao, M. Jin, and Q. Guo, "Beamforming and computing capacity allocation for ISAC-assisted secure mobile edge computing," IEEE Wireless Commun. Lett., pp. 1-1, Sep. 2024.], multiple users transmit uplink information to a base station equipped with a mobile edge computing server for computation, while a potential eavesdropper is simultaneously eavesdropping. The duplex base station sends radar signals to interfere with the potential eavesdropper, thus achieving secure communication.

[0005] The literature [Y. Xu, T. Zhang, Y. Liu, and D. Yang, "UAV-enabled integrated sensing, computing, and communication: A fundamental trade-off," IEEE Wireless Commun. Lett., vol. 12, no. 5, pp. 843-847, May 2023.] describes a UAV that unloads information in the air to a ground access point for computation, while simultaneously sending radar sensing signals to potential targets to complete the sensing task, striking a trade-off between communication rate and sensing beammap gain.

[0006] In [N. Huang, C. Dou, Y. Wu, L. Qian, B. Lin, and H. Zhou, "Unmanned-aerial-vehicle-aided integrated sensing and computation with mobile-edge computing," IEEE Internet Things J., vol. 10, no. 19, pp. 16830-16844, Oct. 2023.], the authors used a drone to send radar signals to sense targets and offloaded the received reflected radar information to an edge server for computation.

[0007] In [Y.Liu, B.Zhang, D.Guo, H.Wang, and G.Ding, "Joint precoding design and location optimization in joint communication, sensing and computing of UAV systems," IEEE Trans. Cogn. Commun. Netw., vol. 10, no. 2, pp. 541-552, Apr. 2024.], the UAV sends sensing signals to ground-based sensing targets to collect sensing data, which is then transmitted to a base station equipped with a ground-based server. This paper considers the UAV hovering in the air to send and receive information, and the communication sensing channel is a Ricean channel.

[0008] The above solution has the following problems:

[0009] 1. The security research on the integration of sensing and communication has not been applied to the scenario of drone communication, and only the communication method between ground duplex base stations and ground users has been considered.

[0010] 2. The issue of secure transmission was not considered in the context of drone communication.

[0011] 3. In multi-user mobile edge computing scenarios, user scheduling strategies are not considered.

[0012] For example, patent application number 202411217273.1, entitled "A Mobile Edge Computing System, Its Optimization Model Modeling Method, and Its Optimization Method," exemplifies this. It constructs a mobile edge computing system consisting of N edge servers and M mobile devices, and integrates factors such as channel gain between edge servers and mobile devices, signal-to-noise ratio between edge servers and mobile devices, data transmission rate between edge servers and mobile devices, total latency and total energy consumption of edge computing, and task cost of edge computing to build an optimization problem model to address the problem that existing technologies are not applicable to environments with multiple edge servers and multiple mobile devices. However, this patent does not consider communication security, and the servers are fixed at a certain point, lacking deployment flexibility.

[0013] The main problem this patent addresses is how to fully leverage the flexibility of drone deployment while minimizing user energy consumption, all while ensuring secure communication. Summary of the Invention

[0014] In view of the above problems, the purpose of this invention is to solve some of the problems in the prior art, or at least alleviate these problems.

[0015] A method for designing sensing-integrated waveforms and UAV trajectories to achieve secure mobile edge computing includes:

[0016] S1: Constructing a communication system model:

[0017] The drone S is set up as an edge server base station, and its flight trajectory is initialized to serve K users on the ground. The drone adopts a user scheduling strategy, and the existence of a potential eavesdropper E is considered. In order to ensure the security of information transmission, the drone S emits radar waves to interfere with E, thereby realizing secure mobile edge computing.

[0018] S2: Constructing an optimized mathematical model:

[0019] With the goal of minimizing energy consumption by ground users, an optimization mathematical model is constructed with constraints such as user offloading ratio, user scheduling, integrated sensor waveform, and UAV flight trajectory.

[0020] S3: For the aforementioned optimized mathematical model, the BCD method is used to decouple it into subproblems concerning only the user offloading ratio, user scheduling, synesthetic waveform, and UAV flight trajectory. For each non-convex subproblem, a convex approximation fitting method is used to convert it into a convex problem for solution. Specifically, when solving the non-convex problem concerning the synesthetic waveform, the semidefinite relaxation method (SDR) method, which discards the rank-one constraint, is used to convert the non-convex problem into a convex problem. Then, the convex optimization solution tool CVX toolbox is used for solution. Finally, the obtained solution is randomized through Gaussian to obtain a solution that satisfies the rank-one constraint.

[0021] S4: Use iterative algorithms to solve for user scheduling, offloading ratio, synesthesia waveform, and UAV flight trajectory.

[0022] The optimized mathematical model is as follows:

[0023]

[0024] In the formula Let α represent the set of optimization variables. k [n] is the offloading ratio of user k in time slot n, θ k [n] represents the user scheduling status of the drone, w[n] represents the integrated sensing waveform, and q... s [n] represents the drone trajectory, E k It is the energy consumption of user k. Indicates the starting and ending positions of the drone, δ t The length of each flight time slot of the drone is defined, V max The maximum flight speed of the drone at any given time is defined, and T represents the total flight time of the drone. These represent the time user k spends calculating the remaining data locally, the time user k spends unloading data, and the time the drone server spends calculating the data unloaded by user k, respectively. sk γ ek These are the signal-to-interference-plus-noise ratio (SIR) of the drone receiving data from user k and the SIR of the eavesdropper receiving data from user k, respectively. s ,Γ e ,Γ sen The communication performance threshold, eavesdropping threshold, and radar beam pattern gain threshold of the UAV are defined respectively, where P[n] represents the radar beam pattern gain. P is defined as the square of the distance from the drone to the eavesdropper in time slot n. max , E UAV , E max The maximum transmission power, total energy consumption, and maximum energy of the drone were defined respectively.

[0025] Among them, C1 to C4 are constraints on user scheduling and offloading ratios; C5 to C6 limit the start and end positions and maximum flight distance of the UAV; C7 to C8 are constraints on the time of data calculation and data offloading; C9 to C10 are constraints on the signal-to-noise ratio between the UAV and the eavesdropper, which are requirements for secure transmission; C11 to C12 are constraints on the integrated sensing waveform, requiring the radar signal to have an interference effect on the eavesdropper, but the power cannot exceed the maximum value; C13 is the energy consumption constraint of the UAV, which limits the total energy consumption of the UAV's flight, calculation and radar signal transmission.

[0026] Decoupling the mathematical model yields subproblems. As shown below:

[0027]

[0028] in, This indicates the energy consumption of data that the user calculated but did not unload. This represents the energy consumption when a user uninstalls the program; this problem is a standard convex optimization problem, which can be solved using the interior point method.

[0029]

[0030] in, This represents the energy consumption when the user performs local computation; this problem is a standard convex optimization problem, which can be solved using the interior-point method.

[0031]

[0032] Where Tr represents the operation of finding the trace of a matrix. This indicates an operation to calculate the expected value. Represents a synergistic beam matrix;

[0033]

[0034] Furthermore, regarding the sub-problems The objective function is defined as minimizing the eavesdropping signal-to-noise ratio, while introducing the auxiliary variable η. k After optimization and improvement, the following mathematical model was obtained:

[0035]

[0036] Among them, H se [n] represents the channel coefficient matrix from the drone to the eavesdropper. g represents the noise power received by the eavesdropper. i Let C6 represent the channel coefficient from the eavesdropper to the i-th user. When the rank-one constraint C6 is ignored, the entire convex optimization problem is a semidefinite programming (SDP) problem, which can be solved using the SDR optimization method. Then, Gaussian randomization is used to process this non-rank-one solution to obtain an approximate solution that satisfies the rank-one condition.

[0037] Furthermore, regarding the sub-problems The following steps were used for optimization and improvement:

[0038] By introducing the following four auxiliary variables, the required conditions can be met:

[0039]

[0040]

[0041] The fourth one is a non-convex constraint, which can be rewritten using SCA to obtain:

[0042]

[0043] Where v0 represents the average rotor-induced velocity when the UAV is hovering, H is the flight altitude of the UAV, and q k q e Defined as the user's coordinates and the eavesdropper's coordinates, respectively. v1[n] represents the position of the UAV in the appropriate m-th time slot, and v2[n] represents two auxiliary variables introduced.

[0044] The result is obtained by rewriting using the Continuous Convex Approximation (SCA) method:

[0045]

[0046] in, express right Find the partial derivative;

[0047] For other non-convex constraints, the same method is used to obtain:

[0048]

[0049] in, The noise power of the signal received by the UAV is defined, p k This represents the user's transmit power, β0 represents the channel gain at a reference distance of 1 meter, and D... k The total amount of user data is defined, B represents the channel bandwidth, and τ k [n] is an introduced auxiliary variable. The definitions are the UAV propulsion power after introducing auxiliary variables, P0 and P... i These represent the airfoil power and induced power of the UAV in hovering state, respectively. tip The blade tip velocity represents the rotor blade; d0, ρ, s, and A represent the fuselage drag ratio, air density, rotor rigidity, and rotor disk area, respectively; X[n], X k1 [n]、X k3 [n] is a newly introduced slack variable that needs to satisfy:

[0050]

[0051]

[0052] Based on the above transformation, a new optimization problem can be obtained, which can be expressed as:

[0053]

[0054] in, c represents the optimization variable. k [n]、γ0、X k [n] is an introduced auxiliary variable; the problem It is a standard convex optimization problem, which is solved using the convex optimization solution tool CVX.

[0055] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for designing a sensor-integrated waveform and UAV trajectory for secure mobile edge computing.

[0056] The unit for realizing secure mobile edge computing with integrated sensing waveforms and drone trajectories includes:

[0057] The initialization module sets up the drone S as an edge server base station and initializes the flight trajectory of S to serve K users on the ground. In order to better reduce the energy consumption of users, the drone adopts a user scheduling strategy. At the same time, considering the existence of a potential eavesdropper E, in order to ensure the security of information transmission, the drone emits radar waves to interfere with E, thereby realizing secure mobile edge computing.

[0058] The mathematical model module aims to minimize the energy consumption of ground users and constructs an optimized mathematical model constrained by user scheduling, user offloading ratio, integrated sensor waveform, and UAV flight trajectory.

[0059] The mathematical model decoupling module is used to decouple the mathematical model based on the BCD method into subproblems that only concern the unloading ratio, user scheduling, integrated sensing waveform, and UAV flight trajectory. For each non-convex subproblem, a convex approximation fitting method is used to convert it into a convex problem for solution.

[0060] The subproblem-solving module is used to solve the unloading ratio, user scheduling, sensor-integrated waveform, and UAV flight trajectory using iterative algorithms.

[0061] The present invention has the following beneficial effects:

[0062] This invention considers the existence of a potential eavesdropper and sends a communication-sensing integrated waveform as a strong interference signal to the potential eavesdropper to disrupt their eavesdropping behavior. Simultaneously, considering that the UAV carries a server, users can offload data to the server for computation, fully leveraging the UAV's flexible deployment and high mobility. Therefore, this invention optimizes sub-problems 1.1 and 2.1 to obtain the optimal user offloading ratio and user scheduling, achieving a strategy that minimizes user energy consumption. Optimizing the improved sub-problem 3.2 yields the optimal communication-sensing integrated beamforming, minimizing the eavesdropper's signal-to-noise ratio and achieving secure communication. Optimizing the improved sub-problem 4.2 yields the optimal UAV flight trajectory, obtaining better channel conditions and further reducing user energy consumption. Finally, by iteratively optimizing the four sub-problems until convergence, a local optimum and a global suboptimal solution are obtained, representing the minimum user energy consumption. This invention combines the flexibility and maneuverability of UAVs with the ability of strong radar signals to act as interference, constructing a UAV-assisted communication-sensing integrated communication system. Through theoretical derivation and simulation verification, the communication-sensing integrated waveform and UAV flight trajectory are designed to minimize user power. Attached Figure Description

[0063] Figure 1 This is a communication system model of the present invention;

[0064] Figure 2 This is a flowchart illustrating the operation of the present invention.

[0065] Figure 3 The optimal flight path map for the drone;

[0066] Figure 4 This is the convergence graph of the algorithm's iterations;

[0067] Figure 5 User scheduling strategy diagram;

[0068] Figure 6 The simulation shows the beam gain diagrams. In the figure, (a) is the beam gain diagram about the x-axis and (b) is the beam gain diagram about the y-axis. Detailed Implementation

[0069] The present invention will be further described below with reference to the accompanying drawings. The embodiments of the present invention are only used to illustrate the present invention and not to limit the present invention. Various substitutions and modifications made based on ordinary technical knowledge and common practices in the art without departing from the technical concept of the present invention should be included within the scope of the present invention.

[0070] like Figure 2 As shown, a method for designing sensing-integrated waveforms and drone trajectories to achieve secure mobile edge computing includes:

[0071] S1: Construct a communication system model.

[0072] The drone S is set up as an edge server base station, and its flight trajectory is initialized to serve K users on the ground. In order to better reduce the energy consumption of users, the drone adopts a user scheduling strategy. At the same time, considering the existence of a potential eavesdropper E, in order to ensure the security of information transmission, the drone S emits radar waves to interfere with E, thereby realizing secure mobile edge computing.

[0073] like Figure 1 As shown, this invention mainly studies a sensor-integrated waveform and UAV trajectory design method for secure mobile edge computing. The UAV S flies from a starting point to a destination, carrying an edge computing server and receiving uplink offload information from K users on the ground. Considering that the UAV knows the specific location information of all users and the eavesdropper, the eavesdropper's two-dimensional position coordinates are q... e =[x e ,y e ] T The two-dimensional location coordinates of the k-th user are q k =[x k ,y k ] T The horizontal position coordinates of the UAV in time slot n are q s [n] = [x] s [n],y s [n] T The Euclidean distance from the drone to the k-th user is denoted as The Euclidean distance representing the distance from the drone to the eavesdropper is... The drone flies at a fixed altitude H, and the transmitter is equipped with an M=M x ×M y The uniform planar array antenna (UPA) is used to transmit radar sensing signals, and it is also equipped with a single antenna for receiving signals.

[0074] The conjugate transpose of the radar transmitting array steering vector can be expressed as:

[0075]

[0076] In the formula and It depends on the vertex of the departure angle of the signal from the drone to the eavesdropper and the associated azimuth angle. x and d y Let be the distances of the antennas along the x-axis and y-axis, respectively, and take half the wavelength. M x M represents the number of antennas on the x-axis. y This represents the number of antennas on the y-axis, where j is the imaginary number. Kronecker product operation, () H This represents the conjugate transpose operation of a matrix.

[0077] The channel from UAV S to user k can be represented as:

[0078]

[0079] The channel from the drone S to the eavesdropper can be represented as:

[0080]

[0081] The k-channel from the eavesdropper to the user can be represented as:

[0082]

[0083] β0 represents the channel gain at a reference distance of 1 meter.

[0084] w[n]s represents the radar signal emitted by the UAV, where The radar signal is represented by a Gaussian random distribution with a mean of 0 and a variance of 1. This represents the integrated sensing waveform. Assuming the drone has the technology to eliminate self-interference, the information received by the drone can be represented as:

[0085]

[0086] The information received by the eavesdropper can be represented as:

[0087]

[0088] Where, d k This represents the message sent by user k. It is white noise that follows a Gaussian distribution. In this case, the SINR signal received by the UAV can be expressed as:

[0089]

[0090] The SINR signal received by the eavesdropper can be expressed as:

[0091]

[0092] Among them, P k h is the transmit power of user k. i [n] represents the channel coefficient from the UAV to the i-th user, g i This represents the channel coefficient from the eavesdropper to the i-th user. This represents the conjugate transpose of the channel coefficients from the drone to the eavesdropper. The noise power of the signals received by the drone is defined.

[0093] S2: Construct an optimized mathematical model.

[0094] With the goal of minimizing energy consumption by ground users, an optimization mathematical model is constructed with constraints such as user scheduling, user offloading ratio, integrated sensor waveform, and UAV flight trajectory.

[0095] The optimized mathematical model is as follows:

[0096]

[0097] In the formula Let α represent the set of optimization variables. k [n] is the offloading ratio of user k in time slot n, θ k [n] represents the user scheduling status of the drone, w[n] represents the integrated sensing waveform, and q... s [n] represents the drone trajectory, E k It is the energy consumption of user k. Indicates the starting and ending positions of the drone, δ t The length of each flight time slot of the drone is defined, V max The maximum flight speed of the drone at any given time is defined, and T represents the total flight time of the drone. These represent the time user k spends calculating the remaining data locally, the time user k spends unloading data, and the time the drone server spends calculating the data unloaded by user k, respectively. sk γ ek These are the signal-to-interference-plus-noise ratio (SIR) of the drone receiving data from user k and the SIR of the eavesdropper receiving data from user k, respectively. s ,Γ e ,Γ sen The communication performance threshold, eavesdropping threshold, and radar beam pattern gain threshold of the UAV are defined respectively, where P[n] represents the radar beam pattern gain. P is defined as the square of the distance from the drone to the eavesdropper in time slot n. max , E UAV , E max The maximum transmission power, total energy consumption, and maximum energy of the drone were defined respectively.

[0098] Among them, C1 to C4 are constraints on user scheduling and offloading ratios; C5 to C6 limit the start and end positions and maximum flight distance of the UAV; C7 to C8 are constraints on the time of data calculation and data offloading; C9 to C10 are constraints on the signal-to-noise ratio between the UAV and the eavesdropper, which are requirements for secure transmission; C11 to C12 are constraints on the integrated sensing waveform, requiring the radar signal to have an interference effect on the eavesdropper, but the power cannot exceed the maximum value; C13 is the energy consumption constraint of the UAV, which limits the total energy consumption of the UAV's flight, calculation and radar signal transmission.

[0099] The beam pattern gain P[n] of the radar can be expressed as:

[0100] P[n]=a H (q s [n],q e )Q[n]a(q s [n],q e )

[0101] in, w H [n] represents the conjugate transpose of the synesthetic waveform. This indicates the operation of seeking the expected value.

[0102] The energy consumption of a drone can be expressed as:

[0103]

[0104] Where Tr represents the operation of finding the trace of a matrix. P represents the energy consumption of the drone in calculating the data uploaded by user k in time slot n. fly This indicates the propulsion power of the drone.

[0105] S3: For the aforementioned optimized mathematical model, the BCD method is used to decouple it into subproblems concerning only user scheduling, offloading ratio, synesthetic waveform, and UAV flight trajectory. For each non-convex subproblem, a convex approximation fitting method is used to convert it into a convex problem for solution. Specifically, when solving the non-convex problem concerning the synesthetic waveform, the SDR method, which discards the rank-one constraint, is used to convert the non-convex problem into a convex problem. Then, the CVX toolbox is used for solution, and finally, the obtained solution is randomized through Gaussian to obtain a solution that satisfies the rank-one constraint.

[0106] To address the non-convexity and high coupling of the optimization problem corresponding to the system model, an approximate fitting method based on the BlockCoordinate Descent (BCD) algorithm and continuous convex approximation is used to decouple and transform the original optimization problem. The solution to this problem will proceed according to the following steps, specifically as follows:

[0107] Based on the Block Coordinate Descent (BCD) algorithm The problem is decoupled to obtain subproblems. As shown below:

[0108]

[0109] in, This indicates the energy consumption of data that the user calculated but did not unload. This represents the energy consumption when a user uninstalls the program; this problem is a standard convex optimization problem and can be solved using the interior point method.

[0110]

[0111] in, This represents the energy consumption when the user performs local calculations; this problem is a standard convex optimization problem, which can be solved using the interior point method.

[0112] set up This represents a syn-sensory integrated beam matrix.

[0113]

[0114] Where Tr represents the operation of finding the trace of a matrix. This indicates the operation of seeking the expected value.

[0115]

[0116] In response to the problem and the problem Both are standard convex optimization problems, which can be solved using the interior-point method. Regarding the problem... and the problem The SCA method is needed to approximate some non-convex constraints with convex approximation, thereby transforming the original problem into a solvable convex optimization problem.

[0117] Regarding the sub-problems Since radar signals do not directly affect the user's energy consumption, but indirectly improve the drone's signal-to-noise ratio by interfering with eavesdroppers, thereby reducing energy consumption, the objective function can be defined as minimizing the eavesdropping signal-to-noise ratio to expand the solution space for drone trajectory optimization. An auxiliary variable η is also introduced. k The mathematical model is:

[0118]

[0119] When the rank-one constraint C6 is ignored, the entire convex optimization problem is an SDP problem, which can be solved using the SDR optimization method. Then, Gaussian randomization is used to process this non-rank-one solution to obtain an approximate solution that satisfies the rank-one condition.

[0120] Regarding the sub-problems The following steps were used for optimization and improvement:

[0121] By introducing the following four auxiliary variables, the required conditions can be met:

[0122]

[0123] The fourth one is a non-convex constraint, which can be rewritten using SCA to obtain:

[0124]

[0125] Where v0 represents the average rotor-induced velocity when the UAV is hovering, H is the flight altitude of the UAV, and q k q e These are defined as the user's coordinates and the eavesdropper's coordinates, respectively.

[0126] Due to the array guide vector a H (q s [n],q e Since it is not affine, some constraints are non-convex. Therefore, it is rewritten using the SCA method to obtain:

[0127]

[0128] Among them, can The detailed expression is as follows:

[0129]

[0130] For other non-convex constraints, the same method is used to obtain:

[0131]

[0132]

[0133] in, The noise power of the signal received by the UAV is defined, p k This represents the user's transmit power, β0 represents the channel gain at a reference distance of 1 meter, and D... k The total amount of user data, P0 and P1, are defined. i These represent the airfoil power and induced power of the UAV in hovering state, respectively. tip The blade tip velocity represents the rotor blade; d0, ρ, s, and A represent the fuselage drag ratio, air density, rotor rigidity, and rotor disk area, respectively; X[n], X k1 [n]、X k3 [n] is a newly introduced slack variable that needs to satisfy:

[0134]

[0135] Based on the above transformation, a new optimization problem can be obtained, which can be expressed as:

[0136]

[0137] in, Represents the optimization variable; problem It is a standard convex optimization problem, which is solved using the convex optimization solution tool CVX.

[0138] S4: Use iterative algorithms to solve for user scheduling, offloading ratio, synesthesia waveform, and UAV flight trajectory.

[0139] The iterative algorithm flow is designed to connect the above convex optimization problems in sequence. The specific steps are shown in Table 1.

[0140] Table 1

[0141]

[0142] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for designing a sensor-integrated waveform and UAV trajectory for secure mobile edge computing.

[0143] The unit for realizing secure mobile edge computing with integrated sensing waveforms and drone trajectories includes:

[0144] The initialization module sets up the drone S as an edge server base station and initializes the flight trajectory of S to serve K users on the ground. In order to better reduce the energy consumption of users, the drone adopts a user scheduling strategy. At the same time, considering the existence of a potential eavesdropper E, in order to ensure the security of information transmission, the drone emits radar waves to interfere with E, thereby realizing secure mobile edge computing.

[0145] The mathematical model module aims to minimize the energy consumption of ground users and constructs an optimized mathematical model constrained by user scheduling, user offloading ratio, integrated sensor waveform, and UAV flight trajectory.

[0146] The mathematical model decoupling module is used to decouple the mathematical model based on the BCD method into subproblems that only concern the unloading ratio, user scheduling, integrated sensing waveform, and UAV flight trajectory. For each non-convex subproblem, a convex approximation fitting method is used to convert it into a convex problem for solution.

[0147] The subproblem-solving module is used to solve the unloading ratio, user scheduling, sensor-integrated waveform, and UAV flight trajectory using iterative algorithms.

[0148] Figure 1 This is a model diagram of the communication system of the present invention;

[0149] Figure 2 This is a flowchart of the iterative algorithm proposed in this invention for solving the optimization problem. Its detailed process corresponds to the specific steps for solving the aforementioned optimization problem.

[0150] Figure 3 This invention is based on the simulation verification of the proposed scheme, through... Figure 2 The process ultimately yields Figure 3 The optimal flight trajectory for the drone is shown.

[0151] Figure 4 The simulation verified the convergence of the proposed solution. The simulation results show the effectiveness of the algorithm proposed in this invention.

[0152] Figure 5 The simulation verified the user scheduling strategy of the present invention. As can be seen from the simulation diagram, the user scheduling strategy of the present invention first selects user 1, then user 2, and finally user 3. The rationality of the user scheduling strategy can be seen from the relationship between the user positions and the UAV trajectories.

[0153] Figure 6 This is a simulation verification of the proposed scheme. Figure (a) shows the beam gain diagram about the x-axis, and (b) shows the beam gain diagram about the y-axis. The simulation results show the beam gain diagram of the proposed scheme. It can be seen from the diagram that the beam gain is greatest at the x-axis and y-axis positions where the eavesdropper is located. This conforms to the design of the proposed scheme.

[0154] This application proposes a method for designing a sensor-integrated waveform and UAV trajectory to achieve secure mobile edge computing. With the goal of minimizing user energy consumption, it dynamically plans and designs the user offload ratio, user scheduling strategy, sensor-integrated waveform, and UAV trajectory.

[0155] The first step is to construct a model of a drone-assisted wireless communication system based on the basic theories of wireless communication and physical layer security.

[0156] Steps two and three refine the mathematical theoretical derivation and problem-solving analysis in the model building process. Addressing the non-convexity and high coupling of the optimization problem corresponding to the system model, an approximate fitting method based on the Block Coordinate Descent (BCD) algorithm and continuous convex approximation is used to decouple and transform the original optimization problem. (For details, please refer to the specific implementation of this invention).

[0157] Step four: Based on the idea of ​​iteration, the sub-problems obtained in steps two and three are iterated in a loop to make the total user energy consumption continuously approach a fixed value. This fixed value is the minimum total user energy consumption required by the present invention.

[0158] This invention first constructs a system model, including a communication model, a sensing model, and a computing model. Then, it proposes an optimization problem model and uses the Block Coordinating Descent (BCD) method to decouple the original optimization problem into subproblems concerning the integrated sensing waveform and the UAV flight trajectory. For each non-convex subproblem, a convex approximation is performed. Finally, an iterative algorithm is used to solve the problem and find the global suboptimal solution to the entire optimization problem and the optimal integrated sensing waveform and UAV flight trajectory.

[0159] Users can offload data to an edge computing server mounted on the drone, reducing energy consumption. Simultaneously, the drone emits strong radar signals to interfere with potential eavesdroppers, achieving the goal of secure mobile edge computing.

[0160] This invention is primarily applied to UAV-assisted wireless communication scenarios, addressing situations where user resources are limited, latency-sensitive tasks need to be processed, and potential eavesdroppers could affect the quality and security of wireless communication. It combines the flexibility and maneuverability of UAVs with the ability of strong radar signals to act as interference, constructing a UAV-assisted, integrated communication and sensing system. Through theoretical derivation and simulation verification, the integrated communication and sensing waveform and UAV flight trajectory are designed to minimize user power. This invention utilizes the flexible deployment of UAVs while simultaneously transmitting strong radar signals to interfere with potential eavesdroppers, achieving the goal of secure mobile edge computing. The combination of UAVs and integrated communication and sensing demonstrates high innovation and originality.

Claims

1. A method for realizing a sensing-integrated waveform and UAV trajectory design for secure mobile edge computing, characterized in that, Comprise: S1: build a communication system model: Set the unmanned aerial vehicle S as an edge server base station, and initialize the flight trajectory of S to serve K users existing on the ground; The unmanned aerial vehicle adopts a user scheduling strategy, while considering that there is a potential eavesdropper E, in order to ensure the security of information transmission, the unmanned aerial vehicle S transmits radar waves to interfere with E, and realizes safe mobile edge computing; S2: build an optimization mathematical model: With the minimum energy consumption of ground users as the target, an optimization mathematical model is constructed with user scheduling, user offloading ratio, integrated sensing waveform and unmanned aerial vehicle flight trajectory as constraints; The optimization mathematical model is: In the formula denotes a set of optimization variables, wherein is the offloading ratio of user k at n time slots, is the user scheduling of UAV, is the integrated sensing waveform, is the UAV trajectory, is the energy consumption of user k, , denotes the start position and end position of the UAV, defines the length of each flight time slot of the UAV, defines the maximum flight speed of the UAV at any time, denotes the total flight time of the UAV, , , respectively denote the time for user k to locally compute the remaining data, the time for user k to perform data offloading, and the time for the UAV server to compute the data offloaded by user k, , are respectively the signal-to-interference-plus-noise ratio of user k accepted by the UAV and the signal-to-interference-plus-noise ratio of user k accepted by the eavesdropper, , , respectively define the UAV communication performance threshold, the eavesdropping threshold, and the radar beam pattern gain threshold, denotes the beam pattern gain of the radar, defines the square of the distance from the UAV to the eavesdropper at n time slots, , , respectively define the maximum transmission power of the UAV, the total energy consumption of the UAV, and the maximum energy of the UAV; Wherein C1~C4 is the constraint of user scheduling and offloading ratio, C5~C6 limits the start and end positions and the maximum flight distance of the unmanned aerial vehicle, C7~C8 is the constraint of calculating data and offloading data time, C9~C10 is the signal-to-noise ratio constraint of the unmanned aerial vehicle and the eavesdropper, which is the requirement of safe transmission; C11~C12 is the constraint of integrated sensing waveform, which requires that the radar signal has interference effect on the eavesdropper, but the power cannot exceed the maximum value; C13 is the energy consumption constraint of the unmanned aerial vehicle, which limits the total energy consumption of the unmanned aerial vehicle flight, calculation and radar signal transmission; S3: for the optimization mathematical model, based on the BCD method, decoupling into sub-problems only about user offloading ratio, user scheduling, integrated sensing waveform and unmanned aerial vehicle flight trajectory, and for each non-convex sub-problem, the fitting method of convex approximation is adopted to convert it into a convex problem for solving; Wherein, when solving the non-convex problem about the integrated sensing waveform, the semi-definite relaxation method (SDR) which discards the rank one constraint is used to convert the non-convex problem into a convex problem, and then the convex optimization solving tool CVX is used for solving, finally the obtained solution is obtained by Gaussian randomization to satisfy the solution of the rank one constraint; S4: solve the user scheduling, offloading ratio, integrated sensing waveform and unmanned aerial vehicle flight trajectory by using iterative algorithm.

2. The method of claim 1, wherein the method is characterized by, decoupling the mathematical model to obtain sub-problems 、 、 、 as follows: wherein, denotes the energy consumption of the user computing the data without offloading, denotes the energy consumption of the user performing the offloading; this problem is a standard convex optimization problem that can be solved using interior point methods; where, represents the energy consumption of the user when performing local computation; this problem is a standard convex optimization problem that can be solved using interior point methods; wherein Tr denotes an operation of taking the trace of a matrix, denotes an operation of taking the expectation; denotes a sensory integrated beamforming matrix; 。 3.The method of claim 2, wherein, For the sub-problem The objective function is defined as minimizing the eavesdropping signal-to-noise ratio while introducing auxiliary variables Optimization improvement is made to obtain the following mathematical model: wherein, denotes the channel coefficient matrix from the UAV to the eavesdropper, denotes the noise power received by the eavesdropper, denotes the channel coefficient from the eavesdropper to the i-th user; when ignoring the rank-one constraint of C6, the whole convex optimization problem is a semi-definite programming (SDP) problem, which can be solved by SDR optimization method, and then the non-rank-one solution is processed by Gaussian randomization to obtain an approximate solution satisfying the rank-one condition. 4.The method of claim 2, wherein, For the sub-problem Optimization improvements are made using the following steps: By introducing the following four auxiliary variables that meet the needs: Wherein, the fourth one is a non-convex constraint, which is rewritten by SCA to obtain: wherein, represents the average rotor induced velocity when the UAV is hovering, is the flight altitude of the UAV, , are defined as the coordinate position of the user and the coordinate position of the eavesdropper, respectively, represents the position of the UAV at the appropriate mth time slot, and represent two auxiliary variables introduced; By continuous convex approximation (SCA) method, the following is obtained: wherein represents to partial derivatives; For other non-convex constraints, the same method is adopted to obtain: wherein, defines the noise power of the signal received by the UAV, is the user transmit power, denotes the channel gain when the reference distance is 1 meter, defines the total amount of data of the user, denotes the channel bandwidth, is an auxiliary variable introduced, defines the UAV propulsion power after the auxiliary variable is introduced, and are respectively the profile power and the induced power of the rotor blade in the hovering state of the UAV, denotes the tip speed of the rotor blade; , , and are respectively the fuselage drag ratio, the air density, the rotor solidity and the rotor disc area; , , is a newly introduced slack variable, which needs to satisfy: Based on the above transformation, a new optimization problem can be obtained, which is expressed as: where, denotes the optimization variable, , , is an auxiliary variable introduced; the problem is a standard convex optimization problem, which is solved using the convex optimization solver CVX.

5. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the integrated sensing waveform and unmanned aerial vehicle trajectory design method for safe mobile edge computing in any one of claims 1 to 4.

6. A communication system implementing the method of any one of claims 1-4, characterized by Comprise: Initialization module, set the unmanned aerial vehicle S as an edge server base station, and initialize the flight trajectory of S to serve K users existing on the ground; In order to better reduce the energy consumption of users, the unmanned aerial vehicle adopts a user scheduling strategy, while considering that there is a potential eavesdropper E, in order to ensure the security of information transmission, the unmanned aerial vehicle transmits radar waves to interfere with E, and realizes safe mobile edge computing; Mathematical model module, with the minimum energy consumption of ground users as the target, an optimization mathematical model is constructed with user scheduling, user offloading ratio, integrated sensing waveform and unmanned aerial vehicle flight trajectory as constraints; The mathematical model decoupling module is configured to decouple the mathematical model based on the BCD method into sub-problems only related to the unloading ratio, the user scheduling, the integrated waveform of the sensory and the flight trajectory of the unmanned aerial vehicle, and to convert each non-convex sub-problem into a convex problem for solving by using a convex approximation fitting method. The sub-problem solving module is configured to solve the unloading ratio, the user scheduling, the integrated waveform of the sensory and the flight trajectory of the unmanned aerial vehicle by using an iterative algorithm.

Images