Unmanned aerial vehicle data security acquisition method and system based on multi-dimensional optimization

Abstract

The invention provides a multi-dimensional optimization-based unmanned aerial vehicle data security acquisition method and system, and the method comprises the steps: constructing an unmanned aerial vehicle-assisted Internet of Things data acquisition system which comprises an unmanned aerial vehicle, a plurality of sensors, a plurality of auxiliary devices and a potential eavesdropper, discretizing a task acquisition period into a plurality of time slots; introducing a sensor scheduling variable to perform dormancy wakeup management on the ground sensor, and introducing an interferer selection variable to select an auxiliary device as a friendly interferer based on a channel condition; establishing a path loss model of a ground-air channel and a ground-ground channel to obtain a channel gain, defining a transmission rate and an eavesdropping rate, considering the uncertainty of the position of an eavesdropper to construct a worst-case secrecy rate model of the unmanned aerial vehicle, and setting constraint conditions; a data acquisition problem is formalized into a secrecy rate maximization problem and decomposed into three sub-problems, the sub-problems are solved, and an optimized unmanned aerial vehicle trajectory, sensor scheduling and jammer selection strategy is output.

Description

Technical Field

[0001] The present invention belongs to the technical field of UAV communication, and relates to a method and system for secure data acquisition of UAVs, and particularly to a method and system for secure data acquisition of UAVs based on multi-dimensional optimization. Background Art

[0002] In recent years, with the development of the sixth-generation mobile communication, the Internet of Things (IoT) has shown great potential in supporting applications such as smart cities, intelligent transportation, and smart healthcare. In these network applications, data in the environment needs to be transmitted from numerous ground sensor nodes (SNs) to a data center for further processing and analysis. However, due to the wide distribution of SNs and limited transmission power, long-distance data acquisition methods are not always feasible. Therefore, unmanned aerial vehicles (UAVs), with their high flexibility and the ability to establish line-of-sight (LoS) transmission links, have become an effective solution for handling data acquisition tasks. In the scenario of UAV-assisted data acquisition, the UAV can fly close to the SNs to shorten the data acquisition distance, enabling the SNs to transmit data at a lower power, thereby extending their battery life.

[0003] On the one hand, although UAVs offer many advantages, ensuring secure and efficient data acquisition in IoT networks assisted by them remains challenging. And under the threat of malicious eavesdropping, the inherent openness of the wireless channel makes private data very easy to leak. Therefore, cooperative jamming (CJ), as a highly potential physical-layer security (PLS) technology, has emerged. The basic idea of CJ is to send artificial noise by using friendly jammers to interfere with the eavesdropper (Eve), thereby greatly increasing the difference between the legitimate channel and the eavesdropping channel. And compared with other PLS technologies, the lightweight characteristic of CJ makes it a more suitable choice.

[0004] On the other hand, energy is also an issue that cannot be ignored in IoT networks assisted by UAVs. First, the limited battery capacity of SNs directly affects the data transmission process. Second, the limited energy of the UAV also restricts its flight speed and path during the data acquisition process. Finally, considering that friendly jammers for sending artificial noise are generally powered by batteries with limited capacity, the limited battery capacity greatly affects the role of cooperative jamming. Therefore, considering the energy consumption of all three simultaneously can ensure more secure and efficient data acquisition in practical situations.

[0005] Existing studies have considered the problems of limited energy consumption and secure data collection in UAV-assisted IoT networks. For example, the study [Y. Li, et al., Data collection maximization in IoT-sensor networks via an energy-constrained UAV (IEEE Trans. Mobile Comput), 2023] investigated the problem of maximizing the data collection volume under the constraint of UAV energy consumption, but did not consider security-related issues; [X. Xiong, C. Sun, W. Ni, and X. Wang, Three-dimensional trajectory design for unmanned aerial vehicle-based secure and energy-efficient data collection (IEEE Trans. Veh. Technol), 2023] studied the methods for bandwidth allocation and UAV trajectory optimization to maximize the system secrecy rate while adopting the PLS technology and considering the energy consumption of the UAV; [R. Zhang, X. Pang, W. Lu, N. Zhao, Y. Chen, and D. Niyato, Dual-UAV enabled secure data collection with propulsion limitation (IEEE Trans. Wireless Commun), 2021] proposed a dual-UAV-assisted secure data collection scheme considering the unknown location of eavesdroppers, but increased the infrastructure deployment cost due to the adoption of dedicated jammers.

[0006] The existing technologies mainly only focus on the energy consumption problems of UAVs or sensor nodes, without considering the importance of the triple energy constraints of UAVs, SNs, and jammers in secure data collection, and usually adopt dedicated jammers to generate artificial noise, thus greatly increasing the deployment cost. At the same time, many studies assume that the location information of eavesdroppers is completely clear, which is obviously unrealistic in practical applications. Therefore, considering the triple energy constraints, using idle IoT auxiliary devices as jammers to transmit interference signals, and ensuring secure data collection at low cost by coordinating UAV trajectory design, sensor scheduling, and jammer selection is more in line with the actual situation. Summary of the Invention

[0007] To solve the above-mentioned technical problems existing in the prior art, the present invention provides a method and system for secure data acquisition of unmanned aerial vehicles (UAVs) based on multi-dimensional optimization. The present invention realizes efficient and secure data acquisition under triple energy constraints by establishing a multi-dimensional optimization framework for maximizing the system secrecy rate and jointly optimizing the UAV trajectory, sensor scheduling, and interferer selection. The technical solutions adopted by the present invention are as follows:

[0008] A method for secure data acquisition of UAVs based on multi-dimensional optimization, comprising the following steps:

[0009] S1. Construct a UAV-assisted Internet of Things (IoT) data acquisition system, where the IoT data acquisition system includes a UAV, several sensors, several auxiliary devices, and a potential eavesdropper, and discretize the task acquisition period into multiple time slots;

[0010] S2. Introduce a sensor scheduling variable to manage the sleep and wake-up of ground sensors, and introduce an interferer selection variable to select an auxiliary device as a friendly interferer based on the channel condition;

[0011] S3. Establish path loss models for the ground-air channel and the ground-ground channel to obtain the channel gain, define the transmission rate and the eavesdropping rate, construct a worst-case secrecy rate model for the UAV considering the uncertainty of the eavesdropper's position, and set constraint conditions;

[0012] S4. Formalize the data acquisition problem as a problem of maximizing the secrecy rate, and decompose it into three sub-problems: UAV trajectory optimization, sensor scheduling optimization, and interferer selection optimization, solve the sub-problems, and output the optimized UAV trajectory, sensor scheduling, and interferer selection strategies.

[0013] Further, the IoT data acquisition system uses a three-dimensional Cartesian coordinate system to describe the position information of all nodes, where the nodes include UAV nodes, sensor nodes, auxiliary device nodes, and potential eavesdropper nodes; the UAV is set to fly at a fixed height to collect data, and the position of the UAV remains unchanged within each time slot.

[0014] Further, a bounded eavesdropper position error model is used to describe the uncertainty of the position of the potential eavesdropper.

[0015] Further, the sensor scheduling variable represents the wake-up or sleep state of the sensor in each time slot, and at most one sensor is in the wake-up state in each time slot; the interferer selection variable represents whether the auxiliary device sends an interference signal in each time slot to confuse the eavesdropper.

[0016] Furthermore, the path loss model of the air - ground channel uses the free - space path loss model to describe the channel gain between the UAV and the sensor; the path loss model of the ground - ground channel uses a combination of large - scale path loss and small - scale Rayleigh fading to describe the channel gains from the sensor to the eavesdropper and from the auxiliary device to the eavesdropper.

[0017] Furthermore, the UAV and the auxiliary device store the same interference signal generator and seed table, and the UAV can eliminate the interference signal through the two - step phase - shift modulation method.

[0018] Furthermore, the construction of the worst - case secrecy rate model of the UAV considering the uncertainty of the eavesdropper's position is specifically as follows:

[0019] Considering the inaccuracy of the eavesdropper's position information, the triangular inequality criterion is used to obtain the lower bound of the distance between the sensor and the eavesdropper and the upper bound of the distance between the auxiliary device and the eavesdropper.

[0020] Based on the lower bound of the distance between the sensor and the eavesdropper and the upper bound of the distance between the auxiliary device and the eavesdropper, the worst - case eavesdropping rate calculated based on the eavesdropper's position estimation error is obtained, and then the worst - case secrecy rate model from the sensor to the UAV is obtained.

[0021] Furthermore, the constraint conditions include data acquisition requirement constraints, sensor energy constraints, auxiliary device energy constraints, UAV energy constraints, UAV maximum speed constraints, UAV initial and final position constraints, sensor mutual - exclusion wake - up constraints, sensor wake - up mode binary variable constraints, and interference selection binary variable constraints.

[0022] Furthermore, the problem of maximizing the secrecy rate is decomposed into three sub - problems: UAV trajectory optimization, sensor scheduling optimization, and interferer selection optimization, by the block coordinate descent method.

[0023] For the UAV trajectory optimization sub - problem, fixing the sensor scheduling variables and the interferer selection variables, by introducing auxiliary variables and applying the successive convex approximation technique, a convex optimization problem is obtained, and then the CVX solver is used to optimize the UAV trajectory variables.

[0024] For the sensor scheduling optimization sub - problem, fixing the UAV trajectory variables and the interferer selection variables, the sensor scheduling variables are relaxed to continuous variables, and by dividing the time slots into several sub - time slots, the continuous solution is rounded to obtain a binary solution that satisfies the constraint conditions.

[0025] For the interferer selection optimization sub - problem, fixing the UAV trajectory variables and the sensor scheduling variables, the continuous convex approximation technique is applied to the objective function to obtain a convex optimization problem, and the obtained continuous solution is reconstructed into a binary solution that satisfies the constraint conditions through a rounding algorithm based on the greedy strategy.

[0026] A UAV data security acquisition system based on multi-dimensional optimization, comprising:

[0027] One or more processors;

[0028] A memory for storing one or more programs;

[0029] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned UAV data security acquisition method based on multi-dimensional optimization.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] 1. The present invention innovatively uses idle Internet of Things devices as friendly jammers, and effectively enhances the security of data transmission through the CJ method, which is more flexible and economical than the method of deploying dedicated jammers;

[0032] 2. The present invention comprehensively considers the triple energy constraints of UAVs, sensor nodes and jammers for the first time, and maximizes the secrecy rate of the system while ensuring the data acquisition requirements by jointly optimizing the UAV trajectory design, sensor scheduling and jammer selection;

[0033] 3. In view of the situation that it is difficult to obtain the accurate position of Eve in the actual scenario, the present invention proposes a new robust optimization method, which effectively solves the problems caused by the uncertainty of Eve's position;

[0034] 4. By innovatively combining the BCD and SCA methods, the present invention successfully transforms the complex mixed-integer non-linear programming problem into a solvable convex optimization problem, and provides an efficient iterative algorithm, which solves the security and efficiency problems faced in the existing UAV-assisted data acquisition. Description of the Drawings

[0035] Figure 1 Schematic diagram of the UAV data security acquisition system based on multi-dimensional optimization in the embodiment of the present invention.

[0036] Figure 2 Algorithm convergence graph of the UAV data security acquisition based on multi-dimensional optimization in the embodiment of the present invention.

[0037] Figure 3 Comparison graph of UAV trajectories under different UAV energy constraints in the embodiment of the present invention.

[0038] Figure 4 Relationship graph between the system secrecy rate and the sensor energy constraint in the embodiment of the present invention.

[0039] Figure 5This is the relationship diagram between the system secrecy rate and the total number of time slots in the embodiments of the present invention.

[0040] Figure 6 This is the relationship diagram between the system secrecy rate and the maximum speed constraint of the UAV in the embodiments of the present invention.

[0041] Figure 7 This is the relationship diagram between the system secrecy rate and the maximum energy constraint of the UAV in the embodiments of the present invention. Detailed implementation manners

[0042] The following will describe in detail the specific implementation manners of the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific implementation manners described herein are only used to illustrate and explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention.

[0043] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0044] Embodiment

[0045] Refer to Figure 1 , a method for secure data collection of UAVs based on multi-dimensional optimization of the present invention includes the following steps:

[0046] S1: Establish a UAV-assisted Internet of Things data collection system, and deploy a UAV (Unmanned Aerial Vehicle, UAV) in it to be responsible for data collection. The system also includes several ground sensor nodes (Sensor Nodes, SNs), several auxiliary devices, and a potential eavesdropper (Eavesdropper, Eve). Construct a three-dimensional Cartesian coordinate system to describe the position information of the four types of nodes, namely UAV, sensor, auxiliary device, and potential eavesdropper, in the Internet of Things data collection system. Set the UAV to fly at a fixed height to collect data; the positions of all nodes are represented by horizontal coordinates and height together. Considering that the eavesdropper usually hides its position, a bounded eavesdropper position error model is used to describe its position uncertainty, and the entire data collection task cycle is discretized into several time slots, so that the UAV remains stationary within each time slot; through time slot division and precise representation of node positions, it lays a foundation for multi-dimensional optimization of subsequent trajectory planning, sensor scheduling, and interference selection.

[0047] Step S1 specifically includes:

[0048] S11: Use a three-dimensional Cartesian coordinate system to describe the position information of all nodes in the Internet of Things data collection network model, including one UAV, M sensors, K auxiliary devices, and one eavesdropper; the origin of the coordinate system can be set at the reference position of the monitoring area, and any point in space is represented by (x, y, z).

[0049] S12: Set the UAV to fly at a fixed altitude of H meters to ensure obstacle-free flight. The two-dimensional coordinates of the UAV at time slot n are represented as:

[0050]

[0051] where q u [n] are the two-dimensional coordinates of the UAV at time slot n, and x u [n] and y u [n] represent the horizontal and vertical coordinates of the UAV at time slot n respectively, and n ∈ {1, 2,..., N} is the time slot index.

[0052] S13: Set both the sensor and the auxiliary device to be deployed on the ground. Therefore, the positions of sensor node m and auxiliary device k are defined as:

[0053]

[0054] where q m is the sensor position, q k is the auxiliary device position, m ∈ {1, 2,..., M} is the sensor node index, k ∈ {1, 2,..., K} is the auxiliary device index, and x m , x k , y m , y k represent the coordinate values of each node in the horizontal and vertical directions respectively.

[0055] S14: Set the eavesdropper to be on the ground and use a bounded eavesdropper position error model to describe the position information of the eavesdropper. Therefore, the estimated position of the eavesdropper is expressed as:

[0056]

[0057] The true position q e of the eavesdropper and the estimated position satisfy:

[0058]

[0059] where χ represents the maximum error range of position estimation, and ||·|| represents the Euclidean distance.

[0060] S15: Define the time required for the entire data collection process as T, and discretize the data collection task cycle T into N equal-length time slots, each time slot having a length of τ t :

[0061] T = N × τ t ;

[0062] Divide time slots to ensure that the UAV maintains its position unchanged within each time slot.

[0063] S2: Construct a node scheduling and interference selection mechanism for UAV-assisted data collection, and introduce sensor scheduling variables and interferer selection variables to describe the system working mode; the sensor nodes adopt a sleep and wake-up mechanism to save energy, ensuring that at most one sensor node transmits data to the UAV in each time slot; meanwhile, the system selects a suitable device from multiple auxiliary devices as a cooperative jammer to send artificial noise to the eavesdropper, increasing the difference between the legitimate communication channel and the eavesdropping channel and improving the confidentiality of data transmission; through the joint design of sensor scheduling and interferer selection, efficient and secure data collection is achieved under the condition of meeting the energy constraint.

[0064] Step S2 specifically includes:

[0065] S21: Introduce the sensor scheduling variable a m [n] represents the wake-up and sleep modes of the sensor node in the time slot:

[0066]

[0067] where a m [n]=1 means that the UAV collects data from the sensor node m in the time slot n, and a m [n]=0 means that the sensor node m is in the sleep state in the time slot n, is the set of sensor nodes, is the set of time slots.

[0068] S22: Introduce the interferer selection variable o k [n] represents the interference decision of the auxiliary device in the time slot:

[0069]

[0070] where o k [n]=1 means that the auxiliary device k is selected as a friendly jammer to send interference signals to confuse the eavesdropper in the time slot n, and o k [n]=0 means that the auxiliary device k does not perform interference operations in the time slot n, is the set of auxiliary devices.

[0071] S23: Set the mutually exclusive wake-up constraint condition of the sensor nodes to ensure that at most one sensor node is in the wake-up state in each time slot to save energy:

[0072]

[0073] This constraint ensures that at most one sensor node transmits data to the UAV in any time slot, avoiding conflicts and energy waste caused by multiple sensors transmitting simultaneously.

[0074] S24: Strategically select ground idle auxiliary devices as friendly jammers in each time slot according to the channel conditions and the available energy budget of the device to achieve cooperative jamming to prevent eavesdropping; when the UAV collects data from a specific sensor node in time slot n, the system will select a suitable auxiliary device to send a jamming signal according to the channel state and energy status, and the jamming signal is synchronized with the sensor data transmission to form an effective jammer for the eavesdropper.

[0075] S3: Establish the channel model and energy constraints of the UAV-assisted secure data acquisition system, and adopt different path loss models for the air-ground channel and the ground-ground channel respectively; considering the uncertainty of the eavesdropper's position, deduce the boundary conditions of the channel gain through the triangle inequality; construct the legal transmission rate model between the sensor node and the UAV and the eavesdropping rate model between the sensor node and the eavesdropper; introduce the worst-case system secrecy rate as the security performance evaluation index; at the same time, accurately model the propulsion energy consumption of the UAV and set the energy constraints of various nodes to ensure the long-term working ability of the system under energy constraints.

[0076] Step S3 specifically includes:

[0077] S31: Use the free space path loss model to describe the channel characteristics between the UAV and the ground sensor node, and the channel gain h m,u [n] is expressed as:

[0078]

[0079] where β0 represents the path loss coefficient at a reference distance of 1 meter, and ||q m -q u [n]|| represents the horizontal distance between the sensor node m and the UAV in time slot n.

[0080] S32: Establish a channel model in a complex ground environment, and use a combination of large-scale path loss and small-scale Rayleigh fading to describe the channel gains h from the sensor node to the eavesdropper and from the auxiliary device to the eavesdropper i,e :

[0081]

[0082] where ||q i -q e || represents the distance between node i and the eavesdropper, α i,e > 2 is the path loss exponent between node i and the eavesdropper, and ξ is a random variable subject to a unit mean exponential distribution, representing the small-scale fading characteristics of the channel.

[0083] S33: When the UAV and the auxiliary device store the same interference signal generator and seed table, the UAV can effectively eliminate the interference signal using the two-step phase shift modulation method. Therefore, define the achievable data transmission rate R m,u [n] from sensor node m to the UAV in time slot n:

[0084]

[0085] where B is the system transmission bandwidth, P m is the transmission power of sensor node m, and σ 2 is the receiver noise power.

[0086] S34: Define the achievable data transmission rate R m,e [n] from sensor node m to the eavesdropper in time slot n:

[0087]

[0088] where h m,e is the channel gain from sensor node m to the eavesdropper, o k [n] is the selection decision variable of auxiliary device k in time slot n, P k is the transmission power of auxiliary device k, and h k,e is the channel gain from auxiliary device k to the eavesdropper.

[0089] S35: Using the triangle inequality criterion, for the uncertainty of the eavesdropper's location information, derive the lower bound of the distance between sensor node m and the eavesdropper:

[0090]

[0091] and the upper bound of the distance between auxiliary device k and the eavesdropper:

[0092]

[0093] where represents the distance between sensor node m and the estimated location of the eavesdropper, represents the distance between auxiliary device k and the estimated location of the eavesdropper.

[0094] S36: Construct the worst-case system secure transmission secrecy rate model R sec :

[0095]

[0096] where, [·] + represents max{0,·}, represents the worst-case eavesdropping rate calculated based on the eavesdropper location estimation error, specifically expressed as:

[0097]

[0098] Among them,

[0099] S37: Establish the energy consumption constraint conditions for sensor nodes:

[0100]

[0101] Among them, is the maximum available energy of sensor node m, is the set of sensor nodes.

[0102] S38: Set the energy consumption constraint conditions for auxiliary devices:

[0103]

[0104] Among them, is the maximum available energy of auxiliary device k, is the set of auxiliary devices.

[0105] S39: Construct an accurate model for the propulsion energy consumption of UAVs:

[0106]

[0107] Among them

[0108]

[0109] The energy consumption of the UAV includes two parts, namely communication energy consumption and propulsion energy consumption. Since in this system, the communication energy consumption of the UAV is negligible compared to the propulsion energy consumption, only the propulsion energy consumption is considered. Among them, P0 represents the blade profile power of the UAV, Δ[n] = ||q u [n] - q u [n - 1]|| represents the flight distance of the UAV between adjacent time slots, U tip represents the tip speed of the rotor blade, d0 represents the fuselage drag ratio, ρ represents the air density, A represents the rotor disk area, s represents the rotor solidity, P I represents the induced power of the UAV, and v0 represents the average rotor induced speed in the hover state.

[0110] S310: Set the energy constraint conditions for UAVs:

[0111]

[0112] Among them, Denote the maximum available energy of the UAV, and ensure that the energy consumption of the UAV during the entire data collection task does not exceed its maximum available energy.

[0113] S4: Model the UAV data secure collection problem based on multi-dimensional optimization and design the algorithm. Formalize the secure data collection problem as a problem of maximizing the system secrecy rate. Under the multi-dimensional optimization framework considering UAV trajectory design, sensor scheduling, and interferer selection, simultaneously satisfy multiple constraints such as data collection requirements, energy consumption limits, and flight speed constraints. Since the original problem is a mixed-integer non-linear programming problem with a high solution complexity, a decomposition and iteration method is used for solution. Transform the complex problem into multiple tractable sub-problems through techniques such as block coordinate descent and successive convex approximation.

[0114] Step S4 specifically includes:

[0115] S41: Construct the problem of maximizing the system secrecy rate:

[0116]

[0117] where denotes the UAV trajectory design variable, denotes the sensor scheduling variable, denotes the interferer selection variable.

[0118] S42: Set the constraint conditions including:

[0119] Data collection requirement constraint: where L m denotes the amount of data to be collected from sensor node m, and ensure that the data collection task of each sensor node is completed.

[0120] Sensor node energy constraint: Ensure that the total energy consumption of each sensor node does not exceed its maximum available energy

[0121] Auxiliary device energy constraint: Ensure that the total energy consumption of each auxiliary device does not exceed its maximum available energy

[0122] UAV energy constraint: Ensure that the total energy consumption of the UAV does not exceed its maximum available energy

[0123] UAV maximum speed constraint: where V max denotes the maximum flight speed of the UAV, and ensure that the flight speed of the UAV is within the physically feasible range.

[0124] UAV Initial and Final Position Constraints: q u [1] = q0, q u [N] = q0, where q0 represents the initial position of the UAV, ensuring that the UAV returns to the starting point after the mission is completed.

[0125] Mutual Exclusion Wake-up Constraint for Sensor Nodes: Ensure that at most one sensor node is in the wake-up state in each time slot to avoid transmission conflicts.

[0126] Wake-up Mode Binary Variable Constraint: Specify the value range of the sensor scheduling variable.

[0127] Interference Selection Binary Variable Constraint: Specify the value range of the interferer selection variable.

[0128] S43: By setting a m [n] = 0, the leakage of privacy information can be avoided, so [·] + operation can be removed from the original problem in the.

[0129] S44: Adopt the Block Coordinate Descent (BCD) method to decompose the original problem into three sub-problems, namely UAV trajectory optimization, sensor scheduling, and interferer selection, and solve them through an alternating iteration method until the objective function converges. The advantage of the BCD method is to transform a high-dimensional complex problem into multiple low-dimensional sub-problems, reducing the difficulty of solution, and is suitable for optimization problems with strong coupling relationships between variables.

[0130] S45: For the UAV trajectory optimization sub-problem

[0131]

[0132] By defining introduce an auxiliary variable γ[n] to handle the non-convex term in the UAV propulsion energy consumption constraint:

[0133]

[0134] Thus, the sub-problem is transformed into

[0135]

[0136] Apply the Successive Convex Approximation (SCA) technique, and use the first-order Taylor expansion to Linearize at the feasible point as follows:

[0137]

[0138] where γ (i) [n] and represent the auxiliary variable and the UAV position at the n-th time slot of the i-th iteration during the iterative process, respectively;

[0139] Perform a similar treatment on the non-convex term R m,u [n] to obtain its linear approximation lower bound:

[0140]

[0141] Solve this sub-problem by iteratively updating the Taylor expansion point.

[0142] S46: Relax the binary variable a m [n] ∈ {0, 1} to a continuous variable a m [n] ∈ [0, 1] to obtain the sensor scheduling optimization sub-problem

[0143]

[0144] This problem is a convex optimization problem and can be solved by CVX. CVX is a powerful tool for convex optimization problems and can efficiently handle problems such as linear programming, quadratic programming, and semidefinite programming; the obtained continuous solution is reconstructed into a binary solution using an improved rounding method: each time slot is divided into G sub-time slots, and the corresponding number of sub-time slots is allocated to each sensor node according to the value of a m [n]. The formula for calculating the number of sub-time slots is ψ m = Ω(G · a m [n]), where Ω(·) represents the operation of taking the closest integer; this improved rounding method can better meet the data acquisition and energy constraints compared to simple rounding.

[0145] S47: For the jammer selection optimization sub-problem

[0146]

[0147] Rewrite as:

[0148]

[0149] Apply the first-order Taylor expansion to the first term for approximation to obtain the upper bound of v m [n]

[0150]

[0151] where represents the interferer selection variable at the \(i\)-th iteration and the \(n\)-th time slot. The binary variable \(o k [n]\in\{0,1\} is relaxed to a continuous variable \(o k [n]\in[0,1]\) to obtain the sub-problem

[0152]

[0153] After relaxing the binary variable to a continuous variable and solving it to obtain a continuous solution, a rounding algorithm based on a greedy strategy is used to reconstruct the continuous solution into a binary solution. Subsequently, it is checked whether each auxiliary device satisfies the energy constraint If not, the number of time slots allocated to this device to send interference signals is reduced until all energy constraints are satisfied.

[0154] S48: By iteratively repeating the above three sub-problems, the UAV trajectory design, sensor scheduling, and interferer selection are continuously optimized until the system secrecy rate converges or reaches the preset maximum number of iterations; finally, the optimized UAV trajectory sensor scheduling strategy and interferer selection decision are obtained to achieve efficient and secure data collection under energy-constrained conditions. The convergence judgment condition is that the change in the objective function value between two consecutive iterations is less than the preset threshold \(\epsilon\), that is where represents the system secrecy rate at the \(i\)-th iteration.

[0155] As Figure 2 shown, in different scenarios, the proposed algorithm shows a trend of faster convergence within a limited number of iterations. In addition, Figure 2 the subgraph in

[0156] Figure 3 also illustrates the convergence process of the algorithm. It can be clearly seen that the system secrecy rate achieved by the algorithm continuously increases until convergence.

[0157] Figure 4Shows the secrecy rate of the proposed scheme and the baseline scheme under the maximum energy of different sensor nodes. It can be seen that as the energy budget of each sensor node increases, the secrecy rate of all schemes improves, except for the pre-arranged data transmission scheme. This improvement is because more available energy allows each sensor node to have more time to transmit data, thus enhancing the system's performance in terms of secrecy rate. However, in the pre-arranged data transmission scheme, each sensor node follows a predetermined order and occupies a specific number of time slots for data transmission. Therefore, changing the maximum energy of each sensor node does not affect the secrecy rate under this scheme. Secondly, the pre-arranged data transmission scheme exhibits the worst performance, which means that in order to prevent unnecessary energy consumption, it is essential to carefully arrange sensor nodes to transmit data. Moreover, as the available energy of each sensor node increases, the curve of the proposed scheme in this invention rises the fastest. This is because the multi-dimensional optimization method of this invention endeavors to make full use of the available energy of each sensor node, thereby minimizing the impact of invalid UAV trajectories and jammer selections on UAV-assisted secure data collection.

[0158] In Figure 5 , the secrecy rate under different numbers of time slots is evaluated. It is observed that when the number of time slots reaches 100, the performance of the random jammer selection scheme deteriorates. The reason is that increasing the number of time slots may lead to invalid jamming attempts, especially when the UAV does not collect data from the SNs. Except for the random jammer selection scheme, the secrecy rate of other schemes improves as the number of time slots increases. This improvement can be attributed to the fact that increasing the number of time slots provides the UAV with more flexibility to get closer to the sensor nodes to collect data. At the same time, as the number of time slots increases, the sensor nodes can occupy more time to transmit data to the UAV, thereby further enhancing the secrecy rate. In addition, the proposed scheme in this invention shows better performance than the baseline scheme, which verifies that the synergistic effect brought by multi-dimensional optimization is valuable.

[0159] Figure 6 The impact of the maximum speed of the UAV on the secrecy rate is studied. From Figure 6Three observations can be drawn. First, as the maximum speed of the UAV increases, the secrecy rate of all schemes except the fixed UAV flight trajectory scheme increases. This is mainly attributed to the fact that the improvement of the UAV's flight ability provides flexibility for designing the UAV's trajectory. In contrast, the fixed UAV flight trajectory scheme sets a predetermined trajectory for the UAV, which means that the speed of the UAV in each time slot is pre-arranged. Therefore, the secrecy rate of this scheme is not affected by changing the maximum speed of the UAV. Second, as the maximum speed of the UAV continues to increase, the growth rate of the secrecy rate of all schemes except the fixed UAV flight trajectory scheme decreases. This trend can be attributed to the fact that the speed of the UAV is not the main factor affecting the data collection efficiency. Third, due to the multi-dimensional coordination in the process of trajectory planning, sensor wake-up scheduling, and jammer selection, the proposed scheme in the present invention improves by approximately 59.68%, 210.91%, 13.90%, and 25.48% compared with the four baseline schemes respectively, which verifies its superior performance in terms of secrecy rate.

[0160] In Figure 7 , we show the secrecy rate under different maximum energy budgets of the UAV. It can be observed that in all schemes, except the fixed UAV flight trajectory scheme, as the maximum energy budget of the UAV is higher, the secrecy rate also becomes higher. The reason is that a higher energy supply makes the UAV more flexible in trajectory planning, thus further improving its data collection performance. Although the UAV has sufficient energy, the energy budgets of the sensor hijacking points and auxiliary devices limit the performance improvement. Therefore, the growth rate of the secrecy rate in all schemes except the fixed UAV flight trajectory scheme shows a downward trend. Similarly, due to the reasons mentioned in the above analysis, the secrecy rate in the fixed UAV flight trajectory scheme remains unchanged. In addition, thanks to the fact that the UAV trajectory, sensor node scheduling, and jammer selection can be flexibly adjusted according to different available energy budgets, the proposed scheme shows the best performance compared with the baseline schemes.

[0161] Referring to Figures 2 to 7 , the simulation results show that for a UAV data security acquisition system based on multi-dimensional optimization, after adopting the joint optimization method of the present invention, within the specified time, the data of all devices can be completely acquired, the total secrecy rate of the system increases, and the system performance is improved, significantly superior to other comparison schemes.

[0162] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of an all-hardware embodiment, an all-software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0163] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for realizing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of multiple flows and / or blocks.

[0164] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means realizes the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of multiple flows and / or blocks.

[0165] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in Figure 1 one or more of the flows Figure 1 or blocks or the combination of multiple flows and / or blocks.

[0166] The above are only the preferred embodiments of the present invention. Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make many possible changes and modifications to the technical solution of the present invention, or modify it into equivalent embodiments with equivalent changes, without departing from the scope of the technical solution of the present invention. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for securely collecting drone data based on multi-dimensional optimization, characterized in that: The following steps are involved: S1. Construct a drone-assisted IoT data collection system, which includes a drone, several sensors, several auxiliary devices and a potential eavesdropper, and discretizes the task collection cycle into multiple time slots; S2. Introduce sensor scheduling variables to manage the sleep and wake-up of ground sensors, and introduce interferer selection variables to select auxiliary devices as friendly interferers based on channel conditions; S3. Establish the path loss model of ground-to-air channel and ground-to-ground channel to obtain channel gain, define the transmission rate and eavesdropping rate, consider the uncertainty of the eavesdropper's position to build the worst-case confidentiality rate model of the UAV, and set constraints; S4. The data collection problem is formalized as a confidentiality rate maximization problem, and decomposed into three sub-problems: UAV trajectory optimization, sensor scheduling optimization, and jammer selection optimization. The sub-problems are solved and the optimized UAV trajectory, sensor scheduling, and jammer selection strategies are output.

2. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1 is characterized in that: The Internet of Things data acquisition system uses a three-dimensional Cartesian coordinate system to describe the location information of all nodes, and the nodes include drone nodes, sensor nodes, auxiliary equipment nodes and potential eavesdropper nodes; the drone is set to fly at a fixed altitude to collect data, and the drone's position remains unchanged in each time slot.

3. The method for securely collecting drone data based on multi-dimensional optimization according to claim 2 is characterized in that: A bounded eavesdropper location error model is used to describe the uncertainty of the potential eavesdropper's location.

4. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1 is characterized in that: The sensor scheduling variable indicates the awake or sleep state of the sensor in each time slot, and at most one sensor is in the awake state in each time slot; the interferer selection variable indicates whether the auxiliary device sends an interference signal in each time slot to confuse the eavesdropper.

5. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1 is characterized in that: The path loss model of the ground-to-air channel adopts a free space path loss model to describe the channel gain between the UAV and the sensor; the path loss model of the ground-to-ground channel adopts a combination of large-scale path loss and small-scale Rayleigh fading to describe the channel gain from the sensor to the eavesdropper and from the auxiliary equipment to the eavesdropper.

6. The method for securely collecting drone data based on multi-dimensional optimization according to claim 5 is characterized in that: The drone and the auxiliary device store the same interference signal generator and seed table, and the drone is able to eliminate the interference signal through a two-step phase shift modulation method.

7. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1, characterized in that: The worst-case confidentiality rate model of the drone considering the uncertainty of the eavesdropper's position is specifically: Considering the inaccuracy of the eavesdropper's location information, the triangle inequality criterion is used to obtain the lower limit of the distance between the sensor and the eavesdropper and the upper limit of the distance between the auxiliary device and the eavesdropper. According to the lower limit of the distance between the sensor and the eavesdropper and the upper limit of the distance between the auxiliary device and the eavesdropper, the worst-case eavesdropping rate calculated based on the eavesdropper position estimation error is obtained, and then the worst-case confidentiality rate model from the sensor to the drone is obtained.

8. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1, characterized in that: The constraints include data collection requirement constraints, sensor energy constraints, auxiliary equipment energy constraints, drone energy constraints, drone maximum speed constraints, drone initial and final position constraints, sensor mutually exclusive wake-up constraints, sensor wake-up mode binary variable constraints, and interference selection binary variable constraints.

9. The method for securely collecting drone data based on multi-dimensional optimization according to claim 1, characterized in that: The confidentiality rate maximization problem is decomposed into three sub-problems: UAV trajectory optimization, sensor scheduling optimization and jammer selection optimization through block coordinate descent method. For the sub-problem of drone trajectory optimization, the sensor scheduling variables and the interferer selection variables are fixed, and by introducing auxiliary variables, the continuous convex approximation technology is applied to obtain the convex optimization problem, and then the CVX solver is used to optimize the drone trajectory variables; For the sensor scheduling optimization subproblem, the drone trajectory variables and the jammer selection variables are fixed, the sensor scheduling variables are relaxed to continuous variables, and the continuous solutions are rounded off to obtain binary solutions that meet the constraints by subdividing the time slot into several sub-time slots. For the jammer selection optimization subproblem, the UAV trajectory variables and sensor scheduling variables are fixed, and the continuous convex approximation technique is applied to the objective function to obtain a convex optimization problem. The obtained continuous solution is reconstructed into a binary solution that satisfies the constraints through a rounding algorithm based on a greedy strategy.

10. A drone data security collection system based on multi-dimensional optimization, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the drone data security collection method based on multi-dimensional optimization as described in any one of claims 1-9.

Images