Multi-uav cooperative secure mobile edge computing privacy throughput efficiency optimization method

Abstract

The present application relates to general aviation equipment mobile edge computing data security technical field, specifically relates to the security mobile edge computing secret throughput efficiency optimization method of multi-unmanned plane cooperation, the method includes: building multi-unmanned plane cooperation security mobile edge computing system;Channel and rate model, task and delay model are constructed, and a secret throughput efficiency model is constructed;Based on the secret throughput efficiency model, a joint optimization problem with the maximum secret throughput efficiency as the target is constructed;The joint optimization problem is solved.The present application strengthens the data transmission security of physical layer by multi-unmanned plane cooperation, realizes the balance of data security and computing efficiency with composite index, and the solving framework is efficient and feasible, significantly improves the effective bit number of safe task completion per unit time of unmanned plane, and is suitable for low-altitude economy and other security sensitive edge computing scenarios.

Description

Technical Field

[0001] This invention relates to the field of mobile edge computing data security technology for general aviation equipment, specifically to a method for optimizing the secure mobile edge computing throughput efficiency of multi-UAV collaboration. Background Technology

[0002] With the rapid development of the Internet of Things (IoT), 5G / 6G mobile communications, and artificial intelligence applications, a large number of computationally intensive and latency-sensitive services are emerging, such as augmented reality, connected vehicles, low-altitude economy, and the industrial internet. Mobile edge computing, by deploying computing servers close to the user side, can significantly reduce task transmission latency and core network load, and has become an important component of the next-generation network architecture. Drones, with their advantages of flexible deployment, high mobility, and high probability of line-of-sight links, can serve as aerial mobile edge servers (MEC UAVs) or collaborative nodes, providing on-demand computing and communication services to ground users, and have broad application prospects in disaster relief, temporary hotspot coverage, and low-altitude logistics. However, the low-altitude open environment in which drones operate also makes wireless links more vulnerable to eavesdropping, and traditional security mechanisms relying solely on upper-layer encryption are insufficient to adequately address physical-layer eavesdropping threats.

[0003] In recent years, physical layer security technologies have garnered widespread attention for their ability to suppress eavesdropping links and improve security rates from an information theory perspective, utilizing methods such as channel randomness, beamforming, and interference cooperation. In UAV-assisted communication and mobile edge computing scenarios, existing research has considered objectives such as maximizing security rates, minimizing energy consumption, or minimizing latency, exploring the joint optimization of user power allocation, task offloading strategies, and UAV trajectories. However, current research typically uses single metrics such as "instantaneous security rate" or "average throughput" as optimization objectives, making it difficult to simultaneously characterize the comprehensive impact of factors such as task data volume, end-to-end latency, and security success probability on system performance.

[0004] Furthermore, existing research largely focuses on scenarios with single UAVs or no interfering nodes, with relatively little research on system modeling and joint optimization of "multi-UAV cooperative" scenarios (including computational UAVs and jamming UAVs). In multi-UAV cooperative scenarios, the flight trajectory of the computational UAV not only affects the channel quality and offloading delay of the legitimate link, but is also limited by physical constraints such as flight speed and energy; the trajectory and power of the jamming UAV have a coupled impact on the eavesdropping link and the security rate. How to jointly design the ground user's transmit power, the allocation of tasks on the local and UAV sides, and the cooperative trajectories of the two types of UAVs within a limited time window to maximize the overall task processing efficiency while satisfying security constraints and system resource constraints is a challenging optimization problem with strong non-convexity and high coupling.

[0005] On the other hand, traditional performance metrics such as "maximizing secure rate" or "minimizing total latency" are insufficient to reflect the overall efficiency of securely completing the amount of data per unit time. In mobile edge computing scenarios with eavesdroppers, both task completion latency and the probability of successful "secure transmission" at the physical layer must be considered. Optimizing only rate or latency may lead to an imbalance between security and efficiency in resource allocation. Therefore, it is necessary to construct a composite metric that balances security and task processing efficiency, and design corresponding efficient solution algorithms.

[0006] In summary, existing technologies generally suffer from the following shortcomings in secure mobile edge computing scenarios involving multi-UAV collaboration: First, there is a lack of comprehensive efficiency indicators focusing on "the number of bits of data security completed per unit time," making it difficult to fully characterize the balance between data link security and computational task latency. Second, there is insufficient research on the joint design of collaborative trajectories between computing UAVs and jamming UAVs, failing to fully utilize jamming collaboration to improve the security performance of the physical layer data link. Third, there is no systematic and efficient solution framework for the unified joint optimization problem of user power, task allocation, and UAV trajectories. There is an urgent need to provide a method for achieving multi-user secure mobile edge computing task offloading and resource joint optimization using multi-UAV collaboration under conditions of eavesdropping threats, especially a joint power allocation, task allocation, and trajectory design method with confidentiality throughput efficiency as the optimization objective. Summary of the Invention

[0007] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a method for optimizing the confidentiality throughput efficiency of secure mobile edge computing in multi-UAV collaborative systems. In a multi-user UAV-assisted mobile edge computing (MEC) system with eavesdroppers, the invention maximizes the confidentiality throughput efficiency (STE) under system resource constraints and data security constraints by jointly optimizing the ground user transmit power, task offload time slot allocation, and the flight trajectories of the computing UAV and the interfering UAV, thereby increasing the number of effective bits for securely completing tasks per unit time.

[0008] To achieve the above objectives, this invention provides a method for optimizing the confidential throughput efficiency of secure mobile edge computing through multi-UAV collaboration. This method employs a dual-UAV collaboration approach and is suitable for systems consisting of multiple ground users, one MEC UAV, one Jammer UAV, and one eavesdropper. The method includes the following steps: In a first aspect, the present invention provides a method for optimizing the confidential throughput efficiency of secure mobile edge computing in multi-UAV collaboration, the method comprising: Establish a multi-UAV collaborative secure mobile edge computing system, the system including... One ground user, one computing drone, one jamming drone, and one fixed eavesdropper; Construct a channel and rate model, a task and delay model, and build a secure throughput efficiency model based on the channel and rate model and the task and delay model; Based on the security throughput efficiency model, a joint optimization problem is constructed with the goal of maximizing security throughput efficiency. The optimization variables include: the transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, and the two-dimensional positions of the computational UAV and the jamming UAV in each time slot. The constraints include: the transmit power of each user in each time slot does not exceed the user's maximum transmit power; the flight speed of the computational UAV and the jamming UAV does not exceed the maximum horizontal flight speed, and the trajectory satisfies the given two-dimensional position conditions of the starting point and the ending point; task bit conservation constraints and upper limits of local and computational UAV computing resources; and a lower limit constraint of basic security rate. Solve the joint optimization problem to obtain the optimal or near-optimal transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, calculate the two-dimensional position of the UAV and the jamming UAV in each time slot, and calculate the final security throughput efficiency.

[0009] In one possible implementation, a channel and rate model is constructed, specifically as follows: Based on the line-of-sight channel model, the legitimate link channel gain of each user to the computing drone in each time slot, the eavesdropping link channel gain of each user to the eavesdropper, and the interference link channel gain of the jamming drone to the eavesdropper in each time slot are established. Based on Shannon's formula, the legitimate link transmission rate and the eavesdropping rate of each user in each time slot are calculated, and the confidential rate of each user in each time slot is also calculated.

[0010] In one possible implementation, the formula for calculating the legitimate link channel gain from each user to the computing drone in each time slot is: in, Indicates the first Individual users in time slots To calculate the legal link channel gain of the drone, Indicates the reference channel gain. This indicates the calculation of the drone in the time slot. Two-dimensional position, Indicates the first The location of each user This indicates the calculated flight altitude of the drone; The formula for calculating the channel gain of the eavesdropping link from each user to the eavesdropper is as follows: in, Indicates the first Channel gain of the eavesdropping link from the user to the eavesdropper Indicates the first Slow fading coefficient for each user Indicates the location of the eavesdropper; The formula for calculating the channel gain of the jamming link from the jamming drone to the eavesdropper in each time slot is as follows: in, This indicates that the jamming drone is in the time slot. To the eavesdropper's interference link channel gain, This indicates that the jamming drone is in the time slot. Two-dimensional position, This indicates the flight altitude of the interfering drone.

[0011] In one possible implementation, the formula for calculating the security rate for each user in each time slot is: in, Indicates the first Individual users in time slots The rate of confidentiality, Indicates the first Individual users in time slots The legal link transmission rate, Indicates the first Individual users in time slots The rate at which it is spied on.

[0012] In one possible implementation, a task and latency model is constructed, specifically as follows: The total number of bits for each user's task is divided into the number of bits for the local execution part and the number of bits offloaded to the computational drone execution part; Calculate the local computing latency, drone-side computing latency, and offloading transmission latency for each user to determine the total task completion latency for each user.

[0013] In one possible implementation, the formula for calculating the total task completion latency for each user is: in, Indicates the first Total task completion time for each user Indicates the first Local computing latency for each user Indicates the first Offload transmission latency for individual users Indicates the first The computational latency of the drone terminal for each user; No. Offload transmission latency for individual users The calculation formula is: in, Indicates the total number of time slots. Indicates the first Individual users in time slots The unload bit allocation variable, Indicates the first Individual users in time slots The legal link transmission rate.

[0014] In one possible implementation, a secure throughput efficiency model is constructed based on the channel and rate model and the task and delay model, specifically as follows: Based on the security rate and offload bit weight of each user in each time slot, the effective security rate of each user is constructed, and the security success probability of each user is obtained. The confidentiality throughput efficiency of the system is obtained based on the effective task completion data volume of all users, the probability of secure success, and the total task completion latency.

[0015] In one possible implementation, the formula for calculating the probability of successful security for each user is: in, Indicates the first The probability of a user's successful security operation. This represents an empirical parameter for adjusting security sensitivity. Indicates the first Effective confidentiality rate per user Indicates the total number of time slots. Indicates the first Individual users in time slots The unloaded bit weight, Indicates the first Individual users in time slots The rate of confidentiality, Indicates the first Individual users in time slots The unload bit allocation variable, This represents the amount of bits that are offloaded to the computational drone execution section.

[0016] In one possible implementation, the formula for calculating the system's confidential throughput efficiency is: in, This indicates the system's secure throughput efficiency. Indicates the total number of users. Indicates the first The amount of valid completed task data per user Indicates the first Total task completion time for each user.

[0017] In one possible implementation, solving the joint optimization problem specifically involves: Step S41: Using a fractional programming method based on the Dinkelbach transform, the joint optimization problem is transformed into an iterative solution of a series of difference form subproblems. Step S42: In each Dinkelbach iteration, the block coordinate descent method is used to divide the optimization variables into three variable blocks: task block, power block, and trajectory block, and perform alternating optimization. Step S43: For the non-convex subproblems generated by the block coordinate descent method, a continuous convex approximation technique is used to perform a local convex approximation, and a convex optimization solver is used to solve the problem. Step S44: Iteratively execute steps S41 to S43 until the entire algorithm exhibits monotonically convergent behavior, obtain the optimal or near-optimal transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, calculate the two-dimensional positions of the UAV and jamming UAV in each time slot, and calculate the final security throughput efficiency.

[0018] Compared with the prior art, the present invention has the following beneficial effects: This invention proposes a Secure Throughput Efficiency (STE) performance metric for secure mobile edge computing systems with dual-UAV collaborative data links. This metric unifies task data volume, data transmission security success probability, and end-to-end task completion latency into a single fractional metric, providing a more comprehensive characterization of the number of securely completed effective task bits per unit time. Compared to optimizing only the security rate or latency, this approach better meets the practical needs of secure mobile edge computing scenarios that balance data security and computational efficiency.

[0019] This invention establishes a dual-UAV secure mobile edge computing system model that includes a computing UAV and a jamming UAV. By rationally modeling legitimate links, eavesdropping links, and jamming links, it fully utilizes the ability of jamming UAVs to collaboratively interfere with eavesdropping links, thereby significantly improving the physical layer data link security performance of the system while ensuring the service quality of legitimate data links.

[0020] This invention proposes a hierarchical solution framework based on Dinkelbach fractional programming, block coordinate descent, and continuous convex approximation. It decomposes the original non-convex fractional joint optimization problem into a series of convex or approximately convex subproblems. The algorithm has a clear structure, is simple to implement, and has good convergence and engineering feasibility.

[0021] Under the same system parameters and resource constraints, the method of the present invention can achieve a higher STE value compared with baseline schemes such as "fixed trajectory + fixed power", "fixed trajectory only", "fixed power only" and "no interference machine". This is manifested in that it can effectively reduce the latency of computing tasks while ensuring or improving the probability of successful data link transmission security, thereby significantly improving the overall security task processing efficiency of the system. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating the secure mobile edge computing confidentiality throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention; Figure 2 This is a schematic diagram of a scenario for a multi-UAV collaborative secure mobile edge computing system provided in an embodiment of the present invention; Figure 3 A comparison of the STE convergence curves of the secure mobile edge computing confidential throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention with the "fixed trajectory + fixed power", "fixed trajectory only", "fixed power only" and "no interference machine" schemes; Figure 4 The curves comparing the final STE (Secure Mobile Efficiency) of the secure mobile edge computing confidential throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention with the "fixed trajectory + fixed power", "fixed trajectory only", "fixed power only" and "no interference machine" schemes under different user number scenarios are shown. Detailed Implementation

[0024] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0025] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0026] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments and claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0027] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0028] See Figure 1 This is a flowchart illustrating the secure mobile edge computing confidentiality throughput efficiency optimization method for multi-UAV collaboration provided in an embodiment of the present invention. Figure 1 As shown, the method includes: Step S1: Build a multi-drone collaborative secure mobile edge computing system. (See [link]) Figure 2 This is a schematic diagram of a scenario for a multi-UAV collaborative secure mobile edge computing system provided in an embodiment of the present invention, such as... Figure 2 As shown, the system includes One ground user, one computational drone (MEC UAV), one jamming drone (Jammer UAV), and one stationary eavesdropper.

[0029] On a two-dimensional plane, the first The location of each user is set as follows The location of the eavesdropper is set as Total service time Discretized There are several time slots of equal length, each time slot having a duration of [missing information]. ; will calculate the drone in time slots The two-dimensional position is denoted as Interfering with drones in time slots The two-dimensional position is denoted as Total system bandwidth Distributed according to the principle of equal distribution Individual users; setting the flight altitude of computational drones and interference drones. , and maximum flight speed And the initial and final positions; setting the reference channel gain. Noise spectral density Maximum transmit power of users jammer power Channel parameters; setting the computational intensity of the user, the computing drone, and the jamming drone; the peak computing power of the user, the computing drone, and the jamming drone; and the total number of bits for each user's task. .

[0030] In this embodiment, the system parameters can be selected as follows: , , Total service time The calculated flight altitude of both the drone and the jamming drone is 20m, and their maximum horizontal flight speed is... The speed is 40 m / s; the total system bandwidth is... The 10MHz band is divided equally among three users; reference channel gain Noise spectral density Maximum transmit power of users jammer power The computational intensity for both the user-compute drone and the jamming drone is 1000 cycles / bit, with peak computing power of 2 GHz and 3 GHz for the user-compute drone and the jamming drone, respectively; the total number of system task bits. The tasks are evenly distributed among 3 users, with each user receiving a total of [number] bits. .

[0031] Step S2: Construct a channel and rate model and a task and delay model, and construct a secure throughput efficiency model based on the channel and rate model and the task and delay model.

[0032] The channel and rate model is constructed as follows: Step S211: Assume that the links between the user and the computing drone and the jamming drone, and between the computing drone, the jamming drone and the eavesdropper are line-of-sight channels, and the path loss is given by the free space model. Under the line-of-sight (LoS) free space path loss model, establish the legitimate link channel gain of each user to the computing drone in each time slot, the eavesdropping link channel gain of each user to the eavesdropper, and the jamming link channel gain of the jamming drone to the eavesdropper in each time slot.

[0033] The formula for calculating the legitimate link channel gain from each user to the drone in each time slot is as follows: in, Indicates the first Individual users in time slots To calculate the legal link channel gain of the drone, Indicates the reference channel gain. This indicates the calculation of the drone in the time slot. Two-dimensional position, Indicates the first The location of each user This indicates the calculated flight altitude of the drone.

[0034] The formula for calculating the channel gain of the eavesdropping link from each user to the eavesdropper is as follows: in, Indicates the first Channel gain of the eavesdropping link from the user to the eavesdropper Indicates the reference channel gain. Indicates the first Slow fading coefficient for each user Indicates the first The location of each user Indicates the location of the eavesdropper.

[0035] The formula for calculating the channel gain of the jamming link from the jamming drone to the eavesdropper in each time slot is as follows: in, This indicates that the jamming drone is in the time slot. To the eavesdropper's interference link channel gain, Indicates the reference channel gain. This indicates that the jamming drone is in the time slot. Two-dimensional position, Indicates the location of the eavesdropper. This indicates the flight altitude of the interfering drone.

[0036] Step S212, the Individual users in time slots The transmission power is bandwidth is The legal link transmission rate for each user in each time slot is calculated based on Shannon's formula. ; Utilizing the interference generated by jamming drones at the eavesdropper's location, combined with and This yields the eavesdropping rate for each user in each time slot. ;according to and Calculate the security rate for each user in each time slot. .

[0037] The formula for calculating the security rate for each user in each time slot is as follows: in, Indicates the first Individual users in time slots The rate of confidentiality, Indicates the first Individual users in time slots The legal link transmission rate, Indicates the first Individual users in time slots The rate at which it is spied on.

[0038] The task and latency model is constructed as follows: Step S221, the first Total bits of tasks for each user Bits allocated to the local execution portion and the amount of bits offloaded to the computational drone execution part ,satisfy .

[0039] No. Total bits of tasks for each user The calculation formula is: in, This indicates the total number of system task bits. This represents the total number of users.

[0040] Step S222: Calculate the local computing latency, drone-side computing latency, and offloading transmission latency for each user, and determine the total task completion latency for each user.

[0041] Specifically, local computation latency is determined by the user's local computational intensity and the local computing power allocated to the task; UAV-side computation latency is determined by the computational intensity of the UAV and the computing power allocated to the user by the UAV; and offload transmission latency is determined by the offload bit allocation variables for each time slot and the legal link transmission rate of the corresponding time slot. Preferably, the offloading transmission delay is obtained by summing the transmission times of each time slot, and the calculation formula is as follows: in, Indicates the first Offload transmission latency for individual users Indicates the total number of time slots. Indicates the first Individual users in time slots The unload bit allocation variable satisfies , Indicates the first Individual users in time slots The legal link transmission rate.

[0042] Combining the local execution and "uplink wireless transmission + drone computing" paths, the formula for calculating the total task completion latency for each user is as follows: in, Indicates the first Total task completion time for each user Indicates the first Local computing latency for each user Indicates the first Offload transmission latency for individual users Indicates the first The computational latency of the drone terminal for each user.

[0043] Preferably, a smooth maximum function is used for differentiability approximation in the algorithm implementation.

[0044] A secure throughput efficiency model is constructed based on the channel and rate model and the task and delay model, specifically as follows: Step S231: Based on the security rate and offload bit weight of each user in each time slot, construct the effective security rate of each user, and then map the security success probability of each user.

[0045] The formula for calculating the offload bit weight for each user in each time slot is as follows: in, Indicates the first Individual users in time slots The unloaded bit weight, Indicates the first Individual users in time slots The unload bit allocation variable, This represents the amount of bits that are offloaded to the computational drone execution section.

[0046] The formula for calculating the effective security rate for each user is: in, Indicates the first Effective confidentiality rate per user Indicates the total number of time slots. Indicates the first Individual users in time slots The unloaded bit weight, Indicates the first Individual users in time slots The rate of confidentiality.

[0047] The formula for calculating the probability of a user's successful security is as follows: in, Indicates the first The probability of a user's successful security operation. This represents an empirical parameter for adjusting security sensitivity. Indicates the first Effective confidentiality rate for each user.

[0048] Step S232: Based on the effective task completion data volume of all users, the probability of secure success, and the total task completion delay, the confidentiality throughput efficiency of the system is obtained.

[0049] The formula for calculating the system's secure throughput efficiency is: in, This indicates the system's secure throughput efficiency. Indicates the total number of users. Indicates the first The amount of valid completed task data per user , Indicates the first The probability of a user's successful security operation. Indicates the first Total task completion time for each user.

[0050] Step S3: Based on the secure throughput efficiency model, construct a joint optimization problem with the objective of maximizing secure throughput efficiency. The optimization variables include: the transmit power of each user in each time slot. Offload bit allocation variables for each user in each time slot Calculate the two-dimensional positions of the UAV and the jamming UAV in each time slot. , The constraints include: user maximum transmit power, UAV flight speed and start / end position constraints, task bit conservation constraints, computing resource constraints, and security rate constraints.

[0051] The specific constraints are as follows: (1) The transmit power of each user in each time slot shall not exceed the user's maximum transmit power. ; (2) Calculate that the flight speed of the UAV and the interfering UAV does not exceed the maximum horizontal flight speed. The trajectory satisfies the given two-dimensional starting and ending positions. (3) Task bit conservation constraints and upper limit constraints on local and compute drone computing resources (e.g., the sum of the number of bits of a single user's local execution part and the number of bits offloaded to the compute drone execution part shall not be less than the total number of bits of its task). Or the sum of the bits offloaded to the computational drone execution unit in each time slot is not less than the total number of bits offloaded to the computational drone execution unit. ); (4) Basic security rate lower limit constraint (e.g.) (Not lower than a preset threshold or its average value not lower than a certain lower bound).

[0052] Step S4: Solve the joint optimization problem to obtain the optimal or near-optimal transmit power for each user in each time slot. Offload bit allocation variables for each user in each time slot Calculate the two-dimensional positions of the UAV and the jamming UAV in each time slot. , And calculate the final confidential throughput efficiency. Specifically: Step S41: Employ a fractional programming method based on the Dinkelbach transform, introducing Dinkelbach parameters. The joint optimization problem is transformed into an iterative solution to maximize... Fixed in each iteration Solve The maximum value is updated based on the obtained solution. Repeat the iteration until convergence.

[0053] Step S42, in each iteration, solve for the maximum During the process, the decision variables are divided into three variable blocks: task block (L block), power block (P block), and trajectory block (Q block). A block coordinate descent method is used to alternately optimize each variable block while keeping other variables fixed. Optimize task block (L block): Fix the transmit power of each user in each time slot. And calculate the two-dimensional position of the drone Interference with the two-dimensional position of the drone Under these conditions, the subproblem degenerates into convex task allocation (including: the number of bits for each user's local execution portion). The amount of bits that each user unloads to the execution part of the computing drone. Offload bit allocation variables for each user in each time slot ) and computing resource allocation (including: local computing power allocated to tasks by each user) Calculate the computing power allocated to each user for the drone. The problem was solved using the CVX tool. Optimized power blocks (P blocks): In fixed task allocation (including the number of bits for each user's local execution portion). The amount of bits that each user unloads to the execution part of the computing drone. Offload bit allocation variables for each user in each time slot ) and the calculation of the two-dimensional position of the drone Interference with the two-dimensional position of the drone Under these conditions, the confidentiality rate for each user in each time slot We perform a first-order convex approximation, construct a linear lower bound, and transform the power optimization problem into convex programming. Optimized trajectory block (Q block): In fixed task allocation (including the number of bits for each user's local execution portion). The amount of bits that each user unloads to the execution part of the computing drone. Offload bit allocation variables for each user in each time slot ) and the transmit power of each user in each time slot Under the condition of continuous convex approximation, the two-dimensional position of the UAV is calculated by iterative update. Interference with the two-dimensional position of the drone This is transformed into a constrained second-order cone programming problem to be solved.

[0054] Step S43: For the non-convex subproblems generated by the block coordinate descent method, a continuous convex approximation technique is used to perform a local convex approximation, and a convex optimization solver is used to solve the problem. For the non-convex terms in the optimized power block (P block) and optimized trajectory block (Q block), the security rate of each user in each time slot. The fractional and difference structures in the problem are used, and a first-order Taylor expansion is employed to construct a local linear lower bound or a convex surrogate function to approximate the original non-convex subproblem as a convex problem; for the optimization task block (L block), the given rate (including: the legal link transmission rate of each user in each time slot) is... The security rate of each user in each time slot ) and calculating the two-dimensional position of the drone Interference with the two-dimensional position of the drone Under these conditions, the offload bit allocation variables for each user in each time slot can be modeled as a convex optimization or second-order cone programming problem. Each subproblem can be solved using existing convex optimization tools. Through multiple BCD cycles and SCA iterations, the inner-layer problem converges.

[0055] Step S44, iteratively execute steps S41 to S43, as... The algorithm is updated until it exhibits monotonically convergent behavior, yielding the optimal or near-optimal transmit power for each user in each time slot. Offload bit allocation variables for each user in each time slot Calculate the two-dimensional positions of the UAV and the jamming UAV in each time slot. , And calculate the final confidential throughput efficiency.

[0056] See Figure 3 The figure shows a comparison of the STE convergence curves between the secure mobile edge computing confidential throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention and the "fixed trajectory + fixed power", "fixed trajectory only", "fixed power only", and "no interference machine" schemes. (See also...) Figure 4This is a curve comparison of the final STE (Secure Mobile Efficiency) curves of the secure mobile edge computing confidential throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention, and the "fixed trajectory + fixed power", "fixed trajectory only", "fixed power only", and "no interference UAV" schemes under different user number scenarios. Wherein, "OURS" represents the secure mobile edge computing confidential throughput efficiency optimization method for multi-UAV collaboration provided in this embodiment of the invention, "FTP" represents the "fixed power only" scheme, "NJU" represents the "no interference UAV" scheme, "FTP+FT" represents the "fixed trajectory + fixed power" scheme, and "FT" represents the "fixed trajectory only" scheme. Figure 3 , Figure 4 As shown, under the same system parameters and initial trajectory conditions, the solution of the present invention is significantly superior to other existing technical solutions in terms of STE index.

[0057] Those skilled in the art will understand that the method of the present invention can be implemented by a software program on a computing UAV controller, ground server, or edge cloud platform equipped with a processor, memory, and wireless communication module. This program can be stored on a computer-readable storage medium and, when executed by a processor, implements the steps described above.

[0058] It should be noted that the specific values ​​in the above embodiments (such as...) =3、 =10、 =0.2 s、 , , The above description is merely a typical configuration of the present invention. In practical applications, it can be adjusted according to the needs of the scenario, such as expanding to more users, more time slots, multiple eavesdropper scenarios, or introducing energy constraints, no-fly zone constraints, etc. These scenarios can still use the methods proposed in this invention and should all be considered to fall within the protection scope of this invention.

[0059] The above description is merely a specific embodiment of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing the secure mobile edge computing throughput efficiency of multi-UAV collaboration, characterized in that, The method includes: Establish a multi-UAV collaborative secure mobile edge computing system, the system including... One ground user, one computing drone, one jamming drone, and one fixed eavesdropper; Construct a channel and rate model, a task and delay model, and build a secure throughput efficiency model based on the channel and rate model and the task and delay model; Based on the security throughput efficiency model, a joint optimization problem is constructed with the goal of maximizing security throughput efficiency. The optimization variables include: the transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, and the two-dimensional positions of the computational UAV and the jamming UAV in each time slot. The constraints include: the transmit power of each user in each time slot does not exceed the user's maximum transmit power; the flight speed of the computational UAV and the jamming UAV does not exceed the maximum horizontal flight speed, and the trajectory satisfies the given two-dimensional position conditions of the starting point and the ending point; task bit conservation constraints and upper limits of local and computational UAV computing resources; and a lower limit constraint of basic security rate. Solve the joint optimization problem to obtain the optimal or near-optimal transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, calculate the two-dimensional position of the UAV and the jamming UAV in each time slot, and calculate the final security throughput efficiency.

2. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 1, characterized in that, The channel and rate model is constructed as follows: Based on the line-of-sight channel model, the legitimate link channel gain of each user to the computing drone in each time slot, the eavesdropping link channel gain of each user to the eavesdropper, and the interference link channel gain of the jamming drone to the eavesdropper in each time slot are established. Based on Shannon's formula, the legitimate link transmission rate and the eavesdropping rate of each user in each time slot are calculated, and the confidential rate of each user in each time slot is also calculated.

3. The method for optimizing the secure mobile edge computing throughput efficiency of multi-UAV collaboration according to claim 2, characterized in that, The formula for calculating the legitimate link channel gain from each user to the drone in each time slot is as follows: in, Indicates the first Individual users in time slots To calculate the legal link channel gain of the drone, Indicates the reference channel gain. This indicates the calculation of the drone in the time slot. Two-dimensional position, Indicates the first The location of each user This indicates the calculated flight altitude of the drone; The formula for calculating the channel gain of the eavesdropping link from each user to the eavesdropper is as follows: in, Indicates the first Channel gain of the eavesdropping link from the user to the eavesdropper Indicates the first Slow fading coefficient for each user Indicates the location of the eavesdropper; The formula for calculating the channel gain of the jamming link from the jamming drone to the eavesdropper in each time slot is as follows: in, This indicates that the jamming drone is in the time slot. To the eavesdropper's interference link channel gain, This indicates that the jamming drone is in the time slot. Two-dimensional position, This indicates the flight altitude of the interfering drone.

4. The method for optimizing the secure mobile edge computing throughput efficiency of multi-UAV collaboration according to claim 3, characterized in that, The formula for calculating the security rate for each user in each time slot is as follows: in, Indicates the first Individual users in time slots The rate of confidentiality, Indicates the first Individual users in time slots The legal link transmission rate, Indicates the first Individual users in time slots The rate at which it is spied on.

5. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 1, characterized in that, The task and latency model is constructed as follows: The total number of bits for each user's task is divided into the number of bits for the local execution part and the number of bits offloaded to the computational drone execution part; Calculate the local computing latency, drone-side computing latency, and offloading transmission latency for each user to determine the total task completion latency for each user.

6. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 5, characterized in that, The formula for calculating the total task completion latency for each user is: in, Indicates the first Total task completion time for each user Indicates the first Local computing latency for each user Indicates the first Offload transmission latency for individual users Indicates the first The computational latency of the drone terminal for each user; No. Offload transmission latency for individual users The calculation formula is: in, Indicates the total number of time slots. Indicates the first Individual users in time slots The unload bit allocation variable, Indicates the first Individual users in time slots The legal link transmission rate.

7. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 1, characterized in that, A secure throughput efficiency model is constructed based on the channel and rate model and the task and delay model, specifically as follows: Based on the security rate and offload bit weight of each user in each time slot, the effective security rate of each user is constructed, and the security success probability of each user is obtained. The confidentiality throughput efficiency of the system is obtained based on the effective task completion data volume of all users, the probability of secure success, and the total task completion latency.

8. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 7, characterized in that, The formula for calculating the probability of a user's successful security is as follows: in, Indicates the first The probability of a user's successful security operation. This represents an empirical parameter for adjusting security sensitivity. Indicates the first Effective confidentiality rate per user Indicates the total number of time slots. Indicates the first Individual users in time slots The unloaded bit weight, Indicates the first Individual users in time slots The rate of confidentiality, Indicates the first Individual users in time slots The unload bit allocation variable, This represents the amount of bits that are offloaded to the computational drone execution section.

9. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 8, characterized in that, The formula for calculating the system's secure throughput efficiency is: in, This indicates the system's secure throughput efficiency. Indicates the total number of users. Indicates the first The amount of valid completed task data per user Indicates the first Total task completion time for each user.

10. The method for optimizing the secure mobile edge computing confidentiality throughput efficiency of multi-UAV collaboration according to claim 1, characterized in that, The solution to the joint optimization problem is as follows: Step S41: Using a fractional programming method based on the Dinkelbach transform, the joint optimization problem is transformed into an iterative solution of a series of difference form subproblems. Step S42: In each Dinkelbach iteration, the block coordinate descent method is used to divide the optimization variables into three variable blocks: task block, power block, and trajectory block, and perform alternating optimization. Step S43: For the non-convex subproblems generated by the block coordinate descent method, a continuous convex approximation technique is used to perform a local convex approximation, and a convex optimization solver is used to solve the problem. Step S44: Iteratively execute steps S41 to S43 until the entire algorithm exhibits monotonically convergent behavior, obtain the optimal or near-optimal transmit power of each user in each time slot, the offload bit allocation variable of each user in each time slot, calculate the two-dimensional positions of the UAV and jamming UAV in each time slot, and calculate the final security throughput efficiency.

Images