Multi-unmanned aerial vehicle MEC secure communication deployment method supporting dynamic role switching

Abstract

The invention discloses a multi-unmanned aerial vehicle MEC secure communication deployment method supporting dynamic role switching, and relates to the technical field of unmanned aerial vehicle mobile edge computing, and the method comprises the steps: S1, setting an unmanned aerial vehicle role; s2, calculating a data unloading rate; s3, calculating a leakage rate; s4, calculating communication secrecy capacity based on the unloading rate and the leakage rate; calculating edge calculation time delay; s5, calculating an optimized objective function constructed based on the communication secrecy capacity and the edge calculation time delay, and obtaining an objective function value; s6, optimizing all variables by using an algorithm, and repeating the steps S2-S5 by using the optimized variable values to obtain a new target function value; and S7, calculating a difference value between the new target function value and the target function value calculated last time, stopping iteration when a condition is met, and otherwise, repeating the steps S6-S7. According to the method, the unmanned aerial vehicle supporting dynamic role switching is introduced, and the role switching decision and related variables of the unmanned aerial vehicle are optimized by adopting an algorithm, so that the edge calculation time delay is reduced while the communication secrecy capacity is improved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) mobile edge computing technology, and in particular to a multi-UAV MEC secure communication deployment method supporting dynamic role switching. Background Art

[0002] Mobile edge computing is widely used in the field of the Internet of Things by offloading computationally intensive and latency-critical data to the edge of the network close to the terminal device to enhance device performance and save battery resources. However, due to long distances or damage caused by natural disasters, some areas have limited infrastructure coverage. Drones with computing resources bring computing services to areas that cannot be covered by fixed mobile edge computing infrastructure due to their rapid deployment and low cost. Therefore, drones are considered to be mobile edge computing platforms with great development potential and broad application prospects. For example, the Chinese patent with authorization publication number CN114372612B proposes a path planning and task offloading method for drone mobile edge computing scenarios based on deep reinforcement learning. This method uses the complementarity of deep reinforcement learning and convex optimization in complexity and accuracy to minimize drone energy consumption and task completion time, and improve the energy efficiency of performing ground terminal tasks. Z.Yang et al. [Z.Yang, S.Bi and Y.-JA Zhang,"Online Trajectory and Resource Optimizationfor Stochastic UAV-Enabled MEC Systems," in IEEE Transactions on Wireless Communications, vol.21, no.7, pp.5629-5643, July 2022, doi:10.1109 / TWC.2022.3142365.] proposed a resource allocation and trajectory planning method for UAV mobile edge computing platform based on Lyapunov optimization and continuous convex approximation, which minimizes the average weighted total energy consumption of ground users under the constraints of energy consumption and queue stability.

[0003] However, due to the line-of-sight wireless channel between the drone and the ground terminal, the data unloaded at the edge in the mobile edge computing network for drones can be easily intercepted by potential illegal devices. The Chinese patent with the authorization publication number CN114070451B provides a method for secure communication of drone mobile edge computing based on non-orthogonal multiple access, which involves the field of physical layer secure communication of mobile edge computing, and realizes secure communication by broadcasting artificial noise from ground jammers to interfere with illegal drones. R.Li et al. [R.Li, Z.Wei, L.Yang, DWKNg, J.Yuan and J.An, "Resource Allocationfor Secure Multi-UAV Communication Systems With Multi-Eavesdropper," in IEEE Transactions on Communications, vol.68, no.7, pp.4490-4506, July 2020, doi:10.1109 / TCOMM.2020.2983040.] proposed a secure communication scheme based on a new multi-purpose drone. Idle drone base stations can be used as friendly jammers to transmit noise signals to illegal devices. The physical layer confidentiality capability of the communication system is improved by jointly optimizing the resource allocation strategy and drone flight trajectory. A.Gao et al. [A.Gao, Q.Wang, Y.Hu, W.Liang and J.Zhang,"Dynamic Role Switching Scheme With Joint Trajectory andPower Control for Multi-UAV Cooperative Secure Communication,"in IEEE Transactions on Wireless Communications,vol.23,no.2,pp.1260-1275,Feb.2024,doi:10.1109 / TWC.2023.3287849.] proposed a joint optimization method of UAV flight trajectory and transmission power based on role switching to ensure the security of multi-UAV cooperative communication. This method also allows each UAV to switch roles as a data collector or a friendly jammer.

[0004] In addition, the limited battery capacity of drones makes it difficult to provide long-term and stable edge computing services for large-scale IoT devices. The Chinese patent with the authorization publication number CN115134370B discloses a multi-drone-assisted mobile edge computing offloading method, which centrally reduces the energy consumption of multi-drone and multi-user data processing by optimizing the drone flight trajectory, the amount of computing tasks in each time slot, and the partial offloading variables. J. Li et al. [J. Li, C. Yi, J. Chen, K. Zhu and J. Cai,"Joint

[0005] Trajectory Planning, Application Placement, and Energy Renewal for UAV-Assisted MEC: A Triple-Learner-Based Approach," in IEEE Internet of ThingsJournal, vol.10, no.15, pp.

[0006] 13622-13636, 1Aug.1, 2023, doi:10.1109 / JIOT.2023.3262687.] The energy-saving scheduling problem of multi-UAV mobile edge computing is studied, and a novel triple learner-based reinforcement learning method is proposed to maximize the long-term energy efficiency of all UAVs.

[0007] In response to the physical layer secure communication problem of multi-UAV mobile edge computing, the present invention also takes into account the energy problem and computing delay problem of UAVs, aiming to dynamically switch the three roles of UAVs, so as to update its edge application service layout, the transmission power of UAVs and ground equipment, and the flight trajectory of UAVs when the UAVs themselves have sufficient power, so as to improve the system communication confidentiality capacity and reduce the edge computing delay. Specifically, by obtaining the latest variable information, UAVs can flexibly switch their roles in the environment to achieve the goals of secure communication and reduced latency. The present invention iteratively updates the role switching decisions of UAVs and the edge application service layout, the transmission power of UAVs and ground equipment, and the flight trajectory of UAVs, providing real-time and secure mobile edge computing services for multiple users in an environment with potential eavesdroppers. Summary of the invention

[0008] In view of the above problems, the present invention proposes a multi-UAV MEC (Mobile Edge Computing) secure communication deployment method that supports dynamic role switching. By introducing multiple UAVs that support dynamic role switching as communication receivers or friendly jammers, it is considered to coordinate with ground equipment to perform secure edge data unloading when the UAV itself has sufficient power, thereby improving the safety communication efficiency of the UAV. The weighted average of the communication confidentiality capacity and edge computing delay is used as the target to optimize the target value in the worst case, thereby realizing the collaborative optimization of the multi-UAV secure communication deployment method in the mobile edge computing system. This method uses block coordinate descent and continuous convex approximation algorithms to jointly optimize the role switching decision of the UAV and the edge application service layout, the transmission power of the UAV and ground equipment, and the flight trajectory of the UAV, thereby improving the communication confidentiality capacity while reducing the edge computing delay.

[0009] The multi-UAV MEC secure communication deployment method supporting dynamic role switching is based on a multi-UAV system including a plurality of UAVs, a plurality of ground devices, a station and a plurality of eavesdroppers, and includes the following steps:

[0010] S1, setting the roles of the drone; the roles include communication receiver, friendly jammer, site updater and idle drone; each drone holds one role and supports dynamic role switching;

[0011] S2, calculating the data unloading rate of the ground device based on the transmission power of the ground device;

[0012] S3, calculates the leakage rate of the data intercepted by the eavesdropper based on the UAV transmission power, the ground equipment transmission power and the friendly jammer Boolean variable;

[0013] S4, calculating the communication confidentiality capacity based on the data unloading rate, the leakage rate, the edge application service type and the communication receiver Boolean variable; calculating the edge computing delay based on the edge application service type and the communication receiver Boolean variable;

[0014] S5, calculating the optimization objective function based on the communication confidentiality capacity and the edge computing delay, and obtaining the objective function value;

[0015] S6, using block coordinate descent and continuous convex approximation algorithm to jointly optimize all variables, and repeating S2-S5 with the optimized variable values ​​to obtain a new objective function value; the variables include communication receiver Boolean variables, friendly jammer Boolean variables, site updater Boolean variables, edge application service type, ground equipment transmission power, drone transmission power and drone position;

[0016] S7, calculate the difference between the new objective function value and the last calculated objective function value. If the difference is greater than the preset difference, repeat S6-S7; otherwise, stop the iteration.

[0017] Preferably, the dynamic role switching is specifically as follows: when the remaining power of a drone reaches a preset power threshold, or when a drone has a need to update edge application services, it switches to a site updater; the three roles of communication receiver, idle drone and friendly jammer can be switched among each other.

[0018] Preferably, the remaining battery power of the drone i It is expressed as:

[0019]

[0020] in, represents the propulsion and hovering energy consumption; represents the transmission energy consumption; Indicates computing energy consumption; E CHa Indicates the amount of electricity added at the site; E ret represents the maximum flight energy consumption to the station; a ij [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no. b ij [t] A Boolean variable of the friendly jammer of drone i, indicating whether drone i plays the role of a friendly jammer to jam the nearest eavesdropper of ground device j. The value is 1 for yes and 0 for no; c i [t] represents the Boolean variable of the site update machine of drone i, whether drone i plays the role of the site update machine, the value is 1 for yes, and the value is 0 for no; t represents the tth time slot.

[0021] Preferably, the optimization objective function is a weighted average of the information confidentiality capacity and the edge computing delay, expressed as follows:

[0022]

[0023] Among them, α represents the weighted coefficient of communication confidentiality capacity; β represents the weighted coefficient of edge computing delay; T represents the number of time slots in the system time; R sec [t] represents the communication confidentiality capacity; L com [t] represents the edge computing latency; t represents the tth time slot.

[0024] Preferably, the communication confidentiality capacity is specifically expressed as:

[0025]

[0026] Where N is the number of ground devices; M is the number of drones; A is the type of mission data; E is the number of eavesdroppers; ja[t]∈{0,1}, indicates whether the task data type of ground equipment j is a, with a value of 1 for yes and 0 for no; z ia [t]∈{0,1}, indicates whether the edge application service type of drone i is a, with a value of 1 for yes and 0 for no; l je ∈{0,1}, indicating whether the data unloaded by ground device j is intercepted by eavesdropper e. When the data unloaded by ground device j is intercepted by eavesdropper e, l je [t] = 1, otherwise, l je [t] = 0; a ij [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no. R ji [t] represents the unloading rate per Hz from ground equipment j to UAV i; R je [t] represents the leakage rate per Hz from ground device j to eavesdropper e.

[0027] Preferably, the unloading rate per hertz from the ground equipment j to the drone i is specifically expressed as:

[0028]

[0029] Among them, p ji [t] represents the transmission power of UAV i receiving ground device j when ground device j and UAV i are performing data unloading. Its value is equal to the ground device transmission power p of ground device j. j [t]; g ji [t] represents the channel gain; σ 2 represents the additive white Gaussian noise power.

[0030] Preferably, the leakage rate per hertz from the ground device j to the eavesdropper e is expressed as:

[0031]

[0032] Among them, p je [t] represents the transmission power of ground device j received by eavesdropper e when ground device j is performing data unloading. Its value is equal to the ground device transmission power p of ground device j. j [t]; p i′j [t] indicates drone i ′ The eavesdropper emits artificial noise to interfere with the ground equipment j, and the eavesdropper receives the drone i ′ Transmit power, its value is equal to UAV i ′ The UAV transmission power p i′ [t], drone ′represents an idle drone that can be transformed into a friendly jammer; b i′j [t] represents the friendly jammer Boolean variable of drone i, which represents drone i ′ Whether to act as a friendly jammer to interfere with the nearest eavesdropper of ground equipment j, the value is 1 for yes, and 0 for no; g je represents the channel gain from ground device j to eavesdropper e; g i′e [t] indicates drone i ′ The channel gain to the eavesdropper e; σ 2 represents the additive white Gaussian noise power.

[0033] Preferably, the edge computing delay is specifically expressed as:

[0034]

[0035] Where N is the number of ground equipment; M is the number of drones; A is the type of mission data; ja [t]∈{0,1}, indicates whether the task data type of ground equipment j is a, with a value of 1 for yes and 0 for no; z ia [t]∈{0,1}, indicating whether the edge application service type of drone i is a, with a value of 1 for yes and 0 for no; D a Indicates the task data size of task data type a; f i represents the computing power of UAV i; a ij [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The present invention utilizes the dynamic role switching of drones and the coordination between drones to jointly optimize the role switching decision of drones and the layout of edge application services, the transmission power of drones and ground equipment, and the flight trajectory of drones, thereby improving the system's physical layer communication confidentiality capacity while reducing edge computing latency, and the weights of the two can be manually adjusted to achieve different optimization goals. In addition, the present invention also considers the drone power problem in drone mobile edge computing secure communications, which is rarely mentioned in the prior art, and considers coordinating with ground equipment to perform secure edge data unloading when the drone itself has sufficient power, thereby improving the drone's secure communication efficiency, and innovatively adds the charging function to the drone's dynamic role switching method, which has significant advantages and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings;

[0039] Figure 1 A system model diagram of a multi-UAV MEC secure communication deployment method supporting dynamic role switching according to an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of a process for deploying a multi-UAV MEC secure communication method supporting dynamic role switching according to an embodiment of the present invention;

[0041] Figure 3 It is a UAV working sequence diagram of the multi-UAV MEC secure communication deployment method supporting dynamic role switching according to an embodiment of the present invention;

[0042] Figure 4 It is a schematic diagram comparing the optimization of the objective function between the multi-UAV MEC secure communication deployment method supporting dynamic role switching according to an embodiment of the present invention and the common UAV role fixing method;. DETAILED DESCRIPTION

[0043] The present invention is further described below through specific implementation modes.

[0044] This embodiment is based on Figure 1 The multi-UAV system model diagram is shown; the multi-UAV system model includes several UAVs, several ground equipment, stations and several eavesdroppers.

[0045] See also Figure 2 and Figure 3 As shown, a multi-UAV MEC secure communication deployment method supporting dynamic role switching of the present invention is as follows.

[0046] Step 1: Determine the number of drones, ground devices, and eavesdroppers, and set system parameters appropriately. Specifically, place M drones, N ground devices, and E potential eavesdroppers in a 400m×400m suburban area. The total system time T0 is discretized into T time slots of the same duration δ t =10s. The system uses Orthogonal Frequency Division Multiple Access (OFDMA) technology with a bandwidth of B = 1MHz. All communication devices are equipped with a single antenna, and the drone is equipped with only one transceiver and uses half-duplex operation to avoid self-interference. Additive white Gaussian noise (AWGN) power spectral density σ 2=-120dBm, the reference distance is 1m channel gain β0 = -60dB. The roles that drones can play include communication receivers, friendly jammers, site updaters and idle drones; the communication receiver receives the communication tasks of the ground equipment to unload and calculate the edge data; the friendly jammer transmits interference signals to interfere with eavesdroppers; the site updater flies to the site to charge or update the edge application service type, and returns to the original position; when the drone is not used as a communication receiver, friendly jammer or site updater, it is an idle drone, and the idle drone maintains the original position without action; a drone holds one role and supports dynamic role switching. When the remaining power of each role drone reaches the preset power threshold, it switches to the site updater; when each role drone has an edge application service update requirement, it switches to the site updater. For example, idle drones and friendly jammers may go to the site to update the application type in advance, which is convenient for subsequent use as communication receivers; the three roles of communication receiver, idle drone and friendly jammer can switch with each other to maximize the subsequent objective function value.

[0047] For a rotorcraft with a speed of V, the propulsion power consumption is modeled as:

[0048]

[0049] Step 2: Determine the onboard resources of each drone, including memory size, computing power, and battery capacity, as well as the scale of each edge application service and mission data. Specifically, the ground device randomly generates a mission z in each time slot. ja [t], that is, ∑ a∈A z ja [t] = 1, where a∈{1,...,A} is the index of the task data type, indicating different data types, z ja [t] = 1 means that device j requests to offload task data a, otherwise z ja [t] = 0. When the edge application service of the drone can calculate the data, the data type is the same, that is, ∑ j∈N ∑ i∈M ∑ a∈A z ja [t]z ia When [t] = 1, the two will communicate, and different types of mission data / edge application service scales are different. In addition, the memory size, computing power, and battery capacity of the drones in the scene may also be different. This difference gives the system more room for optimization.

[0050] Step 3: Assume that the transmission power of ground device j is p j [t], ground equipment j transmits with power p j [t] Offload edge data and consider the worst case, i.e. p j[t]>0 data will be intercepted by the nearest eavesdropper. j and w e The position is fixed, so the communication link and channel gain between the ground equipment and the eavesdropper remain unchanged. ij [t]∈{0,1} represents the communication receiver Boolean variable of drone i, that is, whether drone i collects data from device j as a communication receiver; a ij [t] = 1 means that UAV i plays the role of communication receiver and collects data from ground device j; a ij [t] = 0 means that UAV i does not serve as a communication receiver.

[0051] Step 4: Assume the transmission power of drone i is p i [t], UAV i transmits with power p i [t] transmits artificial noise to interfere with the eavesdropper. Since the UAV can eliminate the interference signal from the received signal, the artificial noise only interferes with the eavesdropper. ij [t]∈{0,1} represents the friendly jammer Boolean variable of drone i, that is, whether drone i acts as a friendly jammer to jam the eavesdropper closest to device j. If drone i chooses to jam the eavesdropper closest to ground device j, that is, it plays the role of a friendly jammer, then b ij [t] = 1, otherwise b ij [t] = 0. Assume that the transmission power of UAV i is p i When artificial noise [t] is generated, since other UAVs serving as communication receivers have known the artificial noise in advance, the influence of the artificial noise can be eliminated when receiving mission data, and the artificial noise will only interfere with the eavesdropper.

[0052] The unloading rate per Hz from ground equipment j to UAV i in time slot t is:

[0053]

[0054] Among them, p ji [t] represents the transmission power of drone i receiving device j when device j and drone i are performing data unloading. Its value is equal to the transmission power p of device j. j [t],g ji [t] is the channel gain, σ 2 is the additive white Gaussian noise power.

[0055] Similarly, the leakage rate per Hz from ground device j to eavesdropper e is:

[0056]

[0057] Among them, p je[t] represents the transmit power of device j received by eavesdropper e when device j is performing data unloading, and its value is equal to the transmit power p of device j. j [t]; p i′j [t] indicates drone i ′ The eavesdropper emits artificial noise to interfere with device j, and the eavesdropper receives drone i ′ Transmit power, its value is equal to UAV i ′ The transmission power p i′ [t], drone ′ Indicates that in addition to the drones working as communication receivers, other idle drones can be used as friendly jammers. je and g i′e [t] are device j and drone i respectively ′ The channel gain to the eavesdropper e. It is worth noting that since all devices and eavesdroppers are in fixed positions, g je Is a constant value.

[0058] Step 5: Set z j [t] = {z j1 [t],…,z jA [t]} represents the task offloading request of ground equipment j, where a∈{1,…,A} is the index of the task data type, z ja [t] = 1 means that device j requests to offload task data a, otherwise z ja [t] = 0, in each time slot, device j randomly generates a task offloading request, that is, ∑ a∈A z ja [t]=1.

[0059] Assume that the edge application service layout of drone i is z i [t] = {z i1 [t],…,z iA [t]}, where z ia [t] = 1 means that drone i is equipped with an edge application that can calculate task data a, otherwise z ia [t] = 0. When the edge application service type of the drone is the same as the data type uninstalled by the device, and the uninstalled data is selected by drone i to receive, i.e., a ij When [t] = 1, that is, ∑ j∈N ∑ i∈M ∑ a∈A z ja [t]z ia [t]·a ij [t]=1, the two will communicate, requiring δ t Edge data offloading and computing are completed within the system.

[0060] Considering that the drone i may have insufficient battery, ci [t]∈{0,1} represents the Boolean variable of the station update machine, c i [t] = 1 means that UAV i serves as the updater to the site. UAV i flies to the site to update the battery or update z i [t]; otherwise c i [t]=0; In order to ensure that the UAV can t Complete charging and return within 1 second, c i The flight speed is set to 60m / s when [t] = 1. In addition, the drone can only select one role at most.

[0061] The remaining battery power E of the drone i i rem [t] is modeled as:

[0062]

[0063] in, They are propulsion and hovering energy consumption, transmission energy consumption, and computing energy consumption. The two constant values ​​E cha and E ret They are the power added at the site and the maximum flight energy consumption to the site. Since it is necessary to ensure that the power of drone i is always greater than the maximum flight energy consumption E to the site ret , the drone i will be charged when the battery level is reduced to E ret Choose to fly to the station to charge. Since you can only choose one role, i [t]=1 and a ij [t],b ij [t] = 0, the drone battery is In addition, if UAV i needs to process a certain type of mission data and the edge application service placed in it cannot process this type of mission data, it will also choose to fly to the site to update the edge application service layout z i [t]; When the drone does not choose to fly to the station to charge c i [t] = 0, the drone battery is

[0064] Step 6: Establish a multi-UAV mobile edge computing network and determine the locations of ground device j and eavesdropper e as w j =[x j ,y j ] and w e =[x e ,y e ], let the position of drone i be q i [t]{=x i [t],y i [t],H}, where {xi [t],y i [t]} represents the horizontal position of UAV i, and H represents the fixed flight altitude of the UAV in the system. Specifically, q i The horizontal coordinate of [t] and w j All obey uniform distribution, w e Then they obey the normal distribution in their respective areas. The flight altitude of the drones is kept at H = 180m, ensuring that the drones can receive data from all devices in the area and the line-of-sight probability is close to 1. δ t The position of the drone is considered fixed. It flies in a straight line for a maximum of 100 m at a constant speed of 20 m / s every other time slot, then hovers, and acts as a communication receiver or friendly jammer after hovering.

[0065] Step 7: Calculate the communication confidentiality capacity and edge computing latency based on the existing system model. The optimization goal of the system is modeled as the weighted average of the communication confidentiality capacity and edge computing latency. This is a difficult non-convex mixed integer nonlinear programming problem. Based on the block coordinate descent method, the original problem is divided into four sub-problems. All a are optimized alternately according to the initial values ​​of the optimization variables given above. ij [t], b ij [t]、c i [t]、z i [t]、p j [t]、p i [t] and q i [t], and use the optimized variable value as the initial value of the optimized variable for the next iteration.

[0066] The communication security capacity R of the tth time slot sec [t] is:

[0067]

[0068] where z ja [t],z ia [t] = 1 indicates that the task data type and edge application service type are a, otherwise, z ja [t],z ia [t] = 0. With g je Similarly, l je ={0,1} is a constant value, indicating whether the data unloaded by device j is intercepted by eavesdropper e.

[0069] The edge computing delay L of the tth time slot com [t] is:

[0070]

[0071] Among them, D ais the task data size of type a, f t represents the computing power of UAV i. The optimization objectives are as follows:

[0072]

[0073] Among them, {α, β} are the weighted coefficients of communication confidentiality capacity and edge computing latency, respectively.

[0074] Step 8: Repeat steps 3-7 until the difference between the two objective function values ​​is less than the iteration termination condition (obj-primal obj ≤err th ) stops iteration.

[0075] See also Figure 4 As shown, the optimization convergence process of the objective function value of the dynamic role switching method of the present invention and the existing fixed role method in the same scenario is compared. It is obvious that the dynamic role switching method has better performance in terms of communication confidentiality capacity and edge computing latency. This is because the fixed role method assigns a fixed role to each drone, so it cannot ensure that the drone as a communication receiver in the scene is always closest to the ground equipment, and the drone as a friendly jammer is always closest to the eavesdropper. The dynamic role switching method has greater flexibility, and the drone can fly closer to the ground equipment or eavesdropper with a shorter moving distance. Among them, the weight of the communication confidentiality capacity part of the dynamic role switching method is 0.6, and the weight of the edge computing latency part is 0.4.

[0076] This embodiment innovatively proposes a multi-UAV mobile edge computing secure communication deployment method that supports dynamic role switching. It utilizes the dynamic role switching of UAVs and the coordination between UAVs to jointly optimize the UAV role switching decision and edge application service layout, the transmission power of UAVs and ground equipment, and the flight trajectory of UAVs, thereby improving the system's physical layer communication confidentiality capacity while reducing edge computing latency, and the weights of the two can be manually adjusted to achieve different optimization goals. In addition, this embodiment also considers the UAV power problem in the UAV mobile edge computing secure communication, and innovatively adds the charging function to the UAV's dynamic role switching method, which has significant advantages and application prospects.

[0077] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited thereto. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.

Claims

1. A multi-UAV MEC secure communication deployment method supporting dynamic role switching, based on a multi-UAV system including a plurality of UAVs, a plurality of ground equipment, a station and a plurality of eavesdroppers, characterized in that: The steps include: S1, setting the roles of the drone; the roles include communication receiver, friendly jammer, site updater and idle drone; each drone holds one role and supports dynamic role switching; S2, calculating the data unloading rate of the ground device based on the transmission power of the ground device; S3, calculates the leakage rate of the data intercepted by the eavesdropper based on the UAV transmission power, the ground equipment transmission power and the friendly jammer Boolean variable; S4, calculating the communication confidentiality capacity based on the data offloading rate, the leakage rate, the edge application service type and the communication receiver Boolean variable; Calculate edge computing latency based on edge application service type and communication receiver Boolean variables; S5, calculating the optimization objective function based on the communication confidentiality capacity and the edge computing delay, and obtaining the objective function value; S6, using block coordinate descent and continuous convex approximation algorithm to jointly optimize all variables, and repeating S2-S5 with the optimized variable values ​​to obtain a new objective function value; the variables include communication receiver Boolean variables, friendly jammer Boolean variables, site updater Boolean variables, edge application service type, ground equipment transmission power, drone transmission power and drone position; S7, calculating the difference between the new objective function value and the last calculated objective function value, if the difference is greater than the preset difference, repeating S6-S7; Otherwise, stop iterating.

2. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 1 is characterized in that: The dynamic role switching is specifically as follows: when the remaining power of a drone reaches a preset power threshold, or when a drone has an edge application service update requirement, it switches to a site updater; the three roles of communication receiver, idle drone and friendly jammer can be switched between each other.

3. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 2 is characterized in that: The remaining battery power of the drone i It is expressed as: in, represents the propulsion and hovering energy consumption; represents the transmission energy consumption; Indicates computing energy consumption; E cha Indicates the amount of electricity added at the site; E ret represents the maximum flight energy consumption to the station; a ij [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no. b ij [t] A Boolean variable of the friendly jammer of drone i, indicating whether drone i plays the role of a friendly jammer to jam the nearest eavesdropper of ground device j. The value is 1 for yes and 0 for no; c i [t] represents the Boolean variable of the site update machine of drone i, whether drone i plays the role of the site update machine, the value is 1 for yes, and the value is 0 for no; t represents the tth time slot.

4. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 1 is characterized in that: The optimization objective function is the weighted average of the information confidentiality capacity and the edge computing delay, which is expressed as follows: Among them, α represents the weighted coefficient of communication confidentiality capacity; β represents the weighted coefficient of edge computing delay; T represents the number of time slots in the system time; R sec [t] represents the communication confidentiality capacity; L com [t] represents the edge computing latency; t represents the tth time slot.

5. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 4 is characterized in that: The communication confidentiality capacity is specifically expressed as: Where N is the number of ground devices; M is the number of drones; A is the type of mission data; E is the number of eavesdroppers; ja [t]∈{0,1}, indicates whether the task data type of ground equipment j is a, with a value of 1 for yes and 0 for no; z ia [t]∈{0,1}, indicates whether the edge application service type of drone i is a, with a value of 1 for yes and 0 for no; l je ∈{0,1}, indicating whether the data unloaded by ground device j is intercepted by eavesdropper e. When the data unloaded by ground device j is intercepted by eavesdropper e, l je [t] = 1, otherwise, l je [t] = 0; a ij [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no. R ji [t] represents the unloading rate per Hz from ground equipment j to UAV i; R je [t] represents the leakage rate per Hz from ground device j to eavesdropper e.

6. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 5 is characterized in that: The unloading rate per hertz from the ground equipment j to the UAV i is specifically expressed as: Among them, p ji [t] represents the transmission power of UAV i receiving ground device j when ground device j and UAV i are performing data unloading. Its value is equal to the ground device transmission power p of ground device j. j [t]; g ji [t] represents the channel gain; σ 2 represents the additive white Gaussian noise power.

7. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 5 is characterized in that: The leakage rate per Hertz from the ground device j to the eavesdropper e is expressed as: Among them, p je [t] represents the transmission power of ground device j received by eavesdropper e when ground device j is performing data unloading. Its value is equal to the ground device transmission power p of ground device j. j [t]; p i′j [t] represents the UAV i′ emitting artificial noise to interfere with the eavesdropper of ground equipment j. The eavesdropper receives the UAV i′ transmission power, which is equal to the UAV transmission power p of UAV i′. i′ [t], drone i′ represents an idle drone that can be transformed into a friendly jammer; b i′j [t] represents the friendly jammer Boolean variable of drone i, indicating whether drone i′ plays the role of a friendly jammer to jam the nearest eavesdropper of ground device j. The value is 1 for yes and 0 for no. je represents the channel gain from ground device j to eavesdropper e; g i′e [t] represents the channel gain from UAV i′ to eavesdropper e; σ 2 represents the additive white Gaussian noise power.

8. The multi-UAV MEC secure communication deployment method supporting dynamic role switching according to claim 4 is characterized in that: The edge computing delay is specifically expressed as: Where N is the number of ground equipment; M is the number of drones; A is the type of mission data; ja [t]∈{0,1}, indicates whether the task data type of ground equipment j is a, with a value of 1 for yes and 0 for no; z ia [t]∈{0,1}, indicating whether the edge application service type of drone i is a, with a value of 1 for yes and 0 for no; D a Indicates the task data size of task data type a; f i represents the computing power of UAV i; a iJ [t] represents the communication receiver Boolean variable of drone i, indicating whether drone i plays the role of communication receiver and collects data from device j. The value is 1 for yes and 0 for no.

Images