High-security offloading method, system, terminal and medium thereof using dual edge computing

Abstract

The present invention relates to the field of air, space and sea integrated network communication technology, specifically a high-security offloading method, system, terminal and medium thereof using dual edge computing, which jointly optimizes the transmission power of drones, drone trajectories, access satellite receiving power and MMU transmission power, and minimizes the transmission delay of the mobile terminal computing task offloading to the small base station and satellite edge server, thereby improving the overall transmission efficiency of the system. In addition, compared with the solution that simply pursues the minimum system delay, the system security is fully considered, and the risk of MMUs being eavesdropped when transmitting tasks to satellites is greatly reduced by deploying unmanned jammers to send interference information, so that the overall system has the maximum average confidentiality rate.

Description

Technical Field

[0001] The present invention relates to the field of air-space-sea integrated network communication technology, and specifically to a high-security unloading method, system, terminal and medium thereof using dual-edge computing. Background Art

[0002] The rapid development of wireless communication technology has promoted the integration of multi-dimensional wireless communications based on air, space, sea, and land, which can provide full-time and all-regional data transmission services. As marine integrated wireless networks are of great significance in the development of national defense and intelligent marine Internet of Things, this has greatly promoted the development of Space-Ari-Aqua Integrated Network (SAAIN) based on the deep integration of air-space-sea three-dimensional networks, providing ubiquitous, intelligent, collaborative and efficient wireless communications for various marine applications.

[0003] In recent years, the rapid development of the Internet of Things has led to an explosive growth in the daily data volume of SAAIN. As the requirements of marine mobile users (MMUs) for network performance such as network service quality and service request latency continue to increase, resource-constrained mobile device terminals are difficult to meet. In order to meet the above challenges, Mobile Edge Computing (MEC) has been proposed as a promising technology and architecture in SAAIN. The idea of ​​MEC is to sink the computing functions and services of the cloud to the edge of the network with geographical advantages, provide real-time computing assistance nearby, and reduce the task computing latency and energy consumption of mobile terminals. When the MEC network adopts the operation mode of partial computing offloading, part of the computing tasks of MMUs are executed locally, and the other part is first considered to be offloaded to the base station (BS) edge server on the ground for computing and execution. If the capacity of the server exceeds the limit, the legitimate UAV will relay the task information to the satellite edge server in the low earth orbit (LEO) satellite network for computing and execution.

[0004] On the other hand, due to the broadcast nature of wireless communications, the task of offloading from terminal devices to access points through wireless channels is vulnerable to eavesdroppers, resulting in information leakage, especially in drone networks with light-of-sight (LOS) links. Therefore, the physical layer security technology - cooperative interference is adopted. During the satellite offloading process, unmanned jammers (Jammer UAVs) are deployed to send jamming signals to effectively ensure the security of transmission. Summary of the invention

[0005] In order to overcome the defects of the above-mentioned prior art, the purpose of the present invention is to provide a high-security offloading method, system, terminal and medium thereof using dual edge computing, so as to solve the technical problems in the prior art of low average confidentiality rate of drone security offloading tasks and prolonged secure transmission of the process of offloading mobile terminal computing tasks to small base stations and satellite edge servers.

[0006] The present invention is achieved through the following technical solutions:

[0007] A high-security offloading method using dual edge computing includes the following steps:

[0008] S1. Build an air-space-sea integrated network model, which includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites;

[0009] S2. Constructing a local computing delay model and a task offloading delay model according to the air-space-sea integrated network model obtained in step S1, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server;

[0010] S3. The information security problem in the process of unloading tasks from relay drones to satellites is solved by using a mathematical model to model the task unloading delay model obtained in step S2 as a problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2, and decoupling the problem P1 into four sub-problems.

[0011] S4, solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem of problem P1 obtained by iterative step S3, and solve the problem P2 obtained in step S3 by simulated annealing algorithm to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

[0012] Preferably, in step S1, the specific steps of constructing the air-space-sea integrated network model are as follows:

[0013] S101. Construct an air-space-sea integrated network model that includes 1 small base station equipped with a MEC server, M MMUs, L security drones, J unmanned jammers, and 5 satellites. The 5 satellites belong to the same orbit, one of which is an access satellite that directly transmits data with the drone. The remaining 4 satellites are equipped with MEC servers for access satellite scheduling and computing offloading tasks. N MMUs are set to compute offloading tasks to the small base station, and MN MMUs use security drones as relays to compute offloading tasks on satellites. The set of security drones is represented as The set of unmanned jammers is represented as The MMU set that offloads tasks to the small base station is represented as The set of MMUs that offload tasks to the satellite is represented as

[0014] S102. Establish the communication model between MMU and UAV and UAV and access satellite. Since LOS link is dominant in the link between MMU and security UAV, and the channel gain obeys the free space loss model, the channel power gain between MMU and security UAV is Channel power gain between MMU and UAV The formulas are as follows:

[0015]

[0016] Where β0 is the channel power gain when the reference distance d0=1, and γ is the path loss index; The horizontal position of the MMU for unloading missions to the satellite, To ensure the horizontal position of the drone, is the horizontal position of the UAV jammer, H is the fixed flight altitude of the safety UAV and the UAV jammer;

[0017] Channel power gain between the security drone and the access satellite Channel power gain between the UAV and the access satellite They are respectively expressed as follows:

[0018]

[0019] Among them, G l is the transmitting antenna gain of the safety drone, G j is the transmitting antenna gain of the unmanned jammer, G s is the satellite’s receiving antenna gain, c is the speed of light in vacuum, H s is the vertical height of the satellite, and f is the carrier frequency;

[0020] S103, when the mth MMU offloads the computing task to the ground small base station, the data transmission rate of the uplink between the MMU and the small base station is expressed as:

[0021]

[0022] Where B is the transmission bandwidth of all MMUs, P m is the power value of the NOMA signal transmitted by the MMU to the small base station, is the received additive white Gaussian noise, 0≤|h1| 2 ≤|h2| 2 ≤...|h m | 2 ≤...≤|h N |2 They represent the channel gains of MMUs 1 to N respectively;

[0023] In the UAV-assisted uplink transmission system, when the MMU is powered by P n,u,l Transmit FDMA signals to legitimate drones, and the MMU transmits FDMA signals to legitimate drones at a power of P. n,u,j The FDMA signal is transmitted to the UAV, and the data transmission rate R of the uplink between the MMU and the security UAV is n,u,l , the data transmission rate P of the uplink between MMU and unmanned jammer n,u,j Respectively expressed as:

[0024]

[0025] Where B is the transmission bandwidth of all MMUs, is the self-interference of the transmitting antenna and receiving antenna of the unmanned jammer, is the channel power gain between the MMU and the security drone, is the channel power gain between the MMU and the unmanned jammer, is the received additive white Gaussian noise;

[0026] The average confidentiality rate of the system generated by the process of MMUs transmitting tasks to legitimate drones is expressed as:

[0027]

[0028] Where M is the total number of MMUs, N is the number of MMUs to small base stations for computing offloading tasks, MN is the number of MMUs to satellites for computing offloading tasks, and R n,u,l is the data transmission rate of the uplink between the MMU and the security drone, R n,u,j is the data transmission rate of the uplink between the MMU and the unmanned jammer.

[0029] Preferably, in step S2, constructing the local computing delay model and the task offloading delay model specifically includes:

[0030] S201, the computing task size of the mth MMU is D m Representation, definition where d m is the amount of data required by the MMU to execute the task, c m To calculate the task d m The number of CPU cycles required for each bit of data in the MMU data; the entire calculation process of MMU data includes local calculation and partial offloading; the amount of task data for MMU local calculation is (1-x m )d m, the amount of data that MMU offloads to the ground MEC server is x m d m , where x is the offload variable; therefore, the latency of MMU local calculation is expressed as:

[0031]

[0032] in The CPU calculation frequency of the MMU with serial number m, x m The proportion of data volume offloaded from MMU to the ground MEC server;

[0033] S202, the delay of computing tasks to be unloaded to the ground. The delay generated when N MMUs unload tasks to the ground is mainly the delay of MMUs unloading tasks to the ground small base station with MEC server and the delay of computing tasks executed by the small base station.

[0034] S203, calculate the delay of task offloading to the satellite. When the remaining MN MMUs face delay-sensitive tasks, the MMUs choose to offload the tasks to the space satellite for execution. The total delay caused by MMUs in space offloading is composed of the delays caused by the following three stages: the delay of MMUs transmitting tasks to the security drone, the delay of the security drone forwarding tasks to the access satellite, and the delay of the access satellite scheduling the MEC satellite.

[0035] Furthermore, in step S202, the delay process of calculating the task offloading to the ground is as follows:

[0036] First, the delay of offloading the mth MMU task to the ground small base station is defined as:

[0037]

[0038] Among them, R m is the data transmission rate of the uplink between the MMU and the small base station;

[0039] Secondly, define the delay caused by the ground small base station executing the mth MMU task as:

[0040]

[0041] where f s It represents the CPU computing frequency allocated to the mth MMU of the ground small base station;

[0042] The total delay of N MMUs in computing offloading tasks in the ground small base station is expressed as follows:

[0043]

[0044] in, The latency of offloading the mth MMU task to the ground small base station; The delay caused by the execution of the MMU task with sequence number m by the ground small base station.

[0045] Furthermore, in step S203, the process of calculating the time delay of offloading the task to the satellite is as follows:

[0046] First, define the delay of the mth MMU transmitting the task to the lth safety drone as:

[0047]

[0048] in, is the horizontal position of the safety drone; H is the height above the ground where the safety drone is fixed; P n,u,j for; is the channel power gain between the MMU and the unmanned jammer;

[0049] Secondly, define the delay of the lth safety UAV forwarding task to the access satellite as:

[0050]

[0051] Where M is the number of subchannels, which is obtained by evenly dividing the bandwidth resource B; B is the transmission bandwidth of all MMUs; G s P a for; f for; is the received additive white Gaussian noise; P n,j,s G l G j H s for;

[0052] After receiving the task forwarded by the UAV, the access satellite first separates the unloading information from the interference information, and then dispatches a MEC satellite for unloading; the five satellites in the system are located in the same orbit, so the propagation delay between two satellites can be estimated using parameters such as the radius of the earth, the altitude of the satellite, and the angle between the orbital planes; when the latitude is above 66°, the satellite link will be interrupted, so the shortest propagation delay occurs at a latitude of about 66°; according to the cosine theorem, the propagation delay between two satellites in the same orbit is:

[0053]

[0054] Among them, R p is the radius of the Earth, n y is the number of satellites on the orbital plane; H s for;

[0055] Finally, the delay caused by the MEC satellite executing the MMU task with sequence number m is:

[0056]

[0057] Among them, W u is the size of the offload task u (u=1,2,...,U), in bits.

[0058] Preferably, in step S3, the task offloading model is modeled as a problem P1 of maximizing the average confidentiality rate and a problem P2 of minimizing the system delay through a mathematical model, and the problem P1 is decoupled into four sub-problems, specifically including:

[0059] S301, jointly optimize the horizontal positions of the security drone and the jammer, and the power sent by the MMU to the legitimate drone, and model the average confidentiality rate maximization problem of the SAAIN system as problem P1:

[0060] P1:

[0061] stC1:

[0062] C2:

[0063] C3:

[0064] C4:

[0065] C5:

[0066] C6:

[0067] in, They represent the initial positions of the security UAV and the jammer respectively; q LF ,q JF They represent the final positions of the safety UAV and the jammer respectively; D represents the maximum horizontal distance flown by the UAV in each time interval;

[0068] Among them, constraints C1-C4 ensure that the horizontal distance moved by the drone in each interval does not exceed D, and constraints C5-C6 indicate that the transmission power of the drone does not exceed the average and peak values;

[0069] S302, using the block coordinate descent method to decouple the above non-convex problem P1 into four sub-problems, including the power optimization problem of the MMU sending to the legitimate UAV, the power optimization problem of the MMU sending to the unmanned jammer, the trajectory optimization problem of the legitimate UAV, and the trajectory optimization problem of the unmanned jammer;

[0070] S303, based on solving P1, the transmission power of the UAV and the receiving power of the access satellite are jointly optimized, and the total transmission delay minimization problem of the SAAIN system is modeled as problem P2:

[0071] P2:

[0072] stC1:

[0073] C2:

[0074] C3:α1+α2=1

[0075] C4:f s ≤f max ,t m ≤T

[0076] Among them, C1 represents the system's constraint on the average confidentiality rate, C2 represents the constraint on the system's task volume, and C4 represents that the computing power of the MEC server in the system is limited, and the local computing time of the MMU is not allowed to exceed the total delay.

[0077] Preferably, in step S4, the optimal transmission power of the UAV and the optimal receiving power of the satellite accessed are obtained by iterating to solve the transmission power optimization problem of the MMU to the legitimate UAV and the unmanned jammer and the trajectory optimization problem of the legitimate UAV and the unmanned jammer respectively, and solving the P2 problem by simulated annealing algorithm, specifically including:

[0078] S401, when solving the MMU transmission power sub-problem and the drone trajectory optimization sub-problem, give the initial values ​​of each variable, bring them into the solution of the power approximate solution of the signal sent by the MMU to the legal drone in the i-th iteration, then substitute the power approximate solution of the signal sent by the MMU to the legal drone and other initial values ​​into the power optimization sub-problem of the signal sent by the MMU to the unmanned jammer in the i-th iteration, solve the local approximate solution of the power of the signal sent by the MMU to the unmanned jammer in the i-th iteration, then substitute the power value of the signal sent by the MMU to the drone and other initial values ​​into the legal drone trajectory optimization sub-problem, solve the i-th iteration local approximate solution of the legal drone trajectory, then substitute the i-th iteration drone transmission power solution and the legal drone trajectory solution into the unmanned jammer trajectory sub-problem to find the local approximate solution, and continuously increase the number of iterations until the difference between the P1 objective function values ​​of the two iterations is within a certain range, then the maximum average confidentiality rate can be obtained, and the process ends;

[0079] S402. Based on the similarity between the solution process of the optimization problem P2 and metal annealing, the simulated annealing algorithm is used to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite. Given the initial temperature T0, annealing rate r, termination temperature T and the values ​​of each optimization variable in the initial state, the transmission power of the legal UAV under the i+1th iteration is solved. If the difference between the i+1th and the i-th iteration is less than 0, the new solution under the i+1th iteration is accepted. At the same time, the transmission power of the unmanned jammer under the i+1th iteration is continued to be solved. If the difference between the i+1th and the i-th iteration is less than 0, the new solution under the i+1th iteration is accepted. The new solution under the i+1th iteration is obtained, and the transmission power of the access satellite under the i+1th iteration is continued to be solved. If the difference between the i+1th and i-th times is less than 0, the new solution under the i+1th iteration is accepted; if the difference between the transmission power of the legal drones under the i+1th and i-th times is not less than 0, first determine whether the relationship between the difference satisfies the requirement of being greater than the random number. If so, the new solution under the i+1th iteration is accepted, otherwise, the new solution under the i+1th iteration is not accepted; after the judgment of the i+1th iteration is completed, the temperature value is attenuated until the termination temperature is reached, the minimum transmission delay of the system can be obtained, and the algorithm ends.

[0080] A high-security offloading system using dual edge computing, comprising

[0081] The first model creation module is used to build an air-space-sea integrated network model, where the air-space-sea integrated network model includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites;

[0082] The second model creation module is used to construct a local computing delay model and a task offloading delay model according to the obtained air-space-sea integrated network model, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server;

[0083] The first data processing module is used to solve the information security problem in the process of unloading tasks from the relay drone to the satellite. The obtained task unloading delay model is modeled into the problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2 through a mathematical model, and the problem P1 is decoupled into four sub-problems.

[0084] The second data processing module is used to solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem through iteration of problem P1, and solve problem P2 obtained in step S3 through simulated annealing algorithm, so as to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

[0085] A mobile terminal comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of a high-security unloading method utilizing dual-edge computing as described above are implemented.

[0086] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a high-security unloading method using dual-edge computing as described above.

[0087] Compared with the prior art, the present invention has the following beneficial technical effects:

[0088] The present invention provides a high-security offloading method using dual-edge computing, which jointly optimizes the UAV transmission power, UAV trajectory, access satellite receiving power and MMU transmission power, and minimizes the transmission delay of the mobile terminal computing task offloading to the small base station and satellite edge server, thereby improving the overall transmission efficiency of the system. In addition, compared with the solution that simply pursues the minimum system delay, the system security is fully considered, and the risk of MMUs being eavesdropped when transmitting tasks to satellites is greatly reduced by deploying unmanned jammers to send interference information, so that the overall system has the maximum average confidentiality rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 It is a flow chart of a high-security offloading method using dual edge computing in the present invention;

[0090] Figure 2 It is a system model diagram of the high-security uninstallation solution in the solution of the present invention.

[0091] Figure 3 It is a flow chart of the simulated annealing algorithm of the present invention.

[0092] Figure 4 It is a comparison chart of system delay between the solution of the present invention and the baseline solution in terms of satellite receiving power.

[0093] Figure 5 It is a comparison chart of system delays between the solution of the present invention and the baseline solution in terms of average confidentiality rate.

[0094] Figure 6 It is the system delay diagram of the solution of the present invention and the baseline solution in terms of the legal UAV transmission power. DETAILED DESCRIPTION

[0095] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0096] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0097] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0098] The purpose of the present invention is to provide a high-security offloading method, system, terminal and medium thereof using dual edge computing to solve the technical problems in the prior art of low average confidentiality rate of drone security offloading tasks and prolonged secure transmission of the process of offloading mobile terminal computing tasks to small base stations and satellite edge servers.

[0099] This high-security offloading method using dual edge computing, such as Figure 1 As shown, the following steps are included:

[0100] S1. Build an air-space-sea integrated network model, which includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites;

[0101] Among them, the specific steps of building an integrated air, space and sea network model are as follows:

[0102] S101. Construct an air-space-sea integrated network model that includes 1 small base station equipped with a MEC server, M MMUs, L security drones, J unmanned jammers, and 5 satellites. The 5 satellites belong to the same orbit, one of which is an access satellite that directly transmits data with the drone. The remaining 4 satellites are equipped with MEC servers for access satellite scheduling and computing offloading tasks. N MMUs are set to compute offloading tasks to the small base station, and MN MMUs use security drones as relays to compute offloading tasks on satellites. The set of security drones is represented as The set of unmanned jammers is represented as The MMU set that offloads tasks to the small base station is represented as The set of MMUs that offload tasks to the satellite is represented as

[0103] S102. Establish the communication model between MMU and UAV and UAV and access satellite. Since LOS link is dominant in the link between MMU and security UAV, and the channel gain obeys the free space loss model, the channel power gain between MMU and security UAV is Channel power gain between MMU and UAV The formulas are as follows:

[0104]

[0105] Where β0 is the channel power gain when the reference distance d0=1, and γ is the path loss index; The horizontal position of the MMU for unloading missions to the satellite, To ensure the horizontal position of the drone, is the horizontal position of the UAV jammer, H is the fixed flight altitude of the safety UAV and the UAV jammer;

[0106] Channel power gain between the security drone and the access satellite Channel power gain between the UAV and the access satellite They are respectively expressed as follows:

[0107]

[0108] Among them, G l is the transmitting antenna gain of the safety drone, G j is the transmitting antenna gain of the unmanned jammer, G s is the satellite’s receiving antenna gain, c is the speed of light in vacuum, H s is the vertical height of the satellite, and f is the carrier frequency;

[0109] S103, when the mth MMU offloads the computing task to the ground small base station, the data transmission rate of the uplink between the MMU and the small base station is expressed as:

[0110]

[0111] Where B is the transmission bandwidth of all MMUs, P m is the power value of the NOMA signal transmitted by the MMU to the small base station, is the received additive white Gaussian noise, 0≤|h1| 2 ≤|h2| 2 ≤...|h m | 2 ≤...≤|h N | 2 They represent the channel gains of MMUs 1 to N respectively;

[0112] In the UAV-assisted uplink transmission system, when the MMU is powered by P n,u,l Transmit FDMA signals to legitimate drones, and the MMU transmits FDMA signals to legitimate drones at a power of P. n,u,j The FDMA signal is transmitted to the UAV, and the data transmission rate R of the uplink between the MMU and the security UAV is n,u,l , the data transmission rate P of the uplink between MMU and unmanned jammer n,u,j Respectively expressed as:

[0113]

[0114] Where B is the transmission bandwidth of all MMUs, is the self-interference of the transmitting antenna and receiving antenna of the unmanned jammer, is the channel power gain between the MMU and the security drone, is the channel power gain between the MMU and the unmanned jammer, is the received additive white Gaussian noise;

[0115] The average confidentiality rate of the system generated by the process of MMUs transmitting tasks to legitimate drones is expressed as:

[0116]

[0117] Where M is the total number of MMUs, N is the number of MMUs to small base stations for computing offloading tasks, MN is the number of MMUs to satellites for computing offloading tasks, and R n,u,l is the data transmission rate of the uplink between the MMU and the security drone, R n,u,j is the data transmission rate of the uplink between the MMU and the unmanned jammer.

[0118] S2. Constructing a local computing delay model and a task offloading delay model according to the air-space-sea integrated network model obtained in step S1, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server;

[0119] Specifically, building a local computing latency model and a task offloading latency model includes:

[0120] S201, the computing task size of the mth MMU is D m Representation, definition where d m is the amount of data required by the MMU to execute the task, c m To calculate the task d m The number of CPU cycles required for each bit of data in the MMU data; the entire calculation process of MMU data includes local calculation and partial offloading; the amount of task data for MMU local calculation is (1-x m )d m , the amount of data that MMU offloads to the ground MEC server is x m d m , where x is the offload variable; therefore, the latency of MMU local calculation is expressed as:

[0121]

[0122] in The CPU calculation frequency of the MMU with serial number m, x m The proportion of data volume offloaded from MMU to the ground MEC server;

[0123] S202, the delay of computing tasks to be unloaded to the ground. The delay generated when N MMUs unload tasks to the ground is mainly the delay of MMUs unloading tasks to the ground small base station with MEC server and the delay of computing tasks executed by the small base station.

[0124] Among them, the delay process of unloading the computing task to the ground is as follows:

[0125] First, the delay of offloading the mth MMU task to the ground small base station is defined as:

[0126]

[0127] Among them, R m is the data transmission rate of the uplink between the MMU and the small base station;

[0128] Secondly, define the delay caused by the ground small base station executing the mth MMU task as:

[0129]

[0130] where f s It represents the CPU computing frequency allocated to the mth MMU of the ground small base station;

[0131] The total delay of N MMUs in computing offloading tasks in the ground small base station is expressed as follows:

[0132]

[0133] in, The latency of offloading the mth MMU task to the ground small base station; The delay caused by the execution of the MMU task with sequence number m by the ground small base station.

[0134] S203, calculate the delay of task offloading to the satellite. When the remaining MN MMUs face delay-sensitive tasks, the MMUs choose to offload the tasks to the space satellite for execution. The total delay caused by MMUs in space offloading is composed of the delays caused by the following three stages: the delay of MMUs transmitting tasks to the security drone, the delay of the security drone forwarding tasks to the access satellite, and the delay of the access satellite scheduling the MEC satellite.

[0135] The process of calculating the delay of offloading the task to the satellite is as follows:

[0136] First, define the delay of the mth MMU transmitting the task to the lth safety drone as:

[0137]

[0138] in, is the horizontal position of the safety drone; H is the height above the ground where the safety drone is fixed; P n,u,j for; is the channel power gain between the MMU and the unmanned jammer;

[0139] Secondly, define the delay of the lth safety UAV forwarding task to the access satellite as:

[0140]

[0141] Where M is the number of subchannels, which is obtained by evenly dividing the bandwidth resource B; B is the transmission bandwidth of all MMUs; G s P a for; f for; is the received additive white Gaussian noise; P n,j,s G l G j H s for;

[0142] After receiving the task forwarded by the UAV, the access satellite first separates the unloading information from the interference information, and then dispatches a MEC satellite for unloading; the five satellites in the system are located in the same orbit, so the propagation delay between two satellites can be estimated using parameters such as the radius of the earth, the altitude of the satellite, and the angle between the orbital planes; when the latitude is above 66°, the satellite link will be interrupted, so the shortest propagation delay occurs at a latitude of about 66°; according to the cosine theorem, the propagation delay between two satellites in the same orbit is:

[0143]

[0144] Among them, R p is the radius of the Earth, n y is the number of satellites on the orbital plane; H s for;

[0145] Finally, the delay caused by the MEC satellite executing the MMU task with sequence number m is:

[0146]

[0147] Among them, W u is the size of the offload task u (u=1,2,...,U), in bits.

[0148] S3. The information security problem in the process of unloading tasks from relay drones to satellites is solved by using a mathematical model to model the task unloading delay model obtained in step S2 as a problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2, and decoupling the problem P1 into four sub-problems.

[0149] In step S3, the above task offloading model is modeled as a problem P1 of maximizing the average confidentiality rate and a problem P2 of minimizing the system delay through a mathematical model, and the problem P1 is decoupled into four sub-problems, including:

[0150] S301, jointly optimize the horizontal positions of the security drone and the jammer, and the power sent by the MMU to the legitimate drone, and model the average confidentiality rate maximization problem of the SAAIN system as problem P1:

[0151] P1:

[0152] stC1:

[0153] C2:

[0154] C3:

[0155] C4:

[0156] C5:

[0157] C6:

[0158] in, They represent the initial positions of the security UAV and the jammer respectively; q LF ,q JF They represent the final positions of the safety UAV and the jammer respectively; D represents the maximum horizontal distance flown by the UAV in each time interval;

[0159] Among them, constraints C1-C4 ensure that the horizontal distance moved by the drone in each interval does not exceed D, and constraints C5-C6 indicate that the transmission power of the drone does not exceed the average and peak values;

[0160] S302, using the block coordinate descent method to decouple the above non-convex problem P1 into four sub-problems, including the power optimization problem of the MMU sending to the legitimate UAV, the power optimization problem of the MMU sending to the unmanned jammer, the trajectory optimization problem of the legitimate UAV, and the trajectory optimization problem of the unmanned jammer;

[0161] Among them, the non-convex problem P1 is decoupled into four sub-optimization problems, as follows:

[0162] a) Only the constraints related to the power sent by the MMU to the legal UAV are retained. The power sub-optimization problem of the MMU sending power to the legal UAV is expressed as follows:

[0163] P1.1:

[0164] stC1:

[0165] Similarly, the non-convex sub-optimization problem of the power sent by the MMU to the unmanned jammer is expressed as:

[0166] P1.2:

[0167] stC1:

[0168] in,

[0169] b) Only the constraints related to the legal UAV trajectory are retained. The trajectory optimization of the legal UAV is non-convex.

[0170] The sub-problem is expressed as:

[0171] P1.3:

[0172] stC1:

[0173] C2:

[0174] C3:t[n]≥0,s[n]≥0.

[0175] in,

[0176]

[0177]

[0178] Similarly, the trajectory optimization non-convex subproblem of the unmanned jammer is expressed as:

[0179] P1.4:

[0180] stC1:

[0181] C2:

[0182] C3:t[n]≥0,s[n]≥0.

[0183] S303, based on solving P1, the transmission power of the UAV and the receiving power of the access satellite are jointly optimized, and the total transmission delay minimization problem of the SAAIN system is modeled as problem P2:

[0184] P2:

[0185] stC1:

[0186] C2:

[0187] C3:α1+α2=1

[0188] C4:f s ≤f max ,t m ≤T

[0189] Among them, C1 represents the system's constraint on the average confidentiality rate, C2 represents the constraint on the system's task volume, and C4 represents that the computing power of the MEC server in the system is limited, and the local computing time of the MMU is not allowed to exceed the total delay.

[0190] S4, solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem of problem P1 obtained by iterative step S3, and solve the problem P2 obtained in step S3 by simulated annealing algorithm to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

[0191] Among them, the optimal transmission power of the UAV and the optimal receiving power of the satellite are obtained by iteratively solving the MMU transmission power optimization problem to the legitimate UAV and the UAV jammer, and the trajectory optimization problem of the legitimate UAV and the UAV jammer. The optimal solution is approximate, and the P2 problem is solved by the simulated annealing algorithm. Specifically, the following are obtained:

[0192] S401, when solving the MMU transmission power sub-problem and the drone trajectory optimization sub-problem, give the initial values ​​of each variable, bring them into the solution of the power approximate solution of the signal sent by the MMU to the legal drone in the i-th iteration, then substitute the power approximate solution of the signal sent by the MMU to the legal drone and other initial values ​​into the power optimization sub-problem of the signal sent by the MMU to the unmanned jammer in the i-th iteration, solve the local approximate solution of the power of the signal sent by the MMU to the unmanned jammer in the i-th iteration, then substitute the power value of the signal sent by the MMU to the drone and other initial values ​​into the legal drone trajectory optimization sub-problem, solve the i-th iteration local approximate solution of the legal drone trajectory, then substitute the i-th iteration drone transmission power solution and the legal drone trajectory solution into the unmanned jammer trajectory sub-problem to find the local approximate solution, and continuously increase the number of iterations until the difference between the P1 objective function values ​​of the two iterations is within a certain range, then the maximum average confidentiality rate can be obtained, and the process ends;

[0193] S402. Based on the similarity between the solution process of the optimization problem P2 and metal annealing, the simulated annealing algorithm is used to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite. Given the initial temperature T0, annealing rate r, termination temperature T and the values ​​of each optimization variable in the initial state, the transmission power of the legal UAV under the i+1th iteration is solved. If the difference between the i+1th and the ith iteration is less than 0, the new solution under the i+1th iteration is accepted, and the unmanned jammer under the i+1th iteration is continued to be solved. Transmit power, if the difference between the i+1th and i-th iterations is less than 0, accept the new solution under the i+1th iteration, and continue to solve the transmit power of the access satellite under the i+1th iteration. If the difference between the i+1th and i-th transmission power of the legal drones is not less than 0, first determine whether the relationship between the difference satisfies the requirement of being greater than the random number. If so, accept the new solution under the i+1th iteration. Otherwise, do not accept the new solution under the i+1th iteration. After the judgment of the i+1th iteration, attenuate the temperature value until the termination temperature is reached, and the minimum transmission delay of the system can be obtained, and the algorithm ends.

[0194] Example

[0195] In the embodiment, each MMU terminal has a certain computing power to meet the simple needs of local execution. However, due to the local computing power and latency requirements, the MMU will offload some tasks to ground small base stations and aerial satellites used as relays by legal drones for processing. Small base stations and satellites equipped with MEC servers have strong computing power and can meet the low latency requirements of terminals.

[0196] The high-security offloading solution using dual-edge computing and physical layer security technology under the SAAIN system includes the following specific steps:

[0197] The first step is to build an air-space-sea integrated network model that includes 1 small base station equipped with a MEC server, M MMUs, L security drones, J unmanned jammers, and 5 satellites;

[0198] Step (1.1) builds an air-space-sea integrated network model that includes 1 small base station equipped with a MEC server, M MMUs, L security drones, J unmanned jammers, and 5 satellites, as shown in Figure 2 shown.

[0199] The five satellites are in the same orbit, one of which is used as an access satellite to directly transmit tasks to the drone, and the other four satellites are equipped with MEC servers for the access satellite to schedule computing offload tasks. Assume that there are N MMUs to offload computing tasks to the small base station, which are distributed in W cells according to their locations; MN MMUs use security drones as relays to offload computing tasks to the satellite. The set of security drones is represented as The set of unmanned jammers is represented as The MMU set that offloads tasks to the small base station is represented as The set of MMUs that offload tasks to the satellite is represented as

[0200] Step (1.2), the UAVs are fixed at a height H above the ground, and the MMUs of the satellite unloading mission are horizontally positioned. The horizontal position of the safety drone is indicated by The horizontal position of the unmanned jammer is represented by Therefore, the Euclidean distance d between the MMU with serial number m and the safety drone with serial number l is m,l use The Euclidean distance between the MMU with serial number m and the unmanned jammer with serial number j is expressed as express.

[0201] In step (1.3), since the LOS link is dominant in the link between the MMU and the UAV, the channel gain obeys the free space loss model. The channel power gains between the MMU and the security UAV and the MMU and the UAV jammer are expressed as follows:

[0202]

[0203] Wherein, β0 is the channel power gain when the reference distance d0=1, and γ is the path loss index.

[0204] Compared with the vertical distance, the horizontal distance between the UAV and the access satellite is negligible. Therefore, the channel power gain between the security UAV and the access satellite and the UAV and the access satellite is expressed as follows:

[0205]

[0206] Among them, G l is the transmitting antenna gain of the safety drone, G j is the transmitting antenna gain of the unmanned jammer, G s is the satellite’s receiving antenna gain, c is the speed of light in vacuum, H s is the vertical height of the satellite, and f is the carrier frequency.

[0207] Step (1.4), according to Shannon's formula, when the MMU with sequence number m offloads the computing task to the small base station on the ground, the transmission rate between the MMU and the small base station is expressed as:

[0208]

[0209] Where B is the transmission bandwidth of all MMUs, P m is the power value of the NOMA signal transmitted by the MMU to the small base station, is the received additive white Gaussian noise, 0≤|h1| 2 ≤|h2| 2 ≤...|h m | 2 ≤...≤|h N | 2 They represent the channel gains of MMUs numbered from 1 to N respectively.

[0210] In the UAV-assisted uplink transmission system, when the MMU is powered by P n,u,l Transmit FDMA signals to legitimate drones, and the MMU transmits FDMA signals to legitimate drones at a power of P. n,u,j When transmitting FDMA signals to the UAV jammer, the data transmission rates of the uplinks between the MMU and the legitimate UAV, and between the MMU and the UAV jammer are expressed as:

[0211]

[0212] in, It is the self-interference of the transmitting antenna and receiving antenna of the unmanned jammer.

[0213] Therefore, the system average confidentiality rate generated by the process of MMUs transmitting tasks to legitimate drones can be expressed as:

[0214]

[0215] where [x] + =max(x,0).

[0216] The second step is to build a local computing delay model and a task offloading delay model based on the air-space-sea integrated network model. The specific steps include:

[0217] Step (2.1), the computational task size of the MMU with sequence number m is represented by D m Representation, definition where d m is the amount of data required by the MMU to execute the task, c m To calculate the task d mThe number of CPU cycles required for each bit of data in the MMU. The entire calculation process of MMU data can be divided into two parts: local calculation and partial offloading. Therefore, the amount of task data for MMU local calculation is (1-x m )d m , the amount of data that MMU offloads to the ground MEC server is x m d m , where x is the offload variable. Based on the above analysis, the latency of MMU local calculation can be expressed as,

[0218]

[0219] in Calculate the CPU frequency for the MMU with index m.

[0220] Step (2.3), the task offloading model proposed in this scheme involves two scenarios: task offloading to ground small base station computing and task offloading to satellite computing.

[0221] In step (2.3), the delay generated when N MMUs offload tasks to the ground is mainly the delay of MMUs offloading tasks to the ground small base stations (equipped with MEC servers) and the delay generated by the calculation of small base stations executing tasks.

[0222] First, according to formula (3), the delay of unloading the MMU task with sequence number m to the ground small base station is defined as:

[0223]

[0224] Secondly, define the delay caused by the execution of the MMU task with sequence number m by the ground small base station as:

[0225]

[0226] where f s Indicates the CPU computing frequency of the MMU with sequence number m assigned to the ground small base station.

[0227] Finally, according to equations (6) and (7), the total delay of N MMUs in calculating the offloading task in the ground small base station is expressed as follows:

[0228]

[0229] Step (2.4), when the remaining MN MMUs face delay-sensitive tasks, the MMUs choose to offload the tasks to the space satellite for execution. The total delay caused by MMUs unloading in space is composed of the delays caused by the following three stages: the delay of MMUs transmitting tasks to the security drone, the delay of the security drone forwarding tasks to the access satellite, and the delay of the access satellite scheduling the MEC satellite. The present invention deploys an unmanned jammer as a friendly jammer, sending a known jamming signal to protect the security drone to safely offload tasks. Assuming that there is a free optical link in the transmission process from the drone to the access satellite, it can not only provide a better transmission rate, but also ensure more reliable and secure transmission.

[0230] First, define the delay of the MMU with sequence number m and sequence number l transmitting tasks to the safety drone as:

[0231]

[0232] Secondly, define the delay of the security drone with sequence number l forwarding the task to the access satellite as:

[0233]

[0234] Where M is the number of sub-channels, which is obtained by evenly dividing the bandwidth resource B.

[0235] After the access satellite receives the task forwarded from the drone, it first separates the unloading information from the interference information, and then dispatches a MEC satellite for unloading. The present invention assumes that the five satellites in the system are in the same orbit, so the propagation delay between two satellites can be estimated using parameters such as the radius of the earth, the altitude of the satellite, and the angle between the orbital planes. When the latitude is above 66°, the satellite link will be interrupted, so the shortest propagation delay occurs at a latitude of about 66°. According to the law of cosines, the propagation delay between two satellites in the same orbit is:

[0236]

[0237] Among them, R p is the radius of the Earth, n y is the number of satellites in the orbital plane.

[0238] Finally, the delay caused by the MEC satellite executing the MMU task with sequence number m is:

[0239]

[0240] Where W u is the size of the offload task u (u=1,2,...,U), in bits.

[0241] The third step is to model the above task offloading model into two formulated optimization problems P1 and P2 through mathematical models, and decouple the non-convex problem P1 into four sub-problems, including:

[0242] In step (3.1), the horizontal positions of the security drone and the jammer are jointly optimized, as well as the power transmitted by the MMU to the legitimate drone. The average confidentiality rate maximization problem of the SAAIN system is modeled as a non-convex optimization problem P1:

[0243]

[0244] in q LF ,q JF denote the initial and final positions of the safety UAV and the jammer, respectively. D denotes the maximum horizontal distance that the UAV flies in each time interval. The operation [·] is ignored. + .

[0245] In the formula, constraints C1-C4 ensure that the horizontal distance moved by the drone in each interval does not exceed D, and constraints C5-C6 indicate that the transmission power of the drone does not exceed the average and peak values.

[0246] Step (3.2) decouples the non-convex problem P1 into four sub-optimization problems, as follows:

[0247] a) Only the constraints related to the power sent by the MMU to the legal UAV are retained. The power sub-optimization problem of the MMU sending power to the legal UAV is expressed as follows:

[0248]

[0249] in,

[0250] The power sub-optimization problem of the MMU sending to the legal UAV is a convex optimization problem. The optimal power value sent by the MMU to the legal UAV can be obtained by directly taking the derivative of the optimization variable as follows:

[0251]

[0252] in λ is not only a non-negative parameter for equal-division search, but also ensures that Established.

[0253] Since the objective function is a convex function and the constraints are convex sets, the CVX tool can be used to solve it through the standard convex optimization algorithm to obtain the optimal power sent by the MMU to the legal drone under the current number of iterations.

[0254] Similarly, the non-convex sub-optimization problem of the power sent by the MMU to the unmanned jammer is expressed as:

[0255]

[0256] in,

[0257] The non-convex optimization problem of the power transmitted by the MMU to the unmanned jammer is treated as an approximate convex optimization problem, and the first-order Taylor expansion is used to process it. Assume is the nth iteration value, and the first lower bound after Taylor expansion can be expressed as:

[0258]

[0259] in

[0260] Therefore, the problem can be transformed into solving the following convex problem:

[0261]

[0262] Since this problem is a convex problem, it can also be solved using the CVX tool through standard convex optimization techniques to obtain the local optimal power sent by the MMU to the unmanned jammer under the current number of iterations.

[0263] b) Only the constraints related to the legal UAV trajectory are retained. The trajectory optimization of the legal UAV is non-convex.

[0264] The sub-problem is expressed as:

[0265]

[0266] in,

[0267]

[0268] Since the trajectory optimization sub-problem of the legal drone is a non-convex problem, the continuous convex approximation method is still used. is the local optimal value obtained after the nth iteration. Therefore, the following inequality exists:

[0269]

[0270]

[0271] Therefore, the above non-convex problem can be transformed into a convex problem, expressed as follows:

[0272]

[0273] Similarly, the trajectory optimization subproblem of the unmanned jammer is expressed as:

[0274] P1.4:

[0275] stC1:

[0276] C2:

[0277] C3:t[n]≥0,s[n]≥0. (24)

[0278] Since the problem has been transformed into a convex problem, the CVX tool can also be used.

[0279] Step (3.3), based on P1, jointly optimize the transmission power of the drone and the receiving power of the access satellite, and model the total transmission delay minimization problem of the SAAIN system as optimization problem P2:

[0280]

[0281] In the formula, C1 represents the system's constraint on the average confidentiality rate, C2 represents the constraint on the system's task volume, and C4 represents that the computing power of the MEC server in the system is limited, and the local computing time of the MMU is not allowed to exceed the total delay.

[0282] P2 is solved based on P1. First, replace the optimization variables that have been determined for the optimization problem P2. The problem is expressed as follows:

[0283] P2.1:

[0284] stC1:

[0285] C2:

[0286] C3:f s ≤f max ,t m ≤T (26)

[0288] in,

[0289]

[0290] Based on the similarity between the solution process of optimization problem P1 and near-ripe annealing, simulated annealing algorithm is used here to solve P1. The optimization target is regarded as energy function, the solution space is regarded as state space, and the optimization problem is solved by simulating the thermal motion of particles through random perturbations.

[0291] The fourth step is to solve the MMU transmit power sub-problem, the drone trajectory optimization sub-problem approximate optimal solution through iteration, and solve the P2 problem through simulated annealing algorithm to obtain the optimal transmit power of the drone and the optimal receiving power of the access satellite. Specifically, it includes:

[0292] Step (4.1), the process of using the CVX tool to find approximate solutions to the MMU transmission power sub-problem and the UAV trajectory optimization sub-problem is as follows: first, give the initial values ​​of each variable, and bring them into the solution to the approximate solution of the power of the MMU sending signals to the legal UAV in the i-th iteration, then substitute the approximate solution of the power of the MMU sending signals to the legal UAV and other initial values ​​into the power optimization sub-problem of the MMU sending signals to the unmanned jammer, and solve the local approximate solution of the power of the MMU sending signals to the unmanned jammer in the i-th iteration, then substitute the power value of the MMU sending signals to the UAV and other initial values ​​in the i-th iteration into the legal UAV trajectory optimization sub-problem, and solve the i-th iteration local approximate solution of the legal UAV trajectory, then substitute the i-th iteration UAV transmission power solution and the legal UAV trajectory solution into the unmanned jammer trajectory sub-problem to find the local approximate solution, and continuously increase the number of iterations until the difference between the P1 objective function values ​​of the two iterations is within a certain range, then the maximum average confidentiality rate, that is, the optimal MMU transmission power and the optimal UAV trajectory can be obtained.

[0293] Step (4.2), based on the similarity between the solution process of the optimization problem P2 and metal annealing, it is proposed to use the simulated annealing algorithm to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, such as Figure 3 Given the initial temperature, annealing rate, termination temperature and the values ​​of each optimization variable in the initial state, the iterative process of "generating a new solution -> calculating the difference between the variable values ​​before and after -> accepting or discarding the new solution" is repeated for the three optimization variables of legal UAV transmission power, UAV jammer transmission power and access satellite receiving power, and the temperature value is gradually attenuated until the termination temperature is reached, so as to obtain the minimum transmission delay of the system, that is, the optimal UAV transmission power and the optimal access satellite receiving power.

[0294] The following is a detailed description of the technical effects of the present invention. The simulation is carried out in a network environment with a fixed number of MMUs, base stations, drones, and satellites, and the two performance indicators of system delay and average confidentiality rate are statistically analyzed. Main network simulation parameters: additive white Gaussian noise power is 100dB, the transmission power of the small base station is 30dB, the bandwidth resource is 100MHz, the computing capacity of the local CPU is 900Mbps, the computing capacity of the ocean edge server is 6400Mbps, and the computing capacity of the satellite edge server is 8000Mbps.

[0295] The performance of the present invention is compared with baseline solution 1 (all MMUs perform local calculations), baseline solution 2 (part of the MMUs use ocean edge servers to offload calculations), and baseline solution 3 (part of the MMUs use satellite edge servers to offload calculations). Figure 4-6 shown.

[0296] In summary, the embodiment of the present invention provides a high-security offloading solution using dual-edge computing and physical layer security technology under the SAAIN system, using ocean edge servers and satellite edge servers, effectively ensuring the high requirements of MMU for network performance and network service quality, while fully considering the security of the system, and adopting physical layer security technology to greatly reduce the risk of MMUs being eavesdropped when transmitting tasks to satellites. Compared with existing mechanisms, the present invention can effectively reduce the transmission delay of the system and improve the overall computing efficiency and security of the system.

[0297] The present invention also provides a high-security unloading system using dual edge computing, comprising

[0298] The first model creation module is used to build an air-space-sea integrated network model, where the air-space-sea integrated network model includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites;

[0299] The second model creation module is used to construct a local computing delay model and a task offloading delay model according to the obtained air-space-sea integrated network model, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server;

[0300] The first data processing module is used to solve the information security problem in the process of unloading tasks from the relay drone to the satellite. The obtained task unloading delay model is modeled into the problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2 through a mathematical model, and the problem P1 is decoupled into four sub-problems.

[0301] The second data processing module is used to solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem through iteration of problem P1, and solve problem P2 obtained in step S3 through simulated annealing algorithm, so as to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

[0302] The present invention also provides a mobile terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, such as a high-security uninstallation program utilizing dual-edge computing.

[0303] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the high-security unloading method using dual-edge computing.

[0304] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A high-security offloading method using dual edge computing, characterized in that: The steps include: S1. Build an air-space-sea integrated network model, which includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites; S2. Constructing a local computing delay model and a task offloading delay model according to the air-space-sea integrated network model obtained in step S1, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server; S3. The information security problem in the process of unloading tasks from relay drones to satellites is solved by using a mathematical model to model the task unloading delay model obtained in step S2 as a problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2, and decoupling the problem P1 into four sub-problems. In step S3, the above task offloading delay model is modeled into a formulated problem P1 of maximizing the average confidentiality rate and a problem P2 of minimizing the system delay through a mathematical model, and the problem P1 is decoupled into four sub-problems, including: S301, jointly optimize the horizontal positions of the security drone and the jammer, and the power sent by the MMU to the legitimate drone, and model the average confidentiality rate maximization problem of the SAAIN system as problem P1: in, They represent the initial positions of the security UAV and the jammer respectively; q LF ,q Jf They represent the final positions of the safety UAV and the jammer respectively; D represents the maximum horizontal distance flown by the UAV in each time interval; is the self-interference of the transmitting antenna and receiving antenna of the unmanned jammer, is the channel power gain between the MMU and the security drone, is the channel power gain between the MMU and the unmanned jammer, is the received additive white Gaussian noise; Among them, constraints C1-C4 ensure that the horizontal distance moved by the drone in each interval does not exceed D, and constraints C5-C6 indicate that the transmission power of the drone does not exceed the average and peak values; S302, using the block coordinate descent method to decouple the above non-convex problem P1 into four sub-problems, including the power optimization problem of the MMU sending to the legitimate UAV, the power optimization problem of the MMU sending to the unmanned jammer, the trajectory optimization problem of the legitimate UAV, and the trajectory optimization problem of the unmanned jammer; S303, based on solving P1, the transmission power of the UAV and the receiving power of the access satellite are jointly optimized, and the total transmission delay minimization problem of the SAAIN system is modeled as problem P2: Among them, C1 represents the system's constraint on the average confidentiality rate, C2 represents the constraint on the system's task volume, and C4 represents that the computing power of the MEC server in the system is limited, and the local computing time of the MMU is not allowed to exceed the total delay; The delay caused by the MEC satellite executing the MMU task with sequence number m, is the time delay of the lth security UAV forwarding the task to the access satellite, is the propagation delay between two satellites in the same orbit, is the time delay of the mth MMU transmitting the task to the lth safety UAV; R n,u,l is the data transmission rate of the uplink between the MMU and the security drone, R n,u,j is the data transmission rate of the uplink between the MMU and the unmanned jammer, d m is the amount of data required by the MMU to execute the task, c m To calculate the task d m The number of CPU cycles required for each bit of data; x m The proportion of data volume offloaded from MMU to the ground MEC server; S4, solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem of problem P1 obtained by iterative step S3, and solve the problem P2 obtained in step S3 by simulated annealing algorithm to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

2. A high-security offloading method using dual edge computing according to claim 1, characterized in that: In step S1, the specific steps of constructing the air-space-sea integrated network model are as follows: S101. Construct an air-space-sea integrated network model that includes 1 small base station equipped with a MEC server, M MMUs, L security drones, J unmanned jammers, and 5 satellites. The 5 satellites belong to the same orbit, one of which is an access satellite that directly transmits data with the drone. The remaining 4 satellites are equipped with MEC servers for access satellite scheduling and computing offloading tasks. N MMUs are set to compute offloading tasks to the small base station, and MN MMUs use security drones as relays to compute offloading tasks on satellites. The set of security drones is represented as The set of unmanned jammers is represented as The MMU set that offloads tasks to the small base station is represented as The set of MMUs that offload tasks to the satellite is represented as S102. Establish the communication model between MMU and UAV and UAV and access satellite. Since LOS link is dominant in the link between MMU and security UAV, and the channel gain obeys the free space loss model, the channel power gain between MMU and security UAV is Channel power gain between MMU and UAV The formulas are as follows: Where β0 is the channel power gain when the reference distance d0=1, and γ is the path loss index; The horizontal position of the MMU for unloading missions to the satellite, To ensure the horizontal position of the drone, is the horizontal position of the UAV jammer, H is the fixed flight altitude of the safety UAV and the UAV jammer; Channel power gain between the security drone and the access satellite Channel power gain between the UAV and the access satellite They are respectively expressed as follows: Among them, G l is the transmitting antenna gain of the safety drone, G j is the transmitting antenna gain of the unmanned jammer, G s is the satellite’s receiving antenna gain, c is the speed of light in vacuum, H s is the vertical height of the satellite, and f is the carrier frequency; S103, when the mth MMU offloads the computing task to the ground small base station, the data transmission rate of the uplink between the MMU and the small base station is expressed as: Where B is the transmission bandwidth of all MMUs, P m is the power value of the NOMA signal transmitted by the MMU to the small base station, is the received additive white Gaussian noise, 0≤|h1| 2 ≤|h2| 2 ≤...|h m | 2 ≤...≤|h N | 2 They represent the channel gains of MMUs 1 to N respectively; In the UAV-assisted uplink transmission system, when the MMU is powered by P n,u,l Transmit FDMA signals to legitimate drones, and the MMU transmits FDMA signals to legitimate drones at a power of P. n,u,j The FDMA signal is transmitted to the UAV, and the data transmission rate R of the uplink between the MMU and the security UAV is n,u,l , the data transmission rate P of the uplink between MMU and unmanned jammer n,u,j Respectively expressed as: Where B is the transmission bandwidth of all MMUs, is the self-interference of the transmitting antenna and receiving antenna of the unmanned jammer, is the channel power gain between the MMU and the security drone, is the channel power gain between the MMU and the unmanned jammer, is the received additive white Gaussian noise; The average confidentiality rate of the system generated by the process of MMUs transmitting tasks to legitimate drones is expressed as: Where M is the total number of MMUs, N is the number of MMUs to small base stations for computing offloading tasks, MN is the number of MMUs to satellites for computing offloading tasks, and R n,u,l is the data transmission rate of the uplink between the MMU and the security drone, R n,u,j is the data transmission rate of the uplink between the MMU and the unmanned jammer.

3. A high-security offloading method using dual edge computing according to claim 1, characterized in that: In step S2, building a local computing delay model and a task offloading delay model specifically includes: S201, the computing task size of the mth MMU is D m Representation, definition where d m is the amount of data required by the MMU to execute the task, c m To calculate the task d m The number of CPU cycles required for each bit of data in the MMU data; the entire calculation process of MMU data includes local calculation and partial offloading; the amount of task data for MMU local calculation is (1-x m )d m , the amount of data that MMU offloads to the ground MEC server is x m d m , where x is the offload variable; therefore, the latency of MMU local calculation is expressed as: in The CPU calculation frequency of the MMU with serial number m, x m The proportion of data volume offloaded from MMU to the ground MEC server; S202, the delay of computing the task to be unloaded to the ground. The delay generated when N MMUs unload the task to the ground is the delay of MMUs unloading the task to the ground small base station with MEC server and the delay generated by the calculation of the small base station executing the task. S203, calculate the delay of task offloading to the satellite. When the remaining MN MMUs face delay-sensitive tasks, the MMUs choose to offload the tasks to the space satellite for execution. The total delay caused by MMUs in space offloading is composed of the delays caused by the following three stages: the delay of MMUs transmitting tasks to the security drone, the delay of the security drone forwarding tasks to the access satellite, and the delay of the access satellite scheduling the MEC satellite.

4. A high-security offloading method using dual edge computing according to claim 3, characterized in that: In step S202, the time delay process of calculating the task offloading to the ground is as follows: First, the delay of offloading the mth MMU task to the ground small base station is defined as: Among them, R m is the data transmission rate of the uplink between the MMU and the small base station; Secondly, define the delay caused by the ground small base station executing the mth MMU task as: where f s It represents the CPU computing frequency allocated to the mth MMU of the ground small base station; The total delay of N MMUs in computing offloading tasks in the ground small base station is expressed as follows: in, The latency of offloading the mth MMU task to the ground small base station; The delay caused by the execution of the MMU task with sequence number m by the ground small base station.

5. A high-security offloading method using dual edge computing according to claim 3, characterized in that: In step S203, the process of calculating the time delay of offloading the task to the satellite is as follows: First, define the delay of the mth MMU transmitting the task to the lth safety drone as: in, is the horizontal position of the safety drone; H is the height above the ground at which the safety drones are fixed; is the channel power gain between the MMU and the unmanned jammer; Secondly, define the delay of the lth safety UAV forwarding task to the access satellite as: Where M is the number of subchannels, which is obtained by evenly dividing the bandwidth resource B; B is the transmission bandwidth of all MMUs; G s is the satellite’s receiving antenna gain; f is the carrier frequency; is the received additive white Gaussian noise; H s is the vertical height of the satellite; After receiving the task forwarded by the UAV, the access satellite first separates the unloading information from the interference information, and then dispatches a MEC satellite for unloading; the five satellites in the system are located in the same orbit, and the propagation delay between the two satellites is estimated using the radius of the earth, the altitude of the satellite, and the angle between the orbital planes; when the latitude is above 66°, the satellite link will be interrupted, so the shortest propagation delay occurs at the latitude of 66°; according to the cosine theorem, the propagation delay between two satellites in the same orbit is: Among them, R p is the radius of the Earth, n y is the number of satellites on the orbital plane; Finally, the delay caused by the MEC satellite executing the MMU task with sequence number m is: Among them, W u is the size of the offload task, u=1,2,...,U, unit: bits.

6. A high-security offloading method using dual edge computing according to claim 1, characterized in that: In step S4, the optimal transmission power of the UAV and the optimal receiving power of the satellite are obtained by iterating to solve the transmission power optimization problem of the MMU to the legitimate UAV and the UAV jammer, and the trajectory optimization problem of the legitimate UAV and the UAV jammer respectively, and solving the P2 problem by simulated annealing algorithm. Specifically, the following are included: S401, when solving the MMU transmission power sub-problem and the drone trajectory optimization sub-problem, give the initial values ​​of each variable, bring them into the solution of the power approximate solution of the signal sent by the MMU to the legal drone in the i-th iteration, then substitute the power approximate solution of the signal sent by the MMU to the legal drone and other initial values ​​into the power optimization sub-problem of the signal sent by the MMU to the unmanned jammer in the i-th iteration, solve the local approximate solution of the power of the signal sent by the MMU to the unmanned jammer in the i-th iteration, then substitute the power value of the signal sent by the MMU to the drone and other initial values ​​into the legal drone trajectory optimization sub-problem, solve the i-th iteration local approximate solution of the legal drone trajectory, then substitute the i-th iteration drone transmission power solution and the legal drone trajectory solution into the unmanned jammer trajectory sub-problem to find the local approximate solution, and continuously increase the number of iterations until the difference between the P1 objective function values ​​of the two iterations is within a certain range, then the maximum average confidentiality rate can be obtained, and the process ends; S402. Based on the similarity between the solution process of the optimization problem P2 and metal annealing, the simulated annealing algorithm is used to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite. Given the initial temperature T0, annealing rate r, termination temperature T and the values ​​of each optimization variable in the initial state, the transmission power of the legal UAV under the i+1th iteration is solved. If the difference between the i+1th and the i-th iteration is less than 0, the new solution under the i+1th iteration is accepted. At the same time, the transmission power of the unmanned jammer under the i+1th iteration is continued to be solved. If the difference between the i+1th and the i-th iteration is less than 0, the new solution under the i+1th iteration is accepted. The new solution under the i+1th iteration is obtained, and the transmission power of the access satellite under the i+1th iteration is continued to be solved. If the difference between the i+1th and i-th times is less than 0, the new solution under the i+1th iteration is accepted; if the difference between the transmission power of the legal drones under the i+1th and i-th times is not less than 0, first determine whether the relationship between the difference satisfies the requirement of being greater than the random number. If so, the new solution under the i+1th iteration is accepted, otherwise, the new solution under the i+1th iteration is not accepted; after the judgment of the i+1th iteration is completed, the temperature value is attenuated until the termination temperature is reached, the minimum transmission delay of the system can be obtained, and the algorithm ends.

7. A high-security unloading system using dual-edge computing, used to implement a high-security unloading method using dual-edge computing as described in any one of claims 1 to 6, characterized in that: include The first model creation module is used to build an air-space-sea integrated network model, where the air-space-sea integrated network model includes several small base stations, maritime mobile user MMUs, security drones, unmanned jammers and satellites; The second model creation module is used to construct a local computing delay model and a task offloading delay model according to the obtained air-space-sea integrated network model, wherein the task offloading delay model includes task offloading to a ground edge server and task offloading to a satellite edge server; The first data processing module is used to solve the information security problem in the process of unloading tasks from the relay drone to the satellite. The obtained task unloading delay model is modeled into the problem of maximizing the average confidentiality rate P1 and minimizing the system delay P2 through a mathematical model, and the problem P1 is decoupled into four sub-problems. The second data processing module is used to solve the MMU transmission power sub-problem and the approximate optimal solution of the UAV trajectory optimization sub-problem through iteration of problem P1, and solve problem P2 obtained in step S3 through simulated annealing algorithm, so as to obtain the optimal transmission power of the UAV and the optimal receiving power of the access satellite, and complete the high-security offloading work using dual-edge computing.

8. A mobile terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a high-security offloading method using dual-edge computing as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a high-security offloading method using dual-edge computing as described in any one of claims 1 to 6 are implemented.

Images