An optimization method for UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission

Abstract

The present invention discloses an optimization method for a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission, including: obtaining system data and pending computing tasks of the drone-assisted mobile edge computing system; solving a preset joint optimization model based on the obtained system data and pending computing tasks to obtain an optimization result; based on the optimization result, controlling the drone to fly according to the optimal drone flight trajectory, unloading the data volume unloaded by the optimal ground active user to the ground idle user and the drone, and allocating the pending computing tasks to the ground active user, the ground idle user, the drone and the satellite for calculation according to the optimal allocation. The present invention can maximize the task completion of all ground users on the basis of meeting the drone energy consumption limit.

Description

Technical Field

[0001] The present invention relates to an optimization method for a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission, and belongs to the technical field of wireless communications and the Internet of Things. Background Art

[0002] With the rapid development of the Internet of Things and 5G communication technologies, applications such as smart homes, face recognition, and augmented reality have appeared in our lives. This has promoted the development of mobile communication devices towards low latency, low power consumption, high reliability, and high-density connections. However, due to the limitations of the battery and computing power of mobile devices, existing communication systems are challenging to perform computationally intensive and delay-sensitive tasks. To address this problem, mobile edge computing is an effective technology. Mobile edge computing refers to servers deployed at the edge of the network to provide services to users. Since edge servers are located near users, they can provide computing services to users at lower energy costs and low latency.

[0003] Traditional mobile edge computing servers (such as ground base stations, access points, etc.) can provide edge computing services for computing-intensive tasks of surrounding devices, thereby reducing the energy consumption and response delay of devices. With the development of industrial technology, the battery capacity of mobile terminals (such as mobile phones, tablets, etc.) is getting larger and larger, and the computing power is getting stronger and stronger. The computing tasks of user devices can be offloaded to nearby mobile terminals. In addition, mobile terminals can also act as relays for user devices, making computing offloading more flexible.

[0004] However, the deployment of traditional edge servers in fixed locations may lead to inflexible and difficult services, which are affected by complex terrain and uncertain demands. UAV-assisted mobile edge computing has flexible and fast deployment capabilities and is considered suitable for providing services for specific areas and special computing needs. UAV-based edge computing servers have high-speed mobility and can freely approach mobile users and provide computing services, thereby significantly improving network performance. In addition, compared with ground communications, air-to-ground communications in UAV-assisted mobile edge computing networks can provide larger link capacity due to line-of-sight link transmission between UAVs and users. Therefore, UAV-assisted mobile edge computing networks are currently a hot topic in mobile edge computing research.

[0005] In the existing scenario of drone-assisted mobile edge computing based on user collaboration, the mode of user collaboration is fixed association, that is, users can only offload tasks to designated users for computing. However, the user distribution is not uniform, so fixed user association cannot better optimize the user resource allocation. Moreover, due to the mobility of drones, the different link performance between drones and users at different times also makes the drone resource allocation uneven. Summary of the invention

[0006] The purpose of the present invention is to overcome the deficiencies in the prior art and provide an optimization method for a UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission, which can optimize the UAV-assisted mobile edge computing system and maximize the task completion of all ground users on the basis of meeting the energy consumption limit of the UAV. To achieve the above purpose, the present invention is implemented by adopting the following technical solutions:

[0007] In a first aspect, the present invention provides an optimization method for a UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission, comprising:

[0008] The obtained system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users;

[0009] Based on the acquired system data and pending computing tasks, the preset joint optimization model is solved to obtain optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies for ground users, UAVs, and satellites;

[0010] Based on the optimization results, the UAV is controlled to fly according to the optimal UAV flight trajectory, and the data volume offloaded by the optimal ground active user is unloaded to the ground idle users and UAVs. According to the optimal allocation, the to-be-processed computing tasks are allocated to the ground active users, ground idle users, UAVs and satellites for calculation.

[0011] In a second aspect, the present invention provides an optimization device for a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission, comprising:

[0012] Acquisition module: used to acquire system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users;

[0013] Optimization calculation module: used to solve the preset joint optimization model based on the acquired system data and the computing tasks to be processed to obtain the optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies for ground users, UAVs and satellites;

[0014] Computing allocation module: It is used to control the UAV to fly according to the optimal UAV flight trajectory based on the optimization results, unload the data volume unloaded by the optimal ground active user to the ground idle users and UAVs, and allocate the pending computing tasks to the ground active users, ground idle users, UAVs and satellites for calculation according to the optimal allocation.

[0015] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in the first aspect is implemented.

[0016] In a fourth aspect, the present invention provides a computer device, comprising:

[0017] A memory for storing instructions;

[0018] A processor is used to execute the instructions so that the device performs operations to implement the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in the first aspect.

[0019] Compared with the prior art, the optimization method of a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission provided by an embodiment of the present invention has the following beneficial effects:

[0020] The present invention obtains system data and pending computing tasks of a drone-assisted mobile edge computing system; wherein the drone-assisted mobile edge computing system includes satellites, drones, ground active users and ground idle users; based on the obtained system data and pending computing tasks, a preset joint optimization model is solved to obtain an optimization result; the optimization result includes: an optimal drone flight trajectory, an optimal user-associated variable, an optimal amount of data unloaded by ground active users, and an optimal allocation of CPU computing frequencies of ground users, drones and satellites; the previously proposed drone-assisted mobile edge computing system framework based on user collaboration does not consider dynamic user associations, and the joint optimization model provided by the present invention can better optimize system resource allocation; the present invention considers a time-division multiplexing communication method to maximize the completion of all ground user tasks on the basis of meeting the drone energy consumption limit, and jointly optimizes the drone flight trajectory, the amount of data unloaded by ground active users, the CPU computing frequency of ground users, drones and satellites, and user-associated variables.

[0021] Based on the optimization results, the present invention controls the drone to fly according to the optimal drone flight trajectory, unloads the optimal amount of data unloaded by the ground active user to the ground idle user and the drone, and distributes the to-be-processed computing tasks to the ground active user, the ground idle user, the drone and the satellite for computing according to the optimal allocation. In the present invention, the drone can not only act as a mobile edge computing server to provide computing services for the ground active users, but also act as a relay to further relay the tasks of the ground active users to the remote satellite for processing.

[0022] Since the solution of the joint optimization model is non-convex and difficult to solve, the present invention solves this problem by iteratively solving three sub-problems; the amount of data unloaded by ground active users, the CPU computing frequency of ground users, UAVs and satellites, and user-associated variables are obtained in a closed form through the Lagrange dual method, the Lagrange multipliers related to the inequality constraints can be obtained by the subgradient method, and the Lagrange multipliers related to the equality constraints can be obtained by binary search; the interior point method can effectively solve the sub-problems related to the UAV trajectory optimization.

[0023] The convergence of the optimization method of the drone-assisted mobile edge computing system provided by the present invention can be guaranteed, and the required complexity is low; simulation results show that compared with existing solutions (such as solutions with fixed user associations and solutions with preset drone trajectories), the present invention significantly improves performance. In addition, the present invention can provide more stable performance to adapt to changes in the operating environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a model of a drone-assisted mobile edge computing system in an optimization method of a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission provided in Embodiment 1 of the present invention;

[0025] Figure 2 It is the flight trajectory of the UAV obtained by the method described in the first embodiment under different task completion times provided in the second embodiment of the present invention;

[0026] Figure 3 is a curve showing the task completion amount of all ground activity users under different comparison schemes as a function of task completion time provided in the second embodiment of the present invention;

[0027] Figure 4 It is a curve of the task completion amount of all ground active users provided in the second embodiment of the present invention as the broadcast power of the drone under different comparison schemes. DETAILED DESCRIPTION

[0028] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0029] Embodiment 1:

[0030] This embodiment introduces an optimization method for a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission, including:

[0031] The obtained system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users;

[0032] Based on the acquired system data and pending computing tasks, the preset joint optimization model is solved to obtain optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies for ground users, UAVs, and satellites;

[0033] Based on the optimization results, the UAV is controlled to fly according to the optimal UAV flight trajectory, and the data volume offloaded by the optimal ground active user is unloaded to the ground idle users and UAVs. According to the optimal allocation, the to-be-processed computing tasks are allocated to the ground active users, ground idle users, UAVs and satellites for calculation.

[0034] The specific steps include:

[0035] Step 1: Obtain system data and pending computing tasks of the drone-assisted mobile edge computing system.

[0036] UAV u, satellite s, I ground active users and K ground idle users, I<K. All users, UAVs and satellites are equipped with single antennas, communication circuits for communication and computing processors with limited computing power. In addition, UAVs are equipped with radio frequency energy transmitters. Within the task completion time T, I ground active users have a certain amount of computing tasks to be processed and adopt a partial offloading method, that is, part of the tasks of ground active users are calculated locally, and the rest are offloaded to UAVs, satellites and ground idle users. UAVs use wireless power transmission to send energy to all users, and ground active users use the collected energy for local calculation and offload tasks to ground idle users and UAVs. UAVs can not only calculate user tasks, but also relay user tasks to satellites. Ground idle users use the collected energy to calculate tasks from ground active users, and they are dynamically associated with ground active users.

[0037] The UAV flies at a fixed altitude H within the mission completion time T, and the satellite flies at a fixed altitude H within the mission completion time T. s Flying, all ground users are on level ground.

[0038] The task completion time T is divided into N time slots, and the width of each time slot is τ = T / N; when N is large enough, the drone can be considered static in the n∈{1,...,N}th time slot.

[0039] The ground user keeps the same position within the task completion time T. The position of a ground-based user is s i =(x i ,y i ), represents the set of ground-based users, The position of the idle ground user is s k =(x k ,y k ), Represents the set of idle users on the ground.

[0040] The horizontal coordinate of the drone is expressed as u[n]=(x[n],y[n]), and the speed of the drone is expressed as The drone flies from the initial point u0 to the end point u within the mission completion time T. N , u0=u[0],u N =u[N].

[0041] In the nth time slot, the channel gain between the i-th ground active user and the UAV is β0 is the channel power gain when the air reference distance is 1m; in the nth time slot, the channel gain between the kth ground idle user and the UAV is The channel gain between the i-th active ground user and the k-th idle ground user is h i,k =g0||s i -s k || -θ , g0 is the channel power gain when the ground reference distance is 1m, and θ is the loss constant.

[0042] The channel gain between the drone and the satellite is In the nth time slot, the energy collected by the i-th ground active user is Where 0≤η≤1 is the energy conversion efficiency of converting the received RF signal into a DC signal, P u is the broadcast power of the drone; in the nth time slot, the energy collected by the kth ground idle user is

[0043] User associated variable a i,k [n]∈{0,1}, a i,k [n] = 1 means that the i-th active ground user is served by the k-th idle ground user in the n-th time slot, ai,k [n]=0 means that the i-th ground active user is not served by the k-th ground idle user in the n-th time slot.

[0044] The previously proposed drone-assisted mobile edge computing system framework based on user collaboration did not consider dynamic user associations. The joint optimization model provided by the embodiment of the present invention can better optimize system resource allocation.

[0045] The system communication adopts time division multiplexing. Users can perform local computing and task offloading at the same time. UAVs can perform computing and task relay at the same time. The time for satellites, UAVs and ground idle users to return the computing results to ground active users is ignored. Each time slot τ is divided into I time periods τ / I. All operations related to the i-th ground active user are completed within the time period τ / I. Each time period τ / I is further divided into two parts. The time for the i-th ground active user to offload tasks to the k-th ground idle user is τ0=τ / (2I), and the time related to the UAV is also τ0=τ / (2I). The τ0 related to the UAV is divided into two parts: the time for the i-th ground active user to offload tasks to the UAV is τ u =τ / (4I), the time it takes for the UAV to relay the i-th ground activity user task to the satellite is τ u =τ / (4I); the computational resources required for the i-th ground active user to calculate one input bit are expressed as C i .

[0046] The embodiment of the present invention considers the time-division multiplexing communication mode to maximize the task completion of all ground users on the basis of meeting the energy consumption limit of the UAV, and jointly optimizes the flight trajectory of the UAV, the amount of data unloaded by ground active users, the CPU computing frequency of ground users, UAVs and satellites, and user-related variables.

[0047] Step 2: Build a joint optimization model.

[0048] With the goal of maximizing the task completion of all ground active users, a joint optimization model of UAV flight trajectory, the amount of data unloaded by ground active users, user-related variables, and the distribution of CPU computing frequency of ground users, UAVs and satellites was established under the constraints of task completion time and maximum energy consumption of UAVs.

[0049] It is expressed by the following formula:

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061]

[0062] C12:u0=u[0],u N =u[N]

[0063] C13:f i [n]≤f i,max ,f i,k [n]≤f k,max ,

[0064]

[0065]

[0066]

[0067]

[0068]

[0069] Among them, u[n] represents the trajectory of the UAV, and the optimization variable L in the objective function is expressed as L={L i,k [n],L i,u [n],L i,s [n]}, L i,k [n] is the number of bits that the i-th active ground user unloads to the k-th idle ground user in the n-th time slot, L i,u [n] is the number of bits that the i-th ground active user offloads to the UAV in the n-th time slot, L i,s [n] is the number of bits that the UAV relays the task of the i-th ground active user to the satellite in the n-th time slot, L i [n] is the number of bits calculated locally by the i-th ground active user in the n-th time slot, L i [n] = τf i[n] / C i , C i The computing resources required to calculate one input bit for the ith ground active user; the optimization variable F in the objective function is the CPU computing frequency of the ground user, the UAV and the satellite F = {f i [n],f i,k [n],f i,u [n],f i,s [n]}, f i [n] is the CPU frequency of the i-th ground active user in the n-th time slot, and its maximum value is f i,max , f i,k [n] is the CPU frequency used by the kth ground idle user in the nth time slot to calculate the ith ground active user task, and the maximum value is f k,max , f i,u [n] is the CPU frequency of the drone used to calculate the i-th ground activity user task in the n-th time slot, and the maximum value is f u,max , f i,s [n] is the CPU frequency of the satellite used to calculate the i-th ground active user task of the drone relay in the n-th time slot, and the maximum value is f s,max ;

[0070] C1-C2 is used to ensure that each active ground user is served by at most one idle ground user in each time slot;

[0071] C3-C4 is used to ensure that the energy consumed by ground active users and ground idle users does not exceed the collected energy. m is a temporary variable in the summation calculation. E i [n] is the energy consumption calculated locally by the i-th ground active user in the n-th time slot, E i [n] = τκ i f i 3 [n], κ i is the effective capacitance coefficient of ground active user i; E i,k [n] is the energy consumption of the i-th ground active user offloading the task to the k-th ground idle user in the n-th time slot, B is the communication bandwidth, is the receiving noise power of the kth ground idle user, τ0 = τ / (2I) is the time for the ith ground active user to offload the task to the kth ground idle user; E i,u [n] is the energy consumption of the i-th ground active user offloading the task to the UAV in the n-th time slot, represents the received noise power of the UAV, τ u =τ / (4I) is the time for the i-th ground activity user to offload tasks to the UAV and for the UAV to relay the ground activity user tasks to the satellite; Calculate the corresponding energy consumption from the i-th ground active user for the k-th ground idle user in the n-th time slot, κ k is the effective capacitance coefficient of the ground idle user k;

[0072] C5-C6 is used to ensure that ground idle users can only calculate the task input data that has been received in the previous time slot. Calculate the workload from the i-th ground active user for the k-th ground idle user in the n-th time slot,

[0073] C7-C8 is used to ensure that the UAV can only calculate or unload the task input data that has been received in the previous time slot. Calculate the task load from the i-th ground active user for the UAV in the n-th time slot, κ u is the effective capacitance coefficient of the drone;

[0074] C9-C10 is used to ensure that the satellite can only calculate the mission input data that has been received in the previous time slot. Calculate the workload from the i-th ground active user for the satellite,

[0075] C11 represents the maximum energy consumption limit of the drone. Calculate the energy consumption corresponding to the i-th ground active user for the UAV in the n-th time slot, E i,s [n] is the energy consumption of the UAV relaying the task of ground active user i to the satellite in the nth time slot, Expressed as the receiving noise power at the satellite end; E u is the broadcast energy consumption of the UAV in the nth time slot, E u =τP u ; E fly [n] is the propulsion energy consumption of the UAV in the nth time slot, E fly [n]=c||v[n]|| 2 , c = 0.5Mτ, M is the mass of the drone; E uav is the maximum energy consumption of the human-machine within the task completion time T;

[0076] C 12 means that the UAV flies from the initial point to the end point within the mission completion time;

[0077] C13-C17 are the value restrictions for each variable.

[0078] Step 3: Solve the preset joint optimization model.

[0079] Step a: Initialize the iteration number variable ξ = 1, the iteration termination accuracy ψ, initialize the drone trajectory u[n] and the user association variable a i,k [n].

[0080] Step b: Given the flight trajectory of the UAV and the user-related variables, the joint optimization model is calculated to optimize the amount of data unloaded by ground active users and the distribution of CPU computing frequencies of ground users, UAVs and satellites, and it is iterated until the sub-problem objective function value E meets the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated, and the optimal amount of data offloaded by the ground active users and the optimal allocation of CPU computing frequencies for ground users, UAVs, and satellites are obtained.

[0081] Step b-1: Given the drone flight trajectory u[n] and the user associated variable a i,k Under the condition of [n], the amount of data L unloaded by ground active users and the distribution F of CPU computing frequency of ground users, drones and satellites are optimized. The joint optimization model is transformed into:

[0082]

[0083] stC3-C11,C12-C17.

[0084] Step b-2: Initialize the dual variables γ corresponding to the inequality constraints C3, C4, C5, C7, C9 and C11 = {γ i,n},μ={μ k,n},ζ={ζ i,k,n}, λ={λ i,n}, η={η i,n} and δ, The value of .

[0085] Step b-3: Based on the given values ​​of the dual variables γ, μ, ζ, λ, η, and δ, use the binary search method to determine the dual variable ρ = {ρ i,k},ν={ν i}and The value of .

[0086] Specifically,

[0087] Step b-3-1: Step b-3-2: Step b-4 through the equality constraints C8, C10 and the optimization variable L i,s The closed form of [n] gives the expression Several different equations for , as shown below:

[0088]

[0089]

[0090]

[0091] By using the equality constraint C6 and the optimization variable L in step b-4 i,k The closed-form solution of [n] can be obtained for the expression Several different equations for , as shown below:

[0092]

[0093]

[0094] Step b-3-2: The value range is in

[0095] It can be deduced that ν i The value range is ν i ∈(ν i,min ,ν i,max ),in

[0096] It can also be deduced The value range is in

[0097] When the binary search terminates, we can get i The corresponding condition value, so that

[0098]

[0099] Continuously binary search ν i , and iterate the corresponding value, so that

[0100]

[0101] When the binary search method converges, the dual variable ν i and The optimal value of can be determined.

[0102] Step b-3-3: The value range is in

[0103]

[0104] It can be deduced that i,k The value range is ρ i,k ∈(ρ i,k,min ,ρ i,k,max ),in

[0105] When the binary search terminates, the optimal ρ can be obtained i,k , which satisfies

[0106] Step b-4: Using the Karush-Kuhn-Tucker condition, the closed-form solution of the model subproblem in step b-1 is obtained as f i [n],f i,k [n],f i,u [n],L i,k [n],L i,u [n],L i,s [n],f i,s [n], is represented by the following formula:

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114] Where [x] + =max{x,0},

[0115] Step b-5: Use a subgradient-based algorithm to iteratively update the dual variables γ, μ, ζ, λ, η, and δ as follows:

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122] Among them, j is the iteration index variable of the subgradient algorithm iteration; denote the step sizes for obtaining the dual variables γ, μ, ζ, λ, η, δ in the j-th iteration respectively.

[0123] Step b-6: Repeat steps b-3 to b-5 until the objective function value E of the model subproblem in step b-1 converges, the iteration is terminated, and the optimal amount of data unloaded by the optimal ground active user and the optimal allocation of CPU computing frequencies for ground users, drones, and satellites are obtained.

[0124] Step c: Given the user association variables, the amount of data unloaded by the ground active users, and the distribution of CPU computing frequencies of the ground users, the UAVs, and the satellites, the joint optimization model is calculated to optimize the UAV flight trajectory; iterative execution is performed until the sub-problem objective function value E satisfies the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated and the optimal UAV flight trajectory is obtained.

[0125] Step c-1: Given a user associated variable a i,k Under the conditions of [n], the amount of data unloaded by ground active users L, and the distribution F of CPU computing frequency of ground users, UAVs and satellites, the flight trajectory u[n] of the UAV is optimized. The joint optimization model is transformed into:

[0126]

[0127] stC3,C4,C11,C12.

[0128] Step c-2: For constraints C3 and C4, the non-concave term h i,u [n] and h k,u [n] respectively approximate their concave lower bounds and It is expressed by the following formula:

[0129]

[0130] Among them, r is the iteration index variable, u (r) [n] is a local point of the UAV trajectory.

[0131] Convert the non-convex optimization problem in step c-2 into the following convex optimization problem:

[0132]

[0133] stC11,C12

[0134]

[0135]

[0136] Step c-3: Use the initial UAV trajectory u[n] as the local point and use the interior point method to iterate the UAV trajectory u[n] until the objective function value E of the model subproblem in step c-2 converges, the iteration is terminated, and the optimal UAV flight trajectory is obtained.

[0137] Step d: Given the flight trajectory of the UAV, the amount of data unloaded by the ground active users, and the distribution of CPU computing frequencies of the ground users, UAVs, and satellites, calculate the joint optimization model and optimize the user-related variables; iterate until the sub-problem objective function value E satisfies the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated and the optimal user association variable is obtained.

[0138] Step d-1: Given the UAV flight trajectory u[n], the amount of data unloaded by the ground active users L, and the distribution F of the CPU computing frequency of the ground users, UAVs and satellites, optimize the user association variable a i,k [n], the joint optimization model is transformed into:

[0139]

[0140] stC1-C3,C5.

[0141] Step d-2: Convert the binary variable a i,k [n] Relaxation to continuous variables Then the model sub-problem in step d-1 is transformed into:

[0142]

[0143]

[0144]

[0145]

[0146]

[0147] Step d-3: Initialize the dual variables α corresponding to the inequality constraints C20, C23, and C24 = {α i,n},β={β i,n} and ω={ω i,k,n}, The value of n∈{1,...,N}.

[0148] Step d-4: Using the Karush-Kuhn-Tucker condition, the closed-form solution of the subproblem in step d-2 is obtained: As shown below:

[0149]

[0150] in, is the Lagrangian dual function pair of the subproblem in step d-2 The partial derivative of .

[0151] Step d-5: Use a subgradient-based algorithm to iteratively update the dual variables α, β, and ω as follows:

[0152]

[0153]

[0154]

[0155] Among them, j is the iteration index variable of the subgradient algorithm iteration. denote the step sizes for obtaining the dual variables α, β, ω in the j-th iteration respectively.

[0156] Step d-6: Repeat steps d-4 to d-5 until the objective function value E of the model sub-problem in step d-2 converges and the iteration terminates.

[0157] Step d-7: Set the value at the end of the iteration The value is changed to an integer of 0 or 1, and the The larger Assign a value of 1, which is Assign a value of 0 to the optimal user-associated variable.

[0158] Since the proposed optimization problem is non-convex and difficult to solve, the embodiment of the present invention solves this problem by iteratively solving three sub-problems. The amount of data unloaded by ground active users, the CPU computing frequency of ground users, drones and satellites, and user-related variables are obtained in a closed form by the Lagrange dual method. The Lagrange multipliers associated with the inequality constraints can be obtained by the subgradient method, and the Lagrange multipliers associated with the equality constraints can be obtained by binary search. The interior point method can effectively solve the sub-problems related to the optimization of drone trajectories.

[0159] Step 4: Based on the optimization results, control the UAV to fly according to the optimal UAV flight trajectory, unload the data volume of the optimal ground active user to the ground idle users and UAVs, and allocate the pending computing tasks to the ground active users, ground idle users, UAVs and satellites for calculation according to the optimal allocation.

[0160] The convergence of the optimization method of the drone-assisted mobile edge computing system provided by the embodiment of the present invention can be guaranteed, and the required complexity is low. The simulation results show that compared with the existing solutions (such as the solution with fixed user association and the solution with preset drone trajectory), the performance is significantly improved. In addition, more stable performance can be provided to adapt to changes in the operating environment.

[0161] Embodiment 2:

[0162] The solution of the present invention is described in detail below with reference to a specific example. The specific example is simulated using MATLAB software. The specific parameters are set as shown in Table 1.

[0163]

[0164]

[0165] In addition to the above table, the coordinates of ground active users are set to {s1,s2} = {(-4,0),(4,0)}, and the coordinates of ground idle users are set to {s1,s2,s3,s4} = {(-4,4),(4,-4),(-4,-4),(4,4)}.

[0166] pass Figure 2 Comparison of the UAV trajectories under different task completion times T shows that the UAV uses its mobility to find the best position in each time slot to complete energy broadcasting and computation offloading. Figure 2 It can be seen that as the task completion time T increases, the trajectory tends to approach the active users and idle users on the ground, which strengthens the air-to-ground communication link between the drone and the user, which can not only increase the energy collected by the user, but also increase the amount of tasks unloaded by the user. When T is small, the drone does not have enough time to get close to the user in order to complete the flight. The curve tends to be smooth, which means that in a shorter time, the drone is able to complete the flight and perform unloading, which is a compromise.

[0167] Figure 3 The figure shows the relationship between the task completion amount and the task completion time T of all ground activity users. Figure 3It can be seen that the task completion of all schemes increases with the increase of T. Among all the comparison algorithms, the method proposed in Example 1 is better. In addition, when T is larger, the gap between the method proposed in Example 1 and other comparison algorithms is larger, which means that when the system resources are more abundant, the method proposed in Example 1 has a more obvious advantage in increasing the task completion of ground activity users. This is achieved by utilizing all computing resources in the system and jointly optimizing the computing task allocation, drone trajectory, and user association variables.

[0168] Figure 4 The task completion of all ground-based active users and the broadcast power P of the drone are shown. u The relationship between Figure 4 As shown in Figure 1, the performance of the method proposed in Example 1 is better than other solutions. The figure further shows that the task completion of all algorithms increases with P u In addition, since the performance may be limited by other parameters, such as the processing time and flight speed of the drone, the increase rate of the task completion of the fixed user association algorithm increases with P u Showing a decreasing trend.

[0169] Embodiment three:

[0170] An embodiment of the present invention provides an optimization device for a drone-assisted mobile edge computing system based on user collaboration and wireless power transmission, comprising:

[0171] Acquisition module: used to acquire system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users;

[0172] Optimization calculation module: used to solve the preset joint optimization model based on the acquired system data and the computing tasks to be processed to obtain the optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies for ground users, UAVs and satellites;

[0173] Computing allocation module: It is used to control the UAV to fly according to the optimal UAV flight trajectory based on the optimization results, unload the data volume unloaded by the optimal ground active user to the ground idle users and UAVs, and allocate the pending computing tasks to the ground active users, ground idle users, UAVs and satellites for calculation according to the optimal allocation.

[0174] Embodiment 4:

[0175] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in Embodiment 1 is implemented.

[0176] Embodiment five:

[0177] An embodiment of the present invention provides a computer device, including:

[0178] A memory for storing instructions;

[0179] The processor is used to execute the instruction so that the device performs the operation of implementing the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in Example 1.

[0180] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0181] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0182] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0183] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0184] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An optimization method for a UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission, characterized in that: include: The obtained system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users; wherein the system data of the UAV-assisted mobile edge computing system includes: UAV u, satellite s, I ground active users and K ground idle users, I<K; The UAV flies at a fixed altitude H within the mission completion time T, and the satellite flies at a fixed altitude H within the mission completion time T. s flight; The task completion time T is divided into N time slots, and the width of each time slot is τ = T / N; The ground user keeps the same position within the task completion time T. The position of a ground-based user is s i =(x i ,y i ), represents the set of ground-based users, The location of the idle ground user is s k =(x k ,y k ), represents the set of idle users on the ground; The horizontal coordinate of the drone is expressed as u[n] = (x[n], y[n]), and the speed of the drone is expressed as The drone flies from the initial point u0 to the end point u within the mission completion time T. N , u0=u[0],u N =u[N]; In the nth time slot, the channel gain between the i-th ground active user and the UAV is β0 is the channel power gain when the air reference distance is 1m; in the nth time slot, the channel gain between the kth ground idle user and the UAV is The channel gain between the i-th active ground user and the k-th idle ground user is h ii,k =g0||s i -s k || -θ , g0 is the channel power gain when the ground reference distance is 1m, θ is the loss constant; The channel gain between the drone and the satellite is In the nth time slot, the energy collected by the i-th ground active user is Where 0≤η≤1 is the energy conversion efficiency of converting the received RF signal into a DC signal, P u is the broadcast power of the drone; in the nth time slot, the energy collected by the kth ground idle user is User associated variable a i,k [n]∈{0,1}, a i,k [n] = 1 means that the i-th ground active user is served by the k-th ground idle user in the n-th time slot, a i,k [n] = 0 means that the i-th ground active user is not served by the k-th ground idle user in the n-th time slot; Based on the acquired system data and pending computing tasks, the preset joint optimization model is solved to obtain the optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies of ground users, UAVs and satellites; wherein the preset joint optimization model aims to maximize the amount of task completion of all ground active users, and under the constraints of task completion time and maximum energy consumption of UAVs, the established joint optimization model of UAV flight trajectory, the amount of data unloaded by ground active users, user-related variables, and the allocation of CPU computing frequencies of ground users, UAVs and satellites is expressed by the following formula: C12:u0=u[0],u N =u[N] Among them, u[n] represents the trajectory of the UAV, and the optimization variable L in the objective function is expressed as L={L i，k [n],L i，u [n],L i，s [n]}, L i,k [n] is the number of bits that the i-th active ground user unloads to the k-th idle ground user in the n-th time slot, L i,u [n] is the number of bits that the i-th ground active user offloads to the UAV in the n-th time slot, L i,s [n] is the number of bits that the UAV relays the task of the i-th ground active user to the satellite in the n-th time slot, L i [n] is the number of bits calculated locally by the i-th ground active user in the n-th time slot, L i [n] = τf i [n] / C i , C i The computing resources required to calculate one input bit for the i-th ground active user; the optimization variable F in the objective function is the CPU computing frequency F of the ground user, the UAV and the satellite = {f i [n],f i，k [n],f i，u [n],f i，s [n]}, fi[n] is the CPU frequency of the i-th ground active user in the n-th time slot, and its maximum value is f i,max , f i,k [n] is the CPU frequency used by the kth ground idle user in the nth time slot to calculate the ith ground active user task, and the maximum value is f k,max , f i,u [n] is the CPU frequency of the drone used to calculate the i-th ground activity user task in the n-th time slot, and the maximum value is f u,max , f i,s [n] is the CPU frequency of the satellite used to calculate the i-th ground active user task of the drone relay in the n-th time slot, and the maximum value is f s,max ; C1-C2 is used to ensure that each active ground user is served by at most one idle ground user in each time slot; C3-C4 is used to ensure that the energy consumed by ground active users and ground idle users does not exceed the collected energy. m is a temporary variable in the summation calculation. E i [n] is the energy consumption calculated locally by the i-th ground active user in the n-th time slot, E i [n] = τκ i f i 3 [n], κ i is the effective capacitance coefficient of ground active user i; E i,k [n] is the energy consumption of the i-th ground active user offloading the task to the k-th ground idle user in the n-th time slot, B is the communication bandwidth, is the receiving noise power of the kth ground idle user, τ0 = τ / (2I) is the time for the ith ground active user to offload the task to the kth ground idle user; E i,u [n] is the energy consumption of the i-th ground active user offloading the task to the UAV in the n-th time slot, represents the received noise power of the UAV, τ u =τ / (4I) is the time for the i-th ground activity user to offload tasks to the UAV and for the UAV to relay the ground activity user tasks to the satellite; Calculate the corresponding energy consumption from the i-th ground active user for the k-th ground idle user in the n-th time slot, κ k is the effective capacitance coefficient of the ground idle user k; C5-C6 is used to ensure that ground idle users can only calculate the task input data that has been received in the previous time slot. Calculate the workload from the i-th ground active user for the k-th ground idle user in the n-th time slot, C7-C8 is used to ensure that the UAV can only calculate or unload the task input data that has been received in the previous time slot. Calculate the task load from the i-th ground active user for the UAV in the n-th time slot, κ u is the effective capacitance coefficient of the drone; C9-C10 is used to ensure that the satellite can only calculate the mission input data that has been received in the previous time slot. Calculate the workload from the i-th ground active user for the satellite, C11 represents the maximum energy consumption limit of the drone. Calculate the energy consumption corresponding to the i-th ground active user for the UAV in the n-th time slot, E i,s [n] is the energy consumption of the UAV relaying the task of ground active user i to the satellite in the nth time slot, Expressed as the receiving noise power at the satellite end; E u is the broadcast energy consumption of the UAV in the nth time slot, E u =τP u ; E fly [n] is the propulsion energy consumption of the UAV in the nth time slot, E fly [n]=c||v[n]|| 2 , c = 0.5Mτ, M is the mass of the drone; E uav is the maximum energy consumption of the human-machine within the task completion time T; C12 means that the UAV flies from the initial point to the end point within the mission completion time; C13-C17 are the value restrictions of each variable; Based on the optimization results, the UAV is controlled to fly according to the optimal UAV flight trajectory, and the data volume offloaded by the optimal ground active user is unloaded to the ground idle users and UAVs. According to the optimal allocation, the to-be-processed computing tasks are allocated to the ground active users, ground idle users, UAVs and satellites for calculation.

2. The optimization method of the UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission according to claim 1 is characterized in that: The solving of the preset joint optimization model includes: Step a: Initialize the iteration number variable ξ = 1, the iteration termination accuracy ψ, initialize the drone trajectory u[n] and the user association variable a i,k [n]; Step b: Given the flight trajectory of the UAV and the user-related variables, the joint optimization model is calculated to optimize the amount of data unloaded by ground active users and the distribution of CPU computing frequencies of ground users, UAVs and satellites, and it is iterated until the sub-problem objective function value E meets the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated, and the optimal amount of data unloaded by the ground active users and the optimal allocation of CPU computing frequencies of ground users, UAVs and satellites are obtained; Step c: Given the user association variables, the amount of data unloaded by the ground active users, and the distribution of CPU computing frequencies of the ground users, the UAVs, and the satellites, the joint optimization model is calculated to optimize the UAV flight trajectory; iterative execution is performed until the sub-problem objective function value E satisfies the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated and the optimal UAV flight trajectory is obtained; Step d: Given the flight trajectory of the UAV, the amount of data unloaded by the ground active users, and the distribution of CPU computing frequencies of the ground users, UAVs, and satellites, calculate the joint optimization model and optimize the user-related variables; iterate until the sub-problem objective function value E satisfies the convergence condition |E ξ -E ξ-1 |≤ψ, the iteration is terminated and the optimal user association variable is obtained.

3. The optimization method of the UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission according to claim 2 is characterized in that: The step b comprises: Step b-1: Given the drone flight trajectory u[n] and the user associated variable a i,k Under the condition of [n], the amount of data L unloaded by ground active users and the distribution F of CPU computing frequency of ground users, drones and satellites are optimized. The joint optimization model is transformed into: Step b-2: Initialize the dual variables γ corresponding to the inequality constraints C3, C4, C5, C7, C9 and C11 = {γ i,n },μ={μ k，n },ζ={ζ i,k,n },λ={λ i，n }, η={η i，n } and δ, The value of Step b-3: Based on the given values ​​of the dual variables γ, μ, ζ, λ, η, and δ, use the binary search method to determine the dual variable ρ = {ρ i，k },ν={ν i }and The value of Step b-4: Using the Karush-Kuhn-Tucker condition, the closed-form solution of the model subproblem in step b-1 is obtained as f i [n],f i,k [n],f i,u [n],L i,k [n],L i,u [n],L i,s [n],f i,s [n], is represented by the following formula: Where [x] + =max{x,0}, Step b-5: Use a subgradient-based algorithm to iteratively update the dual variables γ, μ, ζ, λ, η, and δ as follows: Among them, j is the iteration index variable of the subgradient algorithm iteration; denote the step sizes for obtaining the dual variables γ, μ, ζ, λ, η, δ in the jth iteration respectively; Step b-6: Repeat steps b-3 to b-5 until the objective function value E of the model subproblem in step b-1 converges, the iteration is terminated, and the optimal amount of data unloaded by the optimal ground active user and the optimal allocation of CPU computing frequencies for ground users, drones, and satellites are obtained.

4. The optimization method of the UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission according to claim 2 is characterized in that: The step c comprises: Step c-1: Given a user associated variable a i,k Under the conditions of [n], the amount of data unloaded by ground active users L, and the distribution F of CPU computing frequency of ground users, UAVs and satellites, the flight trajectory u[n] of the UAV is optimized. The joint optimization model is transformed into: Step c-2: For constraints C3 and C4, the non-concave term h i,u [n] and h k,u [n] respectively approximate their concave lower bounds and It is expressed by the following formula: Among them, r is the iteration index variable, u (r) [n] is the local point of the UAV trajectory; Convert the non-convex optimization problem in step c-2 into the following convex optimization problem: Step c-3: Use the initial UAV trajectory u[n] as the local point and use the interior point method to iterate the UAV trajectory u[n] until the objective function value E of the model subproblem in step c-2 converges, the iteration is terminated, and the optimal UAV flight trajectory is obtained.

5. The optimization method of the UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission according to claim 2, characterized in that: The step d comprises: Step d-1: Given the UAV flight trajectory u[n], the amount of data unloaded by the ground active users L, and the distribution F of the CPU computing frequency of the ground users, UAVs and satellites, optimize the user association variable a i,k [n], the joint optimization model is transformed into: Step d-2: Convert the binary variable a i,k [n] Relaxation to continuous variables Then the model sub-problem in step d-1 is transformed into: Step d-3: Initialize the dual variables α corresponding to the inequality constraints C20, C23, and C24 = {α i,n },β={β i,n } and ω={ω i,k,n }, The value of Step d-4: Using the Karush-Kuhn-Tucker condition, the closed-form solution of the subproblem in step d-2 is obtained: As shown below: in, is the Lagrangian dual function pair of the subproblem in step d-2 The partial derivative of Step d-5: Use a subgradient-based algorithm to adjust the dual variables α , β and ω Perform iterative updates as follows: Among them, j is the iteration index variable of the subgradient algorithm iteration, Respectively represent the dual variables obtained in the jth iteration α ,β, ω The step length; Step d-6: Repeat steps d-4 to d-5 until the objective function value E of the model sub-problem in step d-2 converges and the iteration terminates; Step d-7: Set the value at the end of the iteration The value is changed to an integer of 0 or 1, and the The larger Assign a value of 1, which is Assign a value of 0 to the optimal user-associated variable.

6. An optimization device for a UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission based on the optimization method for a UAV-assisted mobile edge computing system based on user collaboration and wireless power transmission according to claim 1, characterized in that: include: Acquisition module: used to acquire system data and pending computing tasks of the UAV-assisted mobile edge computing system; wherein the UAV-assisted mobile edge computing system includes satellites, UAVs, ground active users and ground idle users; Optimization calculation module: used to solve the preset joint optimization model based on the acquired system data and the computing tasks to be processed to obtain the optimization results; the optimization results include: the optimal UAV flight trajectory, the optimal user-related variables, the optimal amount of data unloaded by ground active users, and the optimal allocation of CPU computing frequencies for ground users, UAVs and satellites; Computing allocation module: It is used to control the UAV to fly according to the optimal UAV flight trajectory based on the optimization results, unload the data volume unloaded by the optimal ground active user to the ground idle users and UAVs, and allocate the pending computing tasks to the ground active users, ground idle users, UAVs and satellites for calculation according to the optimal allocation.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in any one of claims 1-5.

8. A computer device, characterized in that: include: A memory for storing instructions; A processor is used to execute the instructions so that the device performs operations to implement the optimization method of the drone-assisted mobile edge computing system based on user collaboration and wireless power transmission as described in any one of claims 1 to 5.

Images