Energy consumption optimization method for MEC communication based on drone and backscatter assistance

Abstract

The present invention discloses a method for optimizing energy consumption of MEC communication based on drones and backscattering assistance, which mainly solves the problems of insufficient energy and system security of existing mobile smart devices. It includes: 1) building a system model based on drone-assisted backscattering and its edge computing; 2) the drone flies in a certain area in the air, collecting information from multiple users, and the users obtain energy from the radio frequency signals transmitted by the drone, offloading part of the user's task to the drone for edge computing and the other part for local computing; at the same time, multiple eavesdroppers hijack the user information; 3) minimizing the energy consumption of the drone, finding the optimal user and drone CPU frequency, drone transmission power, user backscatter coefficient and drone trajectory; 4) setting the system operating parameters according to the optimization results to achieve optimization. The present invention can effectively improve the security of the system and the energy collection efficiency of smart devices, so as to optimize the network performance.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication technology and further relates to edge computing and backscattering technology. Specifically, it is a method for optimizing the communication energy consumption of mobile edge computing (MEC) based on drones and backscattering assistance, which can be used for information interaction between users and drones in edge computing systems. Background Art

[0002] The emergence of mobile wireless communication networks and the Internet of Things (IoT) has fueled the development of smart mobile devices (SMDs), providing a powerful platform for many novel intelligent applications. Applications such as facial recognition, interactive gaming, and autonomous navigation typically require significant computing resources, resulting in increased energy consumption. However, SMDs have limited computing resources and battery budgets, leading to a conflict between compute-intensive applications and limited resources. Mobile Edge Computing (MEC) is an effective technology that allows SMDs to offload their compute-intensive tasks to MEC servers located at the edge of the network. The MEC servers then return the results to the SMDs. Therefore, MEC can help resource-constrained users perform complex computations, reducing the computational burden on terminals and lowering data processing latency. Furthermore, in rural and remote areas, due to complex terrain, deploying a large number of static servers to offload tasks is unlikely to help reduce data processing latency. In such scenarios, mobile servers are more resilient to uncertain environments than static servers. Furthermore, existing MEC servers cannot promptly process large amounts of data to handle emergencies such as large-scale events. This places higher demands on the network service architecture in mobile communication systems.

[0003] Unmanned aerial vehicles (UAVs) have excellent maneuverability and low cost, and can be used to support more flexible mobile computing services. Therefore, UAV-assisted MEC has attracted much attention. It can provide a wide coverage range and additional computing power, which is of great significance to the future development of the Internet of Things. In addition, considering that the user's battery energy is limited and information transmission to the UAV requires energy supply, scholars have proposed solutions such as energy harvesting EH (Energy Harvesting) technology. However, sometimes the energy collected from the surrounding environment may not be enough to meet power consumption tasks. Backscatter communication is a low-power technology that transmits data by reflecting and modulating the incident radio frequency waves through the backscatter device, so the backscatter device does not need to generate active radio frequency signals and perform analog-to-digital conversion, thereby reducing energy consumption.

[0004] Shi LQ, Ye YH, Chu XL et al. proposed a backscatter-assisted wirelessly powered MEC network in their paper “Computation Bits Maximization in a Backscatter Assisted Wirelessly Powered MEC Network” (IEEE Communications Letters, 2021, 25(2): 528-532). By jointly optimizing the backscatter coefficient and time, active transmission power and time, local computing frequency and execution time of each edge user, a scheme was proposed to maximize the weighted computing bits of all edge users. Nie YW, Zhao JH, Liu J et al. introduced the energy efficiency problem of energy-limited backscatter communication networks in their paper “Energy-efficient UAV trajectory design for backscatter communication: A deep reinforcement learning approach” (China Communications, 2020, 17(10): 129-141). A deep reinforcement learning algorithm was proposed to design UAV trajectories under the constraints of transmission power, reflection coefficient, transmission power and fairness among users. The above solutions are mainly derived from relevant research in the fields of UAV, MEC and backscatter, and have improved communication quality to a certain extent. However, there is currently little research on combining UAV, MEC and backscatter communication technology, and the security of communication is rarely considered. Summary of the Invention

[0005] The present invention aims to address the shortcomings of the above-mentioned existing technologies and propose a method for optimizing energy consumption in MEC communications based on drones and backscattering. By maximizing drone energy consumption, the method finds the optimal user CPU frequency, drone CPU frequency, drone transmit power, user backscatter coefficient, and drone trajectory, thereby solving problems such as high system energy consumption and low security, and effectively improving system performance.

[0006] The basic idea of ​​implementing the present invention is to build a communication system using multiple user BDs carrying backscatter devices, eavesdroppers EDs, and UAVs equipped with MEC servers. The UAVs are allowed to fly in a certain area in the air. The BDs obtain energy from the radio frequency signals emitted by the UAVs, offload part of the user tasks to the UAVs for edge computing, and perform local computing on the other part.

[0007] To achieve the above object, the technical solution of the present invention includes the following steps:

[0008] (1) Build a communication system consisting of K users (BDs) carrying backscatter devices, M eavesdroppers (EDs), and a UAV (UAV) equipped with an edge computing server, where K ≥ 1, M ≥ 1, and m and k represent the mth ED and kth BD, respectively.

[0009] (2) Assume that a UAV flies in a certain area in the air and collects information from multiple BDs. The BD obtains energy from the radio frequency signal transmitted by the UAV, offloads part of the BD task to the UAV for edge computing, and performs part of the task locally. It is assumed that multiple EDs will hijack the information calculated locally by the BD and the information backscattered by the BD to the UAV.

[0010] (3) Obtain the optimal optimization variables of the system built in step (1):

[0011] (3.1) Discretize the finite task completion time T into N equal time slots, and the duration of each time slot is Let the UAV fly at a fixed altitude H, where H is the minimum altitude to ensure that the UAV can avoid all obstacles in the service area; assume that the position of the UAV remains unchanged during each time slot and flies at a constant speed in each time slot, BD and ED are fixed on the ground respectively, let the channel between the UAV and BD be a line-of-sight communication link LOS, and obtain the channel power gain h between the UAV and the kth BD k [n], the channel power gain g between the kth BD and the mth ED km and the channel power gain z between the UAV and the mth ED m [n];

[0012] (3.2) Use time division multiple access (TDMA) to further divide each time slot into K equal durations. The kth BD reflects its task input data in the kth duration; calculates the local computing task amount of the kth BD in the nth time slot and the energy consumed by local computation Define the reflection coefficient of the kth BD at time slot n as α k [n], obtain the backscattering task volume I of the kth BD at time slot n k [n] and the total energy collected by the kth BD on the nth time slot during the reflection phase

[0013] (3.3) The system confidentiality rate R is obtained according to the following formula ku [n]:

[0014] R ku [n]=min(R k [n]-R m,ED [n],0),

[0015] Among them, R k [n] is the rate of the backscatter channel between BD and UAV, R m,ED [n] is the information leakage rate,

[0016] (3.4) The UAV energy consumption E is obtained according to the following formula U [n]:

[0017]

[0018] Among them, w f is the UAV flight energy consumption weight, E fly 、 and They represent the flight energy consumption, computing energy consumption, and transmission energy consumption of the UAV respectively;

[0019] (3.5) Optimal UAV energy consumption expression:

[0020]

[0021] The optimization variable is the transmission power P of the UAV to the kth BD in the nth time slot. k [n], backscatter coefficient α of the kth BD in the nth time slot k [n], CPU frequency f of the kth BD k [n], UAV CPU frequency f in the nth time slot u,k [n] and the UAV trajectory q[n] at the nth time slot;

[0022] Assume that the communication system meets the following constraints:

[0023] This constraint is used to ensure that the collected energy is sufficient for BD use; This constraint is used to ensure that all tasks are calculated. is the minimum task amount of BDk; This constraint is used to ensure that the amount of tasks that the UAV can calculate within the time period T is less than the amount of tasks reflected by the BD; This constraint indicates that the amount of tasks calculated by the UAV must be greater than or equal to the total amount of tasks offloaded by the BD; R k [n]-R m,ED [n]≥R fair , this constraint is the security and confidentiality rate constraint; This constraint is the UAV CPU clock frequency constraint; 0≤f k [n]≤f m ' ax, which represents the CPU clock frequency constraint of BD; 0≤α k [n]≤1, this constraint is the reflection coefficient constraint; 0≤P k [n]≤P max , which is the UAV maximum transmission power constraint; q[0] = q I ,q[N]=q F , which determines the starting and ending points of the UAV’s flight; ||q[n]-q[n-1]||≤τV max ,This constraint is the speed constraint of the UAV within a time slot.,In the above constraint, i represents the summation variable of the time slot number;

[0024] (3.6) Minimize UAV energy consumption E U [n], get the optimal optimization variable:

[0025] (3.6.1) Fix the UAV trajectory among the optimization variables and optimize the remaining variables:

[0026] By introducing the auxiliary variable β k [n] Process the coupling, and then use the Lagrangian dual decomposition algorithm and the subgradient algorithm to solve the convex expression to obtain the minimum UAV energy consumption The optimal auxiliary variable corresponding to the nth time slot The optimal CPU frequency of the kth BD in the nth time slot Optimal CPU frequency of UAV in the nth time slot The optimal transmission power of UAV to the kth BD The optimal backscatter coefficient for the nth time slot

[0027] (3.6.2) Optimizing UAV trajectories:

[0028] Fix the optimization variables obtained in step (3.6.1), and then use the continuous convex approximation algorithm to transform the non-convex problem into a convex problem for solution, and finally obtain the optimal UAV trajectory q for the nth time slot * [n];

[0029] (4) Set the system operating parameters according to the optimal optimization variables, so that the system operates under the operating parameters to achieve system energy consumption optimization.

[0030] Compared with the prior art, the present invention has the following advantages:

[0031] First, because the present invention uses a UAV-assisted MEC network, it can be used to support more flexible mobile computing services, providing wide coverage and additional computing power. This allows BD tasks to be computed not only locally but also offloaded to UAVs equipped with edge servers, thus solving the problem of limited secondary user computing power.

[0032] Second, the present invention uses backscattering technology, which transmits data by reflecting and modulating incident radio frequency waves through backscattering equipment. Therefore, there is no need for backscattering devices to generate active radio frequency signals and perform analog-to-digital conversion, thereby reducing energy consumption;

[0033] Third, since the present invention ensures the security of the system by adding information about eavesdroppers eavesdropping on the communication link, it optimizes the target problem in the presence of multiple eavesdroppers eavesdropping on information, effectively improving the security of the system. At the same time, it combines the advantages of UAV, MEC and backscattering to further improve the system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A schematic diagram of a communication system model constructed in the present invention;

[0035] Figure 2 Flowchart for realizing the method of the present invention;

[0036] Figure 3 This is a diagram of the UAV trajectory simulation results of the present invention;

[0037] Figure 4 FIG1 is a diagram showing the flight speed simulation results of the UAV of the present invention;

[0038] Figure 5 This is a simulation result diagram of the impact of system bandwidth and time period on UAV energy consumption in the method of the present invention;

[0039] Figure 6 This is a simulation result diagram showing the impact of the fair confidentiality rate and the number of time slots on the energy consumption of the UAV in the method of the present invention;

[0040] Figure 7 This is a simulation result diagram showing the impact of UAV flight altitude and time slot number on UAV energy consumption in the method of the present invention; DETAILED DESCRIPTION

[0041] The following describes the implementation process of the technical solution of the present invention in detail with reference to the accompanying drawings:

[0042] Example 1: Reference Figure 2 The present invention provides a method for optimizing energy consumption in backscattering and edge computing based on drone assistance. The specific implementation steps are as follows:

[0043] Step 1: Reference Figure 1 , build a communication system consisting of K users BDs carrying backscatter devices, M eavesdroppers EDs, and a UAV equipped with an edge computing server, where K ≥ 1, M ≥ 1, m and k represent the mth ED and kth BD respectively;

[0044] Step 2: Assume that a UAV flies over a certain area in the air and collects information from multiple BDs. The BDs draw energy from the radio frequency signals transmitted by the UAVs, offloading part of the BD tasks to the UAV for edge computing and part for local computing. It is assumed that multiple EDs can hijack the information calculated locally by the BDs and the information backscattered by the BDs to the UAVs.

[0045] Step 3: Obtain the optimal optimization variables of the system built in step (1):

[0046] (3.1) Discretize the finite task completion time T into N equal time slots, and the duration of each time slot is where τ is sufficiently small; let the UAV fly at a fixed altitude H, where H is the minimum altitude to ensure that the UAV can avoid all obstacles in the service area; assume that the position of the UAV remains unchanged during each time slot and flies at a constant speed in each time slot, BD and ED are fixed on the ground respectively, let the channel between the UAV and BD be a line-of-sight communication link LOS, and obtain the channel power gain h between the UAV and the kth BD k [n], the channel power gain g between the kth BD and the mth ED km and the channel power gain z between the UAV and the mth ED m [n]; specifically calculated as follows:

[0047]

[0048]

[0049]

[0050] Where ρ0 is the channel power gain at a reference distance of 1 m, ||·|| is the Euclidean distance between a pair of vectors; q[n] = (x[n], y[n]) T represents the horizontal coordinate of the UAV at the nth time slot, v[n] represents the flight speed of the UAV at the nth time slot, n∈Θ N ,Θ N =(1,2,...,N);b k =(x k ,y k ) T and e m =(x m ,ym ) T Represent the coordinates of BD and ED fixed on the ground respectively.

[0051] (3.2) In order to avoid interference, each time slot is further divided into K equal durations using time division multiple access (TDMA). The kth BD reflects its task input data in the kth duration; calculates the local computing task amount of the kth BD in the nth time slot and the energy consumed by local computation Define the reflection coefficient of the kth BD at time slot n as α k [n], obtain the backscattering task volume I of the kth BD at time slot n k [n] and the total energy collected by the kth BD on the nth time slot during the reflection phase

[0052] Since the local calculation of BD does not require bandwidth and other resources, the local calculation and backscattering of each BD can be performed simultaneously. Then, in the nth time slot, the local calculation task of the kth BD is

[0053]

[0054] The energy consumed by the kth BD for local computation is

[0055]

[0056] where f k [n] is the CPU frequency of the kth BD, κ k is the effective capacitance coefficient of the kth BD, which depends on the chip architecture of its processor, C k It is the computing resources required to calculate 1 bit of data, that is, the number of CPU cycles required.

[0057] Part of the energy collected in the backscattering phase is used for backscattering, and the remaining energy is used for local calculation. In addition, this system has an adjustable backscattering coefficient for adaptively switching between backscattering mode and energy collection mode. Therefore, the reflection coefficient of the kth BD is defined as α k , then the backscattering task that can be achieved by the kth BD at time slot n is I k [n]:

[0058]

[0059] Among them, P k [n] is the UAV’s transmission power to the kth BD, σ 2represents the noise power at the UAV.

[0060] The energy collected when the kth BD backscatters at the nth time slot is

[0061]

[0062] The energy collected when the kth BD does not communicate with the UAV in the nth time slot is

[0063]

[0064] The total energy collected by the kth BD in the nth time slot during the reflection phase is obtained according to the following formula:

[0065]

[0066] in,

[0067] (3.3) The system confidentiality rate R is obtained according to the following formula ku [n]:

[0068] R ku [n]=min(R k [n]-R m,ED [n],0),

[0069] Among them, R k [n] is the rate of the backscatter channel between BD and UAV, R m,ED [n] is the information leakage rate,

[0070] The rate and information leakage rate of the backscatter channel between BD and UAV are calculated as follows:

[0071] In the nth time slot, the rate R of the backscatter channel between BD and UAV is k [n], the rate R of the eavesdropping channel m [n] are as follows:

[0072]

[0073] R m [n] = Blog2(1+γ m [n]),

[0074] in, is the noise power at ED;

[0075] The information leakage rate R is obtained according to the following formula m,ED [n]:

[0076] R m,ED [n] = Blog2(1+γ m,ED [n]),

[0077] Among them, γ m,ED [n]=max(γ1[n],γ2[n],…,γ m [n],…γ M [n]) is the equivalent signal-to-noise ratio of ED, is the signal-to-noise ratio of the mth eavesdropping channel.

[0078] (3.4) The UAV energy consumption E is obtained according to the following formula U [n]:

[0079]

[0080] Among them, w f is the UAV flight energy consumption weight, E fly 、 and They represent the flight energy consumption, computing energy consumption, and transmission energy consumption of the UAV, respectively, and are obtained according to the following steps:

[0081] (3.4.1) The flight energy consumption of UAV is obtained according to the following formula: fly :

[0082] E fly =P(||v[n]||)τ,

[0083] Where P(||v[n]||) is the UAV flight power, which is expressed as follows:

[0084]

[0085] Where v[n] is the velocity vector, P0 is used to represent the blade profile power in the UAV hovering state, P H Used to represent the blade induced power of UAV in hovering state, v0,U tip ,d0,π,g and A are used to represent the average blade induced speed, UAV rotor tip speed, UAV airframe drag ratio, air drag, UAV rotor stability and UAV rotor area in the hovering state, respectively.

[0086] (3.4.2) Dynamic voltage frequency scaling (DVFS) technology is used to improve the energy efficiency of UAV computing. The UAV computing energy consumption corresponding to the task of the kth BD at time slot n is obtained.

[0087]

[0088] Among them, f u,k [n] represents the CPU frequency of the UAV during the time period (cycles / second), κ u Represents the effective capacitance coefficient of the UAV, which depends on the chip architecture of its processor.

[0089] (3.4.3) BD and UAV communicate via TDMA, and the transmission energy consumption of UAV in the nth time slot is It is expressed as follows:

[0090]

[0091] Among them, P k [n] is the UAV’s transmit power.

[0092] (3.4) The UAV energy consumption E is obtained according to the following formula U [n]:

[0093]

[0094] Among them, w f is the UAV flight energy consumption weight, E fly 、 and They represent the flight energy consumption, computing energy consumption, and transmission energy consumption of the UAV respectively;

[0095] (3.5) Optimal UAV energy consumption expression:

[0096]

[0097] The optimization variable is the transmission power P of the UAV to the kth BD in the nth time slot. k [n], backscatter coefficient α of the kth BD in the nth time slot k [n], CPU frequency f of the kth BD k [n], UAV CPU frequency f in the nth time slot u,k [n] and the UAV trajectory q[n] at the nth time slot;

[0098] Assume that the communication system meets the following constraints:

[0099] This constraint is used to ensure that the collected energy is sufficient for BD use; This constraint is used to ensure that all tasks are calculated. is the minimum task amount of BDk; This constraint is used to ensure that the amount of tasks that the UAV can calculate within the time period T is less than the amount of tasks reflected by the BD; This constraint indicates that the amount of tasks calculated by the UAV must be greater than or equal to the total amount of tasks offloaded by the BD; R k [n]-R m,ED [n]≥R fair , this constraint is the security and confidentiality rate constraint; This constraint is the UAV CPU clock frequency constraint; 0≤f k [n]≤f m ' ax , which represents the CPU clock frequency constraint of BD; 0≤α k [n]≤1, this constraint is the reflection coefficient constraint; 0≤P k [n]≤P max , which is the UAV maximum transmission power constraint; q[0] = q I ,q[N]=q F , which determines the starting and ending points of the UAV’s flight; ||q[n]-q[n-1]||≤τV max ,This constraint is the speed constraint of the UAV within a time slot.,In the above constraint, i represents the summation variable of the time slot number;

[0100] (3.6) Minimize UAV energy consumption E U [n], get the optimal optimization variable:

[0101] (3.6.1) Fix the UAV trajectory among the optimization variables and optimize the remaining variables:

[0102] By introducing the auxiliary variable β k [n] Process the coupling, and then use the Lagrangian dual decomposition algorithm and the subgradient algorithm to solve the convex expression to obtain the minimum UAV energy consumption The optimal auxiliary variable corresponding to the nth time slot The optimal CPU frequency of the kth BD in the nth time slot Optimal CPU frequency of UAV in the nth time slot The optimal transmission power of UAV to the kth BD The optimal backscatter coefficient for the nth time slot

[0103] (3.6.2) Optimizing UAV trajectories:

[0104] Fix the optimization variables obtained in step (3.6.1), and then use the continuous convex approximation algorithm to transform the non-convex problem into a convex problem for solution, and finally obtain the optimal UAV trajectory q for the nth time slot * [n];

[0105] Step 4. Set the system operating parameters according to the optimal optimization variables, so that the system operates under the operating parameters to achieve system energy consumption optimization.

[0106] Example 2: The overall implementation steps of this example are the same as those of Example 1. The process of minimizing the energy consumption of the UAV is described in further detail as follows:

[0107] The present invention minimizes the energy consumption of UAV E U [n], divided into two sub-problems to solve:

[0108] (1) Fix the UAV trajectory in the optimization variable in step (3.6.1) of Example 1 and optimize the remaining variables as subproblem 1: By introducing the variable β k [n] Process the coupling, and then use the Lagrangian dual decomposition algorithm and the subgradient algorithm to solve the convex expression to obtain the minimum UAV energy consumption The corresponding optimal optimization variable, β k [n], BD and UAV CPU frequencies, UAV transmit power, and BD backscatter coefficient.

[0109] Introducing Beta k [n]=P k [n]·α k [n] to handle the coupling variables, then there is a constraint 0≤β k [n]≤P k [n], constraint R k [n]-R m,ED [n]≥R fair Can be converted into

[0110] Given q, the problem becomes convex and can be solved using standard convex optimization tools. In engineering, the Lagrangian dual algorithm can be used to solve it. Its Lagrangian dual function is:

[0111]

[0112] The corresponding dual function is:

[0113]

[0114] The corresponding dual problem is:

[0115]

[0116] Using dual decomposition theory, decompose the problem:

[0117] a、For any k,n, for β k [n], the subproblem can be written as

[0118]

[0119] Using the KKT condition, we can get β k The optimal solution for [n] for

[0120]

[0121] Where n∈{1,2,…,N-1}.

[0122] b. For any k,n, corresponding to f u,k [n], the sub-problem is

[0123]

[0124] Using the KKT condition, we can get f u,k The optimal solution for [n] for

[0125]

[0126] Where n∈{2,3,…,N}.

[0127] c. For any k,n, for f k [n], the sub-problem is

[0128]

[0129] Using the KKT condition, we can get f k The optimal solution for [n] for

[0130]

[0131] d. For any k, n, corresponding to P k [n], the sub-problem is

[0132]

[0133] Since the target is user k, the above sub-problem can be written as

[0134]

[0135] Solve with the help of convex optimization toolkit and get the optimal solution

[0136] e. For any k,n, corresponding to α k [n], the sub-problem is

[0137]

[0138] Solve with the help of convex optimization toolkit and get the optimal solution

[0139] Get the optimal solution and Finally, since the Salter condition is satisfied, the dual gap between the dual problem and the original problem is zero. The subgradient algorithm can be used to iteratively optimize the dual variables, solve the dual problem, and obtain the inequality constraint The related optimal dual variable, the dual variable obtained at the j+1th iteration is expressed as:

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147] in,

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154]

[0155] (2) Optimizing the UAV trajectory in step (3.6.2) of Example 1 is considered as sub-problem 2: After solving the optimization variables of sub-problem 1, optimize sub-problem 2. Sub-problem 2 is made convex by introducing slack variables and using a continuous convex approximation algorithm, and then solved using a convex optimization toolkit.

[0156] Obviously, subproblem 2 is a non-convex problem. Therefore, the SCA algorithm is used to optimize the non-convex terms in the problem to obtain its approximate convex expression. First, for the non-convex term P(||v[n]||) in the objective function, variables v1[n] and v2[n] are introduced to satisfy v1[n]≥||v[n]||. Then there is

[0157]

[0158] Using the SCA algorithm, given any feasible solution v of v1[n],v2[n] 1,j [n],v 2,j [n], the non-convex constraint above is approximated by the following convex constraint:

[0159]

[0160] in Thus, the non-convex term P(||v[n]||) is approximately replaced by the following convex term Constrained and Non-convex to q[n], but convex to ||q[n]-b k || 2 The whole function is a convex function. Based on this, using the SCA algorithm, any feasible point q given q[n] (j) [n], then the second constraint becomes:

[0161]

[0162] Replace constraint three with:

[0163]

[0164] in,

[0165]

[0166] Introduce slack variables so that

[0167]

[0168]

[0169] Performing a first-order Taylor expansion on the right side of the equation yields

[0170]

[0171] There are constraints

[0172] H 2 +||q[n]-b k || 2 ≤Ω k,j [n]

[0173] Therefore, subproblem 2 becomes a convex problem, which can be solved using the convex optimization toolkit, and finally the optimal UAV trajectory is obtained.

[0174] The effect of the present invention is further described below in conjunction with simulation experiments:

[0175] A. Simulation Conditions

[0176] Computer simulation software is used for simulation. Unless otherwise specified, the simulation parameters used in this section are as follows: UAV starts from the starting point q I =[0,0] to the end point q F =[30,50] flight, given the position distribution of BD and ED, the rest of the simulation parameters are summarized in Table 1.

[0177] Table 1 Simulation parameters

[0178]

[0179]

[0180] B. Simulation Content

[0181] Simulation 1: UAV trajectory simulation of the present invention, the simulation results are as follows Figure 3 As shown;

[0182] Simulation 2: UAV flight speed simulation of the present invention, the simulation results are as follows Figure 4 As shown;

[0183] Simulation 3: The impact of system bandwidth and time period on UAV energy consumption in the method of the present invention. The simulation results are as follows: Figure 5 As shown;

[0184] Simulation 4: The impact of the fair confidentiality rate and the number of time slots on the energy consumption of UAV in the method of the present invention. The simulation results are as follows: Figure 6 As shown;

[0185] Simulation 5: The impact of UAV flight altitude and time slot number on UAV energy consumption in the method of the present invention. The simulation results are as follows: Figure 7 As shown;

[0186] C. Simulation Results

[0187] Depend on Figure 3 As can be seen, as the time period T increases, the UAV will fly along the longest possible trajectory from the starting point to the end point, and this trajectory is the optimal path selected by the system. Conversely, as the time period T decreases, the UAV will fly along the shortest possible trajectory from the starting point to the end point.

[0188] Depend on Figure 4As can be seen, when the horizontal coordinates of the UAV's starting and ending points remain constant, the UAV's flight speed tends to increase with the increase in time period T. This is because in the sixth time slot, the distance between the UAV and the BD is the shortest, the path loss is the lowest, and the UAV will stay at the BD longer to transmit data. Therefore, the UAV's flight speed is the lowest.

[0189] Depend on Figure 5 As can be seen, as system bandwidth increases, UAV energy consumption decreases. With increased system bandwidth, ground users can backscatter more data to the UAV, improving the system's task computation rate and reducing communication resources, thus reducing UAV energy consumption. As Shannon's equation shows, increased bandwidth provides ground users and UAVs with more decision-making space regarding power consumption. Therefore, UAV energy consumption decreases. As the time period increases, the UAV spends longer in the user area and the flight time increases. Consequently, UAV energy consumption increases.

[0190] Depend on Figure 6 As can be seen, as the number of time slots N increases, the UAV spends more time in the BD area, thus increasing UAV energy consumption. Conversely, as the number of time slots N decreases, the UAV spends less time in the BD area, thus decreasing UAV energy consumption. As the fair confidentiality ratio increases, each ED individually decodes fewer messages, strengthening the security of system communications and placing greater demands on the entire system, thus increasing UAV total energy consumption. Conversely, as the fair confidentiality ratio decreases, each ED individually decodes more messages, weakening system communications security and reducing demands on the entire system, thus reducing UAV total energy consumption. The above simulation analysis demonstrates the correctness and effectiveness of the proposed method.

[0191] Depend on Figure 7 As can be seen, as the UAV altitude increases, the channel length between the UAV and the ground BD increases, path loss increases, and UAV energy consumption increases. Conversely, as the UAV altitude decreases, the channel length between the UAV and the ground BD decreases, path loss decreases, and UAV energy consumption decreases. As the number of time slots N increases, the UAV spends more time in the BD area, thereby increasing UAV energy consumption. Conversely, as the number of time slots N decreases, the UAV spends less time in the BD area, thereby decreasing UAV energy consumption.

[0192] The above simulation analysis proves the correctness and effectiveness of the method proposed in the present invention.

[0193] Parts of the present invention that are not described in detail belong to common knowledge among those skilled in the art.

[0194] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Obviously, for professionals in this field, after understanding the content and principles of the present invention, they may make various modifications and changes in form and details without departing from the principles and structure of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing energy consumption of MEC communication based on drones and backscatter assistance, characterized in that: The steps include: (1) Build a communication system consisting of K users (BDs) carrying backscatter devices, M eavesdroppers (EDs), and a UAV (UAV) equipped with an edge computing server, where K ≥ 1, M ≥ 1, and m and k represent the mth ED and kth BD, respectively. (2) Assume that a UAV flies in a certain area in the air and collects information from multiple BDs. The BD obtains energy from the radio frequency signal transmitted by the UAV, offloads part of the BD task to the UAV for edge computing, and performs part of the task locally. It is assumed that multiple EDs will hijack the information calculated locally by the BD and the information backscattered by the BD to the UAV. (3) Obtain the optimal optimization variables of the system built in step (1): (3.1) Discretize the finite task completion time T into N equal time slots, and the duration of each time slot is Let the UAV fly at a fixed altitude H, where H is the minimum altitude to ensure that the UAV can avoid all obstacles in the service area; assume that the position of the UAV remains unchanged during each time slot and flies at a constant speed in each time slot, BD and ED are fixed on the ground respectively, let the channel between the UAV and BD be a line-of-sight communication link LOS, and obtain the channel power gain h between the UAV and the kth BD k [n], the channel power gain g between the kth BD and the mth ED km and the channel power gain z between the UAV and the mth ED m [n]; (3.2) Use time division multiple access (TDMA) to further divide each time slot into K equal durations. The kth BD reflects its task input data in the kth duration; calculates the local computing task amount of the kth BD in the nth time slot and the energy consumed by local computation Define the reflection coefficient of the kth BD at time slot n as α k [n], obtain the backscattering task volume I of the kth BD at time slot n k [n] and the total energy collected by the kth BD on the nth time slot during the reflection phase (3.3) The system confidentiality rate R is obtained according to the following formula ku [n]: R ku [n]=min(R k [n]-R m,ED [n],0), Among them, R k [n] is the rate of the backscatter channel between BD and UAV, R m,ED [n] is the information leakage rate, (3.4) The UAV energy consumption E is obtained according to the following formula U [n]: Among them, w f is the UAV flight energy consumption weight, E fly 、 and They represent the flight energy consumption, computing energy consumption, and transmission energy consumption of the UAV respectively; (3.5) Optimal UAV energy consumption expression: The optimization variable is the transmission power P of the UAV to the kth BD in the nth time slot. k [n], backscatter coefficient α of the kth BD in the nth time slot k [n], CPU frequency f of the kth BD k [n], UAV CPU frequency f in the nth time slot u,k [n] and the UAV trajectory q[n] at the nth time slot; Assume that the communication system meets the following constraints: This constraint is used to ensure that the collected energy is sufficient for BD use; This constraint is used to ensure that all tasks are calculated. is the minimum task amount of BDk; This constraint is used to ensure that the amount of tasks that the UAV can calculate within the time period T is less than the amount of tasks reflected by the BD; This constraint indicates that the amount of tasks calculated by the UAV must be greater than or equal to the total amount of tasks offloaded by the BD; R k [n]-R m,ED [n]≥R fair , this constraint is the security and confidentiality rate constraint; This constraint is the UAV CPU clock frequency constraint; 0≤f k [n]≤f m ' ax , which represents the CPU clock frequency constraint of BD; 0≤α k [n]≤1, this constraint is the reflection coefficient constraint; 0≤P k [n]≤P max , which is the UAV maximum transmission power constraint; q[0] = q I ,q[N]=q F , which determines the starting and ending points of the UAV’s flight; ||q[n]-q[n-1]||≤τV max ,This constraint is the speed constraint of the UAV within a time slot.,In the above constraint, i represents the summation variable of the time slot number; (3.6) Minimize UAV energy consumption E U [n], get the optimal optimization variable: (3.6.1) Fix the UAV trajectory among the optimization variables and optimize the remaining variables: By introducing the auxiliary variable β k [n] Process the coupling, and then use the Lagrangian dual decomposition algorithm and the subgradient algorithm to solve the convex expression to obtain the minimum UAV energy consumption The optimal auxiliary variable corresponding to the nth time slot The optimal CPU frequency of the kth BD in the nth time slot Optimal CPU frequency of UAV in the nth time slot The optimal transmission power of UAV to the kth BD The optimal backscatter coefficient for the nth time slot (3.6.2) Optimizing UAV trajectories: Fix the optimization variables obtained in step (3.6.1), and then use the continuous convex approximation algorithm to transform the non-convex problem into a convex problem for solution, and finally obtain the optimal UAV trajectory q for the nth time slot * [n]; (4) Set the system operating parameters according to the optimal optimization variables, so that the system operates under the operating parameters to achieve system energy consumption optimization.

2. The method according to claim 1, wherein: The channel power gain h between the UAV and the kth BD in step (3.1) is k [n], the channel power gain g between the kth BD and the mth ED km and the channel power gain z between the UAV and the mth ED m [n] is calculated as follows: Where ρ0 is the channel power gain at a reference distance of 1 m, ||·|| is the Euclidean distance between a pair of vectors; q[n] = (x[n], y[n]) T represents the horizontal coordinate of the UAV at the nth time slot, v[n] represents the flight speed of the UAV at the nth time slot, n∈Θ N ,Θ N =(1,2,...,N);b k =(x k ,y k ) T and e m =(x m ,y m ) T Represent the coordinates of BD and ED fixed on the ground respectively.

3. The method according to claim 1, wherein: In step (3.2), in the nth time slot, the local computing task amount of the kth BD is Energy consumed by local computing The backscattering task volume I of the kth BD k [n] and the total energy collected by the kth BD during the reflection phase The formula is as follows: Among them, f k [n] represents the CPU frequency of the kth BD, κ k represents the effective capacitance coefficient of the kth BD, C k The computing resources required to calculate 1 bit of data, that is, the number of CPU cycles required; represents the energy collected during backscattering, represents the energy collected when not communicating with the UAV, P k [n] is the transmission power of UAV to the kth BD, σ 2 represents the noise power at the UAV, ii represents the summation variable of the number of BDs, and η represents the energy harvesting efficiency.

4. The method according to claim 1, wherein: The rate and information leakage rate of the backscatter channel between BD and UAV in step (3.3) are calculated as follows: In the nth time slot, the rate R of the backscatter channel between BD and UAV is k [n], the rate R of the eavesdropping channel m [n] are as follows: R m [n]=Blog2(1+γ m [n]), Among them, σ 2 represents the noise power at the UAV; The information leakage rate R is obtained according to the following formula m,ED [n]: R m,ED [n]=Blog2(1+γ m,ED [n]), Among them, γ m,ED [n]=max(γ1[n],γ2[n],…,γ m [n],…γ M [n]) is the equivalent signal-to-noise ratio of ED, is the signal-to-noise ratio of the mth eavesdropping channel, is the noise power at ED.

5. The method according to claim 1, wherein: The flight energy consumption, computing energy consumption, and transmission energy consumption of the UAV in step (3.4) are obtained according to the following steps: (3.4.1) The flight energy consumption of UAV is obtained according to the following formula: fly : AND fly =P(||v[n]||)τ, Where P(||v[n]||) is the UAV flight power; (3.4.2) Dynamic voltage and frequency adjustment technology is used to improve the energy efficiency of UAV computing, and the UAV computing energy consumption corresponding to the task of the kth BD at time slot n is obtained. Among them, f u,k [n] represents the CPU frequency of the UAV during the time period, κ u represents the effective capacitance coefficient of UAV; (3.4.3) BD and UAV communicate via TDMA, and the transmission energy consumption of UAV in the nth time slot is It is expressed as follows: Among them, P k [n] is the transmit power of the UAV.

6. The method according to claim 5, characterized in that: The UAV flight power described in step (3.4.1) is as follows: Where v[n] is the velocity vector, P0 is used to represent the blade profile power in the UAV hovering state, P H Used to represent the blade induced power of UAV in hovering state, v0,U tip ,d0,π,g and A are used to represent the average blade induced speed, UAV rotor tip speed, UAV airframe drag ratio, air drag, UAV rotor stability and UAV rotor area in the hovering state, respectively.

Images