Method for maximizing the completion rate of UAV backscatter mission

Abstract

The present invention discloses a method for maximizing the completion rate of backscattering tasks of unmanned aerial vehicles, which mainly solves the problem that the requirements of intensive computing and energy consumption cannot be met simultaneously in diversified Internet of Things application scenarios. It includes: 1) building a backscattering and edge computing communication system assisted by unmanned aerial vehicles; 2) dividing the target area and indicating the centroid position of each sub-area; 3) the user uses the received radio frequency signal to backscatter the task data to be processed to the MEC server of the unmanned aerial vehicle, and unloads the remaining task data after backscattering to the unmanned aerial vehicle through active transmission; 4) obtaining the total energy consumption of the unmanned aerial vehicle and the total energy consumption of the user equipment during the operation of the unmanned aerial vehicle in the entire target area; 5) minimizing energy consumption and maximizing the user task completion rate to obtain the optimal optimization variable. The present invention can effectively solve the problems of limited equipment energy and large computing delay in actual complex scenarios, so that the network performance is optimal, and can be used in edge computing systems.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication technology, and relates to edge computing and backscatter communication technology, and specifically is a method for maximizing the completion rate of unmanned aerial vehicle backscattering tasks, which can be used in edge computing systems with energy-saving requirements. Background Art

[0002] With the development of IoT technology and the advent of edge intelligence, emerging diversified applications such as face recognition, virtual reality, smart cities, and some lightweight machine learning tasks are changing people's lives. These applications have put forward higher and higher demands on the computing power of resource-constrained IoT mobile devices. To solve this problem, mobile edge computing MEC (Mobile edge computing) technology is proposed to enhance computing power. It deploys computing resources at the edge of the network to provide computing offload services for IoT devices to reduce computing latency and energy consumption. Some existing studies assume that computing resources are deployed in a fixed manner and cannot reduce the waiting time and energy consumption of terminal devices through mobility. In addition, the explosive growth of mobile data traffic or the sparse distribution of network facilities further brings challenges to MEC technology.

[0003] Due to the characteristics of mobility, flexibility and operability, UAVs (Unmanned Aerial Vehicles) can play an important role in wireless communication systems. The UAV-assisted communication system has the advantages of wide coverage, high mobility, and most air-to-ground communications are line-of-sight links, providing reliable connections for devices in areas with limited infrastructure. Therefore, a UAV-assisted MEC communication system is proposed, in which the UAV is equipped with computing resources to provide offload services for nearby user devices. In addition, the UAV can be close to the terminal device, improve the connectivity of devices in areas with sparse network facilities, and support more flexible mobile computing.

[0004] In view of the limited battery capacity of low-power IoT devices, a wireless power communication network is proposed to solve the energy supply problem. The network follows the HTT (Harvest-then-Transmit) method. The specific process is: the energy device transmits a radio frequency RF (Radio Frequency) signal to the IoT device, which is equipped with an RF component to obtain energy and use the collected energy for wireless information transmission. This process is active communication, and the active RF components integrated into the device still consume a lot of energy for data forwarding. In order to meet the needs of transmission and reduce energy consumption, backscatter communication BackCom (Backscatter Communication) technology is considered to be one of the effective solutions to the above problems. BackCom is a passive communication technology, and the sending device does not need to be equipped with active RF components. The communication process is to modulate the data transmitted by the device onto the received carrier signal to complete the data transmission. In addition, the circuit power consumption of the BackCom network is low and can be supported by the energy it collects.

[0005] Yeh S, Wang Y, Perera TDP et al. used wireless power transmission technology to power IoT devices in their paper "UAV trajectory optimization for data gathering from backscattering sensor networks" (IEEE International Conference on Communications (ICC). IEEE, 2020: 1-6). The device reflects the incident RF signal to the receiving end. In this process, the reflection circuit has low power consumption and is proven to be beneficial to energy saving. SHI L, YE Y, CHU X et al. considered the case of a drone-assisted terminal device unloading bits in their paper "Computation bits maximization in a backscatter assisted wirelessly powered MEC network" (IEEE Wireless Communications Letters, 2020: 1–1). The terminal device can not only calculate locally, but also unload task data to the drone. The above two systems are respectively combining drones with backscatter and drones with edge computing. However, for diverse IoT communication systems, the above models cannot meet the requirements of intensive computing and energy consumption at the same time. Summary of the invention

[0006] The purpose of the present invention is to propose a method for maximizing the completion rate of UAV backscattering tasks in view of the deficiencies of the above-mentioned prior art. Combining the advantages of BackCom and HTT technologies in power consumption and data transmission rate, in the scenario of edge computing with hybrid backscattering and active transmission assisted by UAV, by satisfying the constraints of computing resources and energy consumption of UAVs and users, the time allocation of each target area, the reflection coefficient of the user, and the transmission power and computing power allocation during active transmission are found; the weighted total energy consumption of users and UAVs is minimized while ensuring the maximum completion rate of the task, so that users can meet the requirements of intensive computing and energy consumption.

[0007] The basic idea of ​​implementing the present invention is: divide the target area into several sub-areas, in each sub-area, the UAV hovers above the centroid of the area, and acts as a transmitter to provide RF energy to all user devices; the user is equipped with a backscattering device, and the amplitude and phase of the received RF signal are changed by adjusting the load impedance, and the RF signal carries the user data and sends it to the UAV equipped with an edge computing server. After the UAV completes the calculation, the calculation result is returned to the user; the user can also use the remaining energy for active transmission; in addition, the user can also perform local calculations; after completing the task in one sub-area, the UAV flies to the next sub-area.

[0008] To achieve the above object, the technical solution of the present invention comprises the following steps:

[0009] (1) Using a UAV equipped with an edge computing server and a carrier transmitter and multiple local users equipped with backscattering devices, a backscattering and edge computing communication system is built; wherein the UAV and each local user are equipped with a single antenna;

[0010] (2) Divide the target area into N sub-areas, let Θ i represents the i-th sub-region, i∈{1,2,…,N}; M i represents the number of users in the i-th sub-area; j = 1, 2, ..., M i represents the jth user in the i-th sub-area;

[0011] (3) Obtain the total energy consumption E of the drone during its operation in the entire target area UAV And the total energy consumption of user equipment E User , the implementation steps are as follows:

[0012] (3.1) The drone is hovered above the centroid of the i-th sub-area and transmits RF signals to all users in the area at a specific power to obtain the initial energy of user j after the signal is transmitted through the line-of-sight link.

[0013] (3.2) The UAV continues to transmit RF signals to all users in the current sub-area at a specific power. The users use the received RF signals to backscatter the task data to be processed to the mobile edge computing MEC server of the UAV using time division multiple access, and convert the signals into energy for storage through line-of-sight link transmission in other time slots where backscattering is not performed. Among them, the circuit energy consumption generated by user j during the backscattering process is The number of bits of pending task data offloaded to the UAV via backscattering in the allocated time slot is The energy converted and stored is

[0014] (3.3) The user unloads the remaining mission data after backscattering to the UAV through active transmission. Let the transmission power of user j in this active transmission process be P j [i], the number of bits of remaining task data to be offloaded is The total energy consumption is The task volume and energy consumption of local computing performed by user j are and

[0015] (3.4) The drone carrying the mobile edge computing MEC server calculates all the task data, completes the computing task of the current sub-area, and records the amount of completed tasks

[0016] (3.5) After completing the computational task in the current sub-area, the drone flies to the next sub-area until it completes the tasks in all sub-areas. That is, take i = 1, 2, …, N, and get the total energy consumption E of the drone. UAV And the total energy consumption of user equipment E User ;

[0017] (4) Construct the optimal user task completion rate and total energy consumption weighted sum expression G':

[0018]

[0019] Where, Δ=[t e ',t b ',t a ',t c ',T'] represents the time allocation of each stage in the sub-region, is the duration of the energy collection phase, is the duration of the energy backscattering phase, is the duration of the active transmission phase, Calculate the duration for the drone, is the hovering time of the drone in each area; is the backscattering coefficient, is the user's transmit power during active transmission, The computing power allocation for users during local computing; is the user's task completion rate; the user's task completion rate and the weighted sum of total energy consumption G are obtained according to the following formula:

[0020]

[0021] Among them, ω represents the weight coefficient of total energy consumption, Represents the weight coefficient of UAV energy consumption, μ j [i] represents the task completion rate of user j in the i-th region;

[0022] Set the constraints as follows:

[0023] Used to ensure that each user's energy consumption does not exceed the energy collected by the user; L j [i] represents the number of task bits given by each user; t e [i]+t b [i]+t a [i]+t c [i]≤T[i], 0≤α j [i]≤1,0≤P j [i]≤P max ,in P max They represent the maximum available CPU frequency and maximum transmit power of the user, respectively, and Z represents the estimated total working time;

[0024] (5) Minimize the weighted sum of total energy consumption G under the backscattering and edge computing communication system, and convert the G expression into a convex expression through variable substitution, perspective function and continuous convex approximation algorithm;

[0025] (6) By using standard convex optimization tools, the optimal solution of the user task completion rate and the weighted sum of total energy consumption G is obtained, that is, the optimal time allocation Δ * , optimal reflection coefficient α * , optimal transmission power P * And the optimal computing power allocation f * .

[0026] Compared with the prior art, the present invention has the following advantages:

[0027] First, since the present invention adopts edge computing technology, by deploying computing resources at the edge of the network, users can offload computing tasks to edge servers with stronger computing power, thereby solving the problem of limited user computing power; and the UAV-assisted communication system has the advantages of wide coverage, high mobility, and most air-to-ground communications are line-of-sight links, providing reliable connections for equipment in limited infrastructure areas, and also adopts a UAV-assisted MEC system to support more flexible computing.

[0028] Second, in real scenarios, the battery capacity of low-power terminal devices is limited, which cannot meet the user's requirements for computing rate, time delay, etc., and it is difficult to solve practical problems. Compared with the existing methods, the present invention, on the basis of introducing edge computing, considers backscattering and collect-first-then-transmit technology, and combines the advantages of backscattering and collect-first-then-transmit technology in energy consumption and data transmission rate. In the UAV-assisted multi-user edge computing scenario, with the help of hybrid backscattering and active transmission technology, it effectively solves the problems of limited energy and large computing delay of user devices in actual complex scenarios, so that users can meet the requirements of intensive computing and energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Schematic diagram of application scenario of the method of the present invention;

[0030] Figure 2 It is a schematic diagram of the time slot structure of the present invention;

[0031] Figure 3 Flow chart for realizing the method of the present invention;

[0032] Figure 4 It is a simulation result diagram of the influence of communication bandwidth B and user's maximum transmission power on energy consumption in the method of the present invention; wherein (a), (b), and (c) are simulation result diagrams of the influence of communication bandwidth and user's maximum transmission power on user energy consumption, UAV energy consumption and user's average task completion rate in the method, respectively.

[0033] Figure 5 (a), (b), and (c) are the simulation results of the influence of the UAV weight coefficient ψ and the total energy consumption weight coefficient ω on the user energy consumption, UAV energy consumption, and the user's average task completion rate in this method;

[0034] Figure 6 This is a simulation result diagram of the impact of the method of the present invention and the benchmark solution on the final weighted energy consumption and the overall task completion rate when the task volume L is different. DETAILED DESCRIPTION

[0035] The implementation process of the technical solution of the present invention is described in detail below with reference to the accompanying drawings:

[0036] Example 1: Reference Figure 1 , a schematic diagram of the system model in the present invention, the system of the present invention is a backscattering and edge computing communication system composed of an unmanned aerial vehicle (UAV) equipped with an edge computing server and a carrier transmitter and multiple users equipped with backscattering devices, wherein the UAV and each user are equipped with a single antenna, and the target area is divided into multiple sub-areas.

[0037] Example 2: Reference Figure 2 , the time slot structure diagram of the present invention. In the present invention, the target area is divided into N sub-areas, let Θ i represents the i-th sub-region, i∈{1,2,…,N}; M i represents the number of users in the i-th sub-area; j = 1, 2, ..., M i represents the jth user in the i-th sub-area; then the entire working time Z of the drone consists of N time slots, let T[i] represent the hovering time of the drone in sub-area i; each time slot is divided into four parts: let t e [i] represents the time of the energy collection phase, t b [i] represents the time of the backscattering phase, t a [i] represents the time of active transmission phase, t c [i] represents the time of UAV calculation. Since the size of the calculation result from UAV is much smaller than the size of the offload task from the user, the delay caused by data download from UAV to the user is ignored. In order to reduce the conflict between all users, time division multiple access is used for data offloading.

[0038] Example 3: Reference Figure 3 , the present invention provides a method for maximizing the completion rate of a UAV backscattering task. The specific implementation steps are as follows:

[0039] Step 1: Use a drone equipped with an edge computing server and a carrier transmitter and multiple local users equipped with backscatter devices to build a backscatter and edge computing communication system; the drone and each local user are equipped with a single antenna;

[0040] Step 2: Divide the target area into N sub-areas, let Θ i represents the i-th sub-region, i∈{1,2,…,N}; M i represents the number of users in the i-th sub-area; j = 1, 2, ..., M i represents the jth user in the i-th sub-area;

[0041] Step 3: Obtain the total energy consumption E of the drone during its operation in the entire target area UAV And the total energy consumption of user equipment E User , the implementation steps are as follows:

[0042] (3.1) The drone is hovered above the centroid of the i-th sub-area and transmits RF signals to all users in the area at a specific power to obtain the initial energy of user j after the signal is transmitted through the line-of-sight link. The implementation is as follows:

[0043] The communication link is dominated by the line-of-sight link, and the channel power gain between the UAV and the jth user in the i-th sub-area is h j [i]:

[0044]

[0045]

[0046] Among them, ρ 0 represents the channel power gain when the reference distance is 1 meter, H represents the minimum altitude of the drone above the highest obstacle in the service area, η represents the energy collection efficiency, P 0 Indicates the UAV transmit power.

[0047] (3.2) The UAV continues to transmit RF signals to all users in the current sub-area at a specific power. The users use the received RF signals to backscatter the task data to be processed to the mobile edge computing MEC server of the UAV using time division multiple access, and convert the signals into energy for storage through line-of-sight link transmission in other time slots where backscattering is not performed. Among them, the circuit energy consumption generated by user j during the backscattering process is The number of bits of pending task data offloaded to the UAV via backscattering in the allocated time slot is The energy converted and stored is

[0048] Specifically, according to the following formula and

[0049]

[0050]

[0051]

[0052] Where B is the communication bandwidth, ξ is the performance difference between backscattering and active transmission, and α j [i] represents the reflection coefficient of the jth user, P cb Represents constant circuit consumption during backscattering.

[0053] (3.3) The user unloads the remaining mission data after backscattering to the UAV through active transmission. Let the transmission power of user j in this active transmission process be Pj [i], the number of bits of remaining task data to be offloaded is The total energy consumption is The task volume and energy consumption of local computing performed by user j are and The details are as follows:

[0054]

[0055]

[0056]

[0057]

[0058] Among them, P j [i] represents the transmission power of user j, P ca represents the constant power consumption of the circuit in the active transmission phase, The CPU frequency for each user, C j [i] is the amount of computing resources required to calculate 1 bit of data, κ j [i] is the effective capacitance coefficient of the user device.

[0059] (3.4) The drone carrying the mobile edge computing MEC server calculates all the task data, completes the computing task of the current sub-area, and records the amount of completed tasks

[0060]

[0061] Among them, f u [i] indicates the CPU frequency of the drone.

[0062] (3.5) After completing the computational task in the current sub-area, the drone flies to the next sub-area until it completes the tasks in all sub-areas. That is, take i = 1, 2, …, N, and get the total energy consumption E of the drone. UAV And the total energy consumption of user equipment E User ; They are respectively represented as follows:

[0063]

[0064] Among them, κ u is the effective capacitance coefficient of the UAV, P h is the power consumed by the drone during hovering, P f is the power consumed by the drone during flight, T f It's time for the drone to fly.

[0065] Step 4: Construct the optimal user task completion rate and total energy consumption weighted sum expression G':

[0066]

[0067] Where, Δ=[t e ',t b ',t a ',t c ',T'] represents the time allocation of each stage in the sub-region, is the duration of the energy collection phase, is the duration of the energy backscattering phase, is the duration of the active transmission phase, Calculate the duration for the drone, is the hovering time of the drone in each area; is the backscattering coefficient, is the user's transmit power during active transmission, The computing power allocation for users during local computing; is the user's task completion rate; the user's task completion rate and the weighted sum of total energy consumption G are obtained according to the following formula:

[0068]

[0069] Among them, ω represents the weight coefficient of total energy consumption, Represents the weight coefficient of UAV energy consumption, μ j [i] represents the task completion rate of user j in the i-th region;

[0070] Set the constraints as follows:

[0071] Used to ensure that each user's energy consumption does not exceed the energy collected by the user; L j [i] represents the number of task bits given by each user; t e [i]+t b [i]+t a [i]+t c [i]≤T[i], 0≤α j [i]≤1,0≤P j [i]≤P max ,in P max They represent the maximum available CPU frequency and maximum transmit power of the user, respectively, and Z represents the estimated total working time;

[0072] Step 5: Minimize the weighted sum G of total energy consumption under the backscattering and edge computing communication system, and transform the G expression into a convex expression through variable substitution, perspective function and continuous convex approximation algorithm;

[0073] Step 6: By using standard convex optimization tools, we can obtain the optimal solution of the user task completion rate and the weighted sum of total energy consumption G, i.e., the optimal time allocation Δ * , optimal reflection coefficient α * , optimal transmission power P * And the optimal computing power allocation f * .

[0074] The effect of the present invention is further described below in conjunction with simulation experiments:

[0075] A. Simulation Conditions

[0076] In the network considered by the present invention, all users are located in a 50×50 two-dimensional area, and the target area is grouped into several sub-areas. There are several methods for particle clustering, such as mean shift, K-means, and density-based spatial clustering. The present invention divides all users equally into N areas. The rest of the simulation parameters are summarized in Table 1.

[0077] Table 1 Simulation parameters

[0078]

[0079]

[0080] B. Simulation Content

[0081] Simulation 1: The simulation results of the impact of different communication bandwidths and maximum transmission power of drones. A is the simulation of user energy consumption, B is the simulation of drone energy consumption, and C is the simulation of the user's average task completion rate. The simulation results are shown in Figure 1. Figure 4 As shown;

[0082] Simulation 2: The simulation results of the impact of different drone energy consumption weight coefficients and total energy consumption weight coefficients. Among them, a drone energy consumption weight coefficient and total energy consumption weight coefficient are simulations of user energy consumption, b is a simulation of drone energy consumption, and c is a simulation of the user's average task completion rate. The simulation results are shown in Figure 5 As shown;

[0083] Simulation 3: The impact of the proposed method and the benchmark solution on the final weighted energy consumption and the overall task completion rate when the task volume L is different. The simulation results are as follows: Figure 6 shown.

[0084] C. Simulation Results

[0085] Depend on Figure 4 It can be seen from Figure (a) that when the user's transmission power increases, the user's own energy consumption will also increase; at the same time, the increase in bandwidth will lead to an increase in the optimized user's transmission power, and the user's energy consumption will also tend to rise. In Figure (b), the energy consumption of the drone decreases with the increase of the transmission power. This is because under the same task volume, the increase in transmission power will shorten the time the drone hovers above the area, and the hovering energy consumption will be reduced; in Figure (c), the user's average task completion rate will increase with the increase of transmission power, and when the bandwidth increases, the transmission rate when the user unloads the task increases, the transmission task volume increases, and the task completion rate also increases accordingly.

[0086] Depend on Figure 5 (a), (b), (c) show that as the weight of the drone increases, the system pays more and more attention to the flight and computing energy consumption of the drone. Since the goal is to minimize, the drone will upload the task as quickly as possible. At this time, the drone's hovering time becomes shorter and the energy consumption will decrease, which can be seen in Figure (b). Similarly, in Figure (c), the increase in weight will make the system pay less attention to the user's task completion rate, so there will be a downward trend. In these three figures, as the total energy consumption weight ω increases, it means that the system pays more and more attention to the total energy consumption of the user and the drone. It can be seen that both the user's energy consumption and the drone's energy consumption will decrease.

[0087] Depend on Figure 6 It can be seen that the overall objectives of the present invention and other inventions present the same trend. The hybrid scheme of local computing, backscattering and active transmission in the present invention achieves a weighted combination of lower energy consumption and higher completion rate. Therefore, the present invention is feasible and effective when performing computing task processing.

[0088] The above simulation analysis proves the correctness and effectiveness of the method proposed in the present invention.

[0089] 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 structures 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 to maximize the completion rate of UAV backscatter missions. It is characterized in that The steps include: (1) Using a UAV equipped with an edge computing server and a carrier transmitter and multiple local users equipped with backscattering devices, a backscattering and edge computing communication system is built; wherein the UAV and each local user are equipped with a single antenna; (2) Divide the target area into N sub-areas, let Θ i represents the i-th sub-region, i∈{1,2,…,N}; M i represents the number of users in the i-th sub-area; j = 1, 2, ..., M i represents the jth user in the i-th sub-area; (3) Obtain the total energy consumption E of the drone during its operation in the entire target area UAV And the total energy consumption of user equipment E User , the implementation steps are as follows: (3.1) The drone is hovered above the centroid of the i-th sub-area and transmits RF signals to all users in the area at a specific power to obtain the initial energy of user j after the signal is transmitted through the line-of-sight link. (3.2) The UAV continues to transmit RF signals to all users in the current sub-area at a specific power. The users use the received RF signals to backscatter the task data to be processed to the mobile edge computing MEC server of the UAV using time division multiple access, and convert the signals into energy for storage through line-of-sight link transmission in other time slots where backscattering is not performed. Among them, the circuit energy consumption generated by user j during the backscattering process is The number of bits of pending task data offloaded to the UAV via backscattering in the allocated time slot is The energy converted and stored is (3.3) The user unloads the remaining mission data after backscattering to the UAV through active transmission. Let the transmission power of user j in this active transmission process be P j [i], the number of bits of remaining task data to be offloaded is The total energy consumption is The task volume and energy consumption of local computing performed by user j are and (3.4) The drone carrying the mobile edge computing MEC server calculates all the task data, completes the computing task of the current sub-area, and records the amount of completed tasks (3.5) After completing the computational task in the current sub-area, the drone flies to the next sub-area until it completes the tasks in all sub-areas. That is, take i = 1, 2, …, N, and get the total energy consumption E of the drone. UAV And the total energy consumption of user equipment E User ; (4) Construct the optimal user task completion rate and total energy consumption weighted sum expression G': Where, Δ=[t e ',t b ',t a ',t c ',T'] represents the time allocation of each stage in the sub-region, t e ' is the duration of the energy collection phase, t b ' is the duration of the energy backscattering phase, t a ' is the duration of the active transmission phase, t c ' is the calculation time of the drone, T' is the hovering time of the drone in each area; α' is the backscatter coefficient, P' is the user's transmission power during active transmission, f' is the user's computing power allocation during local calculation; μ' is the user's task completion rate; the weighted sum of the user's task completion rate and total energy consumption G is obtained according to the following formula: Among them, ω represents the weight coefficient of total energy consumption, Represents the weight coefficient of UAV energy consumption, μ j [i] represents the task completion rate of user j in the i-th region; (5) Minimize the weighted sum of total energy consumption G under the backscattering and edge computing communication system, and convert the G expression into a convex expression through variable substitution, perspective function and continuous convex approximation algorithm; (6) By using standard convex optimization tools, the optimal solution of the user task completion rate and the weighted sum of total energy consumption G is obtained, that is, the optimal time allocation Δ * , optimal reflection coefficient α * , optimal transmission power P * And the optimal computing power allocation f * .

Images